Automotive innovation the science and engineering behind cutting edge automotive technology

Automotive Innovation

http://www.taylorandfrancis.com

Automotive Innovation

The Science and Engineering behind

Cutting-Edge Automotive Technology

Patrick Hossay

CRC Press

Taylor & Francis Group

6000 Broken Sound Parkway NW, Suite 300

Boca Raton, FL 33487-2742

CRC Press is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S. Government works

Printed on acid-free paper

International Standard Book Number-13: 978-1-138-61176-4 (Hardback)

This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been

made to publish reliable data and information, but the author and publisher cannot assume responsibility for the

validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the

publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let

us know so we may rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or

utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including

photocopying, microfilming, and recording, or in any information storage or retrieval system, without written

permission from the publishers.

For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://

www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA

01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users.

For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been

arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for

identification and explanation without intent to infringe.

Library of Congress Cataloging‑in‑Publication Data

Names: Hossay, Patrick, 1964- author.

Title: Automotive innovation : the science and engineering behind

cutting-edge automotive technology / Patrick Hossay.

Description: First edition. | Boca Raton, FL : CRC Press/Taylor & Francis

Group, [2020]

Identifiers: LCCN 2019009155 | ISBN 9781138611764 (hardback)

Subjects: LCSH: Automobiles—Technological innovations—Popular works. |

Automobiles—Design and construction—Popular works.

Classification: LCC TL240 .H655 2020 | DDC 629.2—dc23

LC record available at https://lccn.loc.gov/2019009155

Visit the Taylor & Francis Web site at

http://www.taylorandfrancis.com

and the CRC Press Web site at

http://www.crcpress.com

http://www.copyright.com

https://lccn.loc.gov

http://www.taylorandfrancis.com

http://www.crcpress.com

Contents

Preface ..............................................................................................................................................ix

Author ........................................................................................................................................... xiii

1. Bringing the Fire .....................................................................................................................1

What Is Gasoline? ....................................................................................................................2

The Engine ................................................................................................................................3

The Four Strokes ......................................................................................................................4

The Engine Comes Together ..................................................................................................8

Valve Train .............................................................................................................................. 11

Defining the Combustion Chamber .................................................................................... 13

Pistons ..................................................................................................................................... 16

The Head ................................................................................................................................. 19

Ignition ....................................................................................................................................23

Knocking ................................................................................................................................. 24

Fuel Delivery .......................................................................................................................... 27

Low-Temperature Combustion ............................................................................................34

2. The End of Compromise ..................................................................................................... 39

Advanced Digital Control..................................................................................................... 39

Sensor Technology .................................................................................................................40

Engine Control .......................................................................................................................42

Variable Valve Actuation ......................................................................................................45

Induction ................................................................................................................................. 52

Forced Induction ....................................................................................................................55

Compression Ratio .................................................................................................................58

3. Getting Power to the Pavement .........................................................................................65

What Do We Need a Drivetrain to Do? ..............................................................................65

Manual Transmission Coupler .............................................................................................72

Manual Transmission ............................................................................................................ 74

Automated Manual ................................................................................................................ 76

Automatic Transmission Coupler ........................................................................................80

Automatic Transmissions ..................................................................................................... 82

Transmission Control ............................................................................................................86

Continuously Variable Transmissions ................................................................................88

Differentials, AWD, and Torque Vectoring ........................................................................ 92

Advanced Tires and Control ................................................................................................ 96

4. Electric Machines ..................................................................................................................99

The Principles of the Electric Motor .................................................................................. 100

Making an Electric Machine .............................................................................................. 102

Motor Performance ..............................................................................................................

1.13

Advanced piston manufacturing.

Future piston structures may be defined with 3D printing. IAV Automotive Engineering is developing this pis-

ton with a honeycomb lattice to define the piston’s structure. IAV reports exceptional strength, about 20% less

weight than its conventional counterpart, and reduced thermal expansion.

18 Automotive Innovation

printed with a thin graphite-infused, low-friction patch. A recently developed piston coat-

ing that includes graphite, molybdenum disulfide, and carbon fiber promises a full 10%

friction reduction over an uncoated piston (Image 1.14).6

Even with this increasing sophistication and precision, the aluminum piston is now fac-

ing a challenge from a high-tech version of its old iron rival made of new high-strength

steel alloys for diesel passenger car applications. Because of the greater strength of this

new steel, the piston can be made smaller and lighter. In particular, the distance from

the piston pin to the upper surface of the piston, called the crown, can be shortened. This

allows for less mass and potentially a longer connecting rod and thus a larger displace-

ment in the same overall size engine. Or, with the same displacement, a shorter block

could be used, allowing a significant reduction in engine size and weight. Because steel

offers lower heat conductivity than aluminum, the steel piston must make use of cooling

galleries, which significantly complicates the manufacturing process since the piston must

be made in two parts. Because steel expands less with heat than aluminum, the challenge

of accommodating thermal expansion is eased a bit. Whether gasoline engine applications

of an advanced steel piston could be developed is still to be seen. We’ll look more closely

at advanced steel in Chapter 8.

Another piston option that has been touted by producers of aftermarket performance

parts but not by OEM producers is ceramic-coated piston crowns. The idea is to produce

a strong, insulating coating on the top of pistons. Zirconium ceramic fits the bill well, as

it provides good insulation and thermal expansion behavior that matches the metal and

thus allows for an enduring bond.7 The hope is that this insulation will help maintain the

heat energy in the cylinder, increasing thermal efficiency; and evidence indicates this is

6 M. Ross, “Pumped Up: Piston Evolution.” Engine Technology International.com March, 2015.

7 D. Das, K.R. Sharma and G. Majumdar, Review of emission characteristics of low heat rejection internal com-

bustion engines. International Journal of Environmental Engineering and Management 4 (4), 2013, 309–314.

IMAGE 1.14

Steel piston.

Aluminum pistons are now common in all production automobiles, with cast aluminum common in fleet vehi-

cles and hypereutectic or stronger forged aluminum used in performance applications, but pistons manufac-

tured from new steel alloys such as this are set to change this for diesel engines. This monotherm steel piston

made by Mahle was the first made for light passenger cars.

Source: Mahle

http://International.com

19Bringing the Fire

the case.8 However, the process can be expensive, the gains are modest, and the resulting

increased combustion temperature also tends to produce higher NOx emission, making

this a problematic adoption for production cars.9

As goes the piston, so goes the connecting rod. Like the piston, the rod has been exposed

to increasing pressure and heat. Performance rods are often made of forged steel, though

aluminum alloy rods are not unusual. Titanium offers extremely light and strong per-

formance rods. And because of a rod’s simpler design requirements, it’s easier and more

cost-effective to make rods out of titanium than pistons. Aluminum MMCs also offer some

promise. In fact, because it’s not exposed to the same extreme temperatures as the piston,

the connecting rod offers many promising avenues for weight reduction. Newer research is

examining the use of composite materials for connecting rods, with Lamborghini explor-

ing the possibility of a carbon fiber rod that is half the weight of a conventional steel rod.10

Similarly, the design of piston rings has evolved over time. Rings provide the seal

against the cylinder wall, but in defining that tight fit, they also account for a majority

of the friction in the engine. Classically, an upper ring provided a tight seal, called the

compression ring, and a lower ring ensured that engine oil from the crankcase didn’t

make its way in the combustion chamber, called an oil control ring or scraper. This lower

ring allows excess oil to be cleaned off the cylinder walls and returned to the crankcase

through oil drain holes in the ring groove, leaving only a slight lubricating film, enough

to allow movement of the piston without excessive fouling of the combustion chamber.

There are multiple variations in the number, cross-sectional design, and the placement of

the rings, though generally automobile manufacturers have moved to smaller ring packs

to reduce friction loss, with some recent compression rings less than a millimeter thick,

less than half the thickness of typical rings a decade ago.11 Ceramic coating is also used to

reduce ring friction and wear, and new application technologies are significantly reducing

the cost of this once-expensive option (Images 1.15 and 1.16).12

The Head

The head, and the valves it contains, define the last remaining wall of the combustion

chamber. The head must be engineered precisely enough to define the top end of the com-

bustion chamber, ensure a tight seal with the block, and provide precision-machined valve

seats that are able to maintain an exact fit at high temperatures.

8 K.S. Mahajan and S.H. Deshmukh, Structural and thermal analysis of piston. International Journal of

Current Engineering and Technology 5 (June), 2016, 22–29. Available at http://inpressco.com/category/ijcet;

K.Thiruselvam, Thermal barrier coatings in internal combustion engine. Journal of Chemical and Pharmaceutical

Sciences 7, 2015, 413–18; and A. Sh. Khusainov, A.A. Glushchenko, Theoretical prerequisites for lowering pis-

ton temperature in internal combustion engines. International Conference on Industrial Engineering, ICIE 2016

Procedia Engineering 150, 2016, 1363–1367.

9 D. Das, K.R. Sharma and G. Majumdar, Review of emission characteristics of low heat rejection internal

combustion engines. International Journal of Environmental Engineering and Management 4 (4) 2013, 309–314.

Available at http://www.ripublication.com/ ijeem.htm

10 D. Undercoffler, “Lambo Expands Carbon-fiber Footprint.” Automotive News July 4, 2016.

11 V.W. Wong and S.C. Tung, Overview of automotive engine friction and reduction trends—Effects of surface,

material, and lubricant-additive technologies. Friction 4(1), 2016, 1–28.

12 L. Kamo, P. Saad, W. Bryzik and M. Mekari, Ceramic coated piston rings for internal combustion engines.

Proceedings of WTC2005 World Tribology Congress III September 12–16, 2005, Washington, DC, USA.

http://inpressco.com

http://www.ripublication.com

20 Automotive Innovation

Perhaps the most noticeable variation in the design of the head is how many valves are

integrated into the combustion chamber and in exactly what way. While the classic engine

had one intake and one exhaust valve, by the early 1980s this convention was regularly

challenged in an effort to allow greater flow in and out of the cylinder. In fact, in the

IMAGE 1.16

Piston coatings.

Ceramic coatings can reduce surface friction by a third. This advanced coating by FederalMogul includes solid

lubricants and carbon fibers, offering reduced friction and significantly improved wear.

IMAGE 1.15

Advanced piston.

Pistons have advanced significantly. Ring packs have become much smaller and slipperier than they once were.

And cooling channels, such as the one in this Federal-Mogul Elastothermic piston, can reduce piston crown

temperatures by about

a fifth.

21Bringing the Fire

mid-1980s, an experimental Maserati V6 2.0-liter turbo used six valves per cylinder, defin-

ing a very tight fit. While six-valve heads are decidedly not common, four valves per cylin-

der certainly are. Particularly with the greater demands being placed on smaller engines,

multivalve heads accommodate higher rpm, allowing the engine to breath at a much faster

rate. Because the incoming charge is cooler than the high-heat, high-energy exhaust, a

common variation is to increase the size or number of the intake valves specifically. This

allows more area for the relatively cool, heavy and slower-moving intake charge to enter

the cylinder. There’s reason for caution, however. More valves are not always better and

can complicate the valve train and constrain head design options. Similarly, the common

performance upgrade of installing oversized valves in hopes of producing more horse-

power can be a fool’s errand, as the valve size is frequently not the performance bottleneck.

The larger question is how do we integrate these valves into the head, and what geom-

etry will we use to shape the upper combustion chamber? There are three primary con-

siderations at play in answering this question: First, maintaining the high energy and

full emulsification of the incoming fuel–air mixture, principally by producing an agitated

flow into the chamber that keeps the kinetic energy of the incoming charge high. Second,

intensifying the energy of the vaporized charge in the last stage of compression to prepare

the charge for combustion, we call this squish. And, third, removing the exhausted charge

in the chamber quickly and completely after combustion to allow it to be replaced with a

fresh charge, we call this scavenging. Let’s look at each of these.

A priority is to ensure a highly energized flow into the cylinder as this helps produce a

desirable and fast-moving flame front upon combustion. The placement of the valves and

shape of the piston are vital here. Positioning the valves so that the intake occurs obliquely,

with both an end-over-end rotation, called tumble, and a spiral rotation of the mixture,

called swirl, is the goal. Without this energy, the charge might lose the integrity of its tar-

geted mixture profile, and the resulting flame front from a low energy charge may move

too slowly for a high-speed engine (Image 1.17).

A clean and fast flame front is also encouraged by shaping the piston and head to define

a desired squish pattern. Squish is the final squeezing of the charge in the last bit of the

compression stroke. Ideally this injects a burst of energy into the fuel–air mixture. With

contemporary engines designed to make maximum use of this effect, the area between the

IMAGE 1.17

Swirl and tumble.

Proper swirl and tumble patterns are key variables in the vaporization and emulsification of charge delivery

in a high-performing engine. Producing a high-energy charge promotes a faster flame front that can meet the

needs of high-speed performance.

22 Automotive Innovation

head and piston at TDC, called the clearance volume, is tiny by the standards of past gen-

erations. Modern engine pistons come within a single millimeter of the head. This brings

the charge into an unstable and highly energized state, greatly speeding the rate of com-

bustion and allowing the piston to absorb as much of the combustion energy as possible.

The chamber shape and piston head are designed to make maximum use of this energy.

But there’s a trick to this: you need to avoid preignition and maintain low emission of NOx

which tend to increase with rising ignition temperature. We’ll look at just how that’s done

in the next chapter.

Scavenging is yet another consideration. The shape and dynamics that keep an incom-

ing charge energized and turbulent can also be used the help eject the exhaust out of the

cylinder. One method is allowing for some valve overlap, a short period when the exhaust

valve hasn’t yet completely closed and the intake valve begins to open. This can allow for

cross-flow thru the combustion chamber, with some of the energy of the incoming air

being used to push the remaining exhaust out of the chamber. The effect can be facili-

tated by placing the valves opposed to each other so that flow from one to the other drafts

through the cylinder.

So, how do we shape the head to ensure that we achieve all these aims? As you might

expect, there’s no single answer. Over the years, there have been any number of com-

bustion chamber designs. Early ‘flat heads’, for example, had valves either side by side

(called a T head) or across from each other (L head) that faced upward and were adjacent

to the cylinder. Once common, these configurations are now obsolete, as they can’t offer

the sort of compression and control desired in modern, high-performance automobiles.

More recently, engines have tended toward some variation of three head designs. The first

two utilize a so-called I-head that mounts valves facing the piston directly and can thus

allow the upward facing valve stems to connect with one or two camshafts in the head,

thus called an overhead cam. These can commonly entail either an inverted cup-shaped

chamber that has earned it the name bathtub chamber, or an angled design with one side

higher than the other, forming a wedge chamber.

A heart-shaped chamber is a common variation on the bathtub design that is defined

by two squish regions, typically a large circle at the spark plug and a smaller one on the

opposing end. The result forms a crescent or heart shape. The spark plug in the center of

the head favors a desirable flame front. However, having the valves close to each other

makes heat transfer a problem, limiting octane tolerance of the engine (Image 1.18).

A third common configuration is the hemispherical head. This may be the most well-

known head design, largely thanks to Dodge’s promotion of its ‘hemi’ engine. The rounded

chamber top offers good geometry for turbulence and facilitates the use of opposing valve

placement and so allows excellent cross-flow scavenging. In addition, because the valves

are on opposing sides of the head, they are more thermally isolated. This helps keep the

valves cooler and so helps avoiding knocking. A variant of this design can accommodate

multiple valves and is called a pentroof, or penta, design. Because it must provide multiple

valve seats, it tends to have flatter sides.

New innovations in the material and manufacturing have also helped improve valve

design. Stainless steel valves are now common and available in various alloys that can

improve hardness and heat resistance. Recently developed titanium valves can be 40%

lighter than stainless steel valves. As previously discussed, the weight of oscillating com-

ponents can be definitive to performance. Valve weight in particular can be a limiting

factor on maximum engine speed. Valves that are even just a slight bit lighter can signifi-

cantly reduce the energy needed to operate the valves and so allow for more aggressive

valve lift. This can mean significant rewards in engine response and speed. Sodium-filled

23Bringing the Fire

valves offer another interesting example. As the sodium liquefies with heat, it shifts back

and forth through the stem of the valve, transferring heat away from the head, another

example of shaker cooling.

Ignition

With the combustion chamber defined, and the charge primed and ready for ignition, the

obvious question arises: when do we ‘light ‘er up’? The ignition of the charge has to come

at the right time, not too soon, since this might cause the cylinder pressure to rise too high

before the piston completes the compression stroke; and not too late, since this might mean

we’ll lose some of our ability to absorb the combustion energy before the end of the power

stroke.

IMAGE 1.18

Chamber designs.

Just about every manufacturer has used some version

of a wedge design (upper left). With the valves aligned

side by side on the long wall of an asymmetric wedge, the spark plug is placed on the opposing short side. The

walls of this design allow for advantageous tumble. And the intense push of the charge from the narrow to the

full side of the wedge at the late point of compression offers desirable squish. The heart-shaped head (upper

right) represents a common head design. Note the close proximity of the valves.

The bathtub chamber (lover left) defines a symmetrical head shape that can place valves upright or at slight

opposing angles. The hemispherical head (lower right) may be the most well-known head design, largely

thanks to Dodge’s promotion of its ‘hemi’ engine with excellent cross-flow scavenging and good thermal isola-

tion for the valves, the design deserves some praise.

24 Automotive Innovation

Because ignition timing needs to change while the engine operates, it’s not quite as sim-

ple as the timing of the valve and piston movement. These two parts can move in synch

with no trouble; every turn of the crankshaft needs half a turn of the camshaft, no matter

what engine speed, no matter what driving conditions. But, for ignition timing, the situ-

ation changes. As previously noted, combustion is a process. The flame front initiated by

the spark plug is not instantaneous, it takes time to pass through the combustion chamber,

and that time varies as a result of the energy in the charge and the richness of the mixture.

Leaner air–fuel ratios define a slower-moving flame front. High energy due to intense

compression or high heat can speed the flame front. And, of course, the engine moves

at dramatically different speeds, sometimes rotating at an idle, say about 800 rpm, and

sometimes at as much as 6,000 or more rpm. Because we need the flame front to develop

it’s push at just the right time, this means we need to adjust the ignition event to account

for engine speed and mixture, so that the main push of the flame front occurs when the

piston is early in the downward stroke and can absorb the energy effectively.

Consequently, to ensure proper timing of the flame front, we need to initiate ignition

before the piston reaches the top of the compression stroke, or TDC. This is called igni-

tion advance. If we didn’t have an advance, and the spark plug ignited at TDC, the main

push of combustion would occur late in the power stroke, and the piston would reach the

bottom of the stroke before the combustion force was complete. The faster the engine is

rotating, the worse this effect would get, since the flame front moves at a relatively fixed

rate and won’t speed up to match engine speed. We measure the advance of the ignition in

degrees of crankshaft rotation before TDC, or BTDC, and so talk about advance in number

of degrees.

This means that at low rpm, we might want the spark plug to fire 15° BTDC. But at high

rpm, we might need ignition to be 30° or more BTDC, to allow the expanding combustion

gasses to fill the chamber and provide a maximum amount of power to the piston well

before the piston reaches the end of the power stroke (Image 1.19). Older cars had a simple

mechanism that could allow for two timing settings, one for low rpm and one for high

rpm. New cars allow for much better ignition timing because the engine’s operations are

computer controlled (lots more on this in the next chapter). Ideally, we’d like to advance the

ignition timing as much as we can, to allow us to harvest as much of the energy from the

power stroke as possible (Image 1.20). The problem is, if we advance the timing too much,

we can create or exacerbate engine knocking.

Knocking

In a perfect engine, the arcing of the spark plug will ignite a critical flame kernel that will

in turn begin the propagation of an even flame front and a resulting fast and powerful but

smooth push against the piston head. An abrupt, explosive blow, even though powerful,

won’t do. Think of this like a playground merry-go-round. To keep it moving, you need

even and firm pushes. Imagine trying to get that merry-go-round spinning by slamming

it with a sledgehammer. When this sort of uncontrolled or uneven combustion happens in

an engine, we call it knocking. This is a general term that actually refers to two things that

can go wrong: preignition and detonation.

Preignition, as the name implies, is when the ignition process starts early through

autoignition, before the spark plug fires. To fully understand what causes this, we need

25Bringing the Fire

IMAGE 1.19

Ignition advance.

Ensuring that combustion takes place just after the piston reaches top dead center (TDC) requires having initial

ignition take place well before TDC, and be adjusted to accommodate engine speed. The faster the engine speed,

the greater the ignition advance required.

IMAGE 1.20

Combustion chamber pressure.

Ideally, maximum combustion chamber pressure is reached just after TDC, about 15° or so, allowing for maxi-

mum recovery of combustion energy by the piston. To ensure proper scavenging, the exhaust valve opens

(EVO) before combustion pressure has completely dropped.

26 Automotive Innovation

to recall a key characteristic of gasoline—octane rating. Remember that the octane rating

gives us a sense of the fuel’s tolerance to compression. A high octane rating means the fuel

can be compressed more without fear of triggering autoignition, and a low octane rating

indicates that the fuel is more likely to spontaneously ignite under relatively lower pres-

sure. So, if an engine is run on a fuel with an octane rating that is too low, the heat and

pressure caused by the compression stroke could trigger autoignition before the spark

plug fires. This might be made more likely if the spark plugs or valves get overheated, add-

ing excess heat to the compression stroke. This preignition is clearly not a good thing. If

ignition happens too early, the force of combustion could actually push against the piston

while it’s still in the compression stroke, effectively trying to turn the engine against its

direction of rotation.

A more common form of knocking is detonation. Unlike preignition, detonation does

not begin before the spark plug fires. Instead, the spark-initiated ignition increases the heat

and pressure in the combustion chamber, causing spontaneous autoignition in at least one

other location in the cylinder (Image 1.21). So, combustion in the chamber happens in more

than one location at a time, and instead of having a singular kernel that leads to a unified

and even push, we have multiple points of ignition that lead to a more disorganized com-

bustion and a much less even push against the piston. The resulting violent combustion is

a frequent cause of damage to bearings and pistons.

This threat of detonation is the core challenge of ignition timing. One cause of detona-

tion could be excessive ignition advance. If the spark plug fires too early, the cylinder pres-

sure will increase while the combustion chamber is still relatively large and before the

flame front can travel through the chamber, leading to one or more points of unburned

IMAGE 1.21

Knocking.

Desired ignition is defined by a single ignition point and flame front in the combustion chamber. Knocking

can be caused by preignition, defined by the autoignition of the fuel–air mixture before the spark plug fires, or

detonation, defined by multiple ignition points caused by the increased heat and pressure resulting from spark

ignition.

27Bringing the Fire

fuel igniting on their own as the chamber’s pressure and heat increase beyond the fuel’s

autoignition point. The solution is to back off on the ignition advance, and allow the piston

to travel a bit more before initializing combustion. But, if you back off too much, as dis-

cussed, you’ll lose the ability to harvest some of the energy from combustion. The balance

can be tricky.

A central variable in this effort is how much

we compress the charge. This is defined by

the engine’s compression ratio, defining the amount the air is squeezed prior to combus-

tion. This ratio is defined by the proportion of the volume of the combustion chamber

when it’s at it’s largest (bottom dead center) to the volume at its smallest (top dead center).

We called the smallest volume clearance volume (most often written Vc), the change in

volume caused by piston movement is called displacement volume (Vd), and so the total

volume when the piston is at BDC is the two volumes combined. That means the compres-

sion ratio is (Vc + Vd)/Vc.

Taking what we’ve already discussed into account, we can see that defining an engine’s

compression ratio is a balancing act too. A higher compression ratio can allow us to draw

more power from combustion. This could enable us to gain significant improvements in

fuel economy or power, as we’re able to draw a much greater amount of energy from a

given amount of fuel. But, as the compression ratio goes up, so does the need for higher

octane fuel that can withstand the higher pressure without autoignition; and, if we com-

press the charge too much, we invite preignition and detonation. Preignition is more likely

when the engine is operating at high load since the engine temperature goes up. So, a

high-performance engine might have a compression ratio (CR) 8:1 to avoid knocking at

high engine loads, while an engine targeting high efficiency might be closer to 13:1. With

advanced control discussed in the next chapter and appropriate fuel, compression ratios

as high as 14:1 are possible. By comparison, diesel engines start at 14:1 and go to about 23:1.

This adds complexity to much of what we’ve discussed so far, since how we manage

heat, engine speed, and knocking are all shaped by the compression ratio. For example,

a higher compression ratio will increase engine heat, but cooling the piston temperature

with galleries could reduce the threat of detonation and thus allow a higher compression

ratio. Or a richer mixture can speed the rate of combustion, but it could also cool the com-

bustion chamber and may reduce the possibility of detonation.

Fuel Delivery

Delivering fuel and air to the combustion chamber can be tricky too. To be ready for com-

bustion, the fuel must be atomized (suspended in small bits), emulsified (fully blended with

air), and vaporized (changed into a stable and consistent gas). As we discussed previously,

without adequate access to oxygen, gasoline will not burn. So, the delivery of fuel and the

delivery of air are irrevocably linked. Complicating this somewhat is our desire to avoid

the production of nitrogen oxides. NOx is formed when combustion temperatures are high

enough to break apart the nitrogen in the incoming air supply. As a result, exhaust gas

recirculation (EGR) has been used since the 1970s to return a small portion of the exhaust

gases to the combustion chamber in the next intake stroke as a way of cooling the tempera-

ture of combustion and reducing the formation of these undesirable pollutants.

Early engines did not offer any real opportunity for precise control of the mixture. The

engine that powered the Wright brother’s first flight, for example, simply spilled fuel on a

28 Automotive Innovation

manifold and used the engine’s heat to help vaporize it. And into the 1980s, automobiles

utilized carburetors that channeled the engine’s incoming air through a narrow passage, a

venturi. The resulting vacuum simply sucked up gas from an adjacent small reservoir. The

system worked adequately, but the mixture was imperfect, and would not meet the needs

of today’s engines. Carburetors were replaced by more precisely controlled spray nozzles

for gasoline, fuel injection, by the start of the 1990s. The early systems simply sprayed

fuel through a nozzle into the central throttle that replaced the carburetor, called throttle

body injection (TBI). In the end, TBI was not too different from the carbureted systems

they supplanted.

Both of these systems, the TBI and carburetors, presented an inherent challenge.

Because the air and fuel were mixed at the start of the intake manifold, and had to travel

through the intake manifold channels to get into the chamber, it was always a challenge

to ensurethe mixture remained fully emulsified and atomized as it made its way to the

cylinder. Suspended fuel particles are heavier than the surrounding air, and thus inclined

to settle if the flow calms or slows. So, this was particularly a problem at low engine

speeds when induction flow slowed. To address this, the manifold had to be designed to

ensure continued high-energy turbulent flow to maintain suspension of the fuel particles.

Intuitively, you’d want the path of an intake manifold to be as smooth and large as pos-

sible so that plenty of air can reach the cylinders with minimal resistance. But, a large open

intake would mean low-energy, slow-moving air, which could easily cause the fuel to drop

out of the mixture. Similarly, if you need to ensure the air–fuel mixture remains energized,

smooth walls and wide turns are not your friend. You want sharp edges, rough walls

and sharp corners to stimulate turbulence. However, such a design also meant high flow

resistance, and at higher engine speed this meant a significant power loss. So, the result

was a significant compromise in high-end power to ensure continued emulsification of the

air–fuel mixture at low engine speeds.

Current systems, however, get around this challenge by adding fuel just before the cyl-

inder, in the case of port injection, or in the cylinder itself, with direct injection. This

eliminates an extended run of fuel–air mixture through a manifold, and thus allows for

more precise and independent control of both air and fuel. Manifolds for a port or direct-

injected engine carried only air and so could be made smooth and streamlined. This

reduced pumping losses and so improved efficiency. An open chamber at the start of the

induction system can act as an air reservoir to slow the incoming air and allow a resulting

pressure increase. This air reservoir, or plenum, can dampen pressure fluctuations and

provide higher density air to the inlet channels that feed each cylinder with air, called

intake runners.

The most common fuel system, port injection, sprays the fuel on the backside of the valve

just outside the cylinder. Designing the spray pattern can ensure desirable atomization

and a spray and inlet geometry that promotes a beneficial swirl and tumble pattern. The

result is a fuel flow that is far more precisely defined than previous carburetor or central-

ized injection systems. While older designs sprayed fuel continuously, newer sequential

port fuel injection provides pulses timed with the intake cycle.

More recently, direct injection offers a notable uptick in precision and control. Borrowing

once again from the diesel world, rather than mixing the fuel and air before it enters the

cylinder, gasoline direct injection (GDI) injects a precise spray of fuel in the cylinder itself.

The basic idea has been around for a while, but lack of capable control mechanisms and

the challenge of high-pressure precision injection have made this idea problematic and

expensive until very recently. The resulting control over the amount and pattern of the

fuel delivery offers substantial advantages in fuel economy and performance. Like diesel

29Bringing the Fire

engines, GDI provides a major decrease in pumping loss at low speed by providing a very

lean charge and avoiding the energy loss resulting from drawing intake air through a

narrow throttle. Fuel pulses can be altered for a lean burn when acceleration and torque

demand is low, a stoichiometric burn for moderate driving, or a rich burn for cold starts or

when increased power is demanded.

Direct injection not only allows for precise mixture control, but it can also allow us to

more easily vary the distribution

of the mixture throughout the charge, defining rich and

lean areas of the combustion chamber, though not all GDI systems are capable of this. It

is this characteristic that allows for a very lean operation without jeopardizing ignition

(Image 1.22). A charge with a consistent air–fuel mixture throughout its volume is called

a hom*ogeneous charge, and this is the rough characteristic of a conventional engine’s

intake. The challenge is that a very lean hom*ogeneous charge may not provide enough

fuel to initiate combustion at the spark plug. But the richness required to initiate combus-

tion may be excessively rich at low-load conditions and may invite knocking. So, our desire

to provide a much more lean charge at low loads can be achieved with a stratified charge,

defined by a cloud of rich mixture around the sparkplug tapering to a leaner mixture as

you move through the chamber. The ability to lean the mixture outside of the initial igni-

tion region significantly reduces the threat of knocking, allows for improved efficiency,

and so is a major potential benefit of GDI.

Achieving a precise stratified charge is no easy task. Injection takes place in the later

stage of the compression stroke, allowing a very short amount of time to deliver and vapor-

ize the charge. The required injection speed and pressure to make this happen requires

a high-pressure fuel delivery that can provide fuel at nearly 1,500 psi, some 30 times the

40–70 psi typically present in a port injection fuel manifold. Shaping the distribution of

IMAGE 1.22

Advanced ignition systems.

Very lean operation or extra EGR can improve efficiency, but it can also make achieving a reliable ignition of

the charge challenging. Advanced ignition systems can help address this challenge. For example, BorgWarner’s

EcoFlash system utilizes high-frequency ignition (HFI) to produce an extended corona discharge. The system

can ignite the hydrocarbons throughout the combustion chamber virtually simultaneously, allowing for the

precise and reliable ignition of a very lean charge. Comparing the corona image on the left to a conventional

sparkplug on the right gives a clear sense of the difference.

Image: BorgWarner

30 Automotive Innovation

this high-pressure injection is also a challenge, and has generally been accomplished with

three approaches: a wall-guided process that uses a specifically designed piston head and

the chamber wall to shape the injection spray; an air-guided process that relies on swirl

and tumble pattern; or a spray-guided process, using precisely defined mist from an injec-

tor adjacent to the plug to place a rich cloud directly at the plug location.13 While any

direct-injection system will have some characteristics of each of these approaches, each

system typically adopts one as the primary determinant of chamber dynamics.

A wall-guided, or geometry-guided, approach has the advantage of allowing the great-

est amount of time between injection and ignition, since injection begins much earlier in

the compression stroke. Typically, the injector is located at the side of the chamber near

the intake valves, with the spark plug at the center of the chamber. Proximity to the intake

valves allows the fuel spray to mix with the incoming air more easily, and helps keep the

injector cooler. A specially designed piston head receives the injection spray and shapes

it in the combustion chamber as it moves upward (Image 1.23). Often a deep bowl pis-

ton head is used. The challenge is that this piston geometry can make the attainment of

a hom*ogeneous charge more difficult. And, the spraying of fuel directly on the piston

and cylinder walls can lead to incomplete vaporization, or wetting, and resulting higher

emissions of CO and unburnt hydrocarbons.14 As a result, this approach is not typically

used alone.

An air-guided system is similar, though the swirl of the intake is more definitive in

shaping the charge, so the design of the inlet ducts becomes more critical and the shape of

13 G. Fiengo, A. di Gaeta, A. Palladino, and V. Giglio, Common Rail System for GDI Engines. Springer Briefs in

Electrical and Computer Engineering. Springer, London, 2013.

14 G. Fiengo, A. di Gaeta, A. Palladino, and V. Giglio, Common Rail System for GDI Engines. Springer Briefs in

Electrical and Computer Engineering. Springer, London, 2013.

IMAGE 1.23

Piston crown.

A wall-guided GDI system relies on the interaction of the injector spray pattern and piston head geometry

to shape the incoming fuel charge. The shape of the piston crown in this Buick Enclave engine is designed to

receive the injector spray.

31Bringing the Fire

the piston head less so. The intake ports are specifically designed to develop the targeted

swirl and tumble pattern, often oriented more vertically and using baffles or narrowing

to define a distinct tumble pattern that will help concentrate the fuel mixture at the spark-

plug. In some systems with two intake valves, only one value operates in stratified mode,

to allow for a more defined, high-velocity pattern. The challenge is that while the piston

head geometry is not as vital to defining the charge geometry and so piston wetting is

reduced, a flow-guided system does not fully resolve the performance and efficiency limits

at high load previously mentioned.

These challenges can be largely met with more recently developed spray-guided GDI

systems. This approach locates the injector close to the spark plug in the center of the

chamber and can therefore define more precise spray dynamics (Image 1.24). This allows

for a more reliable and symmetrical fuel distribution and more effective utilization of the

air charge, and so results in the best efficiency and performance at differing engine speeds.

Better injector control can also eliminate wetting and allow for a much higher level of EGR,

as the rich cloud around the sparkplug will avoid the misfire otherwise associated with a

higher level of exhaust gas in the chamber. However, the injector location next to the spark

plug presents some challenges as well. It can cause undesirable heating of the injector and

IMAGE 1.24

Mazda SkyActiv fuel injection.

Mazda uses a notable piston cavity to define a stratified air–fuel mixture around the spark plug. This allows

delayed ignition timing during warm-up to speed emissions system warming and helps avoid the flame front

contacting the piston head too soon.

Image: Mazda

32 Automotive Innovation

lead to plug fouling and soot formation. The soot deposit can also lead to disturbed injec-

tor orifices and thus compromised spray patterns. So, while a spray-defined system can

offer greater efficiency, it also requires not only a robust sparkplug that can accommodate

the thermal stress of repeated cooling and heating, but a very precise control of the fuel

delivery (Image 1.25).

The needed precise control is provided by what is called a piezoelectric injector. This

injection technology utilizes a crystalline material that we will see a lot of in advanced

automotive components. Within the crystalline structure are opposing positive and nega-

tive charges, called dipoles. When the crystal is exposed to an electric charge these dipoles

react, and the material physically expands. In this case, this expansion can be used with a

levering mechanism that amplifies the movement to open and close an injector orifice. The

result is far more precise control of the fuel spray, with pulses as short as a few millisec-

onds. Classic solenoid injectors utilize electromagnets to actuate injection pulses. While

cost-effective, this technology has a longer response time and is not capable of graduated

control, making it difficult to provide the precision spray needed for GDI systems. By con-

trast, varying the voltage applied to piezo injector allows for variation in the extent of the

reaction and regulated control of the injector opening. The result is a highly responsive

very precise delivery of

fuel that can accurately control the volume of fuel injection, from

1 to 150 mg/pulse. This also enables the delivery of multiple pulses per cycle, called split

injection, and can enable very lean operation.15

15 C. Park, S. Kim, H. Kim, S. Lee, C. Kim and Y. Moriyoshi, Effect of a split-injection strategy on the per-

formance of stratified lean combustion for a gasoline direct-injection engine. Proceedings of the Institute of

Mechanical Engineers: Journal of Automotive Engineering 225 (10), 2011, 1415–1426.

IMAGE 1.25

The piezoelectric injector.

This piezo common rail injector by Continental offers an impressive example of a cutting-edge piezoelectric

application. With an ability to handle pressures over 36,000 psi (2,500 bar), a piezo-servo drive, and self- adjusting

servo valve and needle control, this unit can provide extremely precise fuel regulation, variable injection pat-

terns and up to ten injections per cycle. All of this means more precise control, which translates to improved

fuel economy and reduced emissions.

Image: Continental AG

33Bringing the Fire

Split injection can enable a small preinjection during the intake stroke, for example,

which can help cool the combustion chamber. This allows for the exploitation of what is

called the heat of vaporization, defined by the heat absorbed by the fuel as it transitions

from suspended liquid droplets to a vapor. As the atomized fuel in this preinjection pulse

vaporizes it absorbs heat energy, not unlike the heat removed from your body as perspi-

ration evaporates. This cooling increases the density of the air in the cylinder and more

importantly helps avoid knock by cooling the combustion chamber.16 The remaining fuel

is delivered later with a longer pulsation. In some diesel engines, piezoelectric injectors

can provide up to seven injections per cycle.

The use of high-pressure direct injection does not exclude the use of lower pressure port

injection; in fact, the two can be used to compliment each other. So, while direct injec-

tion offers a cooler combustion chamber and so helps avoid knocking, a port injection

system can cool the incoming air and thus allow for its increased density and a resulting

improved air intake efficiency (see the next chapter for volumetric efficiency). This allows

the best of both worlds. The cooling effect of GDI on the combustion chamber can reduce

knocking and potentially allow for a higher compression ratio for greater efficiency. And

by allowing more time for the fuel to vaporize, port injection can reduce some of the chal-

lenges of exclusive direct injection. Direct injection can deliver fuel at high engine loads,

when chamber cooling is important, and a combination can be used at lower loads to help

keep injectors clear and efficiency high.

Combining DI and PI can also avoid some of the challenges of direct injection alone.

The need to deliver some 50 times the fuel pressure required by port injection systems

can result in increased engine noise. In addition, while not universal, there is potential

for more pronounced carbon buildup on some GDI engines. By injecting fuel directly on

the backside (or tulip) of the valves, port injection actually helped keep valves clean. This

effect is lost when the fuel is injected directly in the cylinder. More critically, during rapid

acceleration, a GDI system will be challenged to transition quickly from low-end stratified

charge to a rich hom*ogeneous charge as the engine accelerates. Combining GDI with PI

can help address each of these issues. The combination can also allow for other innova-

tions. For example, the Toyota’s D-4S engine couples this with slits on the side of the injec-

tor that are periodically charged to blow the carbon off the outside of the injector as an

automated self-clearing mechanism.17

Mercedes has been using a cutting-edge form of multiple method fuel delivery, called

Turbulent Jet Injection (TJI) in its F1 cars since 2014; Ferrari is looking to follow suit soon.

TJI replaces the spark plug with a jet injector inside a small ignition chamber (Image 1.26).

While the main combustion chamber is supplied with the majority of the charge, about 3%

of the fuel is injected in this small adjacent chamber, providing an excessively rich mixture

around the plug. Ignition triggers a rapid flame jet streaming through precisely defined noz-

zles into the main combustion chamber. The fuel is ignited simultaneously at several points

in the top chamber, causing more effective combustion, more power, and lower emission

while avoiding knocking. This allows for an ultra-lean burn, with a fuel–air ratio exceeding

29:1 (λ = 2), and theoretically approaching 60:1. Several manufacturers have worked over the

past two decades to define an ultra-lean burning engine that would be highly efficient and

clean. The availability of precision fuel injection has now made that possible.

16 S.P. Chincholkara and Dr. J.G. Suryawanshib, Gasoline direct injection: An efficient technology. 5th

International Conference on Advances in Energy Research, ICAER 2015, December 15–17, 2015, Mumbai, India

Energy Procedia 90 2016, 666–672.

17 C. Schweinsberg, “Toyota Advances D4S with Self-Cleaning Feature on Tacoma.” Wardsauto.com August 27,

2015.

http://Wardsauto.com

34 Automotive Innovation

Low-Temperature Combustion

Similar technology is being explored to pursue the possibility of what is called low-

temperature combustion (LTC). The idea is a gasoline-fueled engine that operates with

compression ignition like a diesel, and so burns at a much lower temperature. The goal

is to reduce the heat of combustion so that we can dramatically reduce the generation

of harmful NOx caused by excessive heat, or PM caused by excessive richness, and reap

higher fuel efficiency. Some of these aims are achieved in current production cars with

EGR, recirculating about 10% of the exhaust gas back into the combustion chamber to cool

the combustion chamber. This is a well-established technology. But a promising route for

significant further improvement is to design a gasoline engine that can capture the sort

of thermal efficiency that has long existed in diesel engines, while avoiding the emissions

challenges of diesel fuel. Because gasoline offers a much lower autoignition temperature

than diesel, this sort of design could only be possible with the precise fuel and combustion

control offered by modern fuel systems.

hom*ogeneous charge compression ignition (HCCI) is being examined as a possible

path to this sort of ultra-efficient combustion. A lean, well-mixed fuel–air charge in the

intake stroke is ignited by the heat of compression rather than a spark plug. Again, the

hope is to achieve the efficiency of a diesel engine with the emission of a gasoline engine.

With rising pressure and even mixture distributed throughout the combustion chamber,

ignition takes place as a spontaneous reaction throughout the cylinder, producing a rapid

IMAGE 1.26

Turbulent jet injection (TJI).

The TJI system replaces the spark plug with a small precombustion ignition chamber that can trigger a rapid

flame jet ignition into the main combustion chamber.

Source: Mahle

35Bringing the Fire

and distributed pressure rise rather than a flame front. The low temperature and fast reac-

tion means very low production of NOx.

A key to making this work is in the fuel. In HCCI, all fuel has to fully vaporize prior to

the reaction, or remaining liquid droplets will lead to undesirable emissions. Even more

importantly, fuel characteristics are critical to ignition timing in HCCI, since the system

relies on spontaneous autoignition, without the timing assistance of either a spark plug or

a timed injector. So, in HCCI fuel chemistry plays a determinant role. Since the charge is

definitively uniform, variations in fuel mixture or vaporization distribution that shaped

spark ignition engines do not exist. As you might guess, typical pump gasoline is generally

not precise enough for such a system. The imprecise autoignition point of pump gas makes

it hard to achieve a predictable ignition, particularly at low load. Instead, a precisely engi-

neered fuel with a well-defined octane rating is needed. All of this means HCCI needs very

accurate control. And HCCI experiments in the laboratory, where control is relatively easy,

have at times lead to more CO and more unburned fuel in the emissions.18 A turbocharger,

which we will discuss in the next chapter, can help produce more complete combustion by

providing more air to the combustion chamber to help ensure that all the fuel is completely

burned. But, ensuring a complete and controlled HCCI burn remains a challenge.19

The challenge is that we need to provide a mixture that is rich enough to support autoig-

nition, but not so rich that it can’t all be efficiently burned once combustion has started. A

solution might come from stratified charge compression ignition (SCCI). At partial loads,

the mixture can be leaned to ensure efficiency. If combined with compression ignition

technology, the result could be highly efficient.

The trick is that we need to be able to ensure a desirable burn under varying engine

conditions and loads. The perfect compression ignition engine really needs a fuel that is

somewhere between gasoline and diesel, and the precise point changes as the engine load

changes. A diesel engine defines a precise ignition point by injecting the fuel late in the

process, timed to generate the targeted ignition sequence. But, this generates plenty of NOx

and soot. Gasoline typically produces less NOx and soot, but ignition through compression

is hard to control. Current research in what’s called Reactively Controlled Compression

Ignition (RCCI) tries to bridge this by providing a primary charge of gasoline through port

injection and a smaller secondary injection of diesel through direct injection. Fuel reactivity

could even potentially be precisely controlled with multiple injections of diesel to produce

a specific gradient of fuel mixtures in the squish region and outward. Or an alternative fuel

with a more defined octane rating might be used to more precisely determine the ignition

point.20 The result is high efficiency and low emissions, possibly the best of both worlds.

18 A.A. Hairuddin, A.P. Wandel and T. Yusaf, An introduction to a hom*ogeneous charge compression igni-

tion engine. Journal of Mechanical Engineering and Sciences 7 (December), 2014, 1042–1052; and P. Kumar and

A.Rehman, hom*ogeneous charge compression ignition (HCCI) combustion engine—A review. IOSR Journal

of Mechanical and Civil Engineering 11 (6) Ver. II, 2014, 47–67.

19 S. Saxena and I.D. Bedoya, Fundamental phenomena affecting low temperature combustion and HCCI

engines, high load limits and strategies for extending these limits. Progress in Energy and Combustion

Science 39, 2013, 457–488; M.M. Hasan and M.M. Rahman, hom*ogeneous charge compression ignition

combustion: Advantages over compression ignition combustion, challenges and solutions. Renewable and

Sustainable Energy Reviews 57 (May), 2016, 282–291; and M. Izadi Najafabadi and N.A. Aziz, hom*ogeneous

charge compression ignition combustion: Challenges and proposed solutions. Journal of Combustion

57(May), 2013, 1–14.

20 S. Saxena and I.D. Bedoya, Fundamental phenomena affecting low temperature combustion and HCCI

engines, high load limits and strategies for extending these limits. Progress in Energy and Combustion Science

39, 2013, 457–488; and S.L. Kokjohn, R.M. Hanson, D.A. Splitter and R.D. Reitz, Fuel reactivity controlled com-

pression ignition (RCCI): A pathway to controlled high-efficiency clean combustion. International Journal of

Engine Research 12 Special issue paper 209.

36 Automotive Innovation

If this sounds like science fiction, it’s not. Ford has looked at an ethanol boost ‘Bobcat’

Engine to possibly replace its 6.7 liter diesel in its super duty truck. The ethanol boosting

system (EBS) runs as a typical port injection gasoline engine at modest loads. But to meet

high-load demands, the engine injects E85 ethanol from a separate tank. With the high

octane and cooling capacity of E85, the engine can accommodate more compression and

produce more power on demand. The engine isn’t ready for prime time, but Ford and

others are looking for ways they can provide all the power and torque of diesel, with-

out all the expensive diesel emissions treatment equipment. Just as impressive, Mazda

is placing a variant of an HCCI engine in its next-generation Mazda 3, defining the first

production car with a form of gasoline compression ignition. Its supercharged SkyActiv-X

engine will offer an ultra-lean burn mode that can decrease emission by nearly a third, the

manufacturer claims (Image 1.27).21

21 A. Stoklosa, “Driving Mazda’s Next Mazda 3 with Its Skyactiv-X Compression-Ignition Gas Engine.” Car and

Driver September, 2017.

IMAGE 1.27

Spark-controlled compression ignition.

Mazda’s SkyActiv-X engine offers what the carmaker is calling a spark-controlled compression ignition (SPCCI)

engine. The aim is to combine the reliability and reduced emissions of spark ignition with the fuel economy of

compression ignition. While the idea of spark-controlled compression ignition may seem oxymoronic, the basic

principle is a sort of hybrid of spark and compression ignition that can offer improved efficiency by enabling a

lean mixture to burn more evenly and reliably. Ignition is initiated by the spark plug, which raises the cylinder

pressure so that the remaining lean charge in the cylinder ignites via compression. Under conditions when

compression ignition is not ideal, the engine can shift back to traditional spark ignition.

Image: Mazda

37Bringing the Fire

None of these innovations would have been possible a generation ago. All of them

require a more sophisticated engine management system that can precisely determine the

vehicle’s driving conditions and engine parameters, and deliver fuel exactly as needed.

Taking into account torque demand, engine speed, and other operating parameters, the

engine management system can alter the fuel and type of charge to suit changing con-

ditions. A hom*ogeneous charge could be used for improved emissions throughout the

range. At a modest torque range, a simple stratified charge can be implemented, adding

EGR at low torque range to improve emissions. When more power is desired, a duel injec-

tion could occur, defining a hom*ogeneous charge in most of the cylinder and a localized

enriched charge near the plug. This sort of control capacity has fundamentally altered the

possibilities for the internal combustion engine, and is the subject of the next chapter.

http://www.taylorandfrancis.com

The End of Compromise

For a century, automotive engine design has been a game of compromise. Designing an

engine for high engine speed meant it would be rough and inefficient at low speed. An

engine designed for low-end torque meant limited high-speed performance. Similarly,

anintake designed for efficient cruising often meant compromised acceleration. Or a high

compression ratio meant for efficient cruising could invite knocking in high-end operation.

However, the enhanced control capacity made possible with more sophisticated micro-

processors and advanced materials and design are quickly making such compromises a

thing of the past. Increasingly, we are able to change the fundamental characteristics of

the engine while it’s running, to effectively remake the engine into what we want when

we want it.

Advanced Digital Control

A fundamental element that has made this possible is the use of increasingly powerful

computer systems that manage and actually modify the basic characteristics of the engine

while it operates. This has allowed for an ever-increasing level of control and innovation.

These dedicated computer systems, or embedded systems, function with an array

actuators, sensors, microprocessors, controllers and instruments to replace and enhance

mechanical functions with more precise and smarter electronics. So, they add considerable

functionality and capacity to today’s automobile, but they also add complexity.

While enhanced digital control has reset the bar for automotive performance and rede-

fined what’s possible, it’s also worth noting a few challenges as we begin. Automotive

digital control technology has developed irregularly, and its evolution is far from com-

plete. Multiple stand-alone control units marked the early adoption of embedded systems,

and soon progressed to an array of microcontrollers, sensors, and associated functions.

Like the broader systems overall, the software has developed incrementally, without a

clear overarching architecture or plan.1 The result is somewhat of a technological patch-

work quilt in the modern car. Each manufacturer defines a distinct control architecture,

with their own priorities, approaches, and functions. A typical automobile manufactured

today may have well over 100 electronic control units (ECUs), with multiple intercon-

necting buses that allow devices to communicate with each other. Because each hardware

component typically includes its unique software, every new function or feature of a car

usually means the introduction of an additional control unit. The result is what can be

called a highly heterogeneous system architecture; and the resulting challenge of software

1 R. Hegde, G. Mishra and K.S. Gurumurthy, Software and hardware design challenges in automotive embed-

ded systems. International Journal of VLSI Design and Communication Systems 2 (3), 2011, 164–174.

40 Automotive Innovation

integration and development has been overwhelming, with tens of millions of lines of

code, and thousands of signals travel between different ECUs.2

As cost, complexity, and redundancy increase and ECUs reach the limit of their pro-

cessing power, interest increases in developing a more integrated control architecture for

vehicles with organized interconnected networks and serial sharing of data and infor-

mation.3 The up-integration of control functions, use of more powerful microcontrollers

that can manage multiple functions, multicore processors that can allow one control unit

to take on a greater number of operations, integration of functions located near each into

single controllers, and other innovations promise to cut the inefficiency, complexity, and

cost of current automotive digital systems significantly.4 In the same vein, an international

effort to develop a universal software architecture for vehicle systems, called Automotive

Open System Architecture Consortium (AUTOSAR), could allow for improvements and

additions of functions to a vehicle without the high initial time and cost investments.5 The

appeal of a standardized modular system with standardized interconnects and scalability

is clear.

Still, while there are some challenges to work out, the digitalization of the automobile

also offers unprecedented opportunity. Just about every aspect of the modern automobile

has become ‘smarter’, including not just the engine, but also the transmission, brakes, sus-

pension, steering, safety features, and the increasing sophisticated infotainment system.

Most notably, the engine control module (ECM) of a modern car receives signals from

dozens of sensors to actively adjust ignition timing, intake parameters, fuel mixture, injec-

tion pulses, and other variables to more ideally tune the engine for enhanced performance,

fuel economy, and emissions reduction. This requires ongoing communication between

a processor capable of running millions of processes a second with an array of sensors

that measure the mass of incoming air, the fuel mixture, engine knocking, system voltage,

engine temperature, oil pressure, atmospheric pressure, vehicle speed, exhaust quality,

crankshaft position, camshaft position, and many other factors. It gets complicated for

sure, but the result is a car that is better, cleaner, and more efficient than ever thought

possible.

Sensor Technology

Understanding the new digital landscape requires that we first understand what infor-

mation we’re working with and what sensors are in place. And there are many. Either

optical or magnetic sensors are used as crankshaft position sensors (CKP) and cam-

shaft position sensors (CMP) to allow tracking of rotation within a very few degrees.

Updated versions of engine coolant temp sensors (ECT), throttle position sensors (TPS),

2 R. Coppola and M. Morisio, Connected car: Technologies, issues, future trends. ACM Computing Surveys 49 (3),

2016, 1-36; and A. Sangiovanni-Vincentelli and M. Di Natale, Embedded system design for automotive applica-

tions. IEEE Computer Society 40 (10), 2007, 42–51.

3 “Automotive Technology: Greener Vehicles, Changing Skills.” Electronics, Software and Controls Report,

Center for Automotive Research (CAR), May 2011.

4 Ibid; and R. Hegde, G. Mishra, and K.S. Gurumurthy, Software and hardware design challenges in automotive

embedded systems. International Journal of VLSI design & Communication Systems 2 (3), 2011, 165–174.

5 J. Schlegel, “Technical foundations for the development of automotive embedded systems.” Available at: hpi.

de/fileadmin/user_upload/fachgebiete/giese/Ausarbeitungen_AUTOSAR0809/Technical_Foundations_

schlegel.pdf

hpi.de/fileadmin/user_upload/fachgebiete/giese/Ausarbeitungen_AUTOSAR0809/Technical_Foundations_schlegel.pdf

41The End of Compromise

vehiclespeed sensors (VSS), fuel pressure sensors (FPS), and others provided a more pre-

cise, responsive, digitized version of the mechanical sensors we once relied on. It is no lon-

ger sufficient to simply measure coolant temperature; modern vehicles now continuously

monitor intake air temperature, outside air temperature, engine coolant temperature, oil

temperature, transmission fluid temperature, and exhaust temperature. Other sensors are

more distinctly new and innovative, measuring manifold air pressure, engine knocking,

and emissions quality.

A key variable in engine control is the mass of incoming air. A manifold air pressure

(MAP) sensor functions like a strain gauge, sensing the slightest deformations of a sensing

element to measure air pressure. Older systems used a combination of MAP and the engine

speed to calculate the approximate amount of air entering the engine. More recently, mass

airflow (MAF) sensors allow for a direct and more accurate measure. Because air changes

its density with temperature, simply measuring the volume of air entering an engine

would not be sufficient. It would miss the thinning of the air with rising temperature or

altitude and so be inexact. An MAF sensor suspends a heated wire in the path of incoming

air. The electrical resistance of the wire changes with temperature, so by precisely measur-

ing the change in resistance of this heated wire, we are able to assess the cooling effect of

the mass of air passing the wire. The integrated circuited converts the resistance reading

to a useful signal and the ECM is able to calculate the mass of incoming air quickly and

with precision. Cool, right? (Image 2.1).

As discussed in the previous chapter, the ability to identify detonation or preignition is

vital to engine control, and accomplished with a knock sensor (KS). Typically located in

the engine block or head, this sensor allows the ECM to adjust engine timing to resolve

knocking. Piezo materials make this possible. You’ll remember from Chapter 1 that piezo-

electric materials alter their shape when exposed to a charge. This also works in reverse,

when the crystal experiences a change in its shape, it generates an electric charge. When

squeezed or shaken (or knocked), the dipoles are disturbed, and rearrange, resulting in a

small electric charge across the crystal. That charge sends a signal to the ECM, which then

knows to adjust timing.

IMAGE 2.1

Measuring air.

This mass airflow sensor (MAF) uses a ‘hot wire’, a heated wire with a resistance that changes predictably with

temperature, to measure the cooling effect of incoming airflow. This then allows us to know the air’s mass.

42 Automotive Innovation

A key in defining a cleaner car, and correctly managing the fuel mixture, is the ability to

assess the emissions from the engine, and thus enable responsive management of the fuel

mixture. The oxygen sensor, also called a lambda sensor, does this by evaluating oxygen

content in the exhaust. The sensor usually consists of a thin, thimble-shaped component

of ceramic zirconium dioxide (ZrO2) coated with platinum on either side. The thimble is

placed in the exhaust pipe, so one side is exposed to the exhaust gas and one side is exposed

to outside air. The difference in exposure to oxygen causes a chemical reaction that pro-

duces charged particles of oxygen (ions) traveling through the ceramic membrane. As the

difference in oxygen levels on the two sides varies, so does the flow of ions. This produces a

voltage difference, not unlike a tiny battery, but in this case called a Nernst cell. You might

also think of this as a difference in the pressure of oxygen on either side of the sensor,

leading to a flow of oxygen particles across the zirconium dioxide. The resulting voltage

difference won’t tell you the actual percentage of oxygen, but it does allow you to identify

changes in oxygen concentration compared to the outside air, and that’s good enough to see

if the engine is burning lean or rich. When the mixture is rich, there is very little oxygen left

in the exhaust and the voltage increases. When the mixture is lean, more oxygen is left in

the exhaust and a lower voltage results. Because the membrane is only conductive at high

temperatures (600°F), a heating element is contained in the sensor housing. The exhaust

would eventually heat the sensor, but a heater allows for reliable readings sooner.

Of course, this description is partial, and highly simplified, highlighting a few charac-

teristics of the ECM architecture and some key sensors. Actual operation includes multiple

ongoing diagnostic routines providing continuous analysis, data, and communications

potential through the on-board diagnostic (OBD) system. It includes multiple modes

of communication between the powertrain, chassis, suspension, differential, and other

vehicle sensors and mechanisms. And, as we’ll see in Chapter 9, automotive control sys-

tems increasingly incorporate rapid expansion in the integration of echo detection sensors

such as radar, lidar, and sonar, as well as navigation and communication technology from

advanced GPS to automated vehicle-to-vehicle coordination, and countless additional fea-

tures. We will explore the possibilities and implications of these innovations later, but for

now, let’s look at how enhanced digital control has redefined the operation of the internal

combustion engine.

Engine Control

The automotive advances made possible through digital control are nowhere more impres-

sive than in the performance of the internal combustion engine. Let’s begin with three

primary aims to digital engine control: managing the fuel mixture, ignition timing, and

exhaust gas recirculation in engines so equipped. (We’ll see in a bit that other functions,

such as varying valve operation, are also in play.) Digital control allows for the identifi-

cation of distinct operational modes, based on driving conditions and driver demands,

with targeted engine parameters within each mode. Once again, this allows us to avoid

the compromises of previous generations, and define distinct targeted priorities and cor-

responding control logic for each mode. While manufactures differ in the precise differ-

entiation, most ECMs include some variation of seven basic modes: engine start, engine

warm-up, normal cruising with both an open-loop and closed-loop variation, high load

(acceleration), deceleration (often with a high and low variant), andidle.

43The End of Compromise

Upon engine initial start, the ECM crank control logic is automatically selected and

delivers the rich mixture required to start the engine, pulsing the fuel injectors to promote

a quick start up. The target air–fuel ratio will be determined by engine temperature, and

may vary anywhere from an extreme rich mixture of 2:1 to a modestly rich 12:1, as defined

by data stored in a programmable read-only memory (PROM). Engine characteristics will

change this greatly, of course; with a GDI engine potentially requiring a much less rich

mixture thanks to more precise fuel metering, for example. The ECM may also check bat-

tery voltage, to determine any needed operating compensation, as fuel injector timing and

response speed can be significantly affected by low voltage.

Once idle rpm is achieved, the ECM switches to idle mode, maintaining an air–fuel

ratio that ensures the lowest stable engine idle while avoiding stall. The thermostat will

ensure that the coolant is automatically recirculated through a short loop that bypasses

the radiator to assist in quick engine warm-up, as has been done for decades, but now an

electrically assisted thermostat can also manage the ideal temperature for optimal engine

operation dependent on initial driving demands. Called a MAP-controlled thermostat

since the Manifold Air Pressure sensor provides an indication of engine load to the ECM,

this feature allows the thermostat to alter warm-up parameters to suit engine load while

addressing combustion and engine temperature, NOx reduction, and fuel economy. While

this is done, the oxygen sensor is warmed, and idle speed, injector pulse duration, and

mixture are progressively adjusted to suit engine temperature.

Once the engine has warmed up, the ECM will enter open-loop operation, meaning that

the ECM will rely on limited sensor data and continue to use PROM fuel ratio data, some-

times called a lookup table, to manage fuel mixture until the O2 sensors can provide useful

information. This is not ideal, as the ECM principally uses engine speed to define igni-

tion timing without compensating for the important impact load has on proper timing.

Once the O2 sensors are warm and providing reliable information, signals from the oxygen

sensors in the exhaust can be used to modify mixture for ideal performance and clean

operation. So, the ECM can move from a feed-forward control to closed-loop operation

and begin to reduce emissions levels quickly. Because virtually every car manufacturer in

the US utilizes a three-way catalyst (TWC) for emissions management, which works best

when combustion is stoichiometric, keeping the fuel mixture at or near the stoichiometric

ratio is a key concern (Image 2.2). Using information from the MAF, the controller is able to

determine the airflow into the combustion chamber to manage the amount of fuel through

the injectors.

Unlike open-loop operation, closed-loop control allows for automatic adaptation to

mechanical conditions. The ECM is managing more than just mixture; it may be defin-

ing fuel injector pulse duration, spark timing, and multiple additional variables based

on engine temperature, MAF, barometric pressure, engine speed, knocking, and other

variables. Again, a key concern in this process is knocking. The ECM will automatically

advance spark timing to achieve improved performance. However, the desirability of spark

advance is limited by the threat of denotation. Once the KS identifies excessive knocking,

spark timing is moved slightly later, or retarded. Different manufactures do this differ-

ently; but generally, the ECM will drop the advance until knocking is eliminated, and then

progressively increase it

until the knock point is identified, allowing high-performance

operation at the cusp of the knock limit. The ECM may even make micro adjustments dur-

ing operation based on learned past knocking events, a practice Subaru called Fine Knock

Learning. Like many ECM functions, managing ignition timing while avoiding knocking

requires precise and fast-processing capacity, as the engine rotates a full 18° in a single

millisecond at 3,000 rpm.

44 Automotive Innovation

The ECM’s monitoring and responsive control is not limited to mixture and spark timing

of course. Multiple variables are monitored and the ECM may initiate additional adjust-

ments at this point. The ECM will initialize EGR, allowing a percentage of exhaust gas to

be cooled and recirculated into the combustion chamber in an effort to cool combustion

temperatures and reduce the production of NOx. The ECM may trigger the delivery of

secondary air to the catalytic converter to ensure complete combustion of remaining emis-

sion pollutants if it has a secondary air injection (SAI) system. System voltage may require

adjustment to fuel injector signals. More generally, a range of powertrain variations may

be initiated. For example, if the vehicle is equipped with a lock-up torque converter (we’ll

look at this in the next chapter), it may signal a solenoid to lock the clutch mechanism

and therefore eliminate efficiency-draining slip. In sum, closed-loop operation allows

for sophisticated control logic that ensures engine performance as operating conditions

change, both in the management of combustion and the vehicle more generally.

During high-load demand, signaled by the TPS or possibly an additional sensor that identi-

fies wide-open throttle, the ECM will again switch modes to provide a rich mixture that can

meet the demand for hard acceleration. This rich mixture will temporarily degrade emissions

control and fuel economy; but engineers and regulators view this as an acceptable compro-

mise to provide demanded acceleration, as acceleration capacity is at times needed for safety.

Upon deceleration, the ECM will lean the fuel mixture to allow for reduced emission

and improved fuel economy. This also helps avoid an exhaust backfire caused by excessive

rich mixture. If the deceleration is severe, the lean-out may be extended to a complete fuel

cut-off, allowing for more effective engine braking and reduced emissions. Signals from

the MAP and TPS will bring the ECM back into closed-loop mode quickly to avoid stall-

ing, or the ECM will return to idle control mode when the vehicle is stopped.

This might seem complicated, but it’s actually a highly simplified rendition of the ECM

operation. We’re not even considering the integration with the transmission operation,

IMAGE 2.2

Three-way catalytic converter operation.

Although NOx emission can be higher at a stoichiometric fuel ratio, the catalytic converter works most effec-

tively to reduce unburned hydrocarbons, carbon monoxide, and NOx at stoichiometry. So, in order to ensure a

clean burning car, the ECM targets stoichiometry in closed-loop operation.

45The End of Compromise

fuel system evaporative emissions control, cabin air conditioning and heating demand,

and a host of other variables and functions. But the principle point is clear: ECMs are

defining a move beyond conventional control practices. These are no longer single-input/

single-output systems (SISO). They define increasingly complex, multivariate operations

with sophisticated algorithms and multiple control strategies.6 Data from disparate sen-

sors are combined to create performance information that exceeds the capacity of any one

sensor, a process called sensor fusion. The resulting control structure moves toward a

form of artificial intelligence, with the ability to monitor, adapt, and modify control logic

by ‘learning’ from past outcomes. Future ECMs may determine what ignition timing, fuel

management, or response to knocking work best for your engine, or even your driving,

and adapt accordingly. We’ll see more on these sorts of systems in Chapter 9.

Variable Valve Actuation

All of this control technology has allowed for a level of management of the engine’s pro-

cesses that was unthinkable a generation ago. We’ve already seen the results of this in mix-

ture and ignition timing management, though it is perhaps even more impressive in the

control of valve movement, a variable that was once mechanically fixed with no possible

variation during operation. More varied and precise control of the valves allows us to once

again make an end run around the compromises that were once unavoidable. Where once

we had to make a tough choice in defining valve movement, now we can have our cake and

eat it too. We can design variable valve events that allow our cars to feel like a sports car at

high revs without sacrificing efficiency when cruising.

As discussed in the previous chapter, in the classic internal combustion engine, the

intake valve opens and closes during the intake stroke; and the exhaust valve opens and

closes during the exhaust stroke. The problem with this scenario is that the velocity of

incoming air does not vary to suit the engine speed. As the engine speeds up, like the

flame front, the speed of the incoming charge often cannot keep up. With a conventional

cam, the valve lifts from its seat pretty slowly, so while the valve may be ‘open’ it takes

some time for air to begin to flow freely through the opening. In fact, flow through an open

valve is significantly restricted for perhaps two-thirds of the total valve opening time.7 The

solution may be to keep the valves open longer, but it’s not that easy. The ideal valve opera-

tion changes with engine speed or load. In a perfect world, the opening and closing of each

valve could be timed to the precise needs of the engine at a given load.

A key variation in the operation of valves at high engine speeds and loads is to open the

intake valve earlier and close the exhaust valve later, defining valve overlap (Image2.3). In

addition to allowing more time for air movement, as discussed previously, it can enhance

scavenging, allowing the incoming charge to help push out the outgoing exhaust. Typically,

the opening of the intake valve (IVO) is timed to begin about 10° BTDC, to allow the valve

to fully open for the intake stroke. And, the resulting aggressive pulsation of pressure

in the exhaust manifold as all cylinders exhibit powerful cross flow can create pressure

waves in the exhaust manifold that can help actively draw the exhaust out of the cylinder.

6 H. Morris, Control systems in automobiles. In J. Haappian-Smith (ed) Modern Vehicle Design. Butterworth-

Heinemann, Oxford, 2001.

7 R. Stone and J.K. Ball, Automotive Engineering Fundamentals. SAE International, Warrendale, PA, 2004.

46 Automotive Innovation

At high engine speeds, this is desirable, but the advantages of a large overlap diminish at

low engine speeds. In fact, at low loads, keeping some of the exhaust in the combustion

chamber is actually a good thing. It means less incoming air is needed when the throttle

is more closed, decreasing pumping losses. And, like the effect of the EGR system, more

exhaust gas in the chamber at low loads can mean lower NOx emissions. In effect, this is a

bit like making the combustion chamber smaller at lower engine speeds.

The advantage of longer valve opening isn’t limited to the benefits of overlap. There is

also an advantage to an early exhaust valve opening. Ideally, we’d want to extract all the

power we can out of the expanding combustion gas before we open the exhaust valve.

However, pushing all the exhaust gas out of the combustion chamber also takes energy.

And, opening the exhaust valve before reaching the end of the power stroke can help us

save some of the energy required to push out the still-pressurized exhaust gas at the start

of the exhaust stroke, called blowdown loss. At high load and

108

Torque and Power ................................................................................................................ 110

Cooling .................................................................................................................................. 112

vi Contents

Induction Motor ................................................................................................................... 113

Permanent-Magnet Machines ............................................................................................ 118

Magnets ................................................................................................................................. 121

BLPM Control ....................................................................................................................... 122

Reluctance Machines ........................................................................................................... 124

Advanced Motor Possibilities ............................................................................................ 126

5. Electrified Powertrains ...................................................................................................... 131

Gas versus Electrons? .......................................................................................................... 131

Hybrid Drive ......................................................................................................................... 133

Baby Steps ............................................................................................................................. 135

Mild Hybrid .......................................................................................................................... 137

Full Hybrid ........................................................................................................................... 141

Adding a Plug ....................................................................................................................... 148

Power ..................................................................................................................................... 148

Electric Vehicle ..................................................................................................................... 151

Electric Car Viability ........................................................................................................... 155

Using Energy Effectively .................................................................................................... 160

6. The Electric Fuel Tank ....................................................................................................... 165

What’s a Battery? .................................................................................................................. 166

Battery Performance ............................................................................................................ 171

Battery Management ........................................................................................................... 172

Cell Balancing ....................................................................................................................... 175

Cooling Systems ................................................................................................................... 176

Battery Chemistry ................................................................................................................ 179

Nickel-Based Batteries ......................................................................................................... 181

Lithium .................................................................................................................................. 184

Future Possibilities............................................................................................................... 190

7. Automotive Architecture .................................................................................................. 195

General Chassis Design ...................................................................................................... 195

Frames ................................................................................................................................... 197

Crashworthiness ..................................................................................................................200

Materials ................................................................................................................................ 203

Alternative Metals ............................................................................................................... 210

Manufacturing Metal .......................................................................................................... 216

Plastics ................................................................................................................................... 218

Suspension ............................................................................................................................ 224

Chassis Control .................................................................................................................... 226

Bringing It All Together ......................................................................................................229

Modularity ............................................................................................................................230

8. The Power of Shape ............................................................................................................233

The Nature of Drag ..............................................................................................................234

The Power of Shape .............................................................................................................235

Boundary Layer .................................................................................................................... 237

The Shape of a Car ...............................................................................................................238

The Front of the Car ............................................................................................................. 240

viiContents

Addressing the Rear Wake ................................................................................................. 243

Three Dimensional Flow .................................................................................................... 247

Vortex Generators ................................................................................................................250

Lift .......................................................................................................................................... 251

The Ground ...........................................................................................................................253

Wheels ................................................................................................................................... 257

Bringing the Body Together ............................................................................................... 259

Active Aerodynamics .......................................................................................................... 261

9. Smarter Cars ........................................................................................................................ 265

Smarter Driving ................................................................................................................... 266

Perception.............................................................................................................................. 272

Radar ......................................................................................................................................

high engine speed, an early

opening of the exhaust valve will allow more time for the pressure inside the chamber to

drop before the piston begins to move up. But, in low-load conditions, the early exhaust

opening means lost energy and torque, since there’s plenty of time for the pressure to drop

before the exhaust stroke.

On the other end of the stroke, we can consider the closing of the intake valve. For great-

est power at high load, we would want to time the closing of the intake valve to ensure the

largest mass of incoming charge. You might expect this to be at the bottom of the stroke;

however, the incoming air through the intake manifold develops a momentum that has

the effect of pushing a wave of air into the cylinder, thus favoring a late closing of the

intake valve at high engine speeds. The intake pressure wave tends to maximize late in the

stroke. So, taking full advantage of this effect, and maximizing the mass of incoming air,

requires we close the intake valve a bit after BDC. At high rpm, this might be significantly

longer than at low power where a late valve closing might be detrimental as it diminishes

the compression stroke.

Similar challenges present themselves with defining valve lift. We might be tempted to

presume that maximum valve lift is always desirable, since this would allow for the least

restrictive flow when the valve is open and so faster intake flow and improved exhaust

exit. This is true at high rpm, but, at low rpm, there is a cost to high valve lift. High valve

IMAGE 2.3

Valve overlap.

Delaying the closing of the exhaust valve and advancing the opening of the intake valve create an overlap that

will allow for more effective scavenging of the combustion chamber by allowing cross flow of incoming air.

47The End of Compromise

lift means air can flow into the chamber slowly and calmly at low rpm, losing turbulence

and risking wetting.8 Additionally, at partial loads, a smaller valve opening reduces the

need for a more restrictive throttle, reducing pumping losses. So, while a smaller valve

opening is severely detrimental to the engine’s high-load performance, a high lift opening

can deteriorate low-load performance.9 In the past, we conventionally identified a lift pat-

tern that defined a compromise, but that’s no longer necessary.

Over the past two decades, control of valve timing and lift has become increasingly

more sophisticated. Systems that allow a basic phase change of the camshaft for discrete

high and low speed cam timing are now common. With a single cam, a valve phase change

can retard valve timing by roughly 10°, allowing for significantly improved high-load

operation with an earlier intake opening. If the engine has separate intake and exhaust

cams, each might be adjusted independently, allowing for greater allowance for engine

load and speed. However, while allowing for two or more discreet valve timing settings is

an improvement, the possibility of a progressive variation in valve timing to suit changing

conditions would be better (Image 2.4).

8 T. Wang, D. Liu, G. Wang, B. Tan and Z. Peng, Effects of variable valve lift on in-cylinder air motion. Energies

8, 2015, 13778–13795.

9 F. Uysal, The effects of a pneumatic-driven variable valve timing mechanism on the performance of an otto

engine. Journal of Mechanical Engineering 61 (11), 2015, 632–640.

IMAGE 2.4

Toyota dynamic force engine.

Toyota’s 2.0-liter, direct-injection, in-line, 4-cylinder engine uses advanced combustion control technology and

variable valve timing to achieve an impressive 40% thermal efficiency.

Image: Toyota Motor Company

48 Automotive Innovation

In fact, greater capacity for variation and control is possible. Cam profile switching

(CPS) typically uses two sets of distinct cam lobes, or otherwise shift between two distinct

cam profiles. As a result, both timing and lift can be changed. But once again, only as a

choice between two options, typically a close-to-conventional cam for normal operation,

and a cam designed for performance operation at high rpm. Manufacturers have defined

multiple sophisticated systems that can allow for control of both lift and timing, and can

offer incremental or continuous variable valve timing (VVT), variable valve lift (VVL),

or both. BMW, for example, uses a sprocket on a helical spline of the camshaft to progres-

sively adjust valve timing; and their Valvetronic variable valve system utilizes an electron-

ically adjusted cam and rocker arm to vary the relationship between the camshaft and the

valve stem allowing variation in valve lift from 0.3 to 9.7 mm (Image 2.5). This means valve

lift variation can be used to control engine speed, and once again eliminate throttling loss

from a conventional throttle.

There are nearly as many ways to accomplish VVT as there are manufacturers. Honda

was an early adopter of variable valve profiles with its VTEC engine. The engine switched

between two cam profiles, allowing smooth operation and efficiency at low rpm, with

greater performance possibilities when the engine switched to the more aggressive cam at

4,500 rpm. Honda’s recent engines include a three-stage VTEC with separate cams for low,

IMAGE 2.5

BMW’s VANOS variable valve timing system.

BMW’s VANOS variable valve timing system places a sprocket on a helical spline at the end of the camshaft.

The sprocket moves in and out with oil pressure, causing a progressive shift in the relative alignment of the

camshaft and thus a change in timing.

Source: BMW

49The End of Compromise

medium, and high engine speeds. GM’s Intake Valve Lift Control uses an adjustable cam

to shift the pivot point of the rocker arms and thus vary valve lift. FIAT’s MultiAir sys-

tem utilizes an electrohydraulic system, using pressurized engine oil to actuate the intake

valves, allowing for control of both lift and timing.

As mentioned, more complete control of valve events can allow for a change in the fun-

damental parameters of the engine’s operation, even redefining the basic four-stroke pro-

cess. Atkinson cycle engine operation offers a great example. The idea of an Atkinson cycle

has long defined a variation from the classic Otto cycle four-stroke engine. Dating back to

the mid-1800s, the basic idea is this: since there is more combustion power to be extracted

at the end of the power stroke, the Atkinson Cycle allows for a longer power stroke rather

than the equal length strokes of the Otto Cycle (Image 2.6). The extra piston travel allows

the engine to usefully extract more energy from combustion. The Toyota Prius was notably

the first to mimic this cycle in a production car by delaying the close of the intake valve.

This allows some of the intake charge to return to the manifold, effectively shortening the

intake stroke, and by relation, making the power stroke comparatively longer. The result

is a significant improvement in efficiency, but a loss of power. Compressing the incoming

air with a super or turbo charger can help compensate for this shorter compression stroke

and recover lost power, defining a variant called the Miller Cycle. Multiple car manufac-

turers now use VVT to produce a variant of this Miller/Atkinson Cycle. So, this isn’t just

about fine tuning engine operation; these systems are basically changing the fundamental

operation of the four-stroke engine on the fly as conditions warrant. Is it just me, or is that

really, really cool?

IMAGE 2.6

Atkinson cycle engine.

The Atkinson cycle originally dates to an idea initiated by John Atkinson in the mid-1800s. His design used a

multilink connection to allow for a significantly longer power stroke and a shorter intake stroke, allowing the

piston to harvest a greater proportion of the combustion energy than is possible when the two strokes are equal.

50 Automotive Innovation

Equally amazing, advanced valve control can effectively change the size of the engine

as needed, on the fly. The ability to effectively shut

down cylinders by closing both valves

allows for cylinder deactivation, letting an engine perform as though some of its cylin-

ders are gone, so acting as a form of variable displacement. Fuel delivery is cut off, and

both intake and exhaust valves remain closed on the deactivated cylinders (Image 2.7). The

sealed cylinder then operates as an air spring, with the energy of the upward piston increas-

ing cylinder pressure and redelivering the energy to the piston on the down stroke. Well

over 90% of the energy is recovered, allowing the pistons to simply go along for the ride,

neither helping nor seriously hindering the engine when not needed. When more power

is demanded, the cylinders are seamlessly reactivated. The result is a significant improve-

ment in efficiency at low load while maintaining power capacity when needed, in effect

allowing a V8 to perform like a four cylinder at cruise but like a full V8 when needed.10

These technologies are impressive, but the full potential of variable valve control is

marked by the promise of individualized and precise control for each valve independently.

The idea is to have each individual valve operated discretely, perhaps with an electromag-

net as is done on some large ship engines or with a pneumatic actuator, as is done with a

prototype engine made by the Swedish firm Freevalve. With computer control, timing and

lift of each valve may be altered to suit operating conditions. A key advantage is the imme-

diate opening and closing of the valves. With a camshaft, opening and closing takes time,

defined by the elliptical cam lobe, and this dictates valve timing. With a solenoid-controlled

pneumatic system, moving a valve from closed to fully open is much quicker, defining a

nearly square valve lift profile, and once again requiring less compromise (Image 2.8).

In addition, the weigh and space savings is significant. No camshaft, no pushrods, no

rocker arms, no rollers. The resulting friction reduction is also worth noting. The leading

10 F. Millo, M. Mirzaeian, S. Luisi, V. Doria and A. Stroppiana, Engine displacement modularity for enhancing

automotive s.i. engines efficiency at part load. Fuel 180 (September), 2016, 645–652.

IMAGE 2.7

Mazda cylinder deactivation.

Hydraulic lash adjusters on the first and fourth cylinders are used to shift the pivot point of the rockers and

define variable displacement. Normally, the rocker arm pivots around the center point; but when cylinders are

deactivated, the pivot point becomes the valve side so the valves remain closed. Because a four-cylinder engine

is more likely to experience heightened vibration when only two cylinders are firing, a centrifugal pendulum

damper in the automatic transmission is used to help reduce vibration.

Image: Mazda

51The End of Compromise

manufacturer, Freevalve, has reported a 40% peak power increase from an in-line four-

cylinder turbocharged test engine, with improved fuel economy and lower cold start emis-

sions, in a package that is significantly lighter, smaller, and no more expensive than a

comparable high-end engine (Image 2.9).11

11 Author communications with Andreas Möller at Freevalve.

IMAGE 2.8

Camless valve lift.

Controlling valves with independent solenoids allows for faster valve movement than can be achieved with

rounded cam lobes, defining a more square lift profile and thus the potential for more precise valve timing.

IMAGE 2.9

Freevalve.

The Swedish firm Freevalve has defined a radical departure from con-

ventional engine design with its electrically controlled pneumatic valve

actuators.

Source: Freevalve

52 Automotive Innovation

The possibilities are intriguing, to say the least. This sort of control capacity could allow

for more sophisticated and precise variation in engine operation. Exact valve control can

be used to define a part-time HCCI or Miller-cycle engine. It could allow for multiple cyl-

inder deactivation modes. It might enable an internal EGR mode. Or, it may well do all of

the above.

Induction

If digital control allows for more exact delivery of fuel and spark, it can also allow for

improved delivery of air to the combustion chamber. As described in the previous chapter,

an internal combustion engine can be thought of as a large air pump, with a performance

engine at high speed requiring up to 40 pounds of air, or about 500 ft3, every minute. The

challenge of delivering this air at a pressure and velocity that can readily scavenge and fill

a combustion chamber in the few milliseconds the intake valve is open can determine the

capacity of an engine.

The aim is to provide air to the combustion chamber as effectively as possible to allow

complete combustion, an effort measured as volumetric efficiency (VE). VE is defined as a

ratio of the incoming air volume to the volume of the chamber. So, since each cylinder has

one intake stroke with every two rotations of the crankshaft, the VE of an engine would

be 100% if the air intake were half the engine displacement with every revolution. This

assumes that the displacement volume is the same as the total volume of the cylinder,

which ignores the clearance volume. But, the clearance volume is very small, so for sim-

plicity we use displacement volume. So, an engine operating at a 100% VE is taking in its

total displacement with every two crankshaft revolutions. Of course, it is also important

that this air be evenly allocated with equal distribution of incoming air to each cylinder.

While a typical VE for a conventional, naturally aspirated engine might be 75%–90%,

higher is possible with precise tuning and control.

Pressure is the strongest determinant in defining VE. And pressure in the induction

system is not constant. During an intake stroke, the engine draws in air-reducing induc-

tion pressure. This defines a low-pressure wave that travels back through the runner to

the plenum. In response, the relatively higher pressure in the plenum gets sucked into the

runner and defines a return pulse of high pressure that provides flow into the combustion

chamber. At the end of this stroke, the high-pressure air now flowing into the chamber

meets a pressure backflow as the piston begins its upward travel and the valve closes,

creating a high-pressure shockwave. This wave travels back through the intake path and

can be reflected and travel back toward the valve. The resulting oscillation of pressure has

a frequency defined by the length and volume of the intake chamber and is akin to some-

thing called wind throb or Helmholtz resonation.

These dynamics can benefit or hinder engine performance. For example, higher engine

speeds would benefit from a larger plenum providing dense air to the runners, but this will

produce slow throttle response and reduced low-end torque. At a certain engine speed, the

high-pressure air pulses returning through the runner can arrive with the next opening

of the intake valve helping push the charge into the chamber and possibly increasing VE

to over 100%.12 So, if the intake runner length is right, a synchronization with the engine

12 J.G. Bayas and N.P. Jadhav, Effect of variable length intake manifold on performance of ic engine. International

Journal of Current Engineering and Technology Special Issue 5 (June), 2016, 47–52.

53The End of Compromise

speed is possible and the high-pressure wave arrives just as the intake valve opens, pro-

viding a supercharging effect or ram effect.13 But this only happens in a narrow rpm

range, when the travel time of the air pulses and the engine speed are synchronized. So,

once again, in designing an engine a compromise is often required; acoustic tuning can

provide a beneficial inducting pressure boost and torque increase at a targeted rpm point

through proper intake runner sizing, but at other engine speeds, this benefit deteriorates.

So, as engine operating conditions change, the ideal induction system also changes.

Runners need to be sized short enough to allow

the low-pressure wave to travel down

the runner and draw a high-pressure return pulse within the short intake stroke, and

long enough to allow the high-pressure pulse caused by the valve closing to travel back

through the runner and be reflected back in time to meet the valve as it reopens for the

next intake stroke. At higher engine speed, since the pulse wave travels at a relatively

fixed speed (the speed of sound), a shorter length would be required to allow the wave to

return in time (Image 2.10). At a lower rpm, the length would need to be longer, since there

is more time between valve openings. But, long runners would increase flow restriction at

high engine speeds and decrease VE, and hence, the need for compromise.

Once again, more advanced mechanical designs and digital controls lessen the need

for compromise. Variable intake systems (VIS) can adjust the length of the induction run

to suit engine speed and load. Short runners provide minimal restriction of flow at high

engine speed. With engine speed comes increased induction velocity, again favoring an

13 J. Bayas, A. Wankar and N.P. Jadhav, A review paper on effect of intake manifold geometry on performance

of ic engine. International Journal of Advance Research and Innovative Ideas in Education 2 (2), 2016, 101–106.

IMAGE 2.10

Port pressure pulsing.

Inlet port pressure varies with the opening and closing of each of the engine’s valves. When the intake valve

opens, a significant port pressure drop ensues and is followed by a rapid upward pressure bump.

54 Automotive Innovation

open short run to provide maximum dynamic pressure pushing air into the cylinder. But

long runners can be advantageous at slower speeds, taking advantage of the Helmholtz

resonance and the resulting positive pressure wave into the chamber, synchronizing the

pressure wave oscillation with engine speed.

Variable length intake runners can be defined in stages, with two or more set length

possibilities, or they might allow continuous variation of the runner length. BMW was

the first carmaker to offer a continuously variable intake manifold, back in 2001. The

Differentiated Variable Air Intake (DIVA) system was defined by a rotating inner rotor

that could vary the entry point of the incoming air into a toroidal induction runner. The

effective length the incoming air had to travel could therefore vary from over 26 inches to

as little as 9 inches. The maximum length was used up to 3,500 rpm, after which the run

was gradually reduced. While effective, the necessarily large size of the multiple toroidal

runners serving each cylinder limited its application. Since this early introduction in the

late 1990s, staged and continuous variable length intake manifolds have become common

(Image 2.11).

By ensuring high-velocity airflow into the chamber, a VIS induction can also improve

swirl pattern, promote a more complete and fast combustion, and help avoid engine knock.

The runner may be tapered to encourage acceleration and proper swirl, particularly if

targeting lean burn. The mechanics is straightforward: actuators open and close induc-

tion pathways to lengthen or shorten the path of incoming air while referencing MAP

and throttle position. So induction pressure pulses are effectively tuned to more closely

suit engine speed. This allows positive pressure pulses to contribute positively to engine

performance through a wider range of engine speeds. The result is to replace the previous

peaked torque curve with a ‘thicker’ curve defined by distinct performance peaks for each

runner length (Image 2.12). Of course, continuously variable induction, rather than two or

three length options, can provide a more precise improvement in performance, with the

cost of added mechanical complexity. By varying the induction length continuously to

IMAGE 2.11

Mazda variable induction.

This 2.6-liter, 700-horsepower rotary engine from Mazda is fitted with continuously variable intake stacks,

allowing the runner length to adjust with engine speed for optimal performance.

55The End of Compromise

suit engine speed and load, a continuous VE of over 100% can be achieved in a naturally

aspirated engine.

Forced Induction

Of course, another way to deliver high-pressure air to the combustion chamber is to actively

pressurize it with a pump, called a supercharger or more casually a blower. Designs vary,

but vane and centrifugal designs are common. A chain or gear could be used to drive

the supercharger off the crankshaft, but most commonly a belt is used. Such systems can

increase torque by roughly a third, and horsepower by somewhat more.

This increased power does not come without a cost. A supercharger can increase the pres-

sure in the chamber and therefore the pressure during the compression stroke, increasing

knocking. The simplest solution is to use an engine with a lower compression ratio. And,

because the supercharger increases the MAF to the engine, it also requires a proportional

increase in fuel flow to maintain stoichiometric conditions, and this could mean decreased

fuel economy. However, there is a silver lining; forced induction is so effective at increas-

ing power that it can offer the opportunity for engine downsizing, which of course can

mean improved fuel economy.

Of course, even though superchargers can increase power, they also require some of

the engine’s energy to drive them. This increases the load on the engine, so the benefit is

limited by a trade-off. An alternative is to tap into the tremendous waste energy in the

engine’s exhaust to drive a compressor, defining a turbocharger (Image 2.13). Thisenergy

IMAGE 2.12

Torque with three-stage variable induction.

Each runner length in this idealized variable induction system develops a specific peak torque curve. As the

engine switches from one to the other, it can utilize the most favorable curve for a given engine speed, effec-

tively widening the engine’s peak torque production capacity over a much larger rpm range.

56 Automotive Innovation

is significant; roughly a third of the energy from combustion can be lost through the

exhaust.14 A turbo allows us to harvest some of this waste energy to usefully increase

power. The basic idea is to place a driven turbine in the exhaust system that can be spun

by the escaping hot exhaust gas, and use this energy to drive a coupled compressor in

the induction system. The result is an extremely high-speed turbine/compressor system

reaching speeds upwards of 150,000 rpm and, because the system is driven by hot exhaust

gasses, extremely high temperature. Turbochargers therefore require careful balancing,

robust bearings, and precise engineering. Because the induction air rises in temperature,

a cooling exchanger called an intercooler is used to allow the pressurized induction air to

cool and thus increase its density before entering the engine.

Capable of extreme speeds and heat, turbochargers require careful control as they have

the potential to define an uncontrolled positive feedback loop: The turbo results in dra-

matic increase in engine power, which increases exhaust energy, which in turn feeds the

turbo. Such a system could easily overstress the engine and lead to catastrophe. To avoid

this, a poppet valve, called a wastegate, allows some of the exhaust to bypass the turbo

and therefore limit the boost. Similarly, to allow the boost to diminish quickly, say for

deceleration or gear changes, a recirculation valve, or dump valve, can rapidly drop the

pressure in the intake system and allow engine speed reduction.

When designing a turbocharger for engine performance, we are once again faced with

trade-offs. Turbos take some time to spool up, so there’s a delay between hitting the

14 P. Punov, Numerical study of the waste heat recovery potential of the exhaust gases from a tractor engine.

Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering 230 (1),

2015, 37–48;

and P. Fuc, J. Merkisz, P. Lijewski, A. Merkisz-Guranowska and A. Ziolkowski, The assessment of exhaust sys-

tem energy losses based on the measurements performer under actual conditions. In E.R. Magaril and C.A.

Brebbia (eds) Energy Production and Management in 21st Century—The Quest for Sustainable Energy. WIT Press,

Ashurst, 2014, 369–378.

IMAGE 2.13

Turbocharger.

A turbocharger includes a turbine that is driven by escaping hot exhaust gasses (left) connected to a compres-

sor that boosts the pressure of incoming air (right). The resulting boost can increase horsepower significantly.

57The End of Compromise

accelerator and the generation of boost, called turbo lag. Smaller turbines take less time to

get spinning due to a lower moment of inertia, so a smaller turbine may be used, sacrific-

ing total boost for the sake of reduced lag. Alternatively two small turbines are often used

instead of one large turbine. These can be parallel twin units or alternatively sequential

systems can allow one turbine to be used at lower engine speeds and a larger second tur-

bine, or both, to be used at higher speeds. This has the benefit of managing turbo lag while

making higher boost available.

A more elegant solution comes from the possibility of a variable geometry turbocharger.

Variable geometry turbos (VGT) have been used to improve low-end engine torque in die-

sel engines for a couple of decade, but this technology was not available to gasoline engines

because of the significantly higher exhaust temperatures of gas engines. Manufacturing

a variable mechanism that could withstand the high temperatures of a gasoline exhaust

system simply wasn’t possible. But recent advances in material technology have changed

this. A VGT can vary the effective geometry of the turbine, typically with movable vanes,

and thus adjust the turbine for quick spool up or high boost throughout the engine speed

range. The result is a high total boost capability, with imperceptible turbo lag (Image 2.14).

The potential to integrate new turbocharger technology with other performance fea-

tures is tremendous. For example, the Hyboost project vehicle by British engineering firm

Ricardo achieves radical engine downsizing using an electric supercharger coupled with

a turbocharger and energy capture and storage technology.15 This extremely cost-effective

15 S. Rajoo, A. Romagnoli, R. Martinez-Botas, A. Pesiridis, C. Copeland and A.M.I. Bin Mamat, Automotive

exhaust power and waste heat recovery technologies. In A. Pesiridis (ed) Automotive Exhaust Emissions and

Energy Recovery. Nova Science Publishers, New York, 2014.

IMAGE 2.14

Variable geometry turbo.

The Porsche VGT uses guide vanes on the outside of the turbine. At low engine speeds, the electronically con-

trolled vanes tilt to produce a narrow gap that speeds exhaust gas to spin up the turbine more quickly and avoid

lag; but at high engine speeds, the vanes open allowing a full turbo boost and avoiding the need for a bypass valve.

58 Automotive Innovation

design makes use of technology already on the market, adopting exhaust energy recap-

ture and regenerative braking to drive an electric supercharger on a turbocharged GDI

engine. The result is a three-cylinder, 1.0-liter engine that matches the performance of its

2.0-liter counterpart with a dramatic decrease in carbon emissions.16 Similarly, but with

more dramatic effect, Formula 1 cars and Audi’s diesel R18 LeMans racer have coupled

electric motors with turbochargers. The motor assists with low-speed boost, allowing the

use of a larger turbine without lag and enabling a more significant high-end boost. Volvo’s

production T6 Drive-E engine incorporates a turbocharger and supercharger in a 2.0-liter

package that produces 316 horsepower and an impressive 35 highway mpg.17 Mercedes is

taking this one step further by incorporating Formula 1 Motor Generator Unit (MGU) into

its hypercar designs. The idea is to place a motor/generator onto the turbocharger shaft,

allowing the generator to recover excess energy from the turbo as it spins down, store that

in a battery, and use that energy to get the compressor spinning and eliminate lag.18

Compression Ratio

In the end, the challenge of forced induction, tuned runners, variable valve action, and

just about everything else discussed thus far often comes down to the same issue: avoid-

ing knock. Basically, we want to squeeze as much power as we can out of the fuel, without

squeezing it so much that it ignites on its own. Tricky business. But instead of tweaking all

the peripheral variables to manage this, why not get to the heart of the matter and adjust

the amount the engine squeezes the charge? Advanced engineering and sophisticated

digital control offers us the possibility of doing just that.

Defining the CR of an engine has long been the very epitome of automotive engineering

trade-offs. An engine’s compression ratio is not a variable that can typically be altered, like

timing or fuel mixture; it is fundamental to the engine design. So, it needs to be defined

carefully at the start. At lighter loads and lower engine speeds, a high compression ratio

can provide improved fuel consumption, since higher compression means higher effi-

ciency. An increase in the CR from 8 to 12 might produce a 20% or more improvement

in an engine’s ability to turn combustion into mechanical work, called thermal efficiency

(Image 2.15).19 While you might expect higher pressure to mean increased knocking, the

risk of preignition is low given the low load. However, at high loads, as combustion cham-

ber temperatures rise, the same compression ratio would likely lead to autoignition. And

of course, a lower CR could allow for greater boost at high loads.

So, the resulting CR for a production engine tends to represent a compromise. A higher

compression ratio offers greater thermal efficiency and thus more power and fuel effi-

ciency, but it also invites knocking. So, in the past, designers defined a middling compres-

sion ratio that best suited the intended use of the engine. Compression ratios for gasoline

16 https://ricardo.com/news-and-media/press-releases/hyboost-demonstrates-new-powertrain-architecture.

17 G. Witzenburg, “Volvo’s T6 Engine Part of Bold Powertrain Strategy.” Wards Auto June 15, 2016. Available at

www.wardsauto.com/engines/volvo-s-t6-engine-part-bold-powertrain-strategy

18 A. MacKenzie, “Revealed: 2019 Mercedes-AMG Project One Powertrain.” Motor Trend May 30, 2017. Available

at www.motortrend.com/news/revealed-2019-mercedes-amg-project-one-powertrain/

19 K. Satyanarayana, R.T. Naik and S.V. Uma-Maheswara Rao, Performance and emissions characteristics of

variable compression ignition engine. Advances in Automobile Engineering 5 (2), 2016, 1–5.

https://ricardo.com

http://www.wardsauto.com

http://www.motortrend.com

59The End of Compromise

engines can be as low as 6:1, though 10:1 is more typical, with Mazda offering the highest

production car CR of 14:1.

Defining an engine that can change its compression ratio mid-operation could end this

compromise (Image 2.16). In addition, since the primary limitation to the usage of many

alternative fuels is getting the compression ratio right, a variable CR could allow multiple

fuel usage including everything from straight vegetable oil or diesel, to hydrogen and

methane. With the capacity to offer fuel efficiency and still provide power when needed,

a variable compression ratio (VCR) could also allow for more compelling forced induction

and ‘extreme downsizing’ to very small, light, and powerful engines. But all this requires

a mechanical configuration that can change an engine’s core operation without adding

excessive complexity or weight. Conceptually, there are two approaches that could allow

the CR to be changed: You can change the clearance volume, or you can change the piston

travel.

A simple solution would be to change the crown of the piston and therefore

change the

clearance volume. Performance engine builders have often machined or swapped piston

heads as a way to vary the compression ratio of an engine. In the same vein, Ford and

Daimler Benz have experimented with a piston configuration that has a sliding head and

skirt unit that allowed for variation in the piston crown height. Hydraulic cylinders in the

inner piston slide the outer head up or down relative to the inner piston body, changing

the deck height and thus the compression ratio. The design approach has the advantage of

IMAGE 2.15

Fuel efficiency and compression ratio.

At high torque with a wide-open throttle, the engine’s fuel consumption per unit of power generated, or Brake

Specific Fuel Consumption (BSFC) is lowest at a low compression ratio but increases somewhat exponentially as

CR rises. At a lower load demand, fuel consumption decreases as compression ratio rises. So, it makes sense to

try to vary compression ratio as the load on the engine changes.

Source: D. Tomazic, H. Kleeberg, S. Bowyer, J. Dohmen, K. Wittek, and B. Haake (FEV), Two-Stage Variable

Compression Ratio (VCR) System to Increase Efficiency in Gasoline Powertrains DEER Conference 2012,

Dearborn, October 16th, 2012.

60 Automotive Innovation

not requiring alteration of the basic engine architecture, but it has the significant disadvan-

tage of adding complexity and weight to the engine’s oscillating mass.

Moving the cylinder head could offer the same effect without changing the piston. A

few years ago, SAAB engineers designed an articulated cylinder head that did just that.

The head and cylinder liner were combined into an integrated housing and hinged to

the lower crankcase. By moving the upper cylinder housing only slightly the compres-

sion ratio could be varied from 8:1 to 14:1. With integrated forced induction, the system

promised efficiency, fuel flexibility, and notable power, reaping 225 horsepower from a

1.5-liter engine.20 Unfortunately, production of the engine was determined to be prohibi-

tively costly.

A similar result can be achieved more simply by adding an adjustable piston to the head.

When this piston moves in and out of the combustion chamber, it changes the clearance

volume and thus the CR. This approach is used by Lotus’ so-called Omnivore research

engine. Among other features, the head and block of this two-stroke design would be

cast as one piece. A piston, or ‘puck’, would slide in and out of the combustion chamber to

effectively vary the clearance volume and allow the engine to use multiple fuels, earning

the engine its name.21

The alternative approach is to change the piston stroke. Perhaps the most direct way

to achieve this is to change the connecting rod’s length. By extending the connecting rod

slightly, the clearance volume can be reduced significantly, offering a definitive change in

CR. German engineering firm FEV uses an eccentric piston pin connection that is actu-

ated by two pistons integrated into the connecting rod body (Image 2.17). They are able

to adjust the CR from 8.8:1 to 12:1 in response to inertia forces and combustion pressure,

20 M.P. Joshi and A.V. Kulkarni, Variable compression ratio (vcr) engine: A review of future power plant for

automobile. International Journal of Mechanical Engineering Research and Development 2 (1), 2012, 9–16.

21 Case Study: Omnivore Research Engine. Available at www.lotuscars.com/engineering/case-study-

omnivore-research-engine

IMAGE 2.16

Improved efficiency of VCR engines.

While a VCR engine typically does not provide improved efficiency across its complete operating range, it does

provide notably improved efficiency in the most frequent driving conditions.

Data: A. Baheri and M. Tajvidi, Simulation and analysis of an internal combustion engine with variable com-

pression ratio. International Journal of Engineering Sciences & Research Technology 4(12), 2015, 757–763.

http://www.lotuscars.com

61The End of Compromise

requiring no active control.22 Porsche is looking into a similar design produced by engi-

neering firm Hilite International.

Multiple VCR designs offer mechanical variations in the rod-crank linkage. This seems to

have become the most popular approach among carmakers, perhaps because the actuation

mechanism is stationary, adding little to the engine’s oscillating mass and providing easier

mechanical control. Some use eccentric gearing to alter the crankshaft rotation, while oth-

ers use actuators to adjust the crankshaft relative position. Some have been designed with

a tip-in/tip-out strategy that adopts a defined lower CR only above a preset throttle posi-

tion. Others integrate more continuous variation of CR. Nissan, Peugeot, Acura, and oth-

ers have been playing with such designs for several years. The most promising approach

seems to be the use of a multilink connection between the rod and crankshaft that can

allow for an effective variation of the shaft throw. This could be controlled by an eccentric

rotary actuator, a hydraulic piston, or some other control linkage.

Perhaps the most advanced version of this approach comes from Infiniti (Image 2.18).

The Japanese automaker is planning to provide the first production vehicle with a VCR

engine in the 2019 QX50. The turbocharged, four-cylinder, 2.0-liter engine is innovative.23

The rod is not directly connected to the crankshaft, but to a multilink, with the piston rod

connected on one side, an actuator arm connected on the other, and the crankshaft throw

in between. An electric motor drives an actuator linkage that moves one side of the mul-

tilink up and down. The effect of course is to move the other side, where the piston rod is

connected, relatively up and down in response and thus alter piston height. The engine

can operate in Otto or Atkinson cycle, allowing a longer piston stroke at a high compres-

sion ratio. Because the rods are nearly vertical in the combustion stroke, and the multil-

ink is designed to provide improved reciprocating motion, the engine is designed to run

smoother and quieter than a conventional counterpart.

22 S. Asthana, S. Bansal, S. Jaggi and N. Kumar, A comparative study of recent advancements in the field of vari-

able compression ratio engine technology. SAE Technical Paper 2016-01-0669, 2016.

23 “Infinity Develops a Variable Compression Ratio Engine”. Available at www.ADandP.media

IMAGE 2.17

Variable connecting rod.

This innovative design by FEV uses an eccentric piston pin connection that is

actuated by two pistons integrated into the connecting rod body. This allows a

VCR engine with minimal engine modifications.

Image: FEV

http://www.ADandP.media

62 Automotive Innovation

This design is not entirely dislike the French MCE-5 Intelligent Variable Compression

Ratio, or VCRi, four-cylinder engine (Image 2.19). Initially partnered with Peugeot and now

working with Dongfeng, MCE-5 hopes to have the first variable-CR Miller-cycle engine

ready for production within a few years. A lower linkage is moved with a hydraulic piston

adjacent to the combustion cylinder to vary piston travel.24 So, the MCE design allows for

independent variation of each cylinder. While an impressive feat of engineering, whether

the added value of piston-specific control is worth the additional complexity is not yet

clear. The manufacturer points out that this will allow for faster and more accurate CR

control, helping knock mitigation and facilitating advanced combustion modes like HCCI.

And by reducing the impact of the accumulated small variations in component tolerances

of the cylinder system, called tolerance stack, it permits larger manufacturing tolerances

and so potentially lower production costs.25 The CR can vary continuously along a much

larger range, and the potential efficiency benefit is major.

24 Cost, Effectiveness and Deployment of Fuel Economy Technologies for Light-Duty Vehicles. Committee on the

Assessment of Technologies for Improving

Fuel Economy of Light-Duty Vehicles, Phase 2, Board on Energy

and Environmental Systems Division on Engineering and Physical Sciences National Academy of Sciences,

Washington, DC, 2015.

25 Author communication with Frédéric Dubois, MCE-5 Development S.A.

IMAGE 2.18

Infiniti VCR engine.

Infiniti has produced the first variable compression ratio engine in a production car. The harmonic drive moves

the actuator arm, which in turn rotates the control shaft, modifies the piston travel, and so varies the compres-

sion ratio.

Source: INFINITI

63The End of Compromise

The design of these mechanisms, and the general design of any modern engine, is a

juggling act. Change one thing and three others are affected. Increase the boost and you

have to decrease the compression ratio. Increase the valve lift for improved power and

suffer deteriorated efficiency. Lengthen the induction runner for better cruising and lose

high-end power. The list goes on. But increasingly, the juggling is getting easier and the

trade-offs less troublesome. Advanced vehicle electronic management and innovative

engineering have eliminated many of the compromises that once defined automotive

design. The result is engines that are more powerful, more efficient, and cleaner than ever

thought possible. Pretty cool.

IMAGE 2.19

Intelligent variable compression ratio (VCRi).

French MCE-5 VCRi could vary CR continuously from 8:1 to as high as 18:1. Using the inertia of the engine to

drive the adjustment mechanism allows changes in CR to take place in less than a tenth of a second, and could

allow for compression ignition operational mode.

Source: MCE-5 DEVELOPMENT

http://www.taylorandfrancis.com

Getting Power to the Pavement

Recent advances in engine design and control are impressive by anyone’s measure. But

it’s not just about the engine; it can’t be. The power coming from the engine only maters

if we can effectively transfer it to the wheels. So, the design and capacity of the mecha-

nism that connects the engine to the wheels, called the drivetrain, has to be every bit as

sophisticated and efficient as the engine itself; and it is. As is true for engines, innovations

in drivetrain components and control have revolutionized the industry, and profoundly

change the capacity of the automobile and the nature of driving.

The tasks we expect from a drivetrain are not simple. We need a coupler that can allow

us to connect and disconnect the engine from the rest of the drivetrain. The need for this

should be pretty obvious; without it, the car would have to be moving anytime the engine

is turning. Though, as we’ll see, the coupler does a lot more than just provide stopping

ability. We also need gearing. Again, the reason should be fairly clear. A typical internal

combustion engine can produce useful power between about 1,500 and, let’s say, 6,000

rotations per minute (rpm). Now consider that at 5 miles per hour (mph), an axle on a typi-

cal car is turning at about 80 rpm and at 70 mph, that axle needs to turn at about 1,200 rpm,

or about 15 times faster. Our engine can’t turn 15 times faster than its lowest speed. Much

more importantly, while the engine’s operable speed range is fairly wide, the range at

which it produces desired power and efficiency is much smaller. So, even if we could con-

nect the engine to the axle with no gearing and settle for a car with a 30 mph top speed, the

resulting power, torque, and efficiency would be terrible. Lastly, steering a car, or at least

turning it, requires that the axles be able to rotate at different speeds, as the outside wheels

will strike a longer arc than the inside wheels. If we can’t manage this, the tires will slip

on any turn, leading not only to rapid tire wear but terrible handling. And all of this is just

the tip of the iceberg; we expect much more of a modern drivetrain than these very basic

functions. As a result, there has been plenty of need for innovation.

Once again, it’s about new ideas and materials, but it is also about precise digital control

and the possibilities this offers. We have largely replaced hydraulic control with digital

electronic control in both automatic and manual transmissions. Advanced sensors and

complex control logic, including artificial intelligence, fuzzy logic, and other advanced

algorithms, have changed our understanding of what is possible. Advanced systems now

in production or in the works provide predictive distribution of torque and power in any

driving conditions, instantly identify slip and compensate with a redirection of torque to

any wheel, and even offer the possibility of intelligent tires that can measure, evaluate, and

adapt to road conditions. In short, the drivetrain of the future is here.

What Do We Need a Drivetrain to Do?

The drivetrain links the engine to the vehicle’s motion. So, we might start by considering

what is needed to produce this motion. To propel the car forward we need to overcome the

66 Automotive Innovation

combined rolling resistance of the tires and bearing friction, as well as the effect of any

incline. We also need to overcome the aerodynamic drag of the car. We’ll talk about this

in Chapter 8; but for now, it’s worth noting that drag increases exponentially with speed,

while rolling resistance increases linearly. So, although rolling resistance is the primary

resistive force at low speed, at high speed drag becomes the definitive force (Image 3.1).

The torque produced by the engine has to be able to match the torque required at the

wheels to overcome the forces resisting the car’s movement. So, let’s look at the torque

and power produced by the internal combustion engine. The torque produced by the

engine at any given speed is directly related to the force provided by the piston during

the power stroke. More particularly, it is the product of the force delivered perpendicular

to the crankshaft throw and the throw length. So, we measure it in units of length and

force, such as pound-feet (lb-ft) or newton meters (N-m). An engineer might point out that

this unit might be more precisely expressed as pound force (lbf) to identify it as a unit of

force and distinguish it from a pound of mass. As we’ve already seen, an engine design

is generally tuned to produce performance at a defined speed range; on either side of this

range, the efficiency of the combustion process can be compromised. So, it’s not surprising

that when we graph torque production, we can see that it typically dips at low and high

engine speeds, with maximum production somewhere in the middle. As we discussed

in the previous chapters, innovative technologies are being deployed to widen the range

at which this torque can be produced. So, the torque curve is getting wider, or fatter, and

more consistent, or flatter.

IMAGE 3.1

Power requirement with increasing speed.

As the speed of a vehicle increases, both friction resistance and aerodynamic resistance increase. However,

aerodynamic resistance increases as a function of the square of the speed, making it the dominant resistive

force at higher speeds.

67Getting Power to the Pavement

Torque is notably different than power (Image 3.2). Torque is an expression of rotational

force at a given point in time. On the other hand, as you might recall from basic physics,

work is defined by an application of force over distance. And power is the capacity to do

work over time. Consider an example: If you push on your stalled car with 50 pounds of

force but can’t get it to move, you’ve accomplished no work. (Though technically, your

muscles moved quite a bit in the effort, so they did some work. It just didn’t get you any-

where.) If you then get smart, release the parking brake and try again, you’re able to move

your car 20 ft. So, you have now accomplished 1,000 ft-lbf of work (50 pounds of force times

20 ft). Notice, the unit is foot-pounds for work, to distinguish it from lb-ft of rotational

torque; and once more

we use the unit pounds-force to distinguish it from pounds of mass

or weight. So much for work, but we don’t know how much power you were able to pro-

duce until we know how long it took you to accomplish this herculean task, since power

is work over time. If it took you an hour to get the car to move 20 ft, we can now assess the

power you produced. In this case, you’ve provided 50 lbf × 20 ft/60 min, so about 17 ft-lbf/

min of power. Since one horsepower is defined as the ability to move 550 pounds 1 ft in

1 min (550 ft-lbf/min), you’ve produced a whopping 0.03 HP. Nice try. However, if you were

able to move the car that 20 ft in 5 s, then you would have produced 50 lbf × 20 ft/0.083 min,

or about 12,048 ft-lbf/min, that’s nearly 22 horsepower. Much better.

The key here is that power is an expression not only of how much force you can produce

but how quickly you can produce it. Getting back to engines, torque is generated only

in the power stroke, and roughly maximized when the connecting rod is perpendicular

to the throw of the shaft. The engine’s power, on the other hand, is the result of both the

IMAGE 3.2

Engine power and torque.

In this typical torque and power curve, notice how the horsepower continues to climb after the torque has

plateaued. As engine speed increases, power production will grow up to a point when the engine is no longer

capable of generating more power.

68 Automotive Innovation

force generated by these strokes, and the number of strokes that the engine can produce

in a given amount of time. The faster the engine rotates, the more power strokes in a

given period of time, and so the more power that is being produced. Double the engine

speed, and all things being equal, the power output doubles. The torque, on the other

hand, doesn’t change in a fixed way with engine speed, as we saw in the graph above.

Of course the increase in power cannot continue forever. As we saw in Chapter 2, as the

engine speeds up, its ability to take in air and produce useful combustion can deteriorate

as the engine approaches its maximum speed. So, at some point, we will see a dip in power

production, even though speed continues to increase. All this makes the typical tendency

to compare engines based on their maximum horsepower a bit foolish. Maximum power

is just that, maximum, determined at a particular engine speed. It doesn’t always tell you

very much about the engine’s operation at any other speed (Image 3.3).

Another reason we shouldn’t get too carried away with engine horsepower is because

what really matters isn’t the power of the engine, but the power that can be delivered

through the drivetrain to the pavement. Transforming torque and speed through the gear-

ing of the drivetrain allows us to adjust the power and torque available at the drive wheels.

Gearing allows us to effectively transform torque to speed and vice versa by applying a

form of rotational leverage. As a small gear rotates against a larger gear, it must provide

greater speed, but the result in the larger gear is greater torque. If the larger gear, on the

other hand, drives the small gear, it can move slowly and generate greater speed in the

smaller gear, but the price is torque. In either condition, power at each shaft remains equal;

gears do not create power. This means the actual power and torque we can produce at the

IMAGE 3.3

Comparing engine power.

The common practice of comparing engine capacity by comparing top horsepower production is problematic at

best. Notice how in this idealized case the lower horsepower engine actually generates more horsepower for a

much greater range of operation, even though the other engine would be rated for a higher peak horsepower. So,

looking only at peak horsepower could lead to misleading conclusions about engine performance.

69Getting Power to the Pavement

wheels can vary; and while we only have one power curve for our engine, we can produce

multiple torque curves for our drive axle depending on the gearing we select.

Remember the power needed to accelerate our car increases with speed. In addition, the

more our available power exceeds the resistive force of drag and friction, the more power

available for acceleration. So, if we shifted into fifth gear at 10 mph, we would not be able

to accelerate, since our available tractive force at the drive axle does not exceed the total

resistance to the movement of our car. Dropping the car into a lower gear increases the

tractive force at the wheels and allows acceleration. But if we stay in that low gear, we’ll

reach our maximum speed in that gear pretty quickly. The magic of gear changing allows

us to accelerate quickly at low speeds, and change gears so we can continue to accelerate

at higher speeds.

Understanding all this allows us to appreciate key characteristics of vehicle performance.

For example, at some point the power the engine is able to produce and the power required

for continued forward motion are equal. This defines our maximum vehicle speed (all

other variables like stability and mechanical integrity allowing). Below that speed, a lower

gear can allow for increased surplus power (Image 3.4). So, dropping down a gear can

allow for increased acceleration during overtaking. Conversely, a gear beyond the maxi-

mum speed gear might allow for lower engine speed, and so quieter and more efficient

operation at cruise, even though it would reduce available power somewhat. That is the

principle behind overdrive. Since early performance transmissions often had the fastest

IMAGE 3.4

Performance and gearing.

Accelerative capacity is determined by the power that can be delivered to the wheels in excess of the combined

resistance force or friction and drag. Under-gearing can limit top speed, and over-gearing can limit acceleration

capacity.

70 Automotive Innovation

gearset at direct drive, or a gear ratio of 1:1, overdrive is usually considered to be any ratio

larger than this.

Acceleration typically defines a series of torque peaks and drops with each gear change.

Between each peak as engine speed continues to increase there is a marked decline in

available torque, called torque interruption. Minimizing this interruption and producing

a continuous and smooth transition of torque for power throughout a long acceleration

is a key to performance. The effect can be felt casually in a typical car as you accelerate;

engine speed increases and then drops to accommodate shifts, so acceleration is applied

in steps (Image 3.5). The driver experiences a shift shock with each transition. While this

was once a normal characteristic of driving, it is not ideal, and has decreased significantly

with advanced transmissions.

More available gear ratios can help minimize torque interruption by filling the gaps

between the curves. It also allows the engine to operate more closely to its most efficient

speed, offering improved fuel economy. An engine consumes the least fuel per unit of

power produced, called brake specific fuel consumption (BSFC), on the lower end of it

speed range (Image 3.6). So, a gearbox that allows the engine to stay closer to this effi-

ciency region as the car’s speed varies will produce greater fuel economy. In fact, the desire

for efficiency has resulted in a general increase in the gearing of the typical production

vehicle, allowing the engine to stay in a more efficient operating range and avoid fluctua-

tions in engine speed that can kill efficiency. As a result, the four- or five-speed automatic

transmissions that were typical just a decade ago have been replaced with at least six to

eight gears, and increasingly nine, ten or more.

At any given point in the acceleration process, there is often more than one possible

gear choice, making variations in the timing of the shift an option. As the graph indicates,

IMAGE 3.5

Fuel map.

This typical fuel map denotes BSFC (g/kWh) within differing regions of engine speed and torque. Fuel

con-

sumption dips on the left side of the graph. So, the most efficient operating region for any given torque require-

ment is along this left side where engine speed is low.

71Getting Power to the Pavement

delaying each shift until the maximum tractive force has been reached in that gear ratio

can help keep tractive power high, and acceleration quick. However, the cost of this

strategy—there is always a cost—is higher fuel consumption. On the other hand, early

shifting can significantly increase fuel economy, but the price is accelerative capacity. As

we will see, the gear change timing of automated transmissions depends on the design

priorities of that vehicle, driving conditions, as well as selections made by the driver if the

car is equipped with selectable drive modes.

The basic shift logic that must be managed by any transmission can be understood graph-

ically.1 The two lines on the graph below represent upshift and downshift points. Normal

driving in any given gear is defined between these two lines. The space between the lines

provides a lag, or hysteresis, in gear changes so constant shifting is avoided. When driving

at a given operating point between the lines, a large throttle position increase can lead to

downshift to provide increased power for acceleration. Though the same throttle position

1 M.G. Gabriel, Innovations in Automotive Transmission Engineering. SAE International, Warrendale, PA, 2003.

IMAGE 3.6

Tractive effort.

The tractive force available at the wheels varies with each gear. As vehicle speed increases, gear changes are nec-

essary to continue to generate forward force. However, at some speed, the tractive effort available is exceeded by

the resistive force of air and friction, and no further acceleration is possible. The more gears available, the less

severe the dips in tractive force between gears, and the better the engine is able to remain at a desirable operat-

ing speed for a given vehicle speed.

72 Automotive Innovation

increase with an accompanied acceleration might not lead to a gear change, an increase in

speed with only a minor increase in throttle position could lead to downshift for improved

fuel economy (Image 3.7). A drop in throttle position could lead to a downshift if speed

is maintained, or no shift if some speed is lost. And a drop in speed while maintaining

throttle position, say due to an incline, could lead to a downshift.

The gearing used for any particular vehicle depends on the characteristics of that vehicle.

Greater rolling resistance, say due to high vehicle weight, may require a drop in gearing

at the low end of the range. A large frontal area may dictate lower gearing on the high end

where drag will define a predominant force. A desire for smooth fast cruising may mean

an additional overdrive gear is needed. The performance expectations and priorities of

the vehicle will matter a lot, of course. In addition, expected towing capacity, targeted top

speed, engine characteristics, and a wide array of other concerns will define the optimal

gearing combination. All of that having been said, in general, the essential parameters are

the needed upper and lower gear ratios, with the middle gears largely defined to provide

a reasonably even distribution of gearing between the two.

Manual Transmission Coupler

With these concepts in mind, we can start to think about how it all comes to life in an

actual gearing mechanism. We can start with the most straightforward version, the man-

ual transmission; and a good starting point for that is the basic clutch. A conventional

IMAGE 3.7

Shift control.

Shifting speeds vary with driving conditions and throttle position. With a hard acceleration, up shifting is

delayed to ensure maximum available torque for overtaking. At low throttle position, upshift happens at a

lower engine speed, to help ensure improved fuel economy.

Source: Adapted from M.G. Gabriel, Innovations in Automotive Transmission Engineering. SAE

International, Warrendale, PA, 2003.

73Getting Power to the Pavement

manual transmission, or stick, relies on the driver to physically change gear ratios to suit

driving conditions. To make this possible, we have to enable a manual decoupling of the

engine and transmission. This not only is needed to allow the car to stop without turning

off the engine but to smoothly decouple and then recouple the engine and the driveline to

allow gear changes. The principle way this is accomplished in a manual car is a mechani-

cal friction coupling or clutch. A key challenge is to avoid a strong jerk, or shift shock, both

for initial acceleration and during gear changes, which usually requires an ability to slip

the friction coupling for gradual engagement.

A friction coupler is a pretty simple device. In its most basic form, imagine two rotating

disks placed with their planar surfaces near each other. If the two are pressed together,

then the friction contact means the rotation of one will cause the rotation of the other. If

we release the pressure, the two can rotate independently. An automotive friction clutch

is not that much more complicated. It’s composed of three primary components: a fly-

wheel, a pressure plate, and a disk. A cover is bolted to the engine flywheel on one side

and connected to the pressure plate on the other. The disk, also called a clutch plate, is

placed between the pressure plate and the flywheel (Image 3.8). It has friction surfaces

on both sides, and is splined so it rotates with the transmission input shaft. We can use a

collection of coil springs or a diaphragm spring to squeeze the sandwich together. When

we do, the friction between the flywheel, disk, and pressure plate causes them to all rotate

together. When we release the pressure, the disk can rotate independently of the cover

and pressure plate, decoupling the engine and transmission. When pressure is reapplied,

friction serves to lock all the parts back together and the two shafts rotate at the same

speed. Importantly, if the force is applied gradually, allowing an initial slip, the torque

from the driving shaft will increase proportionally providing a smooth application of

torque to the drivetrain.

IMAGE 3.8

The clutch assembly.

The basic components of a clutch assembly are fairly straightforward. This SACHS performance clutch includes

a flywheel, a clutch plate or disk, and a pressure plate with a diaphragm spring. A clutch release bearing rides

on the diaphragm spring to engage or disengage the grip on the disk, and is actuated by a release mechanism.

Image: ZF Friedrichshafen AG

74 Automotive Innovation

The basic system is pretty straightforward, but some enhancements are possible. A

spring hub is generally used on the friction plate to allow for smoother transitions. A

small braking mechanism can be added to stop the clutch plate from spinning once the

clutch is disengaged, called a clutch stop. And the driven plate can be lightened to reduce

its moment of inertia toward the same end. This enables quicker gear changes, since gear

changes are limited until the transmission input shaft slows. The friction surface is prob-

ably the greatest variable, and can be designed to allow improved wear and smoother

engagement.

Ideally, the friction material on the clutch plate will provide a durable and uniform fric-

tion surface that wears evenly, is able to conduct heat effectively, and demonstrates friction

and strength that is unaffected by high heat. Generally asbestos fiber interwoven with

metal wire has been used, though newer materials are changing this. Kevlar, a polymer

with five times the strength of steel, improved strength, much-improved heat tolerance,

and good friction characteristics that enable a smooth application of braking. Ceramic fac-

ings composed of copper, iron, tin, bronze, and silicon dioxide, or graphite are harder and

can offer excellent heat transfer for high-performance applications, but with a tendency for

sudden engagement that can make them a poor choice for casual driving. Organic compos-

ites of advanced resins, rubber, and metals can offer smooth engagement, durability, and

good heat management. Groove patterns on the clutch material can be designed to allow

improved wear and smoother application without catching. The options are much better

than they were even a decade ago.

Similarly, there are some modifications that are not generally applicable, but make sense

in specific applications. While a single-plate clutch is common, a multiplate clutch is pos-

sible. There are a few reasons this might be desirable. First, the greater the friction surface

area of contact, the more effectively the torque is transmitted. And, adding a number of

plates can allow for a smaller diameter unit while maintaining adequate surface contact.

In addition, to enhance smoothness of interaction as well as cooling, the clutch can be

submersed in a fluid, defining a so-called wet clutch. Wet clutches are not generally used

in conventional manual transmissions where a single-plate dry clutch remains dominant.

However, we will see that recent innovations in transmission designs are making this

option more desirable for certain applications.

Manual Transmission

The basic design of a manual transmission itself is defined by three shafts: an input shaft

connected to the clutch; an output shaft, also called a main shaft; and what is called a

layshaft or countershaft (since it counter rotates). Gears are mounted to the layshaft and

rotate as one unit (Image 3.9). The gears on the output shaft rotate independently but are

always meshed with the layshaft gears. The output shaft is splined, and between each

gear is a ring, called a collar, that incorporates a cone-shaped clutch mechanism, called a

synchronizer and dogteeth. As the collar slides toward a selected gear, initial engagement

of the synchronizer allows a smooth friction contact so both gear and shaft are rotating at

the same speed prior to engagement of the teeth that lock the gear to the shaft. This gear

now defines the driving gear ratio, as other gears on the output shaft spin freely. Placing

an additional gear, called an idler gear, between a layshaft gear and reverse gear on the

output shaft, enables reverse.

75Getting Power to the Pavement

Allowing for multiple gear ratios is then simple, as various gear combinations on the lay-

shaft and output shaft provide differing gearing combinations (Image 3.10). A basic actuator

mechanism controls linkages that enable the movement of each collar, and so the engage-

ment of any gear on the output shaft. Because each collar incorporates a synchronizer, the

gnashing of gears that was common long ago is avoided nearly completely. The driver can

disengage the clutch, and so release tension in the drive train, move the shift lever which

actuates forks that slide a collar disengaging one gear combination and engaging another,

and reengage the clutch by releasing foot pressure from the actuator. Because the gears are

always engaged, there is no need for double clutching, and no damage to gear teeth.

If you live in the US, you might be thinking that this is all tired technology. Manual trans-

missions have lost popularity in the US almost entirely, and now represent less than 5% of

new vehicles. They are declining in Asia as well. However, they remain popular in Europe,

representing about 75% of European auto sales.2 Despite declining interest, there are clear

advantages that make manual transmission worth another look. The solid mechanical con-

nection of manual gearboxes means low gear loss, defining the most efficient way to transfer

engine power to the driveshaft (despite the recent improvements in automatic transmission

we’ll look at soon). A typical manual transmission can provide an overall efficiency percent-

age in the mid to high 90s, a notch above the overall mid to high 80s efficiency of a typical

automatic transmission.3 They also offer lower cost, particularly in Europe where existing

facilities can be used without retooling for a new transmissiontype. Moreover, with the

2 European Vehicle Market Statistics 2016/17. International Council for Clean Transportation. Available at eupock-

etbook.theicct.org; A. Isenstadt, J. German, M. Burd and E. Greif, Transmissions International Council for Clean

Transportation, Working Paper, August, 2016.

3 M.S. Kumbhar and D.R. Panchagade, A literature review on automated manual transmission (AMT). IJSRD—

International Journal for Scientific Research & Development 2(3), 2014, 1236–1239; and M. Kulkarni, T. Shim and Y.S.

Zhang, Dynamics and control of dual-clutch transmissions. Mechanism and Machine Theory 42, 2007, 168–182.

IMAGE 3.9

Manual transmission.

A basic manual transmission includes a layshaft (lower), an input shaft connected to the clutch (left) an output

shaft connected to the drivetrain (upper), collars between the gears on the output shaft, and forks that move the

collars as well as a shift rail (upper) that moves the forks.

http://eupock-etbook.theicct.org

76 Automotive Innovation

performance and automation upgrades now possible with digital control and advanced

manufacturing, a manual transmission no longer means more work behind the wheel, but

it can still mean a more engaged and dynamic driving experience.

Nevertheless, increasing engine speeds and increasing vehicle top speeds are challeng-

ing the capacity of the old manual transmission. Automatic transmissions, as we will see,

have responded with an ever-increasing number of gear ratios, enabling improved fuel

economy and performance. Manuals are somewhat limited here; while seven-speed man-

ual transmissions are available in numerous performance cars, an eight- or nine-speed

manual transmission would be large and offer questionable drivability.

Automated Manual

As is true for so much we have seen so far, modern digital control has enabled a new

phase in the manual transmission’s evolution. A variety of transmission control strate-

gies have been applied to what are called semi-automatic transmissions, or the seemingly

IMAGE 3.10

Toyota six-speed manual transmission.

Toyota has developed an advanced manual transmission primarily for the European market. The design is

notably lightweight and compact, with a total weight of about 88 pounds or 40 kg. Toyota’s iMT (Intelligent

Manual Transmission) controls automatically adjust engine speed when shifting to ensure smooth gear transi-

tions and reduced torque interruption.

Image: Toyota Motor Company

77Getting Power to the Pavement

oxymoronic, automated manual transmission (AMT). By adding hydraulic or electrome-

chanical actuators that allow for the clutch as well as shifting to be controlled digitally,

manufacturers have dispensed with the clutch pedal and developed a transmission that

can incorporate varying degrees of automated function. The digital logic can enable fea-

tures that would normally not be possible with a manual transmission, such as fuel saving

strategies, start-stop technology (to be discussed in Chapter 6), or idling or stopping the

engine when coasting in high gear, called sailing.

The basic idea is to combine the efficiency, simplicity, and affordability of a manual

transmission with the drivability of an automatic. Though it should be said that fuel econ-

omy and manufacturing cost, not necessarily driver preference, have probably been the

strongest motivating factor in the development of the AMT. Early versions of this technol-

ogy simply replaced the clutch pedal with an automated electrohydraulic clutch that was

activated upon shifting. Such systems can still provide an economical alternative that can

make a low- cost manual transmission more attractive, such as Schaeffler’s recent develop-

ment of the eClutch affordable automated clutching mechanism.

Typically, an AMT system will include a fully automatic mode, with no driver

input

needed. So, an AMT can drive a lot like an automatic transmission, although they will

tend to provide a sharper responsiveness because of the direct mechanical engagement of

the drivetrain. The aim is to combine the efficiency and control of a manual transmission

with the drivability of an automatic, in a package that can appeal to a driver’s desire for a

more engaged experience when wanted and simplified driving when not. So, shifting can

take place automatically, with a conventional lever or with paddles on the steering column.

Adopted from Formula One, shift paddles offer a sportier driving experience, typically

with a paddle on the right of the steering wheel shifting to a higher gear and one on the

left dropping to a lower gear.

Clearly, digital control is indispensible to the AMT. A transmission control unit (TCU)

combines torque, vehicle speed, engine speed, and throttle data from the ECU with trans-

mission specific information from gear, clutch, and brake position sensors, as well as

transmission output speed. Moving the shifter sends an electronic signal, not a mechani-

cal transfer, a so-called shift-by-wire system. When a shift is called for by the driver or

indicated by the algorithm, optimal torque, engine speed and shift timing are defined for

a fast and smooth gear transition. Clutch disengagement, shift, and reengagement then all

occur automatically. When downshifting, the control module revs the engine upon disen-

gagement to achieve rev matching (matching the engine speed to the transmission shaft

speed) and so allow for smooth reengagement. In fact, while not done, it could be possible

to operate such as system with no use of a clutch at all given the precise control possible.4

In most systems, gears can be skipped to allow the most effective ratio, while precisely

matching engine and transmission speeds to avoid lurching or lugging the engine.

The control logic can be changed to suit driving needs. In normal drive mode, for exam-

ple, the shifting happens earlier to ensure optimal fuel efficiency and smooth driving. In

sports mode, on the other hand, shifting will happen later to allow the full advantage of

high power operation, and downshifting will become more aggressive to enhance respon-

siveness. More than this, the TCU can include a measure of artificial intelligence, utiliz-

ing throttle position, vehicle speed, engine speed, and other parameters to estimate the

driver’s intention and manage shifting to provide performance characteristics that match

the driver’s intent without even needing to change the drive mode. A speed, throttle, and

4 Z. Zhong, G. Kong, Z. Yu, X. Xin and X. Chen, Shifting control of an automated mechanical transmission

without using the clutch. International Journal of Automotive Technology 13(3), 2012, 487−496.

78 Automotive Innovation

torque combination that indicate uphill driving, for example, might prevent an upshift to

ensure desired acceleration. Downhill driving, on the other hand, might automatically

trigger a downshift to avoid unwanted acceleration.

Even though automation is impressive, in itself it does not solve one of the fundamental

challenges of manual transmissions: the time required to shift gears. Any gear shift has to

entail four steps: a disengagement of the clutch, a move out of one gear ratio, a move into

another gear ratio, and clutch reengagement. That takes time, even if done automatically.

The result is an undesirable torque interruption. But, as always, there is another option:

This challenge can be met by the integration of two clutches, a dual-clutch transmission

(DCT) (Image 3.11). If we use two clutches, the disengagement of one gear and the engage-

ment of another don’t have to happen sequentially, and that means a faster transition.

Originally branded by Volkswagen as the Direct Shift Gearbox (DSG) and Audi as the

S-Tronic transmission, versions of dual-clutch units are provided by just about every major

automaker.

The trick to the DCT is to superimpose two independent main shafts in a single-

transmission housing. Odd numbered gears are placed on one of the coaxial shafts and

even numbered gears on the other. Each shaft is served by a separate clutch mechanism, so

IMAGE 3.11

Dual-clutch module.

Initially designed for European sports cars, this dual clutch module produced by BorgWarner is ideal for high-

rev, high-torque applications and allows performance shifting with minimal torque interruption.

Image: BorgWarner

79Getting Power to the Pavement

a gear on one shaft can be disengaged at the same time that a gear on another is engaged.

This can make shifting much faster and smoother, and nearly eliminate torque interrup-

tion by allowing the subsequent gear to already be selected before a shift takes place.

For example, in shifting from second to third gear, the even-geared clutch is disengaged

to allow exit from second gear while third gear is already selected on the separate odd-

geared clutch. The effect is impressive; in an automated system a DCT can provide shift

speeds in microseconds with minimal torque interruption.5

In order to take advantage of a dual clutch, the digital control logic requires that the con-

trol module predict the direction of the next gear change. So, based on speed, throttle posi-

tion, and driver behavior, the control unit must correctly preselect the likely next gear for

the idle main shaft. When the driver shifts to fifth gear, for example, the control unit algo-

rithm must decide whether the idle shaft should be preselected in fourth or sixth gear. In

normal conditions, the logic is not particularly challenging, with acceleration, the system

preselects the next higher gear in a matter of milliseconds. With deceleration, a lower gear

is selected. An unusual condition, however, can present more of a challenge. For example,

slowing in traffic might trigger a lower gear preselection. As a result, if the traffic suddenly

moves forward, the quick change in throttle position for acceleration can provoke a jerky

response. However, even in this unusual scenario, the response will still be considerably

faster and smoother than a driver-controlled manual transmission.

Clearly, electronic control of the transmission is more complicated than simply iden-

tifying the ideal gear ratio. Remember that neither the shifter nor the accelerator pedal

is a direct mechanical actuator. Both simply convey driver preferences electronically to

a microprocessor. The control logic then synthesizes these demands with engine speed

control, vehicle speed, and other factors to identify shifting points based on identified

driving strategy. For example, a call for hard acceleration needs a delayed shift to maintain

high power. An indication of leisurely acceleration requires early shift to maintain high

fuel economy. Shift points can be informed by a drive mode selection. But they can also

be informed by driver behaviors, driving settings, and even road conditions. An ability to

sense high load, due to climbing or towing for example, or forward acceleration due to a

decline, is indispensible to correct automated control. The system must also interact with

other electronically controlled vehicle systems, engine control, active chassis control, steer-

ing, traction control, electronic stability control, and others. In short, the TCU must be able

to directly or indirectly sense every variable that the most proficient driver would sense

and accommodate while driving. At the same time, the control system needs to protect

against driver error, a call for an inappropriate gear for example. Advanced algorithms,

Kalman filters, and fuzzy logic can allow for a useful synthesis of the data and the iden-

tification of an appropriate control scheme (we’ll cover a bit more on these algorithms in

Chapter 9).

So, while obviously very cool, DCT systems also face a common difficulty, providing

high-torque operation without excess heat. Launch,

273

Lidar ....................................................................................................................................... 275

Optical ...................................................................................................................................277

Sensor Fusion ....................................................................................................................... 281

Driver Monitoring................................................................................................................ 282

Localization ..........................................................................................................................283

Mapping ................................................................................................................................ 286

Communication ....................................................................................................................288

Decision-Making .................................................................................................................. 290

The Road Ahead ................................................................................................................... 293

Index ............................................................................................................................................. 295

http://www.taylorandfrancis.com

Preface

Cars have changed radically over the past few decades, and the pace of change is only

accelerating. Innovations in engine design, fuel systems, digital control, advanced trans-

missions, and a range of other technologies have fundamentally redefined the powertrain.

And advanced electronic control systems, active chassis control, driver assistance systems,

not to mention electrified drivetrains, advanced batteries, and new lightweight materials

are allowing us to profoundly reimagine what is possible. It can be tough to keep up. And

that’s the point of this book.

These exciting changes and innovations in automotive technology are complex, but

they don’t need to be intimidating. Fundamentally, the same basic principles are at work,

whether you’re looking at a Model T or a Tesla. The laws of science and mechanical

principles haven’t changed. Admittedly, the engineering particulars have become more

involved, and there are a lot more of them. But, in the end, all of the technology in the most

advanced vehicles can be understood in principle by anyone with a basic grasp of science

and mechanics.

So, think of this book as a primer, a basic survey of the new automotive landscape with

an eye toward a timely orientation to the most interesting innovations and the most prom-

ising technological advances out there. One aim of this work is to provide a solid intro-

ductory text for an undergraduate course. In particular, the idea is to fill the gap between

a vocational-based automotive repair text and an advanced engineering text. In fact, this

book grew out of an undergraduate survey course on automotive technology and design.

Finding a useful text for this sort of course has always been difficult, as nearly all introduc-

tory texts in automotive technology focus on vocational training for mechanics; and the

only other alternative is often a technically dense engineering text, a rather intimidating

introduction to the field. This text is aimed at the midpoint: a true introductory survey of

the science and engineering in automotive technology that allows a generally informed

reader to develop an understanding of the principles, trends, and challenges in automotive

technology and the possible directions of future developments.

With this in mind, the aim is to keep this work accessible and engaging. A useful orien-

tation to the field should be readable and stimulating for students, mechanics, automotive

enthusiasts, and anyone else who may have an interest in cars, technology, innovation,

or engineering. This stuff is really amazing, exciting, and frequently ingenious; but

all-too-often, the amazing stuff is buried under layers of engineering terminology and

daunting computations that can thoroughly snuff out the flame of enthusiasm in the

uninitiated. Cars can be really exiting; a book on them should be too.

So, this book will be useful for students of automotive engineering and technology that

need an orientation to the field. This book should also be useful to a seasoned automotive

technician trying to stay on top of a rapidly changing field, or a newly minted mechanic

who needs a general orientation to the near future of the automobile, and even an auto-

motive enthusiast who just wants to better understand how recent technological changes

come together.

With luck this text will inspire budding engineers and maybe even motivate a few

mechanics and gear heads to dig deeper, continue to explore the field, and perhaps even

choose to take the next step and select a career that allows them to contribute to redefin-

ing the future of the automobile. This is truly a golden age in automotive design, a time

x Preface

when the future seems up for grabs, and a new possibilities have become not just feasible

but likely.

Remember, this book is intended as a primer. You don’t need a deep background in

automotive technology to keep up. But a basic understanding of science and the funda-

mentals of mechanics won’t hurt. Each topic and each chapter begin at the beginning,

the basic principles that underpin the technology. Subsequently, the chapters move on to

the ideas and engineering that define some of the most exciting innovations in the field;

and in the end, each chapter addresses some of the most promising advances for the near

future. In sum, the chapters offer a basic lay of the land, an orientation to the technology

that is reshaping that field and plenty of real-world examples of remarkable automotive

innovations.

Since the idea is to keep this book accessible, approachable, and short, this text has

defined limits. It is fundamentally about cars, the current cars on the road now, and

the likely changes that will define the cars on the road tomorrow. It’s not about the

automotive industry more generally, or the future of transportation infrastructure, manu-

facturing, or policy. Nor is it a detailed examination of research in science or engineer-

ing. So, automotive-related innovations that could one day reshape vehicles by remaking

transportation infrastructure, such as alternative fuels, fuel cells, or intelligent transpor-

tation systems (ITS), are interesting, but that’s really not what this book is about. Likewise,

this is a primer; so, for a complete presentation of the advanced engineering techniques

and computations needed to design these systems, you’ll need to look elsewhere. In short,

both the scope and depth of this text are intentionally focused. This is an introductory

survey of contemporary automotive innovations for readers with a basic mechanical and

science background.

With all of this in mind, this book addresses four principle areas: first, the technology

of the combustion-based automobile on the road now and in the near future, addressed in

the first three chapters; second, the technology of the electrified drivetrain that’s increas-

ingly present now and very likely to become dominant in the near future, addressed in

the subsequent three chapters; third, innovations in chassis and body design, which are

covered in Chapters 7 and 8; and lastly, a basic introduction to the sensor and navigation

technology that enables advanced driver assistance systems and the possibilities for self-

driving cars, addressed in the final chapter.

So, within this broad framework, Chapter 1 begins with the basics of the internal com-

bustion engine and quickly moves on to review recent innovations in ignition manage-

ment, advanced fuel delivery, combustion chamber design, and moves through to the

basic principles of

for example, needs to take place while

engaging a single clutch. We will see later than an automatic transmission relies on a

torque converter to multiply low-speed torque. But a DCT does not typically have that

option, and particularly in a high-torque condition, say a turbo charged engine or a high-

load diesel, this can lead to problems. The extended clutch slip needed to allow smooth

5 F. Vacca, S. De Pinto, A.E.H. Karci, P. Gruber, F. Viotto, C. Cavallino, J. Rossi and A. Sorniotti, On the energy

efficiency of dual clutch transmissions and automated manual transmissions. Energies 10, 2017, 1562; and

G.Shi, P. Dong, H.Q. Sun, Y.Liu, Y.J. Cheng and Y. Xu, Adaptive control of the shifting process in automatic

transmissions. International Journal of Automotive Technology 18(1), 2017, 179−194.

80 Automotive Innovation

engagement at high engine speed can lead to excess heating, performance deterioration,

and mechanical damage. This brings us back to that idea of submerging the clutch in oil.

A wet clutch can provide smoother engagement and cooling but with some lost efficiency

due to the need for a pump and the losses caused by the resistance of the clutch pack

turning in oil, called churning.6 An innovative option is to just add a torque converter to

a DCT as Honda did with the 2.4 L Acura TLX. As we will see soon, a torque converter

can provide torque multiplication during launch, though lost efficiency and higher costs

are still challenges.

So, the manual transmission has truly come of age. Twin-disk clutches, more closely

spaced gears, improved synchronization, and enhanced automation make manual trans-

missions a promising option for performance, without even considering the improved fuel

efficiency and lower cost. As a result, they are continually relevant in Europe, even if they

have yet to impress American buyers. It is possible that since most US drivers have never

driven a manual transmission they are less favorably disposed to the more abrupt torque

shift of a manual gearbox and less inclined to use the manual feature, making the AMT

feel like it is just a rougher version of an automatic. In addition, their relative complexity

and fragility mean DCTs are increasingly challenged to handle the output of high-power

engines. So, American automakers are largely looking past automated manuals toward

improved automatic transmissions for the US market. The breakthrough may come with

the growth in hybrid drives since these dual-disk systems could be well suited for hybrid

systems, where the additional torque of the electric drive can more easily handle the launch

load. More on this in Chapter 5.

Automatic Transmission Coupler

If we want to understand the automatic transmission systems, we have to start with the

coupler. The friction clutch is efficient, reliable, robust, and inexpensive. But, until recently

it required driver input to operate, and so was not an option for an automated system.

An alternative coupling mechanism called a torque converter provides a fluid rather than

mechanical coupling, and can effectively allow coupling and uncoupling with no driver

input. The basic idea is pretty simple. Imagine a fan blowing on another fan. The rotation

of one causes a rotation of the other. But, if you stopped the rotation of the second fan, it

wouldn’t bother the driving fan. Got that? Okay, now contain the whole thing in a fluid

filled container that can direct and concentrate the flow and provide a viscosity that offers

a much-improved transfer of force. This is your basic fluid coupling.

In practice, the essential mechanism includes three key components: a toroidal impel-

ler that serves as a pump and is fixed to a housing that connects to the engine flywheel;

a turbine that is driven by the pumped fluid, has no mechanical contact with the pump,

and is mounted to the transmission input; and a set of stator blades between the two that

serves to redirect the fluid flow from the turbine to the pump. This last component is key

because it allow for torque multiplication. The impeller is driven by the engine and defines

a rotary flow that produces a centrifugal pressure inside the housing. The resulting pres-

sure causes an outward movement of fluid defining a vortex flow that impacts and rotates

6 F. Vacca, S. De Pinto, A.E.H. Karci, P. Gruber, F. Viotto, C. Cavallino, J. Rossi and A. Sorniotti, On the energy

efficiency of dual clutch transmissions and automated manual transmissions. Energies 10, 2017, 1562.

81Getting Power to the Pavement

the driven turbine. The fluid is then redirected through the stator and back to the impel-

ler. The stator’s capacity to redirect the flow means the return flow from the turbine can

be reversed so it does not work against the rotation of the impeller. This means torque is

increased, making this a more useful device than a simple fluid coupling and earning the

name torque converter (Image 3.12).

Understanding how this works requires a bit more detail. Let’s begin at idle. Because

there is no mechanical connection between the driving and driven members, decoupling

is not a problem and the car sits with the pump turning and the turbine still. As the engine

accelerates from idle, the pump’s speed increases as does the resulting rotational flow and

pressure (indicated below by the arrows). As this pumped fluid impacts the turbine with

growing force, it defines an increasing torque that rotates the turbine and output shaft. The

fluid then continues its helical flow and moves back toward the center of the impeller with

a movement that is opposite the pump rotation. Normally, you would expect this flow to

act against the impeller’s rotation. However, the flow is channeled through the stator and

because a one-way clutch is used to allow the stator to rotate only in the direction of the

impeller, the fluid flowing from the turbine and through the stator is redirected back to

the rotational direction of the pump. The one-way clutch is the key here. By linking the

impeller to the transmission casing, a counter rotational force on the impeller is avoided

and this allows for torque multiplication since the turbine is now being rotated by both the

IMAGE 3.12

The basic torque converter.

A basic torque converter includes a toroidal impeller connected to the housing that serves as a centrifugal

pump, a turbine that is driven by the pumped fluid (indicated by the arrows), and stator blades between the two.

The stator allows for torque multiplication since the high-pressure fluid that returns to the fast-moving pump

can be redetected, ensuring it does not work against the pump.

82 Automotive Innovation

force pushing it due to the pressure of the incoming fluid as well as the benefit of ejecting

the exiting fluid against the fixed stator.

As turbine speed increases, the relative difference in speed of the pump and turbine

decreases, and so the torque multiplication decreases. In this sense, it works a bit like

gearing: the less the speed decreases, the less the torque increases. At higher speed, when

the turbine is turning nearly as fast as the pump, the fluid leaves the turbine with much

less velocity, and the relatively geometry of the turbine and stator change. Effectively, the

turbine is now rotating so quickly that the fluid’s relative transaxial motion changes and

the fluid leaving the turbine begins to hit the backside of the stator. So, the stator no longer

reverses the flow, it now freewheels with the impeller. At this point, called the coupling

point, there is no longer torque multiplication, and the torque converter operates as a basic

fluid coupling. The ability of the stator to freewheel avoids the drop in torque that would

result from the changed geometry at high speeds if the stator were fixed, and it allows a

continued rise in efficiency to perhaps 92% or so.

Fundamentally, the capacity for torque multiplication is possible because the pump

is rotating faster than the

turbine. The greater the difference in the rotation speeds, the

greater the torque. This has the happy effect of providing the greatest torque multiplica-

tion, typically about 2:1, when the car just begins to move. As the speed picks up, the rela-

tive speed difference of the impeller and turbine decrease and the multiplying effect drops

off. The difference between the two speeds is called slip. Fortunately, at higher speed

when no torque multiplication is needed, the slip is reduced. But there is always some slip;

otherwise no torque could be transmitted through flow. At lower speeds, as slip increases,

the efficiency of a torque converter drops notably.

Some of this inefficiency can be addressed by adding a clutch mechanism that can lock

the assembly together at cruise speed called a lock-up clutch. This is particularly advanta-

geous in overdrive, since the negative relative slip could cause fluid cavitation and over-

heating. While this technology has been around for a while, and is nearly universal in cars

manufactured in the last two decades, more precise digital control now allows for earlier

and more frequent engagement of the lock-up mechanism by identifying conditions for

lock-up engagement with low load and high speeds in every gear ratio but the first. The

result is improved fuel efficiency as well as reduced heat accumulation.

Automatic Transmissions

While we now know that a layshaft transmission can be made to operate automatically,

this was not always the case. So, when engineers set out to build an automatic transmis-

sion more than half a century ago, they began with an entirely different starting point.

As a result, the heart of the conventional automatic transmission is not a series of gears

lined up on dual shafts; it is an epicyclic or planetary gearset. The most basic example is

defined by four components: A center gear, called a sun gear, multiple pinion gears that

mesh with the sun gear and can rotate around it, called planet gears, an outer ring with

teeth on the inside, called a ring gear or annulus, and a carrier that connects the planet

gears together and allows them to rotate as a unit (Image 3.13). Each component can be

fixed in position or allowed to spin though the application of clutching mechanisms or

band brakes. Changes in the drive ratio are then simply a matter of changing the driving

and driven components and fixing or releasing others to define various combinations of

83Getting Power to the Pavement

gearing. Thekey characteristic that made this indispensable to automated operation is that

because all gears in the gearset are in constant mesh, torque flow need not be interrupted

to allow changes in gear ratios.

So, a single planetary gearset provides multiple drive ratios by holding one of the three

components in place with a clutching mechanism and engaging the other two to define

a torque transfer. Maximum reduction, or underdrive, can be achieved with the sun gear

as the input and the carrier as the output. The ring gear is held in place and the planet

gears slowly walk around the sun gear as it rotates, defining a reduced rate of rotation.

Maintaining the sun gear as the input, if the carrier is held in place with a clutch pack, the

ring gear can define the output, offering a less reduced underdrive. Alternatively, the ring

gear could be made the input, and while holding the carrier, the planet gears will define a

reverse rotation of the sun gear, providing a reverse output. Variations in increased speed,

or overdrive, can occur if the carrier is used as input and either the sun or ring is output,

with the other held in place. Of course, if all the components are lock together, the gearset

can provide direct drive.

The planetary gearset has a few advantages over the layshaft transmission. Most impor-

tantly, they are compact and because all elements rotate around a single axis they easily

can be placed in series or nested together in differing combinations to define a compound

gearset. The most well-established compound gearset is the Simpson gearset, with appli-

cations in production cars dating back to the 1960s. Effectively, this is simply two plan-

etary gearsets with their sun gears linked. The front and rear planetary components can

have different sized gears and so can define differing gearing options. This basic con-

figuration allowed a significant improvement on the two-speed automatic transmission

that preceded it; and this configuration is still used as a low-cost unit in contemporary

production cars.

IMAGE 3.13

Basic planetary gearset.

84 Automotive Innovation

The commonly used Ravigneaux offers all the basic features of the Simpson gearset,

but in a somewhat more compact package and with increased tooth contact and increased

torque capacity. Unlike the Simpson system with a shared sun gear, two different sized

and independent sun gears are used with a common ring gear (Image 3.14). A set of short

planet pinions mesh with the small sun gear, and a set of longer planet gears connect with

the large sun gear, the three short pinions and the ring gear. The two sets of planet gears

are on a common carrier. Because the carrier is one of the larger and more expensive com-

ponents of a gearset, this typically makes the gearset smaller, lighter and less expensive

than the Simpson gearset.

Once again, four speeds are possible: The lowest gearing ratio can be achieved with the

small sun gear connected to the engine and the planet carrier prevented from rotating.

This results in the small planetary pinions counter rotating against the sun gear and so the

long planetary gears and ring gear turn in the direction of the initial input but at a reduced

speed. Using the same input connection but locking the large sun gear requires the long

pinion gears to walk around the large sun gear producing less of a reduction. Of course,

direct drive is achieved by locking the sun and carrier. An overdrive ratio is attained by

linking the engine to the carrier and allowing the long planetary pinions to walk around

the large sun gear, driving the ring gear output at an increased speed.

A breakthrough in transmission innovation came with the development of the Lepelletier

gearing mechanism, making new configurations and expanded gearing ratios possible.

First proposed in the early 1990s, this system combines a simple planetary gearset with a

Ravigneaux unit on a common central shaft. The input is linked to the simple planetary

ring gear. However, a key feature is the possibility of two inputs, as input can also be con-

currently connected to either the carrier or large sun gear of the Ravigneaux unit or both.

The carrier of the simple gearset is connected by clutches to the large or small sun gear

in the Ravigneaux set, providing a reduced input in underdrive gears. And, typically, the

sun gear of the simple gearset is connected to the housing and does not rotate. The sun

gear of the Ravigneaux gearset is driven by the output of the single planetary gearset.

And the output of the compound gearset is defined by the Ravigneaux ring gear. The key

IMAGE 3.14

Simpson and Ravigneaux gearsets.

The Simpson gearset offers a simple configuration that has made it is the most historically common planetary

gearing configuration. The Ravigneaux offers a more compact unit with improved torque capacity.

85Getting Power to the Pavement

characteristic to note is the sun gears and carrier of the Ravigneaux unit can be driven at

different speeds, allowing for a larger number of possible overall gearing ratios since the

resulting gearing of the compound unit is defined by the combination of the two input

ratios. We’ll talk more about this in Chapter 5.

While the mechanics of the Lepelletier system have always been possible, it is only with

digital control that the complexities of the multiple clutch pack combinations could be prac-

tically managed. The resulting unit is impressive, relatively simple

to control, lightweight,

compact, and robust. The engineering firm ZF used this configuration to produce the first

production six-speed automatic (6 HP 26) in 2001, and since then the number of speeds in

an automatic transmission has increased dramatically (Image 3.15). The Lepelletier gearset

makes 11 speed ratios possible.7

This capacity to increase available drive ratios can define a notable advantage over the

manual transmission. With the ability to nest gearsets and increase gearing combinations

without increasing the number of clutches, more gears does not always mean more weight

or size for an automatic transmission as it generally does for a manual. As a result, choos-

ing a manual transmission often means choosing limited performance. So, for example,

7 E.L. Esmail, Configuration design of ten-speed automatic transmissions with twelve-link three-DOF lepel-

letier gear mechanism. Journal of Mechanical Science and Technology 30 (1), 2016, 211–220.

IMAGE 3.15

Automatic transmission.

The ability to nest gearsets means increased gearing does not always mean significantly increased size or

weight in an automatic transmission. This eight-speed ZF unit makes efficient use of space, fitting clutch packs

and gearsets neatly in a compact design.

Source: ZF Friedrichshafen AG

86 Automotive Innovation

while the 2018 Ford Mustang comes with an impressive seven speed manual as stock, the

optional automatic is able to provide a full ten gear ratios, gear changes that a professional

driver would have trouble matching, and undeniably improved performance. The down-

side is perhaps in driving experience and certainly in cost, in the Mustang’s case adding

$1500 to the sticker price while barely nudging the fuel economy.

Transmission Control

The result of this ever-growing number of gearing ratios can be a mixed bag. Adding

more gear ratios allows the engine and transmission to operate at lower BSFC; and so with

rising efficiency expectations, nine- and ten-speed transmissions are becoming common.

Moreover, more gear options can help achieve specific design goals, such as improved

low-end torque or high-speed cruising. However, more gears can also mean more shifting,

which can effect driver perception of ride quality. So, control systems need to be carefully

designed to exploit the possibilities entailed in more gears without compromising ride

quality or performance. It’s worth noting that it is not only the number of gears but more

importantly the spread of the available gear ratios that can determine ride quality and per-

formance. With a greater spread of gears, the smaller engines being used now can provide

desired launch acceleration with a lower launch gear, while still managing high-speed ride

quality and performance with a higher top end overdrive.

Whatever gearset configuration is used, it is clear that sophisticated computer control

and innovation have dramatically changed the possibilities. Initially, the planetary gear-

set was adopted for automatic transmissions because the goal of automated operation

required it. At the time, it was easier to devise an automated hydraulic mechanism for

engaging and disengaging hydraulic clutches than it was to automate the actuators in

a layshaft gearbox. Until recently, these transmissions were controlled with a network

of hydraulic fluid channels in a complex valve body that actuated the various clutching

mechanisms. However much of these functions have now been integrated into a digital

TCU that uses electronically controlled hydraulic solenoids, expanding design possi-

bilities greatly. Computer control enables enhanced clutch-to-clutch actuation, with a

clutch to one gear disengaged the instant a clutch for the other is engaged. At idle, the

transmission can be automatically placed into neutral, helping manage temperature and

improve fuel efficiency. Multiple variable displacement pumps can be used to produce

just the hydraulic pressure needed, and significantly reduce associated loss. With pump

loss accounting for nearly two-thirds of transmission power loss, this can improve effi-

ciency significantly.8

The new range of possibilities was evident when Ford and GM introduced the result

of their joint effort to produce an advanced 10-speed automatic in 2017. Using sophisti-

cated programing they are able to achieve a high-performance transmission for trucks

and sports cars. The mechanics was kept relatively simple, using as much existing com-

ponentry as possible, the new gearbox links four simple gearsets and six clutches. The

real innovation is in enabling each automaker to develop specific control software for

their applications. Ford can therefore define a transmission that prioritizes towing capac-

ity and fuel economy in its F-150. Closely set overdrive gearing helps improve towing

8 M. Gabriel, Innovations in Automotive Transmission Engineering. SAE International, Warrendale, PA, 2004.

87Getting Power to the Pavement

performance, allowing imperceptible shifting to accommodate higher overdrive torque

needs. At the same time, GM can use the same mechanics to define a sports transmission

for its Camaro. Low-friction components, a bypass that allows for a faster rise to operat-

ing temperature, a capacity to shift without requiring the torque converter to unlock, and

the use of high-performance, low-viscosity automatic transmission fluid (ATF) make this

one of the most efficient transmission ever designed. All in a package that is only an inch

longer and 4 pounds heavier than the six speed it replaced.9

This search for greater transmission efficiency can be helped with careful consideration

of the control logic, but this can also present challenges. Once again, there are trade-offs.

For example, the enhanced efficiency that can be achieved by shifting at an earlier stage

during acceleration, often called shift optimization or aggressive shift logic (ASL), tends

to be tied to reduced acceleration performance. So, often combined with early torque con-

verter lockup, ASL can provide notable efficiency improvements at acceleration and cruise.

And, as mentioned previously, lockup can occur whenever torque multiplication is not

available, in every gear but first. Disengagement occurs when torque demand drops, to

allow coasting with low engine rpm. However, locking the torque converter at a lower

speed can cause shuddering that can impact the driving experience and shape perceptions

of vehicle quality. Improved engineering has resolved some of this, but driver acceptance

of noise is still a concern. And because the effect of downsizing and higher turbocharging

have meant very high torque in the low-speed range, the challenge of addressing noise,

vibration and harshness (NVH) has grown more complex. Features like cylinder deacti-

vation, particularly as it is increasingly being applied to smaller engines, have made this

even more challenging.

To help address the challenge of increased low-speed torque and early shifting in both

manual and automatic cars, the humble flywheel and modest torque converter are getting

a makeover (Image 3.16). Dual-mass flywheels are an established technology that has sig-

nificantly reduced vibration caused by the irregular torque of the crankshaft. However, a

centrifugal pendulum-type absorber goes a step further. By integrating a pendulum mass

that can rotate outward and significantly increase the flywheel’s moment of inertia tem-

porarily, the flywheel can help cope with drivetrain vibration. At low speed, the deflec-

tion of the mass can counteract torsional vibration and dramatically reduce appreciable

shudder and noise in a manual or an automatic application. This makes early shifting

possible while protecting ride quality. So far, this has not been used in dual clutch units,

perhaps because the larger mass needed to provide thermal stability in a dry DCT pro-

vides enough dampening on its own. This works for both AMTs and automatics, though

in automatic transmissions the centrifugal absorber is immersed in oil within the torque

converter.

In fact, changes to the torque converter have defined what is sometimes called a mul-

tifunction torque converter (MFTC). A torsional dampener mechanically insulates the

outer shell of the torque converter from the inner components with a sprung dampener.

This can allow for a more responsive TC while minimizing slippage and dampening the

vibration of early shifting.10 In addition, overall reduced weight from conventional torque

converters can further improve efficiency and responsiveness. The addition of a clutch

on the impeller can be used to allow the engine to spin at a higher speed than the pump,

9 B. Chabot, “The Need for 10-Speeds.” Motor June, 2017. Available at www.motor.com/magazine-summary/

need-10-speeds/

10 A. Isenstadt, J. German, M. Burd and E. Greif, Transmissions. Working Paper, The International Council on

Clean Transportation, August, 2016.

http://www.motor.com

88 Automotive Innovation

enabling higher torque faster, and is particularly useful when trying to address turbo lag.11

So, even the humble torque converter has come of age.

Continuously Variable Transmissions

Conceptually, a more desirable transmission would allow for ideal engine speed and

torque at all times, without compromising for the gap between gears and without disrup-

tion for gear changes. In short, ideally, we would want to select from an infinite possibility

of gear ratios, to choose just the right one for a given condition, and then seamlessly and

continuously adjust to any other drive ratio as needed. Sound impossible? Actually, the

idea of a transmission that can do this isn’t fantasy; it’s not even new. Leonardo DaVinci

sketched one in the fifteenth century, and they have appeared in a variety of automobiles

here and there since the early twentieth century. The 1934 Austin 18 offered an admirable

example. Recently, the application of a continuously variable transmission (CVT) has

enjoyed a revival; but it’s future is less than certain.

The most established CVT type is a variable pulley and belt system called a push belt.

Like any other pulley, these pulleys are defined by two opposing tapered sides. The differ-

ence is that these sides move, opening and closing the width of the pulley, defined by one

11 K. Buchholz, “New-gen Torque Converter Aims at 2017 Vehicle Intro.” Automotive Engineering SAE October,

2014.

IMAGE 3.16

Torque converter with centrifugal pendulum.

This advanced torque converter by LuK includes a centrifugal pendulum absorber that is able to absorb disrup-

tive bumps in rotation and return the energy to the system more smoothly later. This can markedly improve the

operation of engines with fewer cylinders and lower operating speeds that offer improved efficiency.

Image: LuK USA

89Getting Power to the Pavement

fixed and one adjustable pulley half. Called a variator, the sides can be drawn together or

pulled apart to widen or narrow the width of the riding surface for the belt and increase

or decrease the effective radius of the pulley (Image 3.17). Clearly, this system cannot use a

simple V belt. In fact, the belt is defined by several steel bands attached to multiple radial

plates to provide a ridged cross section that can ride on the edges of the variator. This steel

belt is commonly called a Van Doorne belt and is used by nearly every production push

belt CVT. As the variator moves in, the belt is wedged between the narrowing sides of the

pulley and so rides higher up the tapered walls, effectively defining a larger pulley radius.

With the belt having a fixed length, the walls of the other pulley are forced to separate,

allowing the belt to ride lower between the sides and define a smaller radius. This also cre-

ates compression on one side of the belt and slack on the other, hence the name push belt.

The result is seamless adjustment of relative pulley size and drive ratio on the fly. With

smooth acceleration from a stop possible in a low ratio, and modest and reliable low-speed

movement, or creep, a torque converter can be eliminated in favor of a simple starting

clutch. With fuel economy as the target, the CVT typically operates to keep the engine near

a constant speed at low BSFC, shifting drive ratio to accommodate throttle position.

IMAGE 3.17

Toyota direct shift CVT.

Toyota’s new CVT utilizes a launch gear to start from a stop, leaving the belt drive for higher gear ratios. This

enables reduced belt load, a wider gear range and a resulting 6% improvement in efficiency. A reduced belt

angle and smaller pulley allow faster shift speeds.

Image: Toyota Motor Company

90 Automotive Innovation

In the past, the push belt CVT was only an option for low power applications. The lim-

its of the belt’s traction restricted the amount of torque that could be provided. However,

improved chain drive belts developed by Schaeffler Automotive have defied this restric-

tion. Most notably, Audi chose a link plate steel belt approach in its Multitronic CVT.

The result can be as flexible as a V belt and able to transmit force through hard pin ends.

The rolling-pin type pulley contact means less friction, though it can also mean more

slip. Nevertheless, while newer belts such as this offer greater strength and improved

traction and so efficiency, the friction force required to provide belt traction unavoidably

results in energy loss. And performance is heavily dependent on precise lubrication and

accurate control of the clamping force of the pulleys. The result is efficiency somewhere

between 85% and 92%, generally somewhat lower than an automatic transmission and

notably lower than a manual transmission.12 These results have been disappointing to

some manufactures. However, while CVTs used to be limited to smaller cars because

of the torque limitations of belt slip, belt improvement have changed the calculus. The

Nissan Murano, for example, a midsize crossover SUV with a 3.5-liter engine relies on

a CVT.

However, let’s not throw the baby out with the bathwater. If belts are the problem, it is

possible to avoid the use of belts altogether. A notable example is offered by the toroidal

traction drive CVT. Defined by two opposing toroidal disks and two or three oscillating

rollers between them, this system offers variable drive ratios without high friction contact.

If the rollers are made to turn toward the driving disk, the outside of the roller is driven

by the outside of that disk, and the roller then drives the inside of the opposing disk, so a

high ratio is defined. As the rollers are turned toward the other disk, the ratio is increas-

ingly reduced. Because the friction and wear of a system with such high friction contact

between hard surfaces would be extreme, a viscous fluid with high shear strength and

friction characteristics well above ATF is used to transmit the torque between surfaces.

Called elastohydrodynamic lubrication (EHL), this fluid allows the sheer resistance of the

film between the surfaces to transmit force, with the surfaces themselves never touching.

Used by Infinity’s Q35 and Nissan’s Cedric, this drive offered efficiency and a smooth ride

with absolutely no torque interruption. The inherent slip at startup of the toroidal drive

was addressed by adding a torque converter. However, challenges with reliability and

consumer acceptance continue to plague this system and have led Nissan to discontinue

its use.

Hope for the future of the CVT is not lost, but it’s plenty shaky. There’s innovative

work being done to improve on these first efforts. The Torotrak system offers an inter-

esting example. While existing toroidal drive systems utilize a half toroidal geometry,

the Torotrak system used a full-toroidal mechanism, by placing two sets of three orbiting

rollers between mated toroidal disks. The result was a variator with a much wider spec-

trum of potential drive ratios.

However, the system presents some design challenges,

particularly given the large range of roller angles; and before these were resolved, the

maker succumbed to a heavy debt burden and ceased operations.13

A more recent variation on this theme is defined by the engineers at Dana Inc. Their

VariGlide planetary variator could offer the heart of a compact and relatively simple

CVT. The company points out that their system represents a significant departure from

conventional CVTs.14 The key is rolling spheres set between input and output traction

12 H. Heisler, Advanced Vehicle Technology, 2nd edition. SAE International, Warrendale, PA, 2002.

13 M. Gabriel, Innovations in Automotive Transmission Engineering. SAE International, Warrendale, PA, 2004.

14 Communications with Jeff Cole, Senior Director, Corporate Communications Dana Incorporated

91Getting Power to the Pavement

rings(Image 3.18). The spheres can collectively shift their axis of rotation; so, the input and

output rings can be made to run on larger or smaller diameters of the planetary balls’ rota-

tion. The result may be more compact and flexible than the toroidal and more durable and

powerful than the belt drive. With a full ratio sweep taking just two rotations of the unit,

the system can allow a wide range of vehicle calibration options and drive modes. Like

the toroidal drive, an EHL fluid is used, avoiding excess wear and offering efficiencies that

are potentially in the high 90s. While not yet in production, the company hopes to have it

ready for the road by 2022.15

Still, the shift from simple gears to spheres or toroids is anything but certain. When

compared to the four-speed automatics available upon their initial development, CVTs

looked promising. After all, an improvement of as much as 8% could be achieved by

replacing a four-speed automatic with a CVT.16 This is particularly true for low-speed

15 Communications with Jeff Cole, Senior Director, Corporate Communications Dana Incorporated; and

T.Murphy, “Planets Aligning for Dana’s VariGlide Beltless CVT.” WardsAuto August 22, 2017.

16 A. Isenstadt, J. German, M. Burd and E. Greif, Transmissions. Working Paper, The International Council on

Clean Transportation, August, 2016.

IMAGE 3.18

VariGlide system.

The VariGlide planetary variator offers a new option for CVTs. By collectively altering the axis of rotation of

spheres set between traction rings, the input and output rings are made to run on smaller or larger rotational

diameters of the spheres.

Image: Dana Incorporated

92 Automotive Innovation

city driving, when automatic transmission inefficiency can be very high. However, when

now compared to the high-geared transmissions available, the CVT loses a bit of luster.

Moreover, they face a significant challenge of driver acceptance. Used to the experience

of shift shock, and having learned to associate that with performance and acceleration,

drivers can find the experience of CVT cars wanting, particular by those looking for per-

formance over efficiency. As a result, while still maintaining popularity in Japan, CVTs

are losing appeal in the US and Europe. For many, advances in DCTs and high-geared

automatics have made the compromises of the CVT unnecessary. Nevertheless, their rela-

tive simplicity can make a lot of sense for a modest cruiser looking to gain some fuel

economy. It’s likely that it is just that goal that convinced Chevrolet to replace the stan-

dard six-speed automatic with a CVT its 2019 Malibu, the car makers only full size non-

hybrid with a CVT.

Differentials, AWD, and Torque Vectoring

Once shaft speed and torque is established through the transmission, it needs to be dis-

tributed to the wheels. This might seem like a simple task; it is not. There are some tricky

issues. First, the basic point: cars turn. As a result, all wheels on a car strike a distinct arc

during a turn and so need to be able to turn independently. Second, a bit more tricky,

proper handing and safety can be significantly enhanced by the possibility of apply dif-

fering torque to each driven wheel. Third, this is made difficult because each wheel experi-

ences differing downward and lateral forces and differing friction. Lastly, it’s worth noting

that all of this gets harder when we’re using all-wheel drive.

Managing the wheel speed variation needed for turning can be pretty simple. A

basic open differential can allow for differential rotation of the wheels on a single axle,

allowing turns with no necessary tire slip. This is accomplished with a basic pinion

gear and ring gear, coupled with a set of beveled gears to allow differential rotation

at each end. The challenge with this solution is that torque will be lost if either wheel

fails to produce traction. If a wheel slips on ice, for example, it will spin freely, allow-

ing the opposing wheel to remain stationary while the full driveshaft rotation goes to

the wheel with no traction, the opposite of what we want. Alternatively, a limited slip

differential (LSD) can offer the ability to either completely or partially lock the axles

to avoid dumping torque at the least tractive wheel. While there are various designs

out there, the typical system uses clutches to lock the axle whenever the differential’s

ring gear and driveshaft torque are uneven, that is to say when there is acceleration. A

so-called one-way LSD responds only to forward acceleration, and a two way unit also

responds to deceleration.

Electronic stability control (ESC) offers an improvement on this technology. Essentially

an extension of antilock braking (ABS), ESC monitors the vehicle’s motion and slip and

applies individual wheel braking to correct slip and maintain a hold on the road. Basically,

asymmetric braking creates a rotational moment around the car’s vertical axis, called a

yaw moment, and can correct unwanted turning due to slip.

Of course, for this system to work, we need to know when the car is slipping. More pre-

cisely, the yaw rotation of the vehicle relative to its intended direction of travel needs to be

sensed. This can be tricky. A steering angle sensor can be used to determine the intended

93Getting Power to the Pavement

direction of travel. So, that’s simple enough. And wheel speed sensors are used to identify

slip at each wheel. Measuring sideways slide, or lateral acceleration, is a bit more compli-

cated, but not much. A lateral acceleration sensor uses a bending element comprised of

two piezoelectric layers. When a car starts sliding sideways the lateral acceleration pro-

duces a bending force on the element resulting in compression on one side and tension

on the other, and the crystalline sensors produce resulting voltage differences that are

proportional to the rate of acceleration.

We also need to know the yaw rotation of the car; this gets a little thornier. Again,

a piezoelectric element is key; and in this case, it is incorporated into an interesting

example of a quartz rate sensor (QRS). A yaw sensor is composed of a microminature

double-ended tuning fork. Put simply, the tinny two-ended fork is placed vertically, with

two tines pointed upward and two pointed downward. The fork is made of a single

crystalline cell of quartz, so it can be made to vibrate with the application of an alternat-

ing current. The upper two tines of the quartz fork oscillate toward and away from each

other, and since the tines are identical, their vibrations cancel each other out, and there

is no energy transfer to the center of the fork. But when the fork is rotated the symmetry

is disturbed, the rotational effect combines with the linear movement of the car to create

a lateral force, called a Coriolis force, which is proportional to the rate of rotation. The

magnitude of the force is sensed by piezoelectric sensors on the fork, defining a useful

yaw signal.

The ESC control unit monitors yaw acceleration, vehicle speed, and steering angle con-

tinuously, and activates ESC

correction when a loss of steering control is determined. At

this point it actuates asymmetrical braking on wheels to counter the car’s slide. In an over-

steer condition, when the vehicle is rotating too much into the turn, ESC can apply braking

to the outside front wheel to create a counter rotational force. In an understeer condition,

when the car’s rotation is less than needed and the car’s turn is wide, braking can be

applied to the inner rear wheel to increase the rotational force into the turn. The system

can also reduce engine power or even shift the transmission into a low gear to maintain

traction and control.

However, a more effective option comes once again with the possibility of even more

sophisticated digital control. The challenge with ESC is that it relies on the application

of braking. This is fine in a critical cornering, deceleration or crash-avoidance situation.

However, the loss of energy entailed in braking means it is not an ideal way to improve

general cornering performance. If you are trying to take a turn fast, automated braking is

not always your best friend. Moreover, as might be expected, ESC increases fuel consump-

tion.17 A preferable system would actively distribute differential torque to each wheel to

produce the same effect without braking. That’s where torque vectoring comes in. As the

name implies, torque vectoring can deliver discrete torque to each drive wheel using an

active differential, or torque-vectoring differential (TVD). The system compensates for

variations in traction and applies yaw correction to enhance cornering. So, rather than dif-

ferential braking, differential torque is applied, pushing the car forward while enhancing

road holding and cornering. In the simplest example, when slipping is indicated, torque

can be reduced at that wheel and increased at the other. And, in cornering, a TVD can

send power to the outside wheel, avoiding the braking effect of excess torque at the inside

wheel. Using two wheels rather than one, a TVD can mark a noted improvement over

17 D. Piyabongkarn, J.Y. Lew, R. Rajamani and J.A. Grogg, “Active Driveline Torque-Management Systems.”

Control Systems Magazine, IEEE 2010, 30, 86–102.

94 Automotive Innovation

basic ESC. In fact, with the exception of the most critical loss of control situations, torque

vectoring has demonstrated far improved handling over ESC.18 And the efficiency losses

associated with ESC are avoided.19

What’s new in advanced torque vectoring is the capacity for active control. Rather than

rely on engine torque or friction to allocate torque to the wheels, computer control uses

sensor input to provide a wider range of regulation, with the capacity to possibly allocate

all torque to a single wheel, or remove torque from a wheel entirely. Predictive algorithms

can respond to road conditions, driving behavior and tire performance to predictively

adjust torque allocation before slip even occurs.

A typical TVD is composed of differential gearing in the center, very much like the old

open differential (Image 3.19). On either side is a multiplate wet clutch pack, like the LSD.

An electric motor or hydraulic actuator can engage and disengage the clutches and allow

reduced torque to either side as needed. So, when a clutch is actuated on a wheel, the bal-

ance of the available torque is redirected to the alternative wheel. No braking required.

Advanced systems such as Ricardo’s Cross-Axle Torque-Vectoring system used in the

Audi A6 place planetary gearsets inboard of each clutch allowing differential speed at

18 S.M.M. Jaafari and K.H. Shirazi, A comparison on optimal torque vectoring strategies in overall performance

enhancement of a passenger car. Journal of Multi-body Dynamics 230(4), 2016, 469–488.

19 M. Hanco*ck, R. Williams, T. Gordon and M.C. Best. A Comparison of Braking and Differential Control of

Road Vehicle Yaw-sideslip Dynamics. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of

Automobile Engineering 219, 2005, 309–327; and J. Deur, V. Ivanovic, M. Hanco*ck and F. Assadian. Modeling

and analysis of active differential dynamics. Journal of Dynamic Systems, Measure, and Control 132, 2010,

061501–061514.

IMAGE 3.19

Active torque vectoring.

Magna’s TWIN Rear Drive System offers limited slip all-wheel drive with torque vectoring. Torque can be

independently directed to each wheel thanks to twin couplings on the rear axle. Multiplate wet clutches are

controlled by an electric motor and managed by an ECU to offer optimal torque distribution under changing

driving conditions.

Image: Magna

95Getting Power to the Pavement

the clutches. This allows wheel torque to function more independently of the incoming

engine torque. When driving straight, the gears in the gearset are fixed, but in turns,

torque is distributed to each wheel with a response time of less than a tenth of a second

(Image 3.20).

When incorporated in an AWD system, the effect is impressive. For our purposes here,

let’s make a distinction between AWD and 4WD. Four-wheel drive relies on a transfer

case to provide torque to the front axle. However, there is no differential rotation possible

between the two axles. This makes it unsuitable for dry road conditions, since the neces-

sary relative slip of the tires on a turn would significantly increase wear and decrease

road holding. All-wheel drive, on the other hand, utilizes a central differential, allowing

differentiated rotation at each wheel and axle (Image 3.21). So, when combined with torque

vectoring this allows for the delivery of a precise torque at each wheel. During hard accel-

eration or cornering, an AWD system that normally distributes 90% of torque to the front

and 10% to the rear, might even the torque to a 50/50 split; and the bulk of that rear torque

would be sent to the outside wheel, ensuring a more controlled turn. The challenge in

AWD is typically efficiency, as driving both axles entails much more mechanical friction.

However, the addition of torque vectoring can address this by permitting disengagement

of an axle entirely, effectively allowing the drivetrain efficiency to approach that of a two

wheel drive vehicle when the all-wheel drive is not needed.

IMAGE 3.20

Vector drive.

The ZF Vector Drive TVD used by BMW incorporates a planetary gearset that allows wheel torque to function

more independently of engine torque. So, corrective torque can be precisely applied to each wheel even if the

car is decelerating, for example, while descending a winding mountain road.

Image: ZF Friedrichshafen AG

96 Automotive Innovation

Advanced Tires and Control

Lastly, if we’re going to talk about getting ‘power to the pavement’ we need to address

where the ‘rubber meet the road’, literally. The last step in forward traction is the connec-

tion between the tires and the pavement. We have incorporated smart technology into

every step of the powertrain, from generation at the engine to the coupling, transmission,

and differential. It would not do to then have the last and arguably most important step in

this chain left dumb. While the transformation hasn’t happened just yet, tires are on their

way to becoming a critical and integral component of an active and intelligent drivetrain.

Significant enhancements in tire design have already occurred to be sure. Contemporary

tires are more durable, precise, and efficient than ever. Run flat designs not only enable

continued operation with a puncture, but also increasingly provide improved perfor-

mance, ride and affordability, reflected in their growing popularity. However, we have yet

to fully exploit the potential for digital monitoring and control technology in this vital last

step in the powertrain. This is, after all, the only component of the automobile that contacts

the road. But tire engineering is at a turning point; the integration of advanced technology

to define the next generation of intelligent tires is around the

corner.

The possibilities for tire monitoring go well beyond the tire pressure monitoring sys-

tems (TPMS) now standard in all US production vehicles. But it’s worth noting that it was

IMAGE 3.21

Audi all-wheel drive.

The Audi quattro drivetrain offers a base power distribution of 40:60, meaning 40% of the power is delivered

to the front axle and 60% to the rear. However, this distribution is adjusted when needed, with up to 70% at

the front or 85% at the rear to counteract slip. The sports differential further enables torque to be distributed in

continuously variable proportions to each rear wheel based on vehicle dynamics to help ensure optimal road

holding.

Image: Audi

97Getting Power to the Pavement

the development of this system, with the associated capacity to transfer data from inside a

pneumatic tire, which catalyzed the development of more advanced monitoring systems.

Intelligent tire systems are developing the capacity to monitor not only air pressure but

strain, contact patch size, temperature, acceleration, slip, tread depth, and load. This data

can be provided to a tire control system that integrates with the drivetrain.

Measuring tire performance can be difficult, but there are multiple options being

explored. Using tire deformation, vehicle speed, wheel speed, and force on the tire or axle,

algorithms can estimate tire friction and other key factors.20 Data from the yaw sensor,

acceleration sensor, and steering angle sensor can enhance the estimation of wheel grip.

Even the vibration of the wheels as they rotate at different speeds can be used to improve

vehicle control.21 A more direct measurement capacity is being developed that could uti-

lize a strain gauge embedded in the tire. The challenge is to design a gauge material that

is as flexible as the tire itself, so the elongation and flex of the tire is not compromised by

the gauge. Even the possible use of surface acoustic wave (SAW) sensors is being consid-

ered to monitor tire deformation with road contact. Defined by two metallic interlocking

comb shaped electrodes, or interdigital transducers (IDT), on the surface of a piezoelec-

tric material, an ultrasonic sensor mounted inside the tire can measure sidewall deforma-

tion. Alternatively, an ultra-flexible sensor made though photolithography might provide

a thin, flexible sensor that would not interfere with tire function or durability.22 And there

are a variety of other possible microelectomechanical sensors, or MEMS, that are being

considered.23 While not yet ready for mass production, the implementation of smart tie

technology seems imminent.

Powering such systems is also a bit tricky. Of course, intelligent tire sensors could be

may be limited. A passive wireless system using electromagnetic coupling could be used,

and we’ll look at such systems the next chapter; but this would be likely to be inefficient.

A more interesting option is a passive batteryless system that could harvest energy from

tire motion and deformation to energize its sensors. Appropriately placed piezoelectric

materials in the inner-tire could be used to convert the mechanical movement of rota-

tion or deformation into a useful electric charge.24 This may sound crazy, but it’s not. The

energy of low-frequency vibration could potentially be harvested by using piezoelectric

zinc oxide nanowires radially entwined with the tire’s fibers. Or, a capacitive generator

could be used by charging capacitor plates and then moving the plates apart to generate

electrical energy.25 It might also be possible to use the heat generated in tires as a power

source. Work is being done on all these ideas and more.

The ways intelligent tires could be integrated with existing control systems opens a

world of possibilities. Smart tires could inform the traction control systems to provide

20 R. Matsuzaki and A. Todoroki, Wireless monitoring of automobile tires for intelligent tires. Sensors 8, 2008,

8123–8138.

21 T. Umeno, Estimation of tire-road friction by tire rotational vibration model. R&D Review of Toyota CRDL, 37,

2002, 53–58.

22 R. Matsuzaki, T. Keating, A. Todoroki and N. Hiraoka, Rubber-based strain sensor fabricated using pho-

tolithography for intelligent tires. Sensors and Actuators A: Physical, 148, 2008, 1–9; and R. Matsuzaki and

A.Todoroki, Wireless monitoring of automobile tires for intelligent tires. Sensors 8, 2008, 8123–8138.

23 R. Matsuzaki and A. Todoroki, Wireless monitoring of automobile tires for intelligent tires. Sensors, 8, 2008,

8123–8138.

24 A.E. Kubba and K. Jiang, A comprehensive study on technologies of tyre monitoring systems and possible

energy solutions. Sensors 14, 2014, 10306–10345.

25 S. Meninger, J.O. Mur-Miranda, R. Amirtharajah, A.P. Chandrakasan and J.H. Lang, Vibration-to- electric

energy conversion. IEEE Transactions on Very Large Scale Integration (VLSI) Systems, 9, 2001, 64–76.

98 Automotive Innovation

optimal delivery of tractive force to the pavement. They might also communicate with

active chassis, suspension, transmission controls, steering, and a host of other systems to

adjust to changing road conditions. We’ll look at this in Chapter 7. However, more than just

monitoring tire traction, smart tires could actually alter traction to suit driving conditions.

Active inflation management could adjust tire inflation to suit the surface. Tread com-

pounds could adjust to road conditions, defining a more stiff tread when dry to improve

efficiency and handling, and absorbing moisture to become softer when wet, to allow

improved handling. As the tires sense the state of the road, they could potentially adjust

their physical properties to differing road materials, temperature, wetness, and friction,

and report these changes to the rest of the automobile.26 Who says tires aren’t exciting and

drivetrains aren’t innovative?

26 B. Schoettle and M. Sivak, “The Importance of Active and Intelligent Tires for Autonomous Vehicles.”

TheUniversity of Michigan Sustainable Worldwide Transportation Report No. SWT-2017-2 January, 2017.

Electric Machines

Unless you’ve been asleep for the past 10 years, you know that internal combustion engines

are no longer the only way to power an automobile. Increasingly, automakers are turning

to partial or full electric drive to achieve improved fuel efficiency, reduced emissions, and

even upgraded performance. In fact, the potential for the electrification of the drivetrain

is remarkable, leading virtually every major carmaker to declare aggressive targets for the

electrification of their fleet. But, before we get into the cars themselves, we should take a

look at the particulars of the electric machines that power them. Like the internal com-

bustion engine, we can’t appreciate the possibilities for the whole vehicle system until we

understand the particulars behind the propulsion mechanism.

You might think that electrifying a drivetrain is an easy task. After all, electric motors

are a mature technology, with millions of applications. Slapping one on a car can’t be too

difficult, right? However, the demands placed on the traction motors of electric vehicles

(EV) or hybrid electric vehicle (HEV) are very different from the typical demands placed

on most motors. In most cases, whether we are talking about the motor in your washing

machine or a large industrial motor, the unit is fixed, mounted to a stationary structure

or floor, so the weight and size are not particularly important. In addition, typically these

motors operate at a defined speed for a predictable period of time in a relatively predict-

able and controlled environment. However, none of this is true for cars. The speed and

torque demands placed on an automotive motor vary significantly and quickly with no set

pattern. The weight and size and even shape of an EV or HEV motor matter greatly. And

the environmental

conditions vary dramatically, as a car must be able to operate reliably

in any weather condition, from subfreezing to desert hot. And, on top of all this, a motor

must be able to function easily as a generator to allow us to harvest the kinetic energy of

the car during deceleration, called regenerative braking.

So, while there are many differing sorts of electric motors out there, with differing archi-

tecture, differing control requirements, and differing configurations, a relatively few of

them are suitable for use as automotive traction motors. In general, we need motors with a

capacity to produce reliable torque for acceleration and hill climbing as well as an ability

to manage periodic overload for overtaking when needed, ideally in as small a package as

possible, what we can call torque density. And we need a broad speed range that allows

low-speed creeping and high-speed cruising, an ability to produce constant power over a

wide section of this range, and low maintenance and high reliability in differing environ-

mental conditions. Ideally all of this will come in an affordable package. Needless to say,

not all motors fit the bill.

As was true for the internal combustion engine, control is everything. The capacity for

advanced precise digital control has allowed for remarkable innovation in electric machine

design. In fact, it is no exaggeration to say that innovations in the field of power electronics

(the combined electronic, electromagnetic, and electrochemical components that control

and convert power) have fundamentally reshaped what is viable in electric motors and

thus what is achievable in EVs and HEVs. As a result, ideas and concepts that have been

100 Automotive Innovation

known for many decades as conceptually possible but not feasible, have now become tech-

nically practicable options.

The Principles of the Electric Motor

As always, let’s start at the beginning. Motors function by exploiting electromagnetic force,

one of the four fundamental forces of nature. As might be obvious from the name, elec-

tromagnetism defines a relationship between magnetism and electricity. In fact, magnetic

and electric forces are really fundamentally the same sort of thing; both are defined by the

exchange of photons between charged particles, called an exchange force. In our case, the

particles are electrons. You can think of photons as elementary particles or waves that act

as force carriers, traveling between electrons and exerting force, and defining both electric

and magnetic fields.

Of course, to understand magnetic fields we need to understand magnets. Three elements,

iron, nickel, and cobalt, demonstrate the property of ferromagnetism, or the ability to be

permanently magnetized when placed in a magnetic field. This magnetization happens at

the atomic level. The atoms that compose these materials are themselves like tiny magnets

with opposing poles, or magnetic moments, that produce a magnetic field, interact with

other magnetic moments and change their orientation in response to magnetic fields. When

these atoms are similarly oriented throughout the material, their fields combine together and

define a uniform magnetic domain. So, by exposing a ferromagnetic material to a powerful

magnetic field, the orientation of these crystals can be aligned so that all the magnetic axes

point in the same direction, thus creating a magnet. Importantly, with these three materials,

when the magnetic field is removed the polarization remains, defining a permanent magnet

(PM). When certain rare earth elements, in particular neodymium and samarium–cobalt, are

combined with these elements, they can form magnets that are several orders of magnitude

more powerful than the simple ferrite magnets on your refrigerator. In addition, as we will

see later, some metals, such as copper and aluminum, do not become PMs themselves, but

can exhibit magnetic qualities when an electric current is passed through them. This allows

us to define controllable magnets that can be switched on and off, called electromagnets.

As you probably discovered sometime in grammar school, similar magnetic poles repel

each other and opposing magnetic poles attract each other. The force between these poles

can be described as a field shaped by contour lines of equal magnetic force. This magnetic

field can’t really be seen of course; but drawing out the flux lines of a field offers a useful

way of imagining the effect, and will help us understand how motors work (Image 4.1).

Because magnetic flux tends toward the path of least resistance, placing a needle in this

field would cause it to align with the lines of flux. This is because the needle has much

lower resistance to flux, called reluctance, than the surrounding air. The concentration of

flux in a material with low reluctance, like our needle, forms strong temporary poles that

pull the material toward areas of higher flux. So our needle turns, and when aligned with

the magnetic lines of flux, offers a path of lowest reluctance. If this needle were magneti-

cally polarized, this rotational force would be directional, as the north side of the needle

would seek the south pole of the magnet, and vice versa since opposing magnetic forces

attract each other.

All of this may seem a bit dense and academic. But these two dynamics—the attraction

of opposing poles and the force of reluctance in a magnetic circuit—aren’t just interesting,

101Electric Machines

they’re amazing. They allow us to use magnetic fields to create mechanical motion. As we

will see, these two forces are present in any motor or generator and the essential motive

core of every electric or hybrid car on the planet.

We can already see that a motor’s operation is fundamentally defined in the interaction

of movement and magnetism. Now what we need to better shape this interaction is an

ability to define and control the magnetic field, and that requires that we add electricity to

the mix, bringing us to a phenomenon called electromagnetic induction. Induction defines

the relationship between electromotive force (EMF), or voltage, and magnetic fields. Very

simply, if we move an electrical conductor within a magnetic field we will induce a voltage

or EMF within that conductor. This voltage in turn will cause current to flow if connected

to a circuit. So the relative movement of a conductor, let’s say a wire, through a magnetic

field will result in an electric current in the wire. This relationship between changing mag-

netic fields, called magnetic flux, and current works the other way around too. As current

travels through a conductor, it induces a magnetic field around it (Image 4.2).

These two related observations—that electric current can induce a magnetic field and

that magnetic flux can induce current—are critical to understanding electric motors. If we

place a current-caring conductor near a magnet, the magnetic field induced by the current

will interact with the magnetic field of the magnet and define a physical force. This essen-

tial relationship between electricity and force makes all motors possible. After all, that is

the central purpose of a motor, to turn electrical energy into movement, or take physical

movement and generate electricity as a generator. Since the devices we need for cars have

to be able to do both, we can call them electric machines or motor-generators rather than

just motors.

IMAGE 4.1

Magnetic fields.

Lines of flux are (somewhat arbitrarily) said to flow from the north to the south pole of a magnet. As we get

closer to the polls, the lines of flux get closer, indicating more flux per unit area, or that the intensity of the field

gets greater. A needle place in the field would align with the lines of flux to define a path of least reluctance. Two

polarized magnetic fields can interact, defining magnetic attraction and repulsion. The resulting magnetomo-

tive force (MMF) is proportional to the strength of the field

and the distance between the poles.

102 Automotive Innovation

Making an Electric Machine

With this basic information, we could assemble a primitive machine. If we make a loop of

conductive wire, for example, we can place a current through that loop. The current will

induce a magnetic field around the wire. Placing that wire loop between two poles of a

magnet would cause the magnetic field induced by the current and the magnetic field of

the magnets to interact. If the current flow in the loop is oriented properly we could cause

the wire to jump a bit as the two magnetic fields interact (Image 4.3). Like I said, primitive.

This isn’t much of a motor.

We could improve the motor my shaping our wire into a hoop. If our wire hoop can

rotate and put the two poles of the magnet on opposing sides of our axis, we could cause

that jumping motion to push the hoop around, and our wire loop would revolve. By add-

ing slip connectors to provide the current to the loop, we can arrange to mechanically

reverse the flow of current with each half turn of the loop. We call this sort of connection a

slip-ring commutator. Small carbon contact points, called brushes, are used to slide against

the commutator and allow electric current to travel from the stationary frame to the rotat-

ing commutator. The result will be that each time the loop rotates to resolve the force

caused by the magnetic fields, the current is reversed and the rotational torque reappears.

So, the loop remains in constant rotation as it continuously attempts to resolve the force

while the force is continually reestablished with reversal of the current. So, in sum, we’ve

defined a component that rotates, the rotor, a component that stays put and provides a

magnetic field, sometimes called the stator, an axle, and a slip-ring commutator. The term

armature refers to the power-producing component of the machine, in this case the loop.

These are the basic components of any simple direct current motor (Image 4.4).

The same basic principles allow us to understand the function of a generator. However,

to get a sense of how things can differ, let’s stagger the commutator rings instead of

IMAGE 4.2

Current and magnetic fields.

As current travels through a conductor, it induces a magnetic field around it. The direction of the field can be

predicted with the right-hand rule. If you place your right hand with curled fingers and your thumb in the

direction of the current, your fingers will point the direction of the resulting field.

103Electric Machines

IMAGE 4.3

Motive force.

Running a current through a conductor will induce a magnetic field. If we place that conductor within the mag-

netic field of a permanent magnet, the two fields will interact and a force on the wire will be the result.

IMAGE 4.4

Basic DC motor.

As the armature rotates due to the interacting magnetic fields of the permanent-magnet stator and the current

in the armature, the slip-ring commutator reverses the contact with the brushes, and therefore reverse the cur-

rent, allowing the armature to continue to rotate.

104 Automotive Innovation

splitting them. And, in this case, since we want to generate electricity, we can put a handle

on our loop, like Image 4.5.

As we rotate the conductive loop, it swings past the alternating magnetic poles, cutting

across the magnetic field. As a result, an EMF is induced causing a current flow. This

current is quickly reversed as the polarity of the field shifts in the second half of the rota-

tion. The shifting potential back and forth in the loop defines an alternating current (AC),

reaching maximum values as the loop passes each pole with movement perpendicular to

the lines of flux.

With this alternating output, we can now imagine a motor that doesn’t require the same

sort of split commutator we had previously. If the current is consistently shifting from

negative to positive, there’s no need for the commutator to reverse the polarity of the loop,

that’s done automatically by the nature of AC current. In fact, if we took our generator

and placed it on the receiving end of our output, it would now behave as an AC motor

(Image4.6). When it is cranked, it’s a generator providing current; when current is applied,

it is a motor providing rotational force. The resulting motor could be called a synchronous

motor, because its rotation will be defined by the frequency of the AC current supplied to

it. Turn the generator faster and the motor will turn faster.

Don’t run out and make an electric motor for your car from a coat hanger and kitchen

magnet just yet. You might be disappointed in its performance. To have a useful motor, we

need to first produce a strong magnetic field, and second allow it to interact with as much

of a conductor as possible. More windings of the conductor to maximize the conductor’s

cross-sectional exposure to the magnetic field will help. To accomplish this we shape the

conductor into a coil with multiple loops. With each additional loop, we increase the cross-

sectional exposure to the field. The strength of the magnetic field is not shared among the

IMAGE 4.5

Basic AC generator.

When the handcrank is turned, the resulting current through the slip rings alternates as the magnetic field

reverses the induced current, defining an AC output.

105Electric Machines

conductive loops; the force is multiplied. The more the conductor is exposed to the mag-

netic field, the stronger the effect.

We might also add additional poles to our motor, so the armature doesn’t abruptly flip

from one orientation to the opposite (Image 4.7). More poles can allow a smoother rotation

and reduce the rapid fluctuation in torque that follows changing rotor angle, called torque

ripple (Image 4.8).

So far, we’ve relied on magnets to produce our initial field. We called these PMs, not

because they are strictly permanent. As we discussed, they can be made and unmade.

In fact, a serious constraint in the use of these magnets is that too much heat can cause

IMAGE 4.6

Motor-generators.

Turning one motor-generator produces an alternating current that will induce an alternating magnetic field in

the second motor-generator and cause it to turn synchronously.

IMAGE 4.7

Four-pole motor.

The components of this basic motor include a rotor with field windings, a stator with mounted permanent

magnets.

106 Automotive Innovation

a magnet to lose its magnetism by allowing the molecular polarity to reorganize. Rather

they are called permanent because they are magnets all the time, defining a continuous

magnetic field. This can sometimes be troublesome, as by their nature they are always

on, defining a possible safety issue since this means we can never fully shut down our

machine. And, as we will see in a bit, the permanent and fixed nature of PMs can have

undesirable effects at high speed. On top of all this, powerful magnets can be expensive.

IMAGE 4.8

Torque ripple.

Motor torque varies sinusoidally. Reversing the current in a single-phase machine results in torque bumps

called torque ripple. A three-phase motor can help reduce torque ripple by filling in the torque dips with the

additional phases. However, torque ripple is never fully removed.

107Electric Machines

However, the use of PM is not our only option. As discussed earlier, current flowing

through a conductor can define a magnetic field. So, a PM could be replaced by a properly

defined conductor. Once again, a single wire is not going to produce an adequate field for

a useful magnet. We can instead wrap our conductor with a coil that will multiply the

strength of the magnetic field in our core, defining an electromagnet (Image 4.9). Since

a coil’s magnetomotive force is the product of the number of turns and the current, with

each added coil we add to the magnitude of our field.

Placing an iron core in our coil can allow us to shape the magnetic force more effectively.

So far, our magnetic fields have traveled entirely

through air. But air is not a great conduc-

tor of MMF. In fact, it has a very high reluctance. Providing a low-reluctance conductor

for magnetic flux allows us to potentially shape or concentrate the magnetic field. In fact,

the ability of iron to support a magnetic field, called permeability, is roughly 1,000 times

better than air. Consequently, far less of it will be lost as leakage to the surrounding air.

The force produced is proportional to the flux density we can produce. Double the density

and you double the force. Of course, there is a limit. Any given magnetic conductor has a

maximum amount of magnetic flux it can accommodate until saturation. At this point, all

the atomic dipoles in the material are lined up and no further increase in the field is pos-

sible; after that increasing the magnetic field results only in lost energy. When the core is

saturated, the reluctance of the material goes up very quickly. This is typically not a major

factor, as a motor’s core is typically sized to suit its purpose and operate below saturation;

but it is the primary reason a larger power demand generally requires a physically larger

motor.

Since the rotor has to be able to turn, we need to maintain some space between it and

the stator. We call that an air gap. Ideally, this gap will be as small as possible to minimize

loss. Flux is generated in this air gap, so the larger it is the less flux density that can be

achieved; and remember, flux density is what defines the rotational force we can generate.

So, defining this air gap becomes a key step in any motor design.

All of this now allows us to imagine a more sophisticated motor. We can define a coil

armature, to maximize the magnetic flux exposure. We can utilize more poles, to offer

better control and smooth operation of our motor. We might choose to use electromagnets

with iron cores in the stator, to generate a strong and more controlled magnetic field.

IMAGE 4.9

Electromagnet.

108 Automotive Innovation

These basic motors, whether using PM or electromagnets, are an affordable, well-

established technology. For these reasons permanent-magnet DC (PMDC) motors were

once widely used by hobbyists to electrify older cars, so-called de-ICEing (get it?). Motors

like this are still frequently used in cars, to move your window up and down for example.

However it’s not much use as an EV traction motor. The principle problem is that every-

thing we’ve discussed so far requires a commutator. Commutators can cause torque ripples,

they can limit the motor speed, the carbon brushes generate friction that impacts efficiency

and also generates radio frequency interference. Most importantly, the brushes are subject

to wear that seriously impacts reliability and maintenance requirements. No one wants to

swap the brushes on their car’s motor every few thousand miles. So, once we’ve discussed

the basic performance characteristics of electric machines, we will see that while current

technology is based on these same fundamental principles, the actual machines used in

contemporary EVs have moved well beyond these simple motors (Image 4.10).

Motor Performance

As you might expect, the performance of an electric machine depends greatly on the

design, and designs vary enormously. However, these machines do share some general

qualities worth noting. First, they are far more efficient than internal combustion engines.

Most gasoline engines are about 20%–35% efficient in converting the chemical energy

in gasoline to useful torque and speed. A high very performing engine might achieve

IMAGE 4.10

Brushed DC EV.

One of the very few production cars to use a brushed DC motor is the Reva G-Wiz. This much-criticized micro-

car used a 6.4 hp (4.8 kW) motor powered by eight lead-acid batteries under the front seat.

Image: CC BY-SA 2.0

109Electric Machines

efficiency in the mid-40s. However, traction motors in EV regularly achieve efficiencies

orbiting 90% or higher.1

If we’re going to achieve maximum efficiency, we’re going to have to think carefully

about how to control the speed of our motor. For our purposes, this is more challenging

than it might be for a washing machine or blender for at least two reasons: First, unlike

many motor applications, vehicles require a wide range of precise speed control. And

second, we need to do this with as little loss as possible, since any inefficiency reduces

our vehicle’s range. At first blush, we might think speed control is pretty easily achieved.

Varying the voltage at the stator will vary the field magnitude, which should modify the

force on the rotor and so the speed. Putting a variable resistor in series with the stator

would allow us to vary this voltage easily. In fact, this is the approach that was used suc-

cessfully for a long time by analog power systems. The problem is that such an approach

entails significant inefficiency. The combined rheostat and stator circuit draws power con-

sistently. So even when the motor is turning with minimal load, the energy used would

remain high. The excess energy would simply be absorbed by the rheostat and turned to

heat. So, a basic linear power source tends to entail a lot of loss and a fair amount of heat.

A better approach is offered by a modern digital control method called pulse width

modulation (PWM). The core of PWM is essentially a switch that can turn a signal off and

on very quickly. By accurately controlling the amount of time the signal is on as a percent-

age of the total cycle, called duty cycle, and the frequency of the cycle, we can create an

output that behaves like a precisely defined constant voltage. So, while the signal behaves

like low voltage, it is in fact a very rapid succession of full-voltage pulses. As a result,

power loss is small, though loss still occurs during the actual switching, called switching

loss, but this can be reduced with improved control logic. Essentially, the goal of all this

is to digitally encode a precisely modulated analog signal. To get a particular speed at a

given load, a simple lookup table can be used to define the PWM duty cycle needed to

produce the targeted motor speed (Image 4.11).

1 A. Hughes, Electric Motors and Drives: Fundamentals, Types and Application, 3rd edition. Elsevier, London, 2006.

IMAGE 4.11

Pulse width modulation.

By turning the signal off and on quickly, the effective output voltage can be efficiently controlled. The effective

voltage value is the average of the on–off voltage over time.

110 Automotive Innovation

This does not imply that electric motors can now achieve near perfect efficiency. In fact,

inefficiencies in electric motors are a significant concern in the design process. There is, of

course, inefficiency due to friction of the bearing surfaces, and to lesser degree air resis-

tance on the rotor. However, a more pressing design concern is the loss of energy due to

electrical resistance in a machine’s coils. We call this copper loss, and it is principally

lost through the generation of heat. The loss of energy due to resistance in any conduc-

tor is defined by the resistance and the square of the current. This means that even a

small change in current can result in a large change in loss, and high-torque operation

that requires high current is likely to lead to excessive heat. So, efficiency can change sig-

nificantly as operating conditions change. The resulting generation of heat is problematic

for the obvious reason that inefficiency is bad and results in reduced range or diminished

power. But in addition the generation of heat is problematic because PM can suffer demag-

netization when heated, resulting in reduced motor capacity. Moreover, because resistance

of a conductor is defined not only by the conductor’s dimensions and composition but also

by the temperature, increasing temperature can in turn increase copper loss and result in

more heating. Giving us good reason to size electric machines carefully and be concerned

advanced low-temperature combustion possibilities and ingenious

new engine designs. Chapter 2 then builds on this foundation with an examination of the

digital control technology that has redefined the internal combustion engine, from vari-

able valve timing and lift to variable intakes, as well as promising developments in more

advanced active control mechanisms that enable precise on the fly changes in just about

every aspect of the engine, defying the tradeoffs and limits engineers once faced when

designing automotive engines. Chapter 3 then connects these technologies to the road by

examining the rest of the powertrain, beginning with the basic principles of gearing and

moving through to advanced innovations in transmission design including continuously

variable transmissions, automatic manual transmission, dual-clutch systems, torque vec-

toring, and even advances in future tire design, where the rubber literally meets the road.

This then paves the way for an exploration of electrification of the drivetrain. Chapter4

begins with a general introduction to electric motors and their performance advantages and

challenges, with a particular focus on brushless DC and induction AC motors and control

xiPreface

technology. It ends with an introduction to some of the more promising advances in motor

design that may represent the electric machines of future automobiles. Chapter5 examines

the electrified powertrain, beginning with hybrid vehicle engineering, discussing varying

hybrid drive architectures as well as the nature of regenerative braking and recent energy

recovery innovations. It also explores electric vehicle technology, and the challenges and

possibilities for future EVs. Energy storage technology is examined in Chapter6, from basic

battery science to promising developments in advanced battery chemistry and design.

The subsequent two chapters explore advanced vehicle design beyond the powertrain,

beginning with a basic discussion of vehicle structure and handling and moving onto

advanced suspension, active chassis control, new materials, and crashworthiness. Vehicle

aerodynamics is examined in Chapter 8, again beginning with basic concepts of airflow

and bluff bodies and moving onto an examination of recent applications such as air

curtains, active shutters, ground effect management, and other advanced aerodynamic

innovations.

The last chapter examines advanced driver assistance systems. This includes a discus-

sion of sensing technology now in use, such as LIDAR, SONAR, and RADAR, as well as

the benefits and challenges of applying these technologies to advanced vehicle control and

driver assistance features, such as lane keeping, active cruise control, and crash avoid-

ance. The chapter moves on to discuss the potential for a more extensive incorporation of

advanced vehicles into roadway control technology, exploring V2X possibilities, the chal-

lenges of advanced driver assistance and autonomous vehicles, as well a basic introduction

to the artificial intelligence needed for such systems.

In the end, the hope is that this book will help get the reader up to speed, oriented to

the basic science and technology that defines the modern automobile, and its likely future.

Like any book that dares to offer a sweeping survey of a field, it’s likely that a few points

have been missed. Reader’s comments are very welcome and can help improve possible

future versions of this text. It is also certain that this book benefited greatly from the

advice of my colleagues. In particular, I’d like to thank Justine Ciraolo, Jason Shulman,

Marc Richard, and most especially Kristina Lawyer for their very helpful and thought-

ful suggestions. Of course, any errors are entirely my own. I would also like to thank the

array of component suppliers and carmakers that agreed to provide images and insights

for this text and supporting materials. They are identified thorough the text, and I am

mostgrateful.

http://www.taylorandfrancis.com

xiii

Author

Professor Patrick Hossay heads the Energy Studies and Sustainability programs at

Stockton University where he teaches courses in automotive technology, green vehi-

cle innovations, and energy science. He is also an experienced aircraft and automotive

mechanic, and enjoys restoring classic cars and motorcycles.

http://www.taylorandfrancis.com

Bringing the Fire

It makes sense to start our exploration of automotive innovation with what has been

the heart of the automobile for more than a century: the internal combustion engine.

And it makes sense to start a look at the internal combustion engine with the heart of

the process: combustion. The very idea of combustion is usually taken for granted, as

a notion that is self-explanatory. We all know what combustion is, it’s when something

explodes or burns, right? But, thankfully, that’s not exactly what’s taking place in your

engine. We’re going to need a more precise understanding of exactly what and how

something is burned in an engine before we can understand the complete workings of

internal combustion.

So, what exactly is combustion? Put in general terms, it’s a chemical reaction that con-

verts organic material to carbon dioxide while releasing heat energy. The process is called

rapid oxidation because it’s defined by a fast reaction with oxygen. So, at the most basic

level, a chemist would write down a combustion reaction like this:

+ + → +C O heat CO heat2 2

Very simply, this means organic materials made of carbon (C), like wood or paper or in our

case gasoline, react with oxygen (O2) with added heat to create carbon dioxide gas (CO2)

and more heat.

This already allows for some very useful observation: first, combustion needs oxygen

in a most fundamental way. You can think of a car’s engine as a large air pump, drawing

in massive amounts of air, delivering it through a combustion process that consumes the

oxygen, and pushing the deplete product out the other end. In fact, to burn a single gallon

of gasoline completely, an engine needs to draw in nearly 9,000 gallons of air. This need to

provide a ready supply of oxygen for combustion is a critical criterion and definitive chal-

lenge of any engine’s performance.

The second observation is that combustion is a process, not an instantaneous event.

The heat resulting from the reaction, on the right side of the equation above, is the same

heat that feeds the continuation of the reaction on the left. So, what we get in the engine

is absolutely not an explosion, but a rapid combustive expansion that defines a wave of

pressure, more like a push than a bang. This wave is often called a flame front, fed by

the combustion of gasoline, and propagating a swell of pressure that can be converted

by the engine into rotational energy. If we can accelerate that flame front, while keep-

ing it controlled and even, we can achieve greater power from combustion. But, if we

lose control of the combustion process, and perhaps even get more than one flame front

from multiple points of ignition, the result is closer to an explosion, reducing perfor-

mance while potentially harming the engine. Before we see how this all plays out in a

typical engine, we’ll need to understand a bit more about the chemical that fuels this

process—gasoline.

2 Automotive Innovation

What Is Gasoline?

The equation above is a bit too simple. After all, we’re not burning single-carbon molecules.

We’re burning gasoline. Which raises the question, what exactly is gasoline? Gasoline is a

‘hydrocarbon’, which means it is a long chain of carbon connected to hydrogen. Essentially,

the longer the chain, the more potential energy it contains. Crude oil comes in chains that

could be dozens of carbon atoms long. But long molecules like this, while they have plenty

of potential energy in them, are thick, hard to pump, and hard to burn. Think of tar. So, we

need to refine these long, heavy hydrocarbons into shorter chains

about motor temperature control.

In addition to the loss due to resistance in the coil, there is also a loss due to the mag-

netic field. When a fluctuating magnetic field travels through a ferromagnetic material,

loss occurs. As previously discussed, the polarization that defines opposing north and

south charges also occurs at the microscopic level with individual molecules defining

dipoles. These dipole moments respond to the alternating magnetic fields with a torque

force that repeatedly shifts the alignment of the atomic dipoles. This entails internal fric-

tion and requires energy; and it results in the generation of heat, physicists call this hys-

teresis loss.

An additional loss comes from the generation of uncontrolled currents within the mate-

rial. As the machine’s core is exposed to a fluctuating magnetic field, a current is induced

that opposes the external field that created it. As the resulting current flows in a plane that

is perpendicular to the magnetic field, it defines small loops of current. These micro whirl-

pools of lost energy are called eddy currents, and once again result in heat generation.

Combined with hysteresis loss, this effect defines iron loss or core loss.

The balance of these losses varies with condition. The more the rotor turns in synchroni-

zation with the field, the lower these losses, since there is no change in flux at any location

in the rotor and so no induced currents. But, of course, no system is perfect, and neither the

field flux nor the rotor is flawless, so eddy currents are generated even in a synchronous

machine. In general, at higher torque and lower speed, copper losses tend to be the most

significant. And, at high speed and low torque, when the flux in the core is potentially at

its greatest, iron losses are dominant.

Torque and Power

With all these complications, you might be starting to wonder why we want to use elec-

tric machines to power cars at all. However, despite these generally shared challenges,

electric machines also share some very desirable characteristics. Most notably, they pro-

duce maximum torque upon initial start-up, and maintain that capacity until the speed

111Electric Machines

of maximum power generation, or base speed, is attained. As we will see later, this can

make an electric drive an ideal companion to an internal combustion engine that has

limited torque at lower speeds. Once a machine’s base (or rated) speed is achieved, torque

will decrease while maximum power is achieved and maintained (Image 4.12). As we

will see, variations in the type and design of the machine will change these parameters

greatly, but the basic pattern of torque and power generation will hold true for nearly any

machine.

Understanding why this occurs is important. As previously mentioned, the most basic

mode of adjusting speed is by varying the EMF on the rotor. This alters the resulting cur-

rent and thus the strength of the magnetic field. However, as the rotor spins, and the flux

through the rotor changes, it also creates an induced opposing voltage, called a back EMF.

In essence, the machine is also operating as a generator when it is motoring. When this

back EMF nearly equals the supplied voltage this defines the motor’s rated speed. When

a load is applied, the motor slows, reducing back EMF and so increasing the supplied

current, in effect automatically compensating for the change in load. If the load gets high

enough to stop rotation, torque will be at its maximum.

This capacity to provide maximum torque with increasing load is ideal for a traction

device, but it makes speed control tricky. Once at rated speed, getting the motor to spin

faster cannot be accomplished with more voltage, as this would only serve to increase

the back EMF. However, if instead we decrease the magnitude of the stator field by

reducing current, called field weakening, the back EMF will also decrease, and more

current will be able to flow through the rotor resulting in increased speed. The cost,

however, will be reduced torque. The motor now enters the constant power range of

operation. As the torque decreases it is offset with increasing speed to provide constant

power. If you keep decreasing the field, the speed will continue to increase and the

torque will decrease until mechanical failure or excess heat cause the machine to stop.

This once again underscores the importance of precise control technology and thermal

management.

IMAGE 4.12

Ideal power/torque performance for EV electric machine.

112 Automotive Innovation

Cooling

Even at this early point, it is clear that heating is a key limiting factor on the performance of

electric machines (Image 4.13). In fact, the challenge of cooling the machines and the power

electronics that provide control without unnecessary weight, bulk or complexity is a major

concern for manufacturers. Of course, avoiding the generation of heat with efficient and well-

designed machines and power electronics is a good option, and we will see many examples

of this. But even the most efficient machine will generate some heat; and as we will see, the

magnets used in modern machines can be very sensitive to heat, resulting in significant

performance deterioration with inadequate thermal management. As an example, a tem-

perature increase from 20°C to 160°C could result in a loss of nearly half the motor’s torque.2

However, thermal management can be tricky. Any electric machine has multiple com-

ponents made of multiple materials with differing thermal characteristics, all defining

multiple heat transfer pathways. In a hybrid system, where the electric machine is in close

proximity to the internal combustion engine, the task is even thornier. And the cooling

requirements of the power electronics can be equally difficult to address. So, defining a

thermal management system once again presents us with a trade-off between costs, per-

formance, size, and efficiency.

Cooling can generally occur through active or passive methods. Passive cooling sim-

ply uses careful material selection and design to allow ready paths for conductive heat

transfer away from the machine’s core components. In fact, the need to provide thermal

conductivity between heat generating components and the housing and external sinks can

be a key design goal. Active cooling, on the other hand, can be as simple as a cooling jacket

through the stator housing, much like an engine’s cooling systems. More advanced cool-

ing systems place the cooling jackets at the coil rather than the housing. Coolant channels

can run between each winding bundle and along the stator slots. Forced air cooling can

augment the effect. An alternative approach might tap into the cabin AC condenser to cool

the fluid or provide cooled air.

2 B. Bilgin and A. Sathyan, Fundamentals of electric machines. In A. Emadi (ed) Advanced Electric Drive Vehicles.

CRC Press, New York, 2017, 5–27.

Thermally - Limited

Zone of Operation

Continuous

Operation

Transient Operating

Limit

Operating Speed

IMAGE 4.13

Thermal management and motor performance.

A chief constraint for electric machines is the generation of heat. A motor has a defined torque capacity at any

given speed, as discussed previously. That limit defines a region of safe continuous operation. That torque

capacity may be exceeded momentarily, but the excess heat generation limits this to transient burst rather than

continuous operation.

113Electric Machines

Jet impingement cooling is used to spray a coolant either directly on the surface of the

coil or an attached cooling plate, allowing the coolant to evaporate and extract heat. When

materials change phase as a result of cooling or warming they absorb heat energy without

changing temperature. So when a substance changes from liquid to vapor for example,

it absorbs a high amount of heat energy without getting hotter, this is called latent heat.

Impingement takes advantage of this to extract maximum heat

energy. As a result, plates

less than a few square inches are able to absorb impressive amounts of heat energy. And

the addition of fins or microgrooves can increase heat flux several orders of magnitude.3

However, the effective transfer of heat through the machine and onto the cooling plates

can be a challenge.4

Oil spray cooling using a fluid like automatic transmission fluid (ATF) is another option.

The use of oil offers a clear improvement over air cooling.5 The motor can be partially filled

with oil, and splash cooling can be used as the rotor tosses the oil around. Alternatively, oil

can be fed onto the rotor ends, and made to splash across the windings. While not yet used

by a major manufacturer, an emersion method can be imagined that takes its lead from the

cooling of transformers by immersing components in an electrically non-conductive fluid

such as transformer oil. The challenge, of course, is managing the oil’s path and protecting

the air gap.

Whatever the method, the trick is to design something small, which adds neither excess

weight nor size to the motor but provides adequate cooling. There are a lot of innova-

tive ideas out there. A direct winding heat exchanger, developed by Georgia-based DHX,

for example, places a micro-heat exchanger between the stator windings, using tiny cool-

ing channels that significantly increase the cooling area and exploiting an effect they call

localized turbulence to increase the flow rate. The result is a four-fold increase in current

capacity.6 The developers of a cold plate system that uses multiple miniature nozzles and a

plate covered in tiny pin-like copper fins developed for Toyota at Purdue University claim

it can extract a very-impressive 1,000 W/cm2 from power electronics.7 These and other cool-

ing technologies can allow improved reliability and performance from a smaller machine.

Induction Motor

The problem with the motor we’ve discussed up to now is in the commutator. As men-

tioned, it is responsible for inefficiency, wear, limited speed, and other problems. So, why

not get rid of it? In fact, we might rethink the need for a conductive connection with the

3 D. Wadsworth and I. Mudawar, Enhancement of single-phase heat transfer and critical heat flux from an ultra-

high-flux simulated microelectronic heat source to a rectangular impinging jet of dielectric liquid. Journal of

Heat Transfer 114 (3), 1992, 764–768; and H. Sun, C. Ma and Y. Chen, Prandtl number dependence of impinge-

ment heat transfer with circular free-surface liquid jets. International Journal of Heat Mass Transfer 41 (10), 1998,

1360–1363.

4 T. Davin, J. Pelle, S. Harmand and R. Yu, Experimental study of oil cooling systems for electric motors. Applied

Thermal Engineering 75 (1), 2015, 1–13.

5 T. Davin, J. Pelle, S. Harmand and R. Yu, Experimental study of oil cooling systems for electric motors. Applied

Thermal Engineering 75 (1), 2015, 1–13.

6 S. Andrew Semidey and J. Rhett Mayor, Experimentation of an electric machine technology demonstrator

incorporating direct winding heat exchangers. IEEE Transactions on Industrial Electronics 61(10), 2014, 5771–5778.

7 E. Venere, Research team develops new cooling technology for hybrid and electric vehicles. PHYS.ORG, 13

September, 2016. Available at: https://phys.org/news/2016-09-team-cooling-technology-hybrid-electric.html

http://PHYS.ORG

https://phys.org

114 Automotive Innovation

rotor entirely, and instead rely on induction to create an electromagnetic connection with

the rotor. Since induction takes place without physical contact, this eliminates the need for

commutator and brushes.

The basic principle is straightforward. Remember that a current induces a magnetic field,

and as a magnetic field cuts across a conductor, it in turn induces a current. We can then

imagine a chain of current and induction, with the stator current defining a magnetic field

that cuts across the rotor and thus inducing EMF and current in the rotor. This current

flows in the opposite direction of the current in the stator, and induces its own magnetic

field, a process called mutual induction. These two fields will have opposing polarity. So,

we have two magnetic fields that are opposing each other, one on the rotor and one at the

stator; the result is a rotational force on the rotor. Because it is the change in magnetic field,

or the flux, that induces current, the stator would have to be provided with AC current, as

DC would define a steady magnetic field rather than needed changing flux.

The induction motor is a well-established and mature technology, and has earned a rep-

utation for reliability. Unlike the PM machine that we have discussed so far, this approach

has the advantage of not requiring brushes to transfer current to the rotor. With no com-

mutator and no brushes, this can be a simpler, reliable, and efficient machine for EV.

For EV applications, we need to define a rotor that is mechanically durable, a good con-

ductor, and can offer a maximum exposure to the magnetic field of the stator. While wound

armatures can and are used in induction machines of this sort, because of the high speed

and reliability expected of automotive traction machines, a more robust design is needed.

Coils can have trouble enduring the centrifugal force of a high-speed automotive applica-

tion. Instead, we start with a set of conductive rods that define a maximum cross-sectional

exposure to the stator field. Often aluminum is used in such motors; however, once again

the demands of an automotive traction motor are high, so more conductive copper is used

despite the added expense in automotive applications. The ends of the bars are connected

together by rings to define a closed circuit, allowing the current to operate throughout the

copper cage (Image 4.14).

This squirrel cage armature is embedded in iron to provide minimum reluctance, but

the iron core is assembled of stacked laminations separated by insulation to reduce eddy

currents. Ideally the air gap between the rotor and stator is minimal, usually a couple of

hundredths of an inch. Variations in the impedance, resistance, and size and shape of the

rotor can define the generated torque and speed of the motor.

The stator is defined by stacked laminations wrapped in coils, similar to our previ-

ous machines; but the fundamental architecture of the coils is significantly changed to

allow the use of three-phase AC. This entailed three sequential phases of AC, each set

off by a third of the time it takes the sinusoidal current to complete a cycle, known as

the period. So, current is provided in three sine waves in sequence, offset by 120°. The

sequential fluctuations are used to define motor rotational speed by providing each

phase in sequence to successive coils, and so defining a rotating field of flux across the

rotor (Image 4.15).

To do this, the coils of an induction machine are defined in paired poles, with each pair

in opposite polarity. When the poles are energized, or switched, in sequence with each

phase of current, a rotating sinusoidal flux in the stator windings is defined. With the cage

bars shorted on either side, the relative motion between this rotating field and rotor cage

induces a current in the rotor with the same number of poles as the stator’s field. This cur-

rent in turn induces a field that is opposed to the stator’s field. The two fields meet in the

air gap to define torque.

115Electric Machines

An induction machine’s rotor therefore must always rotate at a slower speed than the

stator field since it’s the relative motion between the two that defines the production of

torque. The difference is known as slip. The greater the slip, the greater the torque. At zero

slip, the rotor and stator fields rotate at the same speed, and therefore the stator field does

not cut across the rotor bars, and no torque is produced. This turns out to offer an advan-

tage as it allows a motor controller to shift to a zero

slip condition quickly to reduce torque,

say when tires slip.

On the other hand, when using the machine as a generator, the rotor needs to lead the

stator. This defines a negative slip condition, and reverses the direction of energy flow.

However, the stator must continue to receive current, since there are no magnets in the

rotor and so no way for it to generate flux if it does not receive current induced from the

stator windings. But with the rotor leading the stator, the rotor’s magnetic field now cuts

across the stator and induces a voltage back into the stator coils, enabling the recovery of

energy during regenerative braking.

As you might guess, controlling an induction motor entails somewhat more complexity

than a simple DC motor, and is typically provided by a voltage-fed pulse width modu-

lated inverter. An inverter is the power electronic circuit that can convert the DC from

the battery into the multiphase sinusoidal signal needed by the motor. Inverters generally

IMAGE 4.14

Basic squirrel cage armature.

It’s clear how the basic squirrel cage armature configuration gets its name. With the stacked laminations

removed, the rotor bars and end rings on either side define a conductive cage. The rotor bars are typically

slightly skewed to reduce torque ripples by distributing them over a wider degree of rotation. Rotor losses can

be high in an induction machine, leading to possible heating of the rotor, which is far more difficult to cool than

the stator. As a result, Tesla Motors, the sole current user of induction motors in EVs, uses copper exclusively.

116 Automotive Innovation

hold either current or voltage constant when defining sinusoidal output. To provide con-

trol, PWM is typically used to adjust the width of the switching pulse based on feed-

back voltage. This provides narrow or wide pulses for low or high voltage, respectively,

and adjusts for changes in power consumption while providing a smooth and harmonic

waveform. Soft switching can be added so switching takes place at points of zero current

or zero voltage. This prevents switching during simultaneous high current and high volt-

age, or high-power points, to minimize switching loss and enable a smaller and quieter

inverter.

The embedded control functions required to manage an induction motor are challeng-

ing, requiring complex mathematical models and high-performance control algorithms.

At first glance, the task might look easy. After all, more current to the stator increases the

magnetic field strength at each pole, which increases the resulting magnetic field of the

rotor, in turn increasing the rotational force and thus the motor’s torque. However, unlike

the DC motor, simple PWM or varying the voltage alone will not offer good speed control,

since the frequency of the changing magnetic field is the principle determinant of rota-

tional speed. At the same time, changing the frequency of the AC current alone is prob-

lematic as it combines with voltage to define the air gap flux, and shifting one without the

other would have undesirable effects. So, we can adjust frequency and voltage together,

maintaining a constant ratio between the two to alter the speed of the motor, a method

called variable voltage variable frequency (VVVF). However, as discussed with DC

machines, above the rated speed we can no longer usefully increase voltage. So, we increase

IMAGE 4.15

Three-phase AC.

Three-phase AC provides a more even distribution of torque. Each phase is offset by 120°, and maximizes when

the rotor is aligned with that pole on the stator.

117Electric Machines

frequency alone and experience reduced torque with decreased flux. Unfortunately, this

method is imperfect, control is imprecise and response can be sluggish. So, it is not good

for multi-motor EV configurations that require precise motor speed coordination, and it is

similarly not good for hybrids that need to be accurately controlled to work with the ICE.

So, in general VVVF is not suitable for advanced vehicle applications.

A preferred method of speed control is called Field Oriented Control (FOC), also called

vector control. To understand FOC, we need to recall that the magnetic field generated by

the stator is constantly changing; and at any point in time, the field has a distinct magni-

tude and orientation. So we can understand the stator field as having a direction and mag-

nitude, what physicists call a vector. We can resolve this vector into two components, one

operating perpendicular to the axis of rotation, and so defining rotational force or torque,

and the other operating parallel to the axis and defining magnetic flux. Since these two are

orthogonal components, which means they operate perpendicular to each other in three

dimensions, they are also independent of each other; so changing one has no effect on

the other. FOC uses this principle as the basis of a control logic that allows us to vary the

torque without changing the magnitude of the magnetic field, enabling a more dynamic

torque response.

To accomplish this, you might expect that a precise and instantaneous measure of the

stator field is needed to adjust winding current and thus define a current vector that can

define torque exclusively. But placing a precision sensor in the air gap to measure the stator

field is expensive and not easy. So in EVs we largely rely on an indirect vector control that

uses measured rotor speed and current to calculate a predicted flux angle and magnitude

and identify a reference angle of slip. The actual rotor flux interaction does not need to

be fully identified or measured; instead we only need to identify the slip speed to deduce

rotor flux position.8 A slip lookup table is used to define the generation of torque. Not

unlike the tables used in ICE control, the desired number is based on parameters such as

load, motor speed, and throttle position. The result is that, in effect, the torque-producing

component of the stator flux is managed separately, allowing us to maintain torque at a

near optimal level throughout the operating range. However, FOC is hardly perfect, most

notably, dependent on a fixed lookup table, it can have trouble accommodating changes

in operating temperature and magnetic saturation, and solving this only compounds an

already-complex control operation.

As a result Direct Torque Control (DTC) is gaining interest. This advanced control

scheme directly controls the stator flux linkage and torque by controlling the switching

modes of the constant voltage PWM inverter. Basically, DTC estimates the magnetic flux

and torque based on the current and voltage of the motor, and when these estimated val-

ues vary too far from set reference values, switching is used to bring them back into the

targeted band. The process is quick, simple, and efficient, allowing torque to be directly

controlled, offering faster torque response, while simplifying computational require-

ments. However, this control logic can present challenging torque ripple and sluggish

response. Recent work on improved algorithms and the application of fuzzy logic may

offer solutions.9

8 K.T. Chau, Electric Vehicle Machines and Drives: Design, Analysis and Application. Wiley-IEEE Press, Singapore,

2016.

9 F. Korkmaz, İ. Topaloğlu and H. Mamur, Fuzzy logic based direct torque control of induction motor with space

vector modulation. International Journal on Soft Computing, Artificial Intelligence and Applications (IJSCAI) 2 (5/6),

2013, 31–40; and Y. Bendaha and M. Benyounes, Fuzzy direct torque control of induction motor with sensorless

speed control using parameters machine estimation. 2015 3rd International Conference on Control, Engineering &

Information Technology (CEIT), Tlemcen, Algeria, 2015.

118 Automotive Innovation

Control complexities aside, induction motors may offer a promising option for automo-

tive applications. While a notable improvement on the previous DC machines, the induc-

tion machine’s

power density and efficiency tend to be somewhat lower than their PM

counterparts we will examine next. So, an induction machine is likely to be larger and

heavier than a PM machine with the same performance. Typically a high-speed design is

preferred, as this allows for more power from a smaller and lighter machine. They gener-

ally provide a high starting torque and robust configuration that offers good reliability;

while overall efficiency tends to be lower than the PM motors we will discuss next, they

still operate at about 85% efficient even at maximum load.

The basic induction machine is a mature and established brushless motor technology,

dating back to Nikola Tesla’s introduction in 1887. So, it is not likely that we will see dra-

matic improvements in performance or efficiency in the future. Faced with rapidly devel-

oping innovations in other areas of electric machine technology, future applications for

induction motors in production vehicles may be limited.

Permanent-Magnet Machines

While induction machines clearly still have a place in EVs, the development of high-

performance magnetic material and innovations in machine design are making PM

motors much more promising in most electric and hybrid vehicle applications. You might

be inclined to think that this will raise that ugly problem we discussed previously, the

negative effects of a commutator and brushes. However, the solution is fairly straight-

forward: instead of placing the magnets on the stator and necessitating commutation to

the rotor, we can place them at the rotor. The stator windings define a shifting magnetic

field that crosses the fields of the magnets in the rotor, resulting in rotational force and

defining the basic operation of a brushless permanent-magnet (BLPM) machine. No com-

mutator required.

The stator windings can either be concentrated or distributed. Distributed winding

spreads the armature windings across the stator evenly; this can offer smoother opera-

tion and higher efficiency because of the ability to harvest reluctance torque (something

we’ll discuss soon). In addition, the distribution of the conductor means less heat buildup.

Concentrated windings, on the other hand, place only one phase coil at each slot on the

stator. The advantage is this can be smaller, lighter, and easier to manufacture. While cop-

per loss can be lower in concentrated windings, this configuration generally requires more

poles to avoid torque ripple, and that complicates manufacturing as well as high-speed

operation. In addition, these machines tend toward greater heating and vibration.10 As

a result, with the exception of the Hyundai Sonata hybrid, all major production cars use

distributed windings.

Two key aims in defining the stator are effective heat shedding and the reduction of

resistance. But these two are linked since the generation of heat in the coil is a function of

resistance. Resistance is defined by the cross section of the wire, the material used, and the

10 B. Sarlioglu, C.T. Morris, D. Han and S. Li, Benchmarking of Electric and Hybrid Vehicle Electric Machines,

Power Electronics, and Batteries. 2015 International Aegean Conference on Electrical Machines & Power Electronics

(ACEMP), 2015 International Conference on Optimization of Electrical & Electronic Equipment (OPTIM), 2015; and

Y.Y. Choe, S.Y. Oh, S.H. Ham, I.S. Jang, S.Y. Cho, J. Lee and K.C. Koa, Comparison of concentrated and distrib-

uted winding in an IPMSM for vehicle traction. Energy Procedia 14, 2012, 1368–1373.

119Electric Machines

square of the current. So, to allow high current capacity without excess heating, the cross

section can be adjusted. One option is using wire with a rectangular cross section to allow

the wire wraps to fit together with less wasted space, enabling more coils in a given cross

section. The result can maximize current density, offering greater continuous performance

in the same sized package (Image 4.16). Using rectangular wire in a simplified wave-style

winding with exposed ends, called a bar wound stator, can offer greater surface area expo-

sure and result in improved heat rejection. These exposed turns can be cooled with oil

flow to improve heat rejection by 50% or more.11

A more advanced alternative may soon be defined by extremely low resistance materi-

als. A true superconductor operates with zero resistance but only at extremely cold tem-

peratures. However, work on so-called high-temperature superconducting materials is

improving and could soon redefine the electric machine. For example, a single layer of

carbon shaped in an extremely fine tube, called carbon nanotubes (CNTs) can be used to

define a web of fine woven CNT threads called nanotube yarn, to replace copper wind-

ings. This could potentially halve the resistance of a coil and dramatically lower weight.12

As a bonus, heat shedding would increase notably.

11 S. Jurkovic, K.M. Rahman, J.C. Morgante and P.J. Savagian, Induction machine design and analysis for general

motors e-assist electrification technology. IEEE Transactions on Industry Applications 51(1), 2015, 631–639.

12 D. Johnson, Carbon nanotube yarns could replace copper windings in electric motors. IEEE Spectrum

3 October 2014. Available at https://spectrum.ieee.org/nanoclast/semiconductors/nanotechnology/

carbon-nanotube-yarns-set-to-replace-copper-windings-in-electric-motors

IMAGE 4.16

Square wire.

A square wire profile can allow more efficient windings.

https://spectrum.ieee.org

120 Automotive Innovation

The fundamental mechanical difference between a PM and induction machine is in the

rotor. Rather than a squirrel cage, the rotor is essentially a low-reluctance housing for PMs.

A common version of this places the magnets in a V shape within the rotor (Image 4.17).

Used in the Prius hybrid and multiple other production vehicles, this configuration defines

a more significant space between the magnets and can incorporate a designed air pocket

in the core to change saliency and help reduce counter EMF during regenerative braking.

An advantage of this approach is the angulation of the magnets allows for a more gradual

onset and release of maximum torque, diminishing torque ripple which can be prominent

at low speed and high current.

An alternative would be to place the magnets at the surface. This can provide higher

efficiency because the rotor magnets are closer to the stator coils, so less energy is lost.

However, gluing magnets to the surface of the rotor presents a mechanical weakness that

can limit the upper speed of the machine. This is a bigger deal than may first be appar-

ent. Automotive electric machines can run at well over 10,000 rpm, with speeds likely to

continue to increase to perhaps well over 20,000 rpm. As a result, the durability of the rotor

is an increasingly important concern. In addition, because surface-mounted magnets have

IMAGE 4.17

Rotor magnet configurations.

Reduced torque ripple in PM machines can be achieved by skewing the magnets, effectively offering the same

effect as a skew in the induction squirrel cage. However, this requires a compromise in torque. As an alternative,

the Chevy Bolt utilizes dual layer V-shaped configuration to offer smoother operation.13

13 F. Momen, K. Rahman, Y. Son and P. Savagian, Electrical propulsion system design of Chevrolet Bolt battery

electric vehicle. Energy Conversion Congress and Exposition (ECCE), Milwaukee, WI, 2016.

121Electric Machines

about the same reluctance as air, the space taken up by the magnets is effectively like an

increase in the air gap, and this larger gap can cause a loss of flux density. These may be

the reasons that the Hyundai Sonata is the only production car to use surface-mounted

magnets.

Magnets

More important than their particular mounting configuration, the composition of the mag-

nets has changed greatly over the past generation. In fact, the development of high-power

PM has made an array

of motor topologies possible that would not have been achievable

previously. In the 1970s and 1980s, rare earth elements were found to be particularly prom-

ising in the creation of PMs when combined with metals such as iron, nickel, or cobalt.

This is because the atomic structure of these elements contains the unusual condition of

multiple orbits containing only a single electron. Unlike the more common paired elec-

trons with opposing spins, these unpaired electrons can be made to all spin in the same

direction, generating an especially strong magnetic field. In addition, the crystalline struc-

ture of these materials defines a high directional property, or anisotropy, which makes a

magnetization that is aligned with the crystalline axis easier to achieve and more durable.

The resulting rare earth magnetic alloys are nothing short of remarkable. Previously,

magnets were made either of a ceramic compound of iron oxide and other metals, called

ferrite, or an alloy of aluminum, nickel, and cobalt that produced a much stronger magnet

called Alnico. Yet compared to current rare earth options, both of these have a very poor

specific power, and would need to be many times larger to offer the same performance. In

fact, the samarium–cobalt alloy (SmCo) that defined the first family of rare earth magnets

proved more than twice as powerful as Alnico magnets and many times more power-

ful than ferrite magnets. An important point of comparison is the magnetism remaining

after the external magnetic field that magnetized a material is removed, called remanence

flux density. Also important is a magnet’s resistance to demagnetization, called coerciv-

ity. As the table below indicates, these factors, and the magnetic energy of the material,

are remarkably greater in rare earth magnets. However, the very high cost of these mag-

nets limits their use. An even more impressive and feasible option was defined with the

next generate of rare earth magnets, neodymium–iron–boron (Nd2Fe14B or NIB). With just

about twice the power of SmCo magnets, they are also lower cost. With the price of NIB

magnets having dropped significantly over the past two decades, they are now the go-to

choice for PM machines for everything from cordless tools to automobiles.

However, while impressive, rare earth magnets can be fragile and are susceptible to

accidental demagnification from flux and also from heat. NIB magnets in particular are

vulnerable to high heat, and must be kept at operating temperatures below 300°F to avoid

thermal demagnification, a point defined as the Curie temperature of the material. At

high heat, the magnets are weakened and power is lost. And, because of their low resis-

tance, these magnets can generate high internal eddy currents intensifying the problem.

In fact, frequently the limit to a PM motor’s high-speed operation is heat. This has defined

a continued need for SmCo magnets in high-temperature applications, as they can with-

stand temperatures approaching 800°C (Table 4.1).

Besides heat vulnerability, there is one more challenge to these super magnets. They

require the use of dense rare earth metals. Metals such as dysprosium and terbium are

122 Automotive Innovation

added to improve resistance to heat; if not, the thermal performance of NIBs would be even

worse. However, these metals are largely mined in China, raising concerns over reliable

global access as well as cost. As a result, Honda has developed a commercial motor that uses

no heavy rare earth metals. Instead of the typical process of heating the material and press-

ing it together, called sintering, a hot deformation technique is used that defines much

smaller crystalline grains. These nano-grains are about 100–500 nm across, or about a tenth

of the size resulting from sintering, and allow for greater heat resistance without the use of

heavy rare earth elements.14 The cost and uncertainty of rare earth elements has even made

ferrite magnets of interest, leading GM to use ferrite magnets in one of the motors in the

second generation Chevy Volt. When combined with reluctance torque, the weaker magnets

offered adequate performance and good heat tolerance at a lower price. In fact, ferrite mag-

nets increase coercivity with increasing temperature, just the opposite of the neodymium.

More powerful magnets are not always better. As we know, machines for automotive

application need to perform over a wide speed range. Because PMs are fixed, and can’t

be turned up or down, the designed strength of the magnet represents a trade-off. In par-

ticular, at high speed and low torque, the PM produces a high flux that cannot be turned

off. The back EMF can be accommodated with an opposing stator current, to a point. But

this can seriously degrade efficiency, so it can be better to avoid overly powerful mag-

nets for high-speed machines. Using a weaker magnet means it is principally the rotor’s

saliency that defines torque at low speed and high load. This is a bit more like the reluc-

tance machines we will discuss in a bit.

BLPM Control

Advanced magnets are vital to the performance of modern PM machines, but brushless

PM motors would not be possible without precise digital control. Power electronics provide

each phase of a rotating magnetic field that defines the armature’s rotation. So, although

14 How Honda Developed the World’s First Heavy Rare Earth-Free Hybrid Motor. Available at http://world.

honda.com/environment/face/2016/case59/episode/episode01.html

TABLE 4.1

Comparing Magnet Characteristics

Sources

Remanence

Flux Density Br

(Gauss)

Intrinsic

Coercive Force

Hci (kA/m)

Maximum

Magnetic Energy

BHmax (kJ/m3)

Curie Temperature

(Magnetism Is

Lost) Tc (°C) Cost

Ferrite 200–400 100–300 10–40 570 Very low

Nd2Fe14B 600–1,400 600–1,800 200–440 315 High

SmCo5 900–1,150 450–1,200 60–240 750 Very high

Alnico 600–1,200 50–120 10–80 700–860 Low

Source: S.J. Collocott, J.B. Dunlop, H.C. Lovatt and V.S. Ramsden, Rare-Earth Permanent Magnets: New Magnet

Materials and Application. School of Electrical Engineering, University of Technology, Sydney, NSW, 2007;

J. Liu, “Some Design Considerations Using Permanent Magnets.” Magnetics Magazine March 13, 2016;

J.M.D Coey, Magnetism and Magnetic Material Cambridge. Cambridge University Press, Cambridge, UK,

2009; and J.D. Widmer, R. Martin and M. Kimiabeigi, Electric vehicle traction motors without rare earth

magnets. Sustainable Materials and Technologies 3, 2015, 7–13.

http://world.honda.com

123Electric Machines

unlike the control of a synchronous AC machine. This can be accomplished by the rectan-

gular switching of DC or the generation of a true sinusoidal current.

Rectangular switching defines a permanent-magnet brushless direct current motor

(PM/BLDC). A controller provides pulses of current to the stator windings, shifting the

stator phases off and on. Voltage is provided in two of three phases at any one time, and

switches between phases every 60°. So, each phase is on for 120° of field rotation. This

defines a rotating square wave that produces the rotation of the motor. The resulting

geometry of field interaction with a rectangular signal produces extremely high torque.

If we instead produce a true sinusoidal current, we can define a synchronous machine,

or permanent-magnet synchronous motor (PMSM). The resulting torque and therefore

power density is diminished; but the sinusoidal current smooths the torque, reducing

torque ripple that can be associated with PM/BLDC machines. This is because the syn-

chronous field control allows the stator and rotor to interact at consistent angles throughout

the rotation. A synchronous machine can utilize the same control strategies as induction

machines, since they both run on a sinusoidal waveform. In particular, FOC is commonly

applied to synchronous machines. The high efficiency,

good field weakening capacity, and

high torque have made the PMSM the dominant EV machine.

The PM/BLDC motor offers a good fit for HEVs where the machine principally aug-

ments the primary internal combustion engine. Strong torque density, particularly at low

speeds, and good efficiency makes a good fit with this hybrid architecture. So, for example,

Honda saw this as a good choice for the early Honda Integrated Motor Assist (IMA) sys-

tem. Brushless AC (BLAC) machines, on the other hand, with greater control, smoother

operation, and better power performance are more favorable to being the primary traction

device in a hybrid system or a modest EV. Efficiencies can reach into the mid- and high

90s.15 Consequently, the brushless AC motor has been adopted in cars as varied as the

Nissan Leaf, Chevy Bolt, BMW i3, Mitsubishi i-MiEV, Toyota Prius, and Citroën C-Zero, for

example.

However, brushless sinusoidal machines do require more complicated control. In par-

ticular, the need for precision FOC control of sinusoidal flux at the air gap requires that

the angle of the rotor be exactly known. This necessitates an extremely high-resolution

position sensor.16 The sort of encoder that can convert the angular position and movement

of the shaft to a digital signal with high precision adds significant cost and complexity

that were not necessary for the open-loop control system of the induction machine. And

position sensing for the trapezoidal flux of the BLDC machine only needs to know rotor

position at each commutation point, so each 60°, and can be managed with a less-expensive

Hall effect sensor. However, a variety of sensorless coil control methods that measure the

back EMF and other parameters to determine armature position are being developed for

PMSM applications.

More generally, both PM machines offer good flexibility. Either machine can easily func-

tion as a generator or motor. When a motor, the rotor magnets follow the stator field move-

ment. When a generator, the rotor magnets lead the stator field. Also PM machines allow

for high adaptability and a variety of geometries. An axial flux machine with the air gap

perpendicular to the axis of rotation can be an attractive option, for example. Sometimes

called a pancake motor, this configuration allows for a lot of poles and can provide

15 K.T. Chau and W. Li, Overview of electric machines for electric and hybrid vehicles. International Journal

of Vehicle Design 64(1), 2014, 1–34; and Patrick Hummel et al., UBS Evidence Lab Electric Car Teardown—

Disruption Ahead? UBS Group AG, May 18, 2017.

16 J.X. Shen, Z.Q. Zhu and D. Howe, PM brushless drives with low-cost and low-resolution position sensors. The

4th International Power Electronics and Motion Control Conference, 2004.

124 Automotive Innovation

improved power density over the traditional radial flux PM motors we’ve been discussing,

and improved heat shedding given the large surface area and radius of the rotor. In addi-

tion, the higher moment of inertia can offer smoother operation in certain applications.

This flexibility of design can be useful when trying to integrate a machine into an engine

bay for a hybrid car. Similarly, PM designs can allow for single- or double-sided stators

and multiple rotors. For all these reasons, PM machines are seen as the best option for a

possible wheel-hub configuration to be discussed in Chapter 5.

These PM machines also present challenges related to the use of PM. First, with an

unregulated PM flux, high-speed back EMF can be problematic. Second, sensitivity to high

temperature can be a challenge and significantly reduce performance if not properly man-

aged. Lastly, the rising cost of PM presents a problem for low-cost applications. For these

reasons, there is increasing attention being paid to motors that use reluctance as their driv-

ing force and so do not need to rely on magnets.

Reluctance Machines

Making a motor that relies on reluctance could potentially define a much simpler and

less-expensive option for EVs. Remember that a force is exerted on any low-reluctance

ferromagnetic material in a magnetic field. This force tends to align the material with

the density of magnetic flux, like the needle we considered at the start of the chapter

that turned to define a path of lowest reluctance. This is the basic driving principle of the

switched reluctance machine (SRM). If we properly define a rotor with a low-reluctance

ferromagnetic path, the flux of the stator can be designed to cause the rotor to turn like that

needle to define the path of least reluctance and so exert a rotational force.

In fact, the basic design of a functional SR motor is relatively straightforward. We can

begin with a rotor with projected poles made of laminations of iron, called salient poles.

These will define our low-reluctance paths. We couple this with a stator with similar salient

poles wrapped in coils to generate the flux (Image 4.18). We can then produce a path of

magnetic flux from one pole on a stator to an opposing pole. The rotor will rotate to align

its poles with the flow of flux, providing a path of least reluctance to the field. With proper

digital control, we can rotate this flux around the stator poles, defining significant and

continuous torque on the rotor. So, in sum, current flows through a coil on a salient stator

pole and defines a flux that pulls the salient rotor pole to align with the stator pole and so

reduce the reluctance of the magnetic field. When the poles are aligned, the reluctance is

minimized; and if the pole remained energized, the rotor pole would be held in position,

but instead it is switched to another stator pole that attracts another rotor pole and contin-

ues the rotation. With multiphase operation, as this pole is aligned an alternative pole is

feeling the pull of another stator winding and the rotation continues.

This is not a new idea; in fact, it dates back to early nineteenth century Scotland. However,

because smooth and controlled rotation requires that we precisely define the rotating sta-

tor field, a reliable performance reluctance motor was not possible before the development

of modern advanced power electronics, sensors, and digital control technology.

The torque is defined by the current and speed of the rotating stator field. Unlike PM

machines, only a pulling force is exerted on the rotor pole, so torque is independent of cur-

rent polarity, and an inverter specific to SR machines is needed. Precise switching needs to

be provided to each phase in series but independently, allowing for high-speed operation

125Electric Machines

with adjacent phase currents overlapping. PWM is used to allow control of field magni-

tude; but the discrete excitation of phase windings still makes the SR highly vulnerable to

torque ripple. To address this a torque sharing function (TSF) can be implemented that

distributes torque in relation to armature rotation so that the sum of all phase torques is

defined as the targeted torque production.

Therefore, particularly for SR machines, synchronization of winding excitation and

armature rotation needs to be precise, especially for high-speed operation. A hall effect or

optical sensor can be mounted on the shaft to provide rotor position. In addition, sensor-

less control that estimates rotor position based on known parameters such as current can

improve the fault tolerance of the machine.17

Even with these challenges in mind, the great advantage of a switch reluctance machine

is its simplicity. As demand for electric motors grows and magnets become more expen-

sive, SR offers an inexpensive and robust alternative. The stator has concentrated windings

around each pole. The rotor has fewer poles, a simple laminated structure, and no wind-

ings. With neither magnets nor coils, the rotor can handle high torque and high rpm with

no trouble. Distinct salient poles means the phase windings can operate in isolation from

each

other, reducing the problem of mutual inductance, and offering high fault tolerance.

So, with the rising cost of magnets, the potential for SR motors is being seriously exam-

ined in EV applications and elsewhere (Image 4.19). The machines offer simple control

and cooling, reliability, very high speed operation, excellent heat tolerance, good torque,

and low cost. Although these motors still generally suffer from lower torque density and

high noise, advanced digital control and computational methods are being developed to

address this.18 Increasingly, a well-designed machine with carefully considered pole and

17 M. Yilmaz, Limitations/capabilities of electric machine technologies and modeling approaches for electric

motor design and analysis in plug-in electric vehicle applications. Renewable and Sustainable Energy Reviews 52

(December), 2015, 80–99.

18 M. Cheng, L. Sun, G. Buja and L. Song, Advanced electrical machines and machine-based systems for electric

and hybrid vehicles. Energies, 8, 2015, 9541–9564.

IMAGE 4.18

Switched reluctance machine.

This SR machine has four salient poles on the rotor, defining low-reluctance paths. The stator provides six

salient poles each wrapped in coils. As the stator generates rotating paths of magnetic flux, the rotor rotates to

allow the path of least reluctance.

126 Automotive Innovation

phase configuration can provide control and reduce torque ripple. While still large, with

continued development, increasing power is becoming available from a smaller package.

So, look forward to seeing an SR machine in a production vehicle soon.

A variation on this theme that is somewhat less developed is the synchronous reluctance

machine (SynRM). The basic idea is to devise a rotor with integrated flux barriers and

guides, defining a minimum reluctance in one direction and maximum in the other. The

rotor could then be made to turn synchronously with a sinusoidal voltage. The SynRM

occupies a space somewhere between a PM and induction machine, and offers a new sort

of reluctance drive. With a small number of poles, and the same number of poles on rotor

and stator, it bears only a little resemblance to the SR machine.

This machine is decidedly not ready for deployment in production vehicles. It has high

torque ripple, low power density, and poor efficiency. But, it is still early, and its progress

is interesting. With no heat loss, no eddy currents, and a simple variable frequency drive

managing speed, the idea has clear potential. A PM-assisted SynRM machine may offer

improved efficiency and torque density thanks to the magnets, and might only need a little

assist from modest ferrite magnets.

Advanced Motor Possibilities

The PM motors discussed so far are widely used and demonstrate reliable and impres-

sive performance. However, they are not perfect. In particular, they pose two general

challenges. First, they require PM be mounted on a rapidly spinning and vibrating rotor.

IMAGE 4.19

Switched reluctance in action.

This Land Rover all-electric test vehicle, seen here at the 2013 Geneva Auto Show, is one of the few uses of an

SR machine in an EV. It uses a 70 kW (94 bhp) switched reluctance motor, coupled with a 300-volt lithium-ion

battery pack with a capacity of 27 kWh.

Image: Norbert Aepli, Switzerland

127Electric Machines

Themechanical stress placed on rotors on PM machines due to centripetal force at high

speed is significant. This is particularly true for surface-mounted magnets, but presents a

problem for all PM rotors. Second, buried in the heart of the machine, these magnets can

get easily overheated and are difficult to cool. Heat stressing may be a more significant

challenge than mechanical stress, since it can lead to the partial demagnetization of the

rotor magnets and can have a devastating impact on power capacity. A possible solution

comes from a deceptively simple and radical idea: put the magnets on the stator instead.

An interesting example is based on an ingenious modification of the basic SR machine.

By putting the magnets on the stator of an SR machine, the rotor is left with neither wind-

ings nor magnets, maintaining a robust and simple architecture, and making the machine

easy and inexpensive to manufacture. But the addition of magnets fundamentally shifts

the power capacity of the reluctance drive. The rotor incorporates salient poles coupled

with PM-equipped salient poles on the stator, earning the name: doubly salient PM motor

(DSPM). Operation is somewhat like a square waveform BLDC, but with a much-improved

constant power speed range (CPSR). The result is high efficiency and power density, but

at the cost of some control capacity since the PM flux is uncontrolled.19 And, of course, the

DSPM still entails the cost of PM magnets.

An alternative could replace the PMs with electromagnets to define a doubly salient

electromagnet machine. This promises improved control, as the magnetic flux can now

be regulated. However, the power loss through excitation can tend to be high, so the effi-

ciency drops. And because an electromagnetic coil might require five times the size of an

equivalent neodymium magnet, the motor size increases.20

A promising example of this idea is flux switching PM motors (FSPM). This machine

places magnets between U-shaped magnetic cores on the stator. The magnets are circum-

ferentially magnetized, meaning that one side is north and the other south as you move

along the circumference; and magnets are placed with alternating polarity. Concentrated

windings are then wrapped around the two adjacent segments and the magnet, defining

a single tooth (Image 4.20).

As the rotor coil aligns with the next stator tooth of the same phase, the polarity of the

flux linkage is reversed. By providing opposing direction current at each flux linkage,

torque is generated in both flux locations. With coil excitation, the field on one side is

reduced and on the other side increased, and the rotor moves to the stronger field. As the

rotor travels across the stator field, adjacent rotor poles align with successive stator teeth

and there is a reversal in the PM flux linkage, earning it the name ‘flux switching’. In

essence, this machine operates a bit like a brushless AC motor.21 By using the high-energy

PM excitation and the flux concentration effect, the FSPM provides very good power and

torque density. In addition to the general cooling advantage provided simply by putting

the magnets and coils on the stator, the need for more limited windings means lower cop-

per loss and so less heating. An added advantage of this machine and others like it is the

inherent redundancy, and thus fault tolerance, offered by multiple coils; if a circuit fails,

the motor control can reconfigure and maintain traction capacity.

19 J.T. Shi, Z.Q. Zhu, D. Wu and X. Liu, Comparative study of biased flux permanent magnet machines with dou-

bly salient permanent magnet machines considering with influence of flux focusing. Electric Power Systems

Research 141, 2016, 281–289.

20 J.D. Widmer, R. Martin and M. Kimiabeigi, Electric vehicle traction motors without rare earth magnets.

Sustainable Materials and Technologies 3, 2015, 7–13.

21 M. Cheng, L. Sun, G. Buja and L. Song, Advanced electrical machines and machine-based systems for electric

and hybrid vehicles. Energies, 8, 2015, 9541–9564.

128 Automotive Innovation

But this motor is not perfect. Accidental partial demagnetization can be a concern, as

the PMs are mounted adjacent to the coils. The uncontrolled PM field still presents a chal-

lenging control problem, and it can make the constant power operation range small as the

back EMF is difficult to accommodate. This is a particular challenge for EVs that require a

wide address this CPSR.

This can be addressed by a variant of the doubly salient machine that can be called

hybrid excitation doubly salient machines (HEDSPM). The name may be a mouthful, but

it really says it

all. While maintaining the general doubly salient design, this machine uses

a hybrid combination of both DC field windings and PM to provide improved flux control.

The additional DC winding offers desirable and straightforward controllability, avoiding

the cost and complexity of vector control while widening the CPSR.22 The rotor remains

without coils or magnets, offering motor robustness and improved thermal stability for

the magnets. A positive DC current strengthens the PM flux during high torque and a

negative current can weaken the flux for constant power in high-speed operation. So the

DC coil acts as both an electromagnet and a mechanism of flux control. This allows for

improved motor efficiency over a wider speed range, offering the advantage of a DSPM

while providing magnetic flux control with field windings. But the trade-off is in power

density and efficiency resulting from the added DC coils, as well as added complexity. In

essence, we trade power density for improved controllability and a wider constant power

range.

This hybrid approach has advantages. The flux density at the air gap becomes easily

controllable. This enables flux strengthening for much-improved short periods of high

torque for acceleration or starting. The constant power range is widened greatly by enabling

22 Q. Wang and S. Niu, Overview of flux-controllable machines: Electrically excited machines, hybrid excited

machines and memory machines. Renewable and Sustainable Energy Reviews 68, 2017, 475–491.

IMAGE 4.20

Flux switching permanent-magnet (FSPM) machine.

There are multiple configurations of poles and windings for FSPM machines. For example, a single-phase coil

can be wound around each stator pole with four coils defining one phase. This would define a double-layer coil,

since a coil slot would have two coil phases. Alternatively, each slot could have only one coil, with alternating

poles left without a coil. The torque capacity could be maintained, but this could result in increased torque

ripple.

129Electric Machines

simple flux weakening. Adjusting the flux density also allows for stable voltage output

while operating as a generator, simplifying battery charging.23 It is unfortunate that the

DC field windings of this hybrid system can increase coil loss and reduce power density;

but as you might expect, there is work being done on yet another option.

The option that can preserve some of these advantages while also providing lower cop-

per loss is a specially defined flux memory motor. The key to this design is the use of a

magnet that can be readily re-magnetized, exhibiting what is called a low coercive force,

but is also able to maintain magnetism once polarized, called high remanence. The domi-

nant neodymium magnets do not fit the bill, so Alnico magnets are used. Current pulses

can be used to reset the polarity of these so-called memory magnets during motor opera-

tion. This allows control of the magnetic flux, without requiring the copper losses entailed

in a DC electromagnetic coil, earning the name Flux Modulated Permanent Magnet, or

sometimes Flux Mnemonic Permanent Magnet (FMPM).24

This innovative FMPM offers significant advantages. Since the needed pulse is very brief,

there is minimal coil loss. The controllable magnet also offers a simpler and more efficient

way to provide flux weakening. The additional windings mean reduced power density,

but the added control means torque can be boosted for short periods such as start-up and

overtaking. The only real disadvantage is the complexity of the system, incorporating two

distinct sets of coils, and thus increasing manufacturing difficulty and costs.

If that’s not impressive enough, a variation on this design uses two magnet types to pro-

vide improved power. Since the remanence of Alnico magnets is less than traditional NIB

magnets, power density can suffer in a memory motor. However, by placing two magnets,

one Alnico and one NIB on the rotor, we can have the best of both. The higher remanence

NIB provides needed power density, while the Alnico magnet enables the benefit of flux

memory current pulsation (Table 4.2).25

Before we finish, we should briefly look at one more variation on this theme. Removing

both coils and magnets from the rotor on these advanced machines has resolved the

mechanical issues related to mechanical failure of the rotor at high speed and eased the

problem of cooling. However placing both coils and magnets on the stator can present new

challenges. First, mechanically it can be very difficult to fit multiple coils and PM on the

stator. The stator needs to be made larger and this means a larger motor, which can be a

23 A. Emadi, Advanced Electric Drive Vehicles. CRC Press, Boca Raton, 2017.

24 C. Yu, K.T. Chau, X. Liu and J.Z. Jiang, A flux-mnemonic permanent magnet brushless motor for electric vehi-

cles. Journal of Applied Physics 103(7), 2008, 07F103–07F103-3; and Y. Fan, L. Gu, Y. Luo, X. Han and M. Cheng,

Investigation of a new flux-modulated permanent magnet brushless motor for EVs. The Scientific World Journal

2014, 2014, 1–9.

25 M. Cheng, L. Sun, G. Buja and L. Song, Advanced electrical machines and machine-based systems for electric

and hybrid vehicles. Energies, 8, 2015, 9541–9564.

TABLE 4.2

Comparing PM Machine Characteristics

Power Torque Efficiency Controllability Robustness

DSPM Moderate Moderate Good Moderate High

FRPM Good Good Good Moderate Moderate

FSPM High High Good Moderate Moderate

HEPM Moderate High High Excellent Moderate

FMPM Moderate High High Excellent Moderate

Source: K.T. Chau, Electric Vehicle Machines and Drives: Design, Analysis and Application. Wiley-IEEE Press,

Singapore, 2016.

130 Automotive Innovation

major problem in some automotive applications. Second, having the coils so close to the

outer periphery of the motor invites increased flux leakage.

However, if we swap the location of the rotor and stator, making the outside ring the

rotor and the inside cylinder the stator, these problems can be addressed. Making the

inner cylinder the stator means the inner-core of this component can be used to accom-

modate DC coils and magnets, allowing more efficient use of space and no need to enlarge

the overall motor dimensions. Essentially, this defines two layers to the stator, an outer

layer with armature windings and an inner layer with magnets and possibly DC coils.

With both magnets and DC coils fully incased, flux leakage is minimized. And, because

the armature windings and PMs are located in different layers of the stator, the threat of

accidental demagnetization is reduced. The outer ring can then just provide salient poles,

with no magnets or coils, defining a more robust component with no associated flux leak-

age. Moreover, with an increased radius the moment of inertia of the rotor is now usefully

increased without increasing overall machine size. The higher inertia offers smoother, qui-

eter operation.

Maybe you have already guessed the principle drawback of this design. With the stator

now tucked into the core of the motor, it will be difficult to dissipate heat, making this

motor susceptible to magnet degradation. In fact, the outer magnets can act as an insulator,

making cooling even more challenging. Yet since the inner stator does not rotate, provid-

ing cooling to the core though a fluid or oil system could be a relatively simple task.

In fact, continuing in this same vein, for both PM and induction configurations, there

are any number of possible motor topologies that we could imagine. For example, a dual-

stator–dual-rotor motor can be defined by applying the same techniques as a conventional

motor. The basic idea is to place two concentric stators on either side of a rotor, or vice

versa. Each flux interaction generates torque. The total torque is the sum of the torques

from each stator. While this could add additional weight and complexity, the added reli-

ability and torque capacity are notable,

and excellent for high-power applications.

Nissan’s so-called super motor is an interesting example of such innovation taken to an

impressive finish. Intended to replace the use of two conventional motors, the super motor

has dual rotors, one on the inside and one the outside of the stator. Each rotor is attached

to a separate shaft, and a compound current to the stator is used to control each. In one

application, one rotor could be used for motoring and the second dedicated to generation;

in another, each shaft could be connected to a separate drive axle, allowing independent

control of right and left axle.26

In the end, the choice of technology depends on the application, size requirements, the

needed torque and power, acceptable costs, and any number of other factors. There is no

one right answer. But, the options are getting increasingly exciting. While the potential

cost and limited availability of rare earth elements is likely to continue to drive innovation

in magnet-less options, it is also likely that PM machines will continue to dominate the

industry in the near future. Nevertheless, lower cost and robust options like SRMs are too

promising to ignore. And, most of all, the capacity of precision digital control has rede-

fined our options and catalyzed a wide array of innovative machines with performance

characteristics that were unimaginable just a generation ago. In short, the heart of the elec-

tric car of the future is ready to go.

26 Super Motor | Nissan | Technological Development Activities. Available at www.nissan-global.com/EN/

TECHNOLOGY/OVERVIEW/super_motor.html

http://www.nissan-global.com

131

Electrified Powertrains

Early automobiles included some with internal combustion engines and some with elec-

tric motors. In fact, until the fossil fuel boom that led to low-cost gasoline, it was not clear

which would emerge as dominate, if either. If advanced digital control technology had

been available at the time, it is possible that a combination of the two, a hybrid, would have

led the pack. However, in the absence of the capacity to seamlessly integrate two traction

sources into a single drivetrain, and with battery technology in its infancy, the clear win-

ner at the time was the gasoline ICE. Perhaps this is a pity.

Gas versus Electrons?

The internal combustion engine has the remarkable advantage of tremendous flexibility

of operation and range, but it also has its limits. It is easy to refuel, and the fuel is incred-

ibly energy dense; so the effective range off a small tank of gas is large, and unlimited

in a world that has ready available gasoline virtually on every block. In fact, it is only

thanks to the incredible energy density of gasoline that the basic ICE is able to produce

adequate power, as the typical thermal efficiency of an engine is only about 25%. Toyota

recently unveiled a new engine that can achieve 40% thermal efficiency.1 This is a remark-

able achievement, but it still represents less than half the energy input resulting in useful

power. And no matter how much we improve on engine efficiency, it is by nature based on

combustion and inevitably results in emissions. Some of these emissions cause immediate

health and environmental threats, and others threaten the entirety of the planet through

climate change. The emissions resulting from the manufacture of the gasoline only adds

to the problem.

Conversely, the great advantage of the electric vehicle is that it does not necessarily

entail emissions. There are no emissions from the vehicle itself, and depending on how

the electricity is generated, potentially no or little total effective emission. Even when elec-

tricity is generated from fossil fuels, the well to wheel efficiency of an EV is much higher

than its gasoline-powered counterpart. However, the range and flexibility of current EVs

are limited. Even the best batteries available cannot match the energy density of gasoline,

and charging them takes time.

Combining these two systems, and benefiting from the best either has to offer, can be a

match made in automotive heaven. The low-end torque and efficiency of electric machines

complements the flexibility and power of internal combustion engines. Electrification of

the drivetrain can be seen as an enabler of continued improvement in the internal combus-

tion engine, not necessarily a competitor.

1 D. Carney, “Toyota Unveils More New Gasoline ICEs with 40% Thermal Efficiency.” Automotive Engineering

SAE, April, 2018.

132 Automotive Innovation

However, as technology advances, the option of tossing the ICE completely and going

with a battery electric vehicle can be increasingly viable for certain application. The chal-

lenge of electric vehicles over the past two decades has been twofold: first, the cost of

batteries means a pure electric vehicle that requires higher energy density and greater

capacity has been significantly more expensive. Second, because even the best batteries

have notably lower energy density than gasoline, electric vehicles exhibit a more limited

range. Both of these challenges are real, but both are also diminishing and the pure elec-

tric option is beginning to look more practical.

In any case, the advantages of electrification are certain to define the future of the auto-

mobile. Multiple manufactures have plans to incorporate more electric drives into their

lineup. Volvo has committed to transitioning its entire lineup to electric or hybrid drive

by 2019. GM has announced plans for an ‘all electric future’ with no precise date. Ford

has promised 13 new all-electric models by 2023. Mercedes-Benz has announced plans to

electrify its entire fleet by 2022. In fact, virtually all major manufacturers have committed

to the expansion of their hybrid and electric options over the next decade.2 Nevertheless,

predictions of the death of the ICE can be greatly exaggerated. The energy density of gas-

oline makes the internal combustion engine continuingly vital; and the improvement in

combustion performance and efficiency we talked about in the first two chapters prom-

ises to ensure its relevance into the next generation. But, the advantage of electric drive

makes this an ideal accompaniment; and we are likely to see future fleets of electric and

hybrid vehicles, in a variety of configurations. Technologically, there really is no best

option, just a range of innovative possibilities that can help achieve the targeted goals of

the vehicle.

Nevertheless, we are clearly approaching a tipping point in EV market penetration,

marked by a significant uptick in the rate of EV adoption. Some see this shift coming

globally within the next few years.3 In the US, where fuel prices remain low and eased

emissions standards may be on the political horizon, this point may be a bit further off.

However, within a decade, the average EV is expected to be cheaper in Western countries

than a comparable internal combustion car.4 In fact, globally, electric vehicles are projected

to make up more than half of all light vehicle sales by 2040.5

Electrifying the powertrain isn’t just about mixing and matching traction drives, it is

also about seeking opportunities for synergy and efficiency wherever they lie. This means

capturing the wasted energy of braking to use for subsequent acceleration or maybe just to

power the air conditioner. It can mean downsizing to a smaller, lighter, and more efficient

engine, and allowing it to operate at its peak efficiency. Shutting the ICE down when it’s

not needed, maybe just at stoplights, maybe while cruising. Replacing inefficient engine-

driven pumps with more efficient electric pumps; and perhaps upgrading the entire elec-

tric system to facilitate this and other features such as driver assistance, safety, navigation

and entertainment. And, of course we want to maintain responsiveness, acceleration, han-

dling, and general drivability while we do all this.

2 A.C. Madrigal,

“All the Promises Automakers Have Made about the Future of Cars.” The Atlantic July 7, 2017.

3 “Q-Series: UBS Evidence Lab Electric Car Teardown—Disruption Ahead?” UBS Limited May 18, 2017.

4 N. Soulopoulos, “When Will Electric Vehicles Be Cheaper than Conventional Vehicles?” Bloomberg New Energy

Finance April 12, 2017.

5 “Electric Vehicle Outlook 2017: Bloomberg New Energy Finance’s Annual Long-Term Forecast.” World’s Electric

Vehicle Market July, 2017.

133Electrified Powertrains

Hybrid Drive

The performance characteristics of the ICE and electric motor complement each other

admirably. Electric motors produce their maximum torque at low speeds and are able to

maintain that torque through a significant speed range. As we have seen, as the motor

speeds up, torque production eventually drops off though power is maintained for quite

a while. On the other hand, internal combustion engines have a limited range of opera-

tion. At the lower end, torque is limited, increases with speed to a plateau then drops off

relatively quickly. Power production increases with speed more compellingly, but drops

off quickly once maximum power is achieved. As we have discussed, these characteristics

mean engines require a complex set of gearing to enable efficient operation; and innova-

tive but intricate mechanics are required to allow the engine to operate efficiently in a

wider range of conditions and vehicle speeds. Melding electric motors, with their low-end

performance and wide operating range with the power capacity of a fuel-fed internal com-

bustion engine can offer us the best of both worlds (Image 5.1). In fact, if electric drive and

internal combustion had been melded sooner, it is possible that some of the recent innova-

tions in combustion control, as exciting as they are, might never have been necessary.

Most notably, bringing these two power plants together has clear efficiency advantages.

Internal combustion engines are sized to produce needed power for acceleration, hill

climbing, and overtaking, which means the majority of the time the engine is significantly

oversized for the needs of the car. So, while only 30 horsepower or so are needed to keep a

car cruising down the road, the typical engine has about five times that amount of power

available. By integrating an electric motor that can provide a torque boost when needed,

engines can be downsized significantly, as they no longer need to be sized for the maxi-

mum torque conditions.

IMAGE 5.1

Combined torque.

Combining the low-speed torque of an electric machine with the higher speed torque of an ICE, can offer the

performance advantage of a very fat torque curve.

134 Automotive Innovation

The effect on the combustion engine is called load leveling. Combining the power of

an electric motor and combustion engine means we can manage with a smaller combus-

tion engine with little performance loss. And a smaller engine means less weight and

less friction loss. More than this, taking the peak load burden off the combustion engine

allows it to operate closer to its peak efficiency. At light loads, a given engine may consume

less fuel, but that’s not the way to think about efficiency, because the engine’s production

of power can drop even more than its fuel consumption. As discussed in Chapter 3, an

engine achieves lower BSFC closer to its peak load and typically on the lower side of its

speed range (Image 5.2). So, since a typical car spends most of its time cruising, a smaller

engine can operate much closer to optimal load and speed if an electric motor is available

to help meet momentary high loads, providing an efficiency improvement that is more

than just about a smaller displacement.

There are multiple modes and methods for the integration of the combustion engine

and electric motor. The overall architecture of the hybrid system can be understood as a

power circuit. The ICE and electric machine can be connected in series, parallel, or various

combinations of the two. And, the relative balance of tractive capacity from the two drive

sources can vary.

Similarly, the details of the mechanical connection that allows the traction motor to be

incorporated into the powertrain can vary greatly. Initially, carmakers wondered whether

a new form of transmission was needed to enable hybrid powertrains, or could hybrid

drive simply tap conventional transmissions. The early go-to was the CVT, however, this

quickly gave way to the development of a dedicated hybrid transmission (DHT). DHTs

integrate the electric machines fully into the transmission defining a new sort of power

IMAGE 5.2

Power, torque, and efficiency.

An internal combustion engine can operate closer to its optimal efficiency if an electric motor is used to reduce

momentary high loads, and allow the engine to remain on the lower side of its speed range.

135Electrified Powertrains

couplingunit. While a dedicated unit now represents the majority of current full hybrid

systems, the high cost of reengineering the entire powertrain is making more modest

options attractive as hybrid drives grow more widespread. Many more modest hybrid sys-

tems rely on modified conventional transmissions. For example, dual clutch transmissions

are becoming a more popular choice, with VW, Porsche, and Hyundai leading the way.

The electric drive can help address the lag when starting off that can plague some DCTs.

Alternatively, using a modestly modified conventional automatic transmission can allow

for a simple replacement of the torque converter with a traction motor to define a low-cost

hybrid option, and now probably represents about a third of all hybrid systems. AMTs, on

the other hand, are somewhat problematic, as the torque interruption in low gears can be

challenging to accommodate.6 In sum, which system makes the most sense depends on the

type of hybrid system being considered.

In fact, ‘hybrid’ can mean a great many things: At one end might be a powerful electric

motor that provides all the propulsion for the vehicle, with a small ICE that can drive

a generator and provide extended range. On the other end might be what is essentially

a conventional ICE automobile, but with an added capacity to harvest, store and reuse

wasted kinetic energy through the use of a small motor-generator. To better understand

what this spectrum entails, we need to examine these variations in steps.

Baby Steps

We might start with the most modest electric-drive augmentation of the ICE, so-called

micro hybrid systems. At their most simple, micro hybrids incorporate two systems: one

that allows engine shutdown when it is not needed and another that allows us to capture

lost kinetic energy in braking. The former utilizes a more robust starter and higher capac-

ity battery to allow the vehicle to be automatically shut down when not needed in normal

operation, say at a stop light, and restarted seamlessly when again needed. Called start–

stop technology, this can improve fuel efficiency by as much as 10%, depending on driving

conditions.

This is not as simple as just linking a starter-kill switch to the brake pedal. Shutting the

engine down regularly during normal operation presents some complications. An electri-

cally driven compressor may be needed for cabin cooling, for example, to provide greater

efficiency and to enable operation when engine is off. A small additional electric pump

is needed to keep transmission fluid flowing during engine stop. And power electronics

may need continuous cooling and necessitate an additional electric pump. And there are a

whole lot of functions that are essentially provided as peripheral product of the combus-

tion engine’s inefficiency that need to be otherwise met. Cabin heat for example, is pro-

vided through the engine cooling system. If the engine is stopped, an electric heater may

be needed to ensure cabin heat. Vacuum pressure is used for a great many things, so an

electric vacuum pump may be needed if

of lighter hydrocarbons

that are easier to burn—that’s gasoline (Image 1.1).

To make long carbon chains of crude oil into smaller chains that can be burned by

our engine, we need to break them apart into smaller chains, a process called cracking.

Cracking entails putting these long crude oil molecules in a tall tank, applying heat,

pressure, and a chemical catalyst to encourage a reaction, and breaking them up into

smaller chains of carbon. The sweet spot for our purposes is 4–12 atoms long for gasoline

or 13–20 atoms for diesel. So, gasoline is not really a single chemical compound, it is an

irregular mixture of different lighter and heavier compounds, with chemical character-

istics that vary.

As we start thinking about putting that gasoline in our engine, it’s useful to have a

sense of how the fuel performs, or more precisely how easily it ignites. Not all fuels are the

same. Some might not ignite very easily at normal pressure and temperature. This is the

case with diesel, for example. Others might be more volatile and ignite easily, like butane

lighter fluid. If a fuel is heated or put under pressure, it might even ignite on its own, called

autoignition. If this happens before we want it to in an engine, the result can be very bad.

To help us manage this potential problem, we’ve developed a way to compare the ten-

dency of a fuel to ignite on its own when put under pressure. To do this, we compare it to a

set standard, a defined hydrocarbon chain length that we can use to measure and compare

other hydrocarbons. Since the molecules in gasoline are typically 4–12 carbon atoms long,

we can use a strand that’s eight atoms long as a good comparison. Because a hydrocarbon

with eight carbon atoms is called octane, we call this the octane rating, and use octane’s

ability to withstand compression before spontaneously igniting as a guide. If a given fuel

can be squeezed to 85% of the value of octane before autoignition, we give it an octane rating

of85. If we can squeeze it more, to say 105% of octane, we give it a rating of 105. Prettysimple.

IMAGE 1.1

Hydrocarbon.

A chain of carbon atoms (black) and hydrogen atoms (white) make up a molecule of gasoline. This particular

molecule has eight carbon atoms, so it’s a molecule of ‘octane’. Crude oil is comprised of much longer molecules

that must be broken, or ‘cracked’, into smaller components to make gasoline.

3Bringing the Fire

So, the commonly held misbelief that higher octane fuel contains more energy or power is

simply not correct.

Moving beyond the simple notion of an idealized perfect combustion, what does the

combustion of gasoline really look like? Since gasoline is actually a mixture of many dif-

ferent hydrocarbons, the answer’s tricky. But for the sake of simplicity, let’s pretend it’s all

pure octane. A chemist would write the resulting equation like this:

[ ]+ + → + + +

( )

25 O 3.76N 2C H 16CO 18H O 94N heat2 2

25 units of air oxygen and nitrogen

8 18

gasoline

carbon dioxide

water

nitrogen from the air

This looks complicated, but it’s really not. Notice the eight under the ‘C’ on the left; that

tells us it’s octane. Simple chemistry tells us that a complete reaction, in which every mol-

ecule of gasoline is combined with the right amount of oxygen, would take place with a

mixture ratio of 14.7 to 1. This means that under normal conditions 1 g of gasoline can

combine with 14.7 g of air to produce a perfect oxidation reaction. This ratio is called the

stoichiometric ratio. It’s often represented by the Greek letter lambda—λ. When the mix-

ture has more fuel than the stoichiometric mix, λ is less than 1, and when there’s more air

than the stoichiometric ratio, λ is greater than 1.

But this ideal ratio isn’t always ideal. For example, providing a rich mixture, with more

fuel than the 14.7:1 ratio, can offer more power or easier ignition, since there is more gaso-

line available. At cold start-up, for example, an engine might run more smoothly with a

ratio of 12:1 (or λ = 0.8) until it warms up. Or under high loads or high acceleration, we

may also want a richer mixture to provide more power. Alternatively, when driving on a

highway without much need for power, we would be using much more gas than we need

with such a rich mixture, so we could use a lean mixture with less gasoline to improve

fuel economy, say 24:1 (or λ = 1.55).

An additional complication is that we’re not providing pure oxygen to this reaction,

we’re proving air. While air is about 21% O2, it’s 78% nitrogen, or N2. So far, we’ve ignored

the nitrogen because it’s not part of the oxidizing reaction. But, if the heat and pressure

of combustion rise too high, as a result, for instance, of heavy torque demand, nitrogen

molecules break apart and combine with oxygen, usually producing nitrogen oxide (NO),

but also a bit of NO2, collectively called nitrogen oxides or NOx. Nitrogen oxides are a

primary cause of smog and when combined with water in the atmosphere, can form nitric

acid, a cause of acid rain. Similarly, if we’re providing more fuel than can be effectively

burned, some gas will slip through without burning; we call these unburned hydrocar-

bons (UHCs). An ongoing challenge then is to keep the heat of combustion controlled and

the mixture correct to minimize these undesirable pollutants.

The Engine

The basic components of an internal combustion engine are simple: a cylinder that’s just

a long tube closed on one side and open on the other, and a piston that can slide up and

down in that cylinder. These two components define the combustion chamber. We add a

couple of valves that open and close to let us deliver air and fuel into the cylinder and let

exhaust out. And we can add a means of mechanical connection to the bottom of the pis-

ton, so when it moves up and down, it causes a shaft to rotate (after all, what we’re after is

rotation) (Image 1.2).

4 Automotive Innovation

The piston is connected to a connecting rod by a pin (a piston pin), and the connecting

rod is connected to an off-center, or eccentric, lobe on a shaft. When the piston goes up and

down, the shaft rotates. Rotating that crankshaft is the whole purpose of the engine. So, all

we need to do now is make the piston go up and down; that’s where combustion comes in.

A combustion event in the cylinder when the piston is all the way up will push the piston

down and turn the crankshaft. To keep that process going, in an even and regular cycle so

that the shaft spins evenly and with power, we’ve defined a four-stroke process.

The Four Strokes

The basic up and down movement of the piston is defined in four strokes; that’s why we

call it a ‘four-stroke’ engine. Each stroke is one full movement of the piston up or down. So,

in four strokes, the piston has gone up and down twice. Let’s look at each of these strokes.

IMAGE 1.2

Basic engine components.

A basic internal combustion engine has some key primary components: a piston, a cylinder, a cylinder head,

valves in the head, a connecting rod, and a crankshaft. In this case, our simple engine has dual overhead cams

that operate the valves.

Image: Richard Wheeler

5Bringing the Fire

We can begin with the intake stroke. During this stroke, the piston moves from the top

of the cylinder to the bottom with the intake valve open. This allows an air and fuel mix-

ture to be drawn into the combustion chamber as the crankshaft turns a half rotation. We

call that incoming air–fuel mixture a charge (Image 1.3).

As the crankshaft continues to turn, the intake valve closes and the piston is pushed up,

compressing the charge. This is called the compression stroke. This adds pressure to the

fuel–air mixture, resulting in a rise in heat, preparing the fuel for combustion.

With the fuel–air mixture compressed, and the piston nearing the top of the compres-

sion stroke, a spark plug is used to create a small electric arc that ignites the mixture. With

combustion initiated, a pressure

the engine is going to be shut down while driving.

A simple system might stop the car when the brake application indicates an imminent

stop and restart the engine as the brakes are released. The process needs to be seamless

and quiet, to preserve drive quality. With a low cost, good consumer acceptance, and a

6 C. Guile, “The Effect of Vehicle Electrification on Transmissions and the Transmission Market.” CTI Magazine

December, 2016.

136 Automotive Innovation

good return on investment in fuel economy, the application of basic start–stop systems is

growing, and now used in a wide array of vehicles, not all of which are true hybrids.

Defining a true hybrid entails incorporating regenerative braking capacity. The idea

is so simple, it’s brilliant: capture the excess kinetic energy when braking, and use that

energy to augment acceleration when the vehicle speeds back up. Braking is unavoidable,

of course; but it is also a significant form of energy waste. If you recall the basic equation

for the energy of a moving body, kinetic energy is defined by half the product of the mass

and the square of the velocity. So, with increasing vehicle speed, the energy available for

recovery (or the energy wasted) increases exponentially. Each time the brakes are applied,

valuable forward movement, paid for with expensive fuel and hazardous emissions, is

converted to unwanted heat. In fact, the heat is so excessive and unwanted we design

brakes to dissipate it quickly, literally tossing energy to the wind. Typically, regenerative

braking allows us to capture about a third of that energy and; as a result, in city driving

regenerative braking alone can increase fuel efficiency by about 20% (Image 5.3).7

The regenerative braking system is integrated with friction brakes to ensure braking

performance while maximizing recovered energy. So, with a two-wheel drive vehicle,

regeneration is limited to one drive axle. A brake-by-wire system that replaced hydrau-

lic actuation with electromechanical actuators simplifies the mixing of braking elements,

called blended braking. How much energy is actually recaptured depends on a variety

of factors, including motor and battery capacity, the current state of charge (SOC), and the

rate of braking demanded. In some vehicles, driver input can often vary the regenerative

braking effect. In the Chevy Bolt electric vehicle, for example, a ‘regen’ paddle allows the

driver to apply strong regenerative braking, enough to bring the vehicle to a complete

stop in fact; or drop the transmission into Low to allow what Chevy calls ‘one pedal’ driv-

ing with strong regenerative braking whenever the accelerator is released. Of course, in

7 G. Xu, W. Li, K. Xu and Z. Song, An intelligent regenerative braking strategy for electric vehicles. Energies

4, 2011, 1461–1477; and Cost, Effectiveness, and Deployment of Fuel Economy Technologies for Light-Duty Vehicles.

Committee on the Assessment of Technologies for Improving Fuel Economy of Light-Duty Vehicles, Phase

2; Board on Energy and Environmental Systems; Division on Engineering and Physical Sciences; National

Research Council, National Acadamies Press, Washington, DC, 2015.

IMAGE 5.3

Regenerative braking.

Regenerative braking recaptures a portion of the car’s kinetic energy during deceleration, typically using it to

drive an electric machine so the energy can be saved in a battery and used to supplement acceleration later.

137Electrified Powertrains

the future such features could be applied automatically; but the opportunity for driver

engagement and variable driving modes is considered desirable, particularly as the driv-

ing experience moves away from the conventional.

Defining a modest hybrid system, with start–stop and regenerative capacity, doesn’t

require a total redesign of the powertrain. This makes it an attractive option for carmakers.

A modest version of this might just rely on a hardy starter that can also function as a genera-

tor (Image5.4). General Motors’ belt alternator starter (BAS) system offers a useful example.

Connecting a robust starter/alternator through a high-tension serpentine belt to the crank-

shaft offers a clever way to address braking waste without costly modifications to the ICE or

drivetrain. The belt allows two-way transmission of mechanical power; so power can to be

delivered to the crankshaft and drawn from the crankshaft. A BAS system typically uses a

36–48 volt machine for starting as well as recuperative braking and torque assist. However, the

high torque required at cold starts can often exceed the friction limits of the belt causing slip;

so, the conventional 12-volt starter is retained. Improved belt connections are being developed

to address this; though the BAS system has received limited praise and is unlikely to continue.

Mild Hybrid

A somewhat more aggressive incorporation of electric drive is defined by what is typically

called a mild-hybrid system. By scaling up the coupled electric machine and improving

the mechanical interconnection, an internal combustion engine can be downsized signifi-

cantly and rely on the attached motor to provide augmented torque. Regenerative capacity

is no longer limited to reducing the accessory load and supporting a start–stop capacity, it

can now add some tractive force to the drivetrain. Still, in a mild application, the electric

motor is limited to a supporting role, with no exclusive electric-drive mode. Nevertheless,

IMAGE 5.4

48-volt boost recuperation machine.

This compact boost recuperation machine produced by SEG Automotive provides 12 kW generating power and

up to 40 lb ft (55 N m) of torque. Small electric machines can allow for simple, low-cost hybridization of a con-

ventional ICE with minor modifications to the electrical system and engine, and significant improvement in fuel

economy and engine starting, while enabling the advantages of a 48-volt system.

138 Automotive Innovation

this provides load leveling and allows for more complete regenerative braking that can

apply recovered power to assist the ICE. Start–stop can kick in at idle and potentially while

moving for sailing capacity.

Such functions require a 48-volt system, but we should note that the presence of a 48-volt

system does not necessarily mean the use of a hybrid drive. For example, Porsche relies

on a 48-volt system solely to power the active roll stabilization system on its new Cayenne

(More on this in Chapter 7). Audi offers a more complete example of a 48-volt mild-hybrid

system at work. The carmaker taps a 48-volt system for a mix of efficiency and power, des-

ignating the 48-volt system as primary and using a 12-volt system for low-load accessories

such as the lighting and audio. The 48-volt system uses a belt-driven motor to provide

seamless starting when the engine is warm, offers load leveling acceleration, some sailing

capacity at cruising, and takes on some of the accessory load. The motor and particularly

the power electronics receive active cooling from the engine cooling system. The fuel sav-

ings is modest; however, the system is relatively light and offers smooth operation and

some efficiency gains, all while maintaining power performance. It is now incorporated

into all of Audi’s A8 variants (Image 5.5). In the diesel SQ7, the 48-volt system is used to

operate a modest supercharger that addresses turbo lag.

Such systems may become more common soon. The growing need for higher voltage

capacity to service increasing accessories raises the attractiveness of this simple option.

IMAGE 5.5

Audi mild hybrid.

Audi has announced a plan to incorporate some variant of this 48-volt technology in every car it makes.

Image: Audi

139Electrified Powertrains

Now an established technology, versions of 48-volt systems have been adopted by numer-

ous manufactures; and multiple 48-volt hybrid systems are available as low-cost add-ons

by industry suppliers. So, the impact on fuel economy is modest,

wave begins at the spark plug and rapidly travels through

the combustion chamber. This defines a flame front that pushes strongly down on the pis-

ton, adding energy to the rotation of the crankshaft. This is the power stroke. Later, we’ll

see that we can design engines differently to control and define the propagation of that

flame front through the combustion chamber; in fact, this is a critical element of current

engine research and innovation (Image 1.4).

At the end of the power stroke, the crankshaft has built rotational momentum that will

push the piston back up. The exhaust valve opens, and as the piston moves up, it pushes

out all burnt charge from the combustion chamber, defining the exhaust stroke and allow-

ing the process to begin again. The high-energy exhaust gas moves quickly, initially pass-

ing the exhaust valves at more than 1,500 mph, and marking a significant variation from

the relatively slow-moving incoming charge during the intake stroke. This is why intake

valves are often larger than exhaust valves, as the slow-moving intake air needs a larger

passage to fill the chamber (Image 1.5).

As these four strokes repeat themselves, they define the basic operation of the internal

combustion, four-stroke engine. The fundamentals are pretty simple: suck in a charge,

IMAGE 1.3

Intake and compression stroke.

As the piston moves down the cylinder, the combustion chamber expands, drawing in an air–fuel mixture, or

charge, through the open intake valve as the exhaust valve remains closed. Subsequently, as the air–fuel mix-

ture is compressed, both pressure and temperature rise, preparing the fuel–air charge for combustion. This is

why the octane rating of our fuel is so important. If the octane rating is too low, the fuel might autoignite before

we’re done with the compression stroke, we call that preignition. Or if the rating is too high, the fuel might not

quite be ready for ignition when we want it to be.

Image: Richard Wheeler

6 Automotive Innovation

squeeze it so it’s ready for combustion, ignite it and release energy in the form of heat and

pressure, and clear the cylinder to start the process again.

A diesel engine operates in basically the same manner, but with a few important differ-

ences that are worth noting. Diesel fuel is less volatile than gasoline, and thus harder to

ignite. So, to achieve ignition, we need higher pressure and higher temperature. So, diesel

engines achieve a much higher pressure during the compression stroke, which also pro-

duces a much higher resulting temperature called the heat of compression. While gaso-

line engines reduce volume by about ten to one on average, diesel engines double that,

compressing volume by a ratio of 15–20 to 1. (We’ll talk about the important of compres-

sion ratios in the next chapter.)

At this high pressure, diesel fuels will spontaneously ignite. So, to get a precise igni-

tion point, we can’t have the fuel in the cylinder during the compression stroke or it’s

likely to ignite unpredictably. Consequently, in the intake stroke, the engine takes in only

air, the diesel fuel is then injected into the combustion chamber at the end of the com-

pression stroke, just before the piston reaches the top of its stroke, or top dead center

(TDC). The tremendous heat causes the injected spray to vaporize and ignite quickly, and

a rapid pressure rise occurs by the time the piston reaches just past the top of thestroke.

IMAGE 1.4

Flame front.

The flame front can be visualized as an arc of combustion that propagates away from the point of ignition at

the spark plug and travels through the charge in the chamber, quickly increasing the heat and pressure of the

cylinder and pushing the piston downward. A moving wall of combustion propagated by heat, such as this is,

is called deflagration. Defining the character and speed of this deflagration, and the way it travels through the

combustion chamber, is a central concern of internal combustion innovation.

Image: Rich ard Wheeler

7Bringing the Fire

Unfortunately, spraying diesel fuel late in the compression stage means the fuel mixture

will remain uneven in the chamber when ignition begins, with some lean and some rich

areas. Those rich areas cause more soot (called particulate mater or PM) and the lean

regions cause more nitrogen oxides than a gasoline engine. This typically means a clean

burning diesel requires more complex (and expensive) emissions treatment than its gaso-

line counterpart.

Still, a typical diesel engine has an advantage in economy and torque. Relatively slow

burning diesel operates at a lower speed range than gasoline, but can compensate with

a longer piston travel, or stroke. It is able to take advantage of the resulting leverage

to produce more rotational force or torque. A gasoline engine operates with a shorter

stroke, but higher engine speed, or rotations per minute (rpm), capacity due to the faster

burn rate of gasoline. It’s often assumed that the greater power and efficiency of diesel

is fully explained by the higher compression ratio. And certainly higher compression

allows for higher thermal efficiency. But, a tremendous advantage in efficiency comes

from the fact that varying the injection of the fuel can control engine speed, so die-

sel engines do not need to throttle the air supply to control engine speed like gasoline

engines do. So, unlike conventional gasoline engines, diesels do not expend energy by

drawing air through a restricted intake channel at low throttle speeds, a power reduc-

tion called pumping loss. This and the higher thermal efficiency explain why a diesel

engine can typically achieve about 30% greater efficiency than a gasoline counterpart.

With a lower rpm range, and a heartier construction to handle the increased pressure,

diesel engines also tend to have a longer life than their gasoline counterparts. Of course,

as diesel engines are made smaller and faster to suite automobile applications, and as

gasoline engines borrow ideas from their diesel brethren (more on this later), these dis-

tinctions are less absolute than they once were.

IMAGE 1.5

Exhaust stroke.

The angular momentum of the engine helps push the piston back up

the cylinder. With the exhaust valve now open, this expels the burnt

air–fuel mixture, allowing the cylinder to begin the process again

with another intake stroke.

Image: Richard Wheeler

8 Automotive Innovation

The Engine Comes Together

So, we now have a basic sense of the principal components needed to make all this come

together: cylinders, pistons, valves, connecting rods, a crankshaft, ignition, combustion,

momentum, and all the rest. Let’s look more closely at the components we need to make

this real.

The development of angular momentum keeps the piston moving between power

strokes and the crankshaft rotating evenly. To help maintain this momentum, we put a

large disk on the crankshaft called a flywheel. This disk with a large radius will increase

what physicists call the moment of inertia of the moving parts. Because we want to do

this with the minimal mass necessary (so we don’t slow down the engine), flywheels tend

to be thin disk with a large radius. The inertia is defined by the mass and the radius

squared (i = mr2). So, a small increase in radius can allow a larger decrease in mass. The

flywheel will also provide a contact point for the transmission and a convenient engage-

ment gear for the electric starter. While the flywheel allows for smoother operation due to

the increased moment of inertia, for the same reason it also takes more energy to get the

engine spinning. So, a common performance upgrade is to swap out a heavy flywheel for

a lighter one that will allow for faster engine acceleration.

Because the flywheel rotates at a near-constant speed, and pistons apply torque to the

rest of the crankshaft with each firing impulse, the shaft can whip slightly due to the tor-

sional force of each crank throw. On longer, six-

and eight-cylinder engines, the result can

throw off the valve timing mechanism at the front end of the crankshaft where the tor-

sional oscillation is most severe, or even results in mechanical failure due to severe vibra-

tion. To prevent this, a smaller flywheel with a rubber hub is attached to the front end of

longer crankshafts. While the flywheel favors a constant speed, the rubber serves to absorb

irregular rotational energy of the shaft, operating as a torsional oscillation dampener, or

harmonic balancer, to absorb potentially damaging oscillations (Image 1.6).

IMAGE 1.6

Crankshaft and flywheel.

The large flywheel attached to the right end of the crankshaft increased the assembly’s moment of inertia, thus

improving it’s rotational momentum. Along with the harmonic balancer on the left, this produces more bal-

anced and even operation.

Image: Jaguar MENA

9Bringing the Fire

Besides inertia, the other element that keeps our engine functioning smoothly is the fact

that most engines have more than one piston. So, when one piston is in the compression

stroke, and so not producing any rotational energy, another piston is in the power stroke,

pushing the crankshaft around. Engines can have any number of cylinders, from 1 to as

many as 12 or more. In fact, Cadillac produced a concept car in 2003 called the Sixteen that

had, you guessed it, 16 cylinders. The cylinders are made to begin ignition in succession

at even intervals, called the firing order. This allows for the cylinders to work together

keeping the crankshaft rotating more uniformly, rather than abruptly jerking with each

power stroke or slowing between strokes. It can also allow us to ensure that two adjacent

cylinders aren’t firing at the same time, since they’d both be trying to draw in a charge

simultaneously, potentially causing one to draw fuel and air away from the other, a prob-

lem called induction robbery.

Holding all this together is the main structure of the engine, called the engine block.

An engine block can be made of cast iron, aluminum, or more advanced materials, but it’s

basic character is the same: a large solid structure that contains the cylinders and holds

these key components in place. The bottom of the block is often called the crankcase

because it holds the crankshaft. The cylinders are usually located in the upper portion

of the block. The valves are contained in a separate component, called the cylinder head,

which attaches to the top of the block and completes the combustion chamber.

There are, of course, multiple configurations of cylinders and so multiple shapes to

engine blocks. The common V8 gets its name from the V-like shape defined by the cyl-

inders. This is a common shape because it allows a larger number of cylinders to fit in a

relatively compact package. Alternatively, with fewer cylinders, it is possible to line up the

cylinders in an in-line or ‘straight’ configuration and still fit it under the hood. Of course,

you could line up eight cylinders in a ‘straight 8’ configuration, and enjoy greater power

and smooth operation, if you have the space under the hood, but that would require a

really long hood. Smaller cars might opt for an in-line four cylinder. If the engine is cooled

by air, rather than a fluid cooling system, there’s an advantage to having the cylinders on

opposing sides of the block to achieve maximum airflow over each cylinder. This ‘hori-

zontally opposed’ configuration was used for many years by Volkswagen in its air-cooled

Beetles and buses. It’s the type of engine that powered the ridiculously cool Porsche 911

for more than three decades. And is still the norm in air-cooled small aircraft engines. In

short, there are many variations on cylinder configurations, each presenting trade-offs in

size, fit, power, and other characteristics that define the appropriate engine for any given

application.

Because the engine block contains the combustion chambers, it has to absorb a lot of

heat. To help manage the engine’s temperature, the block is designed with small chan-

nels that allow coolant to flow throughout the block (Image 1.7). The coolant that flows

through these cooling jackets then flows through the radiator, allowing for the dissipa-

tion of excess heat and proper engine temperature regulation. Similarly, in order to both

control temperature and reduce friction and wear, oil is used to provide a lubricating film

between moving parts. The bottom of the block is fitted with an oil pan, to contain this oil

as well as an oil pump to properly circulate the oil under pressure throughout the engine.

To minimize wear and heat buildup in the block, and to keep all the moving parts in

their proper location, all the spots where moving parts would rub against each other are

designed to minimize friction. The walls of the combustion chamber are a key example.

They must not only endure wear and remain dimensionally stable at high temperatures

but also provide minimal friction. Precisely defined patterns of small scratches, called

cross-hatch by mechanics and micro-asperities by engineers (who like to sound smart),

10 Automotive Innovation

provide a surface that is better at allowing oil to cling to the combustion walls and reduc-

ing friction with the piston. This texturing of the surface allows film lubrication to produce

hydrodynamic lift. Recent research indicates that dimpling may provide similar advan-

tages, so look for that as a possible future innovation.

Other junctions of moving parts, for example where the block holds the crankshaft or

camshaft, use specially designed fittings. Because these fittings ‘bear’ the force and fric-

tion required to hold the component in place, they are called bearings. The crankshaft is

held in place with main bearings, with the specialized bearings meant to hold the crank-

shaft from sliding forward called thrust bearings. Similarly, the connecting rod is linked

to the crankshaft with rod bearings. Oil is drawn by the oil pump from the pan and

pumped through channels in the block and into the crankshaft to all these bearings, so

that the moving parts do not actually rub against each other but are suspended by a pres-

surized layer of lubricating oil, called hydrodynamic lubrication. Oil is also delivered to

coat the stems of the valves, and other moving parts in the head. It is estimated that on

average 10%–30% of engine output is lost to internal friction; so, reduction of friction is a

key element of performance and efficiency innovations in modern engines.

As mentioned, the head serves to close off the upper end of the chamber, creating a

tight seal with the block, to ensure a sealed cylinder. It must also define a tight seat for the

valves, allowing them to open and close freely, while ensuring the valves seal tightly when

closed (Image 1.8). It also holds the spark plug in place and defines the basic geometry of

the upper portion of the combustion chamber. As we will see later, variations in the shape

IMAGE 1.7

Cooling the engine.

This cut-away clearly shows the cylinders as well as the oil passages and coolant jackets that channel through

the block around the cylinders to provide cooling and lubrication.

11Bringing the Fire

of the combustion chamber can have a major impact on the performance of the engine.

Because the combustion chamber needs to ensure a pressure tight seal, a head gasket

composed of multiple thin layers of metal is used at the connection of the block and head.

If this gasket fails, it is possible coolant from the coolant jackets in the block could get

into the cylinders. Since coolant cannot be compressed, this can lead to catastrophe for an

engine. When the piston comes up in the compression stroke and is met by an incompress-

ible fluid, something’s got to give.

Valve Train

The key to this whole thing working is having the valves open and close at the right time,

this is called valve timing. It’s pretty clear that the system only works

if the intake valve

is actually open during the intake stroke, or the exhaust valve is open during the exhaust

stroke. Just as obvious, if the intake valve were open during the power stroke, the result

would not be good. To ensure that everything operates in harmony, we connect the valves

to the rotation of the crankshaft through a smaller shaft called a camshaft (Image 1.9).

The camshaft has a series of eccentric lobes, one for each valve. As the shaft rotates, the

lobes will raise and lower each valve a defined amount, called lift, causing them to open

and close (Image 1.10).

The camshaft is connected to the crankshaft so that it rotates in harmony with the crank-

shaft and overall engine operation. Since each valve needs to open and close once with each

four-stroke cycle, the camshaft will need to turn once each time the crankshaft turns twice

to complete a four-stroke cycle. The connection can be made with simple gears, chain, or a

belt. The gears and chain are durable, but they can make an engine nosier (Image 1.11). The

belt is quieter, but it’s more susceptible to wear and needs to be replaced more frequently.

IMAGE 1.8

Valve seats.

The mating surface, or valve face, of intake and exhaust valves are precisely machined to provide a tight seal

with the valve seat. Precisely defined angles can be cut into the face to ensure this.

12 Automotive Innovation

The amount each valve rises is tied to the lift defined by the camshaft and is critical

to the operation of the engine. A small lift will create a small opening, and so allow less

charge to enter the chamber or less exhaust to exit the chamber. A larger opening will

increase the ability of the engine to ‘breath’ by allowing more charge in and more exhaust

out. But this could potentially outstrip the capacity of the engine, or adversely affect fuel

economy or emission by allowing more fuel into the combustion chamber than can be

IMAGE 1.9

Camshafts.

The eccentric lobes on the cams of the camshaft define the movement, or ‘lift’, and timing of the valves.

Source: ThyssenKrupp Presta Chemnitz GmbH

IMAGE 1.10

The valve train.

When the camshaft is located above the valves, it’s called an overhead cam. As the camshafts rotate, they push

on the valve stems, causing the valves to open and close in harmony with the cylinder’s strokes.

Image: Volvo

13Bringing the Fire

effectively burned. It’s not uncommon for gearheads to swap stock ‘cams’ for performance

cams that have a more aggressive profile (more lift); and this can allow an engine to pro-

duce more power. Sometimes this makes sense, sometimes not so much.

Defining the Combustion Chamber

As previously noted, the heart of the internal combustion engine is combustion; and the

shapes and materials that define the combustion chamber in turn define the combustion

itself and fundamentally define the performance of any engine. Like just about any other

component of an automobile, designing the elements of a combustion chamber has been

a game of trade-offs. You want components that are light but also strong, and those two

things don’t always go together. You want parts that are free moving when cool but also

when very hot, again at times a tough requirement. Nevertheless, new technologies are

IMAGE 1.11

Advanced chain drive.

Timing belts typically offer reduced noise when compared to timing

chains, but must be replaced before failure, as the failure of a timing

belt can mean a valve is extended into the combustion chamber when

the piston moves up. If the piston hits the open valve in its upward

stroke, the results won’t be good for either one. This advanced chain

drive by BorgWarner attempts to let you have your cake and eat it

too offering the reliability of a chain and the performance of a belt.

The system promises improved fuel economy with an inverted ‘silent

tooth’ chain.

Image: BorgWarner

14 Automotive Innovation

making these compromises less compromising, offering new possibilities that sometimes

allow us to have our cake and eat it too.

Innovations in design and material have even changed something as fundamental as

the block itself. The block may seem like a fairly simple component at first glance, but

look again. It needs to be strong enough to define the combustion chambers and hold the

engine componentry together under tremendous loads. It must be capable of being formed

and manufactured precisely, to allow for connecting points, coolant and lubricating chan-

nels that permeate the block and allow for proper lubrication and cooling of the engine.

It needs to be able to withstand and transfer intense heat, and endure high loads. And, it

needs to be as light as possible. Because of the relative mass of the block, even a small per-

centage change in weight would mean a significant change in the overall engine weight.

Happily, advances in materials and manufacturing are redefining this once-simple

component. A generation or two ago, blocks were nearly universally made of cast iron, a

durable material that had the strength to hold engine parts together, absorbed high tem-

peratures, and offer a workable, machinable material at low cost. But this is changing.

Aluminum–silicon alloys, magnesium, and advanced composites are changing the way

we think about engine blocks.

In fact, aluminum alloys have been used in engine blocks increasingly since the 1970s,

but not without some challenges. The block is more than simply a container for the other

components of the engine; it defines the major surface area of the combustion chamber.

IMAGE 1.12

Bringing it all together.

This 1.5-liter, three-cylinder, direct-injection engine by Volvo demonstrates modular design, aimed at enabling

greater powertrain options while still benefiting from economies of scale in production. As advanced as it is,

the basic components remain the same. Like any engine, it has a block, head, valves, valve train, and crankcase.

You can’t see the pistons, but they’re in there.

Image: Volvo

15Bringing the Fire

So, it must be thermally stable and able to endure significant wear. Until recently, alu-

minum blocks required the use of a steel sleeve to handle the wear of the piston sliding

against the cylinder. Because this steel sleeve adds to the weight and size of the engine,

numerous alternatives have been used, with varying success. Most of these approaches

try to remove the aluminum near the surface of the cylinder electrochemically; this leaves

a hardwearing silicon layer to define a more capable cylinder wall without adding the

weight of a steel sleeve. While the relative merits of various techniques that accomplish

this are debated, all present some challenges. Another option might be the installation of

a removable cylinder liner that is designed to have coolant flow around it, what’s called a

wet liner. While popular with a few French manufacturers, these still present a significant

increase in engine weight and size, diminishing the intended aim of using an aluminum

block. Recent work with low-friction plasma metal coating may signal the possibility of

aluminum or magnesium blocks with no steel liner, making it much lighter and smaller.

The use of this new technology on the GT500 Shelby Mustang cut the engine weight and

actually used recycled engines to provide the raw material.1 Called Plasma Transferred

Wire Arc (PTWA) technology, the process blows a fine mist of molten steel plasma onto the

cylinder walls to create a hard-wearing finish without a cylinder liner.

As strong as aluminum but lighter, magnesium shows real promise as an engine mate-

rial. This is not a new idea; Volkswagen and Porsche manufactured engine blocks out of

magnesium back in the 1960s, but not without challenges. In fact, more recently, when

BMW chose to use magnesium in the engine block of its N53 engine, they had to couple

it with an aluminum insert forming the cylinders and coolant channels. A newer mag-

nesium alloy, with the catchy name AMC-SC1,

is designed specifically for engine blocks

and may offer a block that is lighter than aluminum but with greater strength and similar

manufacturing requirements and cost.2 This isn’t exactly a fast-track technology; but don’t

be surprised if you see more magnesium powertrain components in the future.

Perhaps more promising are advanced metal composites or metal matrix composites

(MMCs). Combining a metal-binding agent, or matrix, as the major component with a

reinforcing ceramic, organic, or another metal can allow us to precisely engineer specific

characteristics into materials to suit certain purposes. Compacted graphite cast iron (CGI)

is a promising example of this in the search for a superior block material. CGI has been

used previously in the manufacturing of brake components that are strong, lightweight,

low wear, and able to transfer heat better than previous materials. Since these are all desir-

able characteristics of an engine block, the fit seems right. Also called vermicular graphite

iron, CGI used thicker graphite particles than exist in typical cast iron (all iron contains

some graphite). These particles define thick tentacles of graphite that create a tight inter-

woven bond with the surrounding iron matrix. So, while more dense than aluminum, the

resulting superior strength of CGI means it can be made thinner, resulting in a competitive

weight with greater thermal conductivity and excellent internal dampening. The result

is a block that is more compact, up to 75% stronger than gray iron and five times more

fatigue resistant than aluminum.3 Recent advances in manufacturing may allow for the

more widespread use of CGI and similar MMCs, as some have seen limited use due to low

machinability. But high cost is the number one barrier.

1 “Ford Developed High-Tech Plasma Process that can Save an Engine from the Scrapyard while Reducing CO2

Emission.” Ford Media Center December, 2015.

2 C.J. Bettles et al., AMC-SC1: A New Magnesium Alloy Suitable for Powertrain Applications. SAE Technical

Papers, March, 2003.

3 P.K. Mallick, Advanced materials for automotive applications: An overview. In J. Rowe (ed) Advanced Materials

in Automotive Engineering. Woodhead Publishing, Cambridge, 2012, 5–27.

16 Automotive Innovation

Some have explored a far more surprising composite material for an engine block:

carbon fiber. Basically a plastic matrix with long strands of engineered carbon added

for strength, carbon fiber is usually associated with low heat and modest strength. (The

base material is plastic, after all.) However, a carbon fiber block for specialized racing

has seen some preliminary testing.4 This combination of plastic reinforced with carbon

strands is many times lighter than iron, and about half the weight of an aluminum block.

It would be an exceptional material for a block if it could handle the heat and force. If

these issues can be worked out, we may see composite materials used for bocks in pro-

duction cars sometime in the future. And if not, this still gives us an impressive example

of the sorts of innovations being pursued. We’ll look more closely at advanced metals and

composites in Chapter 8.

Pistons

The piston is clearly a central component in any engine. After all, the piston defines the

core function of the engine, converting combustion into rotation. So, even small variations

in the shape, weight, and design of this component alter the performance of an engine in

fundamental ways. Today’s engines are smaller, hotter, and faster than their predecessors.

And today’s piston needs to endure tremendous pressure—over 6 tons of force every two-

tenths of a second at 6,000 rpm—and intense heat—regularly over 500°F—and convert

it as efficiently as possible into linear mechanical movement. And it must be engineered

precisely enough to make an exact fit with the cylinder wall, but not so tight that it can’t

slide. In addition it needs to maintain this precision fit under significant heat variations.

Ifit’s not obvious, achieving all of this is not easy.

At the dawn of the automobile, pistons were large and made of cast iron. With the metal-

lurgy of the time, only a bulky chunk of iron was thought capable of absorbing the force

and wear inside an engine. However, early on it was recognized that iron’s poor ability to

shed heat led to high engine head temperatures and a resulting problem in the combus-

tion chamber as the air fuel mixture expanded with the heat. This expansion meant a less-

dense fuel–air charge and a resulting reduction in the power the engine could produce. So,

by the 1920s, aluminum was the common material for pistons since it has more than three

times iron’s ability to move heat.

To make this a bit tougher, like any oscillating or rotating engine component, piston weight

is a critical element of performance. Because changing the momentum of fast- moving parts

requires significant energy, any weight saved on moving parts such as the crankshaft, rods,

valves and pistons is far more definitive than the same weight reduction on the block or

other stationary parts. At high rpm, a typical piston can accelerate to 50mph, back to zero,

and back to 50 mph in the opposite direction in less than a millisecond. Heavy parts take

more energy and time to get moving and come to a stop, so putting these parts on a diet is

good for acceleration, maximum operating speed, and fuel economy.

As a result, today’s pistons are smaller, lighter, and more durable and precisely engi-

neered than ever. More precise manufacturing has allowed for more exact and complex

head shapes, stronger asymmetric lower walls to the piston, called skirts, and much

4 D. Sherman, “Is This the Engine of the Future? In-depth with Matti Holtzberg and His Composite Engine

Block”. Car and Driver May 6, 2011.

17Bringing the Fire

thinner walls, all while maintaining strength (Image 1.13). Similarly, reducing the piston

mass and improved alloys has also addressed the challenge of thermal expansion, allow-

ing much tighter tolerance, or closer fit, within the cylinder.

Managing piston heat is no easy trick. The piston is exposed to the full heat of combus-

tion most directly, and unlike the block, doesn’t have a large mass to help absorb and

dissipate that heat. What’s more, because all the parts of these lighter and thinner pistons

don’t heat up at the same rate, thermal expansion also requires exact asymmetries and

tappers that allow the maintenance of precise clearances even as extreme and uneven heat

is applied.

Already thinner and lighter, these new pistons need to be made stronger as well. With

rising engine speeds and pressure, ring grooves are hardened with anodizing or laser

hardening to improve strength. The fitting of cast iron or steel insert to reinforce the top

ring groove, a practice often used in diesel engines, is now being considered for aluminum

gasoline pistons. Taking another cue from diesel pistons, high-performance gasoline pis-

tons are now being manufactured as two fused parts to allow for cooling channels in the

piston that can help move heat from the head to the ring pack. Sloshing oil in these cool-

ing galleries creates a cooling effect, sometimes called co*cktail shaker cooling, which can

dramatically reduce piston thermal loading.

Designing a high-performance piston requires that we also think about the friction and

wear on piston surfaces. The sliding of pistons against the cylinder walls is the largest

bearing surface in the engine that is not fed by pressurized oil, and this alone can account

for 5% to nearly 8% of your fuel use.5 This friction loss is as great as the crankshaft and

valve train friction loss combined. So, to reduce energy loss from friction, skirts are now

5 C. Kirner, J. Halbhuber, B. Uhlig, A. Oliva, S. Graf and G. Wachtmeister, Experimental and simulative research

advances in the piston assembly of an internal combustion engine. Tribology International 99, 2016, 159–168.

IMAGE

Automotive innovation the science and engineering behind cutting edge automotive technology - Engenharia Mecânica (2024)