As the aim of any car from a performance standpoint is to be able to go, stop, and handle, "the purpose of this book is to explain some of the chassis engineering aspects related to the design, build, and testing of high performance automobiles" with respect to a number of different subtopics in this particular arena. Such subtopics covered in the book include tire characteristics, weight distribution and dynamics, roll angle and roll force distribution, bushings and deflections, springs and shocks, types of front suspension, designing and building a front suspension, live axle vs. independent rear suspension, building rear suspensions, frame design and building, aerodynamic downforce, rotating inertia, and vehicle testing and tuning. As this essay is being posted on a site for the Toyota Tercel and Paseo, I will be focusing on material relevant to what I know about Toyota's EL chassis and steering comments in that direction. Text with +1 size and bold is the chapter topic. Italics are usually just italic because they were in the book. Outside of a quote it's just for me to stress an idea.
Just to make sure everyone's on the same page, I'd like to include the following definitions: Throughout the essay (and in the racing world as well) use of the word quick refers to acceleration and use of the word fast refers to a specific velocity. In the introduction of the book, Adams describes "good handling" as "going around corners faster while improving driver control." Going and stopping are a bit more straight forward. Going has to do with accelerating quicker and attaining a faster top speed. Stopping in this case could mean slowing or completely stopping. In both cases (going, stopping), as with handling, driver control is an extremely important consideration.
"Because the tires are the only link between the car and the track, they are the key to improved handling. Almost all of the suspension variables are related to how well a car's tires react to the ground." That said, many people (myself as well at one point) fail to remember that "the tires on your car have more effect on its handling than any other component." Adams simplifies the analysis of tires to "their input-output characteristics" with the input being the vertical load, the constantly changing weight on the tire, and the output being traction, how well the tire sticks to the ground which "determines how fast a car can accelerate, brake, and/or corner." With any tire, the performance curve (relationship between vertical load and traction) will result in "a smaller increase in traction as the vertical load is increased. We call this loss of relative traction a loss in the tire's cornering efficiency." Unfortunately, you'll have to buy the book in order to see the graphs, and pictures, but I'll do my best to represent the charts here. It's important to note that since these charts originate from the book, they are not my own work.
| Chart 1-1 Vertical Load vs. Cornering Efficiency |
||||
| Vertical Load (lbs) | Traction available (lbs) | Traction / Vertical Load | Factor | Cornering Efficiency |
| 500 | 700 | 700 500 |
1.40 | 140% |
| 1000 | 1000 | 1000 1000 |
1.00 | 100% |
| 1500 | 1250 | 1250 1500 |
.83 | 83% |
| 2000 | 1500 | 1500 2000 |
.75 | 75% |
A car's, well, a tire's lateral force (measured in g's) is equal to the "factor" so "you can see that with a cornering efficiency of 140%, it would be possible for a car to corner at 1.40 g's. When the cornering efficiency is only 75%, the same car would corner at .75 g's." This is why it's important to have a car as lightweight as possible no matter what the suspension design. Adams notes that "tire factors such as contact patch, tread depth, aspect ratio, etc., must be considered because they... raise or lower the traction curve and they can cause the shape of the curve to change." The alignment settings, camber, caster, and toe, all affect the contact patch. When a car is sitting on an alignment rack (or in the driveway really), the static alignment is set. These static settings change as a car is accelerating forward and back or side to side due to the forces of weight transfer. While the car is in motion these changing alignment settings are called dynamic rather than static. After talking briefly about camber, Adams goes on to talk about the "circle of traction". I'll just say that using the above tire performance curve as an example, say the tire had 500lbs of vertical load, then there would be 700lbs of available traction. This 700lbs can be used entirely to accelerate the car, entirely to decelerate the car, or entirely for cornering. However, in real world track or road conditions, this is only sometimes the case. Often the car is asked to do a combination of cornering and acceleration/deceleration. That 700lbs of traction the tire has to use is split up between the tasks that are asked of it. If you're cornering at the car's capacity and you hit the gas, that's asking it for more than the 700lbs of traction it has available and the tire will brake traction. On a rwd car, this means power oversteer at corner exit and for a fwd car, it'd mean power understeer due to which set of tires is losing traction. Another common scenario is when braking to go into a turn. Most of the car's weight is on the front tires for the purpose of decelerating. If you're already at the car's braking threshold, you can't turn the steering wheel and expect those front tires to keep traction. Braking that (in this case 700lbs of traction) threshold will result in serious understeer or even complete loss of control of the front end. You have to back off of the brakes a little so that the tires have more of their total available traction free to start the turn. This leads into the concept of trail-braking, but that's another story and it's not even included in this book anyway. Adams continues with a discussion on g-forces. "For instance, a 3000lb cornering force acting on a 3000lb car would be a 1.0 g load. By describing cornering forces in g's, various cars can be compared equally, regardless of their individual weights." The chapter on tire characteristics finishes off by touching on the topic of skidpad testing. Instead of describing the process, I'll just insert the equation so that people with a stopwatch, calculator, tape measure, and a couple of cones can go out and... well... nevermind. ;) g=1.225xR/(T^2) where R is the radius of the turn in ft. and T is the time required to complete the circle in seconds.
"A car's weight distribution is determined by how much weight is on each tire. These weights change due to load transfer. The changes in loading are the result of forces acting on the car." Splitting up load to each individual tire, Adams' first example consists of a car weighing 3000lbs. Front to rear weight bias is 50/50 as is side to side bias. Load transfer from cornering in this theoretical example is assumed to be 0. Using the same tire performance curve from the first chapter:
| Chart 2-1 | ||
| Tire Location | Static Weight on Tire | Traction Available |
| Left Front | 750 lbs. | 850 lbs. |
| Right Front | 750 | 850 |
| Left Rear | 750 | 850 |
| Right Rear | 750 | 850 |
| Totals | 3000 lbs. | 3400 lbs. |
"This sounds pretty good until you realize that the weight will transfer from the inside tires to the outside tires as the car develops cornering force going around a corner." Example 2 begins to illustrate this concept as Adams ads the calculation: Lateral Weight Transfer = WH/T where T is the track width (distance between left and right tire centerlines) in inches, H is the height of the center of gravity in inches, and W is the overall weight of the car. In Adams' example we assume that the car has a track width of 60" and the height of the center of gravity is 20". Therefore, the lateral weight transfer would be 1000lbs. "On a car with equal front-to-rear weight distribution, this lateral weight transfer would be split evenly to 500lbs on each axle."
| Chart 2-2 | ||||
| Tire Location | Static Weight on Tire | Lateral Weight Transfer | Weight on Tire During Cornering | Traction Available |
| Left Front | 750 | -500 | 250 | 450 |
| Right Front | 750 | +500 | 1250 | 1130 |
| Left Rear | 750 | -500 | 250 | 450 |
| Right Rear | 750 | +500 | 1250 | 1130 |
| Totals | 3000 | 3000 | 3160 | |
On circle track cars, that only turn left, "one way to help equalize the weight on the tires during cornering is to preload the inside tires. This is done by moving some of the weight from the right side of the car to the left side of the car." Adams' example 3 gives the car a 600lb left side weight bias bringing the total cornering force back up to 1.13 g's. As this doesn't concern any Tercel I've ever heard of, I'm not going to waste time. Example 4 is done with a 60/40 front to rear weight bias and no left to right bias making the static weight on the front tires 900lbs each and the weight on the rear tires 600lbs each. When we calculate the weight transfer of each axle independently, 600lbs is transferred in front and 400lbs in the rear. The total traction available would be 3130 and so with the weight of the car being 3000lbs, the total average cornering force would be 1.04 g's. "This total is misleading, because if you look at just the front-end weights and traction forces in Chart 2-4 (which I have neglected to include), you see that there is 1750lbs of traction for pulling the front-end weight of 1800lbs around corners. The front-end cornering force would then be: 1750/1800 = .97 g's. At the rear... there is 1380lbs of traction to pull 1200lbs of weight around corners. This means the rear cornering force would be: 1380/1200 = 1.15 g's. This analysis shows that the car in this example will not only corner slower than one with equal front-to-rear weight distribution, but it will also understeer in the corners. [...] Even more important is to note that although the total traction will permit the car to corner at 1.04 g's, the front-end won't corner at over .97 g's. This means that because of understeer, the car can only corner at .97 g's." Example 5 combines 2, 3, and 4. Example 6 is the same as example 5, "except the chassis is wedged by adding 200lbs of weight to the right rear tire... to cure understeer." Wedging is rather like putting a book under one table leg of a perfectly level table. Adds weight to said corner as well as the diagonally opposite corner and takes weight off of the remaining two corners. Outside of that, since these examples pertain to circle track cars, I'm going to ignore them. At the end of the chapter, summary points 1, 3, and 5 are good to recap: 1) "The best cornering power is available when front-to-rear weight distribution is equal, assuming the tire size is equal both at the front and rear." 3) "Cars that have front-end weight bias (heavier in front) will tend to understeer while cornering." Of course this depends on suspension set-up as well, but that will be covered a little later. And 5) "In general, the best cornering power will result when all four tires are equally loaded during cornering."
The weight transfer during cornering causes body roll (and/or vice versa), "and the amount that it rolls is called the roll angle. [...] By deciding how much of the roll resistance is on the front and on the rear, you can control the understeer and oversteer characteristics of your car. When a car rolls, the tires change their camber angle to the track surface. Since a tire develops its maximum traction when it runs perpendicular to the track, this positive camber angle results in less cornering power. Less roll angle results in less positive camber, so a car will corner faster if the roll angle is kept small. [...] You can compensate for this loss of camber by settign the car up with static negative camber. [...] Using excessive static negative camber can lead to problems, however. For most street applications, the maximum is about 1.0 degrees, or else the insides of the tires will wear." There are a handful of solutions for limiting roll angle since "there are limits to the amount of negative camber that can be used." One such solution that Adams' lists is lowering the center of gravity. This can be done by adding weight to the bottom of the car (not generally the best idea), removing weight from the top of the car (not generally possible, although carbon fiber or fiberglass would help), moving weight lower (lower seat for example), or moving the entire car body lower (as in lowering springs). Another way to control roll angle is lowering the roll center height, a concept difficult to describe even when looking at the pictures in the book. I'll try later on when I get to chapter 7. Adams' third solution is widening the track. "Because the lateral spring base (how far apart the springs are I'm guessing?) is proportional to the track width, a wider track dimension will reduce the roll angle. [...] Most cars already have as wide a track as practical. This means that for any given car, we can not expect to cause much of a reduction in the roll angle by increasing the track dimension." Keep in mind that adding wheel spacers wears harder on the wheel bearings. Using wheels of a smaller positive offset has the same negative effect on the bearings as well as other negative steering effects which Adams mentions to later on. Controlling cornering force amount is the fourth and last solution Adams proposes which also happens to be the most flexible and common solution. "If you want to go around corners as fast as possible, you will have ever-increasing cornering forces, and therefore ever-increasing roll angles. For example, if a street-driven car with street tires could corner at .75 g's, it might have a roll angle of 3 degrees. This same car on race tires might corner at 1.00 g's. If it did, the roll angle would increase to 4 degrees. This means that cars that corner faster will need more roll stiffness to control the roll angle. [...] The two most common means of controlling the roll stiffness on any given car are via the springs and the stabilizer bars." For the newbies, stabilizer bars are also commonly called (anti)swaybars or (anti)rollbars. "Increasing the spring rates will reduce roll angle. Unfortunately, raising the spring rates can also change other aspects of the car's handling. As an example, if a car had a front spring rate of 700lbs-inch and a roll angle of 2 dgrees, and you wanted to reduce the roll angle to 1 degree, you'd need to install 1400lbs-in. front springs. This would double the roll resistance. But increeasing the spring rates this much would also upset the ride motions and cause the car to understeer." According to Adams, "The best way to increase roll stiffness is to increase the size or effectiveness of the stabilizer bars... If a car is to roll, one wheel will be up in compression (the outside one) and one wheel will be in drooping down (the inside one). Stabilizer bars limit the roll angle of a car by using their torsional stiffness to resist the movement of one wheel up and one wheel down. Connecting both wheels to each end of a stabilizer bar causes this motion to twist the bar. The stiffer the bar, the more resistance to body roll it can provide. Since the forces that cause the car to roll are being absorbed by the stabilizer bar, and these forces are fed into each lower control arm, the outside tire loadings will increase as the stabilizer bar twists." Adams then notes that: Stiffness = Diameter^4. Therefore, as diameter increases a little bit, stiffness increases a lot. All you perverts, keep the jokes to yourselves please. This is why there's a big difference in stock swaybars and ones that are only maybe a millimeter or two larger in diameter. As for swing arm length (parts on either end that connect to the endlinks), "The longer the bar, the less effective it will be." Just when I was starting to feel good about myself! Sorry, had to say it. Anyway, this is how adjustable swaybars work. They'll likely have three holes on the swing arm on each side. If you attach the bar to the endlinks with the holes on the ends, roll stiffness will be less than if you attach the bar to the endlinks with the holes in the middle or the holes closest to the rest of the bar. The caption to a photo is significant in saying, "The effectiveness of a stabilizer bar is also dependent on how well the bar is mounted to the frame and to the control arms. Any lost motion at these connections will result in a loss of bar effectiveness, Make sure you use high-quality, rigid materials for the construction of links and brackets." This is why it's important that when you get a swaybar, it comes with urethane bushings (like the ones from Vectored Velocity) instead of rubber like stock. "By varying the size and effectiveness of the front stabilizer bar vs. the size and effectiveness of the rear stabilizer bar, it is possible to change the understeer/oversteer characteristics of a car." The chapter summary does it's job well: Skipping number 1 because it's pointless... 2) "As a larger rear stabilizer bar will reduce understeer because of its ability to increase the dynamic weight on the outside rear tire, it will also work on cars that turn in both directions." 3) "Because left-side weight bias and weight wedging will not work on a road-race car or a street-driven car, rear roll stiffness changes from a rear stabilizer bar are the primary means (although not the only method) of balancing the understeer/oversteer characteristics of these cars."
The first paragraph of the bushings and deflections chapter is rather enlightening: "Even the best suspension geometry and suspension alignment is only effective if the suspension components do not change shape or position. Suspension pieces in general look strong and rigid, but these pieces must handle loads in the 1000-5000lbs range, so they do bend and deflect. Under high performance driving conditions, the wheels bend, the knuckles bend, the control arms bend and the frame bends. The worst source of deflection on a production car is the suspension bushings. Suspension bushings are deceptively simple devices. Insignificant as they may seem, they have a very imortant effect on your car's handling. Therefore, the construction and quality of your bushings deserves close attention." To really get the most out of this chapter, you really need to see the pictures and diagrams. Suffice it to say that depending on the suspension design, movement due to bushing deflection can easily cause significant changes in alignment, especially camber and to a degree toe. This can dramatically effect steering as well as handling. Not that we notice it in day to day driving since we're used to it, but I bet those cars that have full bushing kits from Energy Suspension available can tell a big difference between before and after installation. According to Adams, there are four different materials used for bushings. The first type of bushing uses a steel-on-steel design. Using steel doesn't do much for dampening vibrations or road noise and these bearings need frequent lubrication, but deflection is practically non-existant. For that reason, steel-on-steel is a common design for race cars. The second type of material is rubber. While deflection is an undesirable characteristic, rubber can twist meaning that it requires no lubrication between the rubber and inner or outer sleeves of the bushing since the inner and outter sleeves can "rotate relative to each other." Also rubber does a great job of absorbing vibrations which is why manufacturers' NVH (noise, vibration, harshness) police love them. Urethane on the other hand, cannot deflect much itself and therefore sliding surfaces exist between material and inner or outer sleeve which causes friction that wears on the bushing and makes it so they require lubrication. "Many people who buy aftermarket urethane bushings believe that the urethane behaves like hard rubber, but most urethane suspension bushings are so hard they have to be considered solid, because they offer littel ability to absorb rotational sheer within themselves. [...] Greasing the bushings before assembly works for a few weeks until the grease is forced out or washed away. Once the grease is gone, the urethane is able to bind to the steel sleeves, and the driver hears squeaks and moans. Without lubrication, the urethane can stick to the steel, and the suspension does not move smoothly." The fourth, and to Adams the best alternative, is nylon. "Nylon inserts are inexpensive and are available from any bearing supply house. These 'nyliner' inserts require sleeves machined to close tolerances for proper operation, so the total cost of installing nylon suspension bushings is slightly higher than urethane bushings. Compared to the cost of labor to remove and replace the control arms, however, the amount of increase is insignificant." Personally, I'd stick to urethane. I'm not a big fan of the deflection in rubber bushings. I have no clue where to get nylon friggin bushings or the budget for proper installation. And I really don't care about squeaky urethane bushings. If it keeps the car solid, then it's doing its job. Of course, speaking of that budget, might as well just replace the old rubber bushings with new rubber bushings. There should be a noticeable difference depending on how old the older bushings were. I guess it depends on prices for factory rubber bushings since I wouldn't bother with junkyard ones. Says Adams, "If you don't drive your car hard, or in competition, you probably do not need solid suspension bushings. However, if you can stand a little more impact harshness, installing suspension bushings will make your car handle like a race car. Both cornering power and steering response will improve greatly."
"Springs and shocks are an integral part of any suspension system. But the total suspension system must be considered as a coordinated package, so just changing springs and/or shocks will not always give the desired results." Adams' discussion of spring rate vs. spring load is a lot better than our floundering definitions in that thread on cutting your springs. "Spring load is the amount of weight it takes to compress the spring one inch expressed in lbs. Spring rate is the amount of weight it takes to compress the spring one inch expressed in lbs-in. [...] Note that the spring rate does not change as the spring is compressed, but the spring load does. In Chart 5-1 (again, not included), the spring compresses 1 inch for every 150 lbs of load. The spring rate determines how much the spring will compress as the loading increases. The spring load determines how much weight the spring can support at a given height. For example, when the spring is compressed at 7 inches, the spring can support 1050lbs. If one corner of your car is sagging, you don't need more spring rate, you need more spring load." Also, here's the difference between sprung and unsprung weight: "Unsprung weight includes parts of the car that are not supported by the springs, which would be the weight of the tire, wheel, knuckle, hub and one-half of the spring, shock and control arms. Sprung weight includes the weight of the body, frame, engine, transmission and one-half of the spring, shock and control arms." Seems to me Adams is forgetting about the brake discs, pads, calipers, and drums (unless you have a Starlet rear disc conversion), all of which would fall under the unsprung category. Another important subject is wheel rate vs. spring rate. "Wheel rate is the actual spring rate at the wheel as opposed to spring rate at the spring. It is nt a simple relationship, so adding 100lbs of spring rate will not add 100lbs of wheel rate unless the spring is mounted directly on the axle. Anytime there are linkages such as control arms involved, you have to consider the linkage ratios." This is going to be extremely tough to visualize without a diagram. "The formula for determining the wheel rate is: Wheel Rate = Spring Rate x [(a/b)^2] x [(c/d)^2]" A is the "distance from control arm inner control arm pivots to the center point where the spring acts on the lower arm." B is "the length of the lower control arm from the ball joint to the inner pivots." C is "the distance from the lower ball joint to the front suspension instant center." And D is "the distance from the center of the tire contact patch to the front suspension instant center." Problem is that the concept of "instant center" is practically impossible to describe. Confused people start reading again here. --> Basically, what all that mathmatical mumbo jumbo means is that the wheel rate is much less than whatever your springs' rate is. To get your wheel rate to be more representative of the spring rate you paid good money for, you can do one (or both) of two things. Bring your tire centerline closer to the spring by using a wheel with a higher positive offset. Or get redesigned control arms which move the lower mounting point of the spring and shock/strut closer to the knuckle/ball joint. Even though all that didn't change how hard the springs are, it'll feel like it did. This might be one way to increase life of those KYB GR-2's as well. If you use wheels with maybe a 40 or 42 offset instead of 38, then you could run stock springs which the KYBs can handle easier while still getting the handling benefits of a higher wheel rate. However, please check threads on what wheel offsets do or don't cause rubbing with what tire widths and overall diameters. About redesigned control arms, the front's off limits anyway since the strut is mounted directly to the knuckle (so it can't be moved any closer). And for the back, I haven't taken a look at the current design, but I have a hunch that the shock mounting points are already as close to the knuckle as possible. Toyota ain't stupid. "There are obvious clearance considerations which must be resolved, but in general, the closer the better. On Winston Cup cars, which still use 5.0 diameter springs, it is impossible to get the springs very close to the ball joint. This results in big linkage ratios and the need for very high spring rates." Adams says mounting rear coil springs directly to the axle housing (or in our case the rear axle bar) will produce "wheel rates equal to the spring rates." I agree with that for a live axle rear suspension (or independent where top and bottom control arms are equal length) where the wheel is just going straight up and straight down, but for a conventional independent suspension that uses unequal length control arms, the wheel follows an arc which means that the wheel is farther out on the "lever" than the spring mounting point. Therefore, there would be a linkage ratio causing the wheel rate to be slightly lower than the spring rate. Adams adds, "It is also a good policy to mount the springs in a nearly vertical position to keep the spring force at its full value. [...] A slight angle has little effect, but anything over 30 degrees will result in a significant reduction in the vertical force on the axle, which reduces the spring's effectiveness." Skimming over irrelevant material such as rear leaf springs and more stuff about circle track cars we come to a quick comment on variable spring rates. "A study of the optimum spring rates for a car eventually come to the conclusion that the best spring rate is a variable one. We would like to have a soft spring rate to absorb the road irregularities and then have a high spring rate to absorb the larger bumps." Adams moves on to selecting spring rates for the street with a baffling discussion that flies in the face aftermarket marketing tradition. "The purpose of suspension springs is to hold the car steady while allowing the wheels to follow road irregularities. In general, the softest possible springs will do this job best. Softer springs will allow each individual wheel to move in relation to the chassis while having the minimum effect on the driver's compartment. This translates into a soft ride, noise isolation and good handling. All stiffer springs do is make the car have a stiff ride. They have no capability to make a significant improvement in handling. As long as the springs on a car are stiff enough to keep the car from bottoming out, they are adequate. If a car is lowered, a slight increase in spring rate can be used to compensate for the reduced ride travel. Some car enthusiasts mistakenly believe that if 300 lbs-in. coil springs are good, then 600 lbs-in. springs have to be better. They're wrong. Optimum road-holding demands that the tires be in contact with the pavement; a soft spring lets the wheels follow road irregularities so that the tires can generate maximum adhesion." Here's the kicker. "Our recommendation for front springs on a street-driven car is to use the standard factory coils. For street use, you can trim one half coil off the top of the spring with an acetylene torch to lower the car slightly. Most front spring rates range between 300 and 350 lbs-in. Trimming the coils as we've recommended will increase the rate approximately 10%. But the true purpose of trimming the front springs is to lower the car for improved aerodynamics and better handling, not to increase the spring rate." Whatever you say Herb. The man makes sense, but I'm still holding out for a set of Eibachs. "Although the spring rate will not change during the life of a spring, the spring load can change. This is commonly called spring sag. Loss of load, or spring sag, can be caused by a variety of reasons, including poor metallurgy, overloading and even fatigue due to high mileage." Rather than increasing the spring rate, Adams suggests installing a "rubber shim on top of the spring to increase the load" which are supposedly available "from dealership parts departments and from auto parts stores." One other thing that's kinda interesting that hadn't occured to me was the idea that "all springs take a certain amount of permanent set after they are installed. [...] This change in height is caused by the inevitable loss in load that any new spring experiences. If you put new springs on your car, they will settle as much as an inch within the first few months." That's enough for springs. How about shocks? "The purpose of shock absorbers is to control the velocity of the suspension. If the shocks don't have enough resistance, the spring will move the suspension too fast and it will have an underdampened motion" which means the car will bounce up and down more than it needs to just from going over a bump. "If the shocks are too firm, the motion will be overdampened" which means that the suspension will feel soft, but bumps still jar your teeth out. In other words, "Extra firm shocks have the same negative effects on ride and handling as extra stiff springs. The tires cannot follow the road irregularities unless they are free to move in relation to the chassis. The relative motion must be as free as possible, but it also must be controlled. [...] We have found that adjustable shocks allow each car to be tuned for critical dampening. Run your shocks as soft as possible--just enough so the car doesn't wallow over bumps." Adams suggests running rubber bushings for shock/strut mounting because this is a critical place for softening road noise. My guess is that deflection here isn't going to matter too much either. Adams also mentions that the same rules about mounting springs also apply to mounting shocks, which is to say mount them as close to the wheel as possible and as vertical as possible.
On to types of front suspension. While a beam axle front suspension ("where both wheels are connected to a rigid axle") is totally dis-similar to the MacPherson strut set-up on our Tercels, Adams still points out an advantage and a strong disadvantage to the beam axle design that tell us a bit. This being the camber control and the heavy unsprung weight. What is meant by camber control is that "there will be no loss of camber because of suspension movement or body roll." Not that there won't be weight transfer, of course, but the vertical position of the wheels never change. As for unsprung weight, "Because both front wheels are attached to the same axle, its weight affects the unsprung weight of each wheel. This inertia, and the interaction between the wheels, reduces road-holding on anything but a smooth surface." Not only handling suffers from the inertia of lots of unsprung weight. "The heavy unsprung weight of the beam axle suspension, and the interconnection of the front wheels, does not allow for as much ride isolation between the wheels and the chassis, so a rougher ride results. Ride quality is also important for road-holding since those factors that improve the ride characteristics also help to improve the road-holding, because they keep the tires in better contact with the road." MacPherson struts are the next design Adams tackles. "The knuckle on a strut-type suspension is mounted on the shock absorber." ...as opposed to mounting the knuckle on an upper control arm. The lower control arm is generally pretty much the same as on a double A-arm suspension, however, the strut no longer needs to be mounted to the lower control arm. "Somtimes they are referred to as MacPherson struts. Strut-type front suspensions became popular on production cars in the '70's because they offered a simple and inexpensive configuration that doesn't take up much space. It is particularly well suited to front-wheel-drive production cars, because it allows room for the front drive-axle to pass through the front hub. Most of today's small cars use this type of front suspension, because it is inexpensive and gives a fairly good ride quality with the compact dimensions needed for front-wheel-drive cars." According to Adams, one complaint racers have with a strut design is that "there isn't much room for wide tires and wheels without increasing the scrub radius." The scrub radius is (roughly) the distance between the wheel's centerline and the ball joints of the steering knuckle. The angle created by the knuckle's upper ball joint not being directly above the lower ball joint has an influence, but we'll ignore that for now. "Increasing the scrub radius causes a big increase in the side loading of the sliding members, which causes bending and higher friction loads." Hence worn wheel bearings, more quickly fatigued bushings, etc. "Another problem that concerns racers is that there is little camber gain possible with this type of front suspension. This means that the outside tire will lose its camber angle to the road as the car rolls during cornering." I don't know where he's getting that since there's no lack of camber adjusting kits for MacPherson equipped cars and I've never heard any complaints before (especially if the car's lowered). I guess he's just comparing to double A-arm or something. Speaking of which, "The front knuckle on a double A-arm front suspension is supported by a triangulated control arm at the top and at the bottom. Early versions of double A-arm front suspension had equal length arms mounted parallel to the ground. [...] Unfortunately for high performance handling, this system does not provide any camber gain, so the front tires lose camber as the body rolls and cornering power is lost. [...] This problem was solved by changing the length of the A-arms. The use of a longer, lower A-arm and a shorter upper A-arm provide a suspension geometry that causes the tires and wheels to gain negative camber as the suspension compresses." One advantage that this type of front suspension has over other types is the use of "rigid control arms to connect the front knuckles to the frame. These arms prevent deflections during hard cornering, which insures that the steering and the wheel laignment stay constant." For the purposes of this design, the terms A-arm and control arm are synonymous.
"A car's suspension design starts at the tires and wheels and moves inboard. The last thing that is designed is the frame." Therefore, we shall start at the tires and wheels. "If you have a choice of wheel offset, it is a design advantage to use wheels that permit the knuckle to be placed inside the wheel." Before we follow that up by talking about scrub radius again, "Since all the loads from the wheel and tire must be fed into the chassis and control arms through the upper and lower ball joints, it is an advantage to place the ball joints as far apart as possible. Any given load results in lower forces if the spread between the points is increased." I'm assuming that refers to the height of the knuckle. Back to scrub radius. "There are significant handling and control advantages in reducing the scrub radius to the minimum. Any bump or cornering force that is applied to the tire can exert a twisting force on the steering that is proportional to the length of the scrub radius. If the scrub radius was zero, these twisting forces would be zero. Cars with zero scrub radius can usually be driven without power steering, because the twisting forces are gone and the tire easily rotates about the steering axis. The factors that increase scrub radius are: positive wheel offset, wheel and tire width, brake rotor width and the design of the knuckle. If any of these components can be made to minimize the scrub radius, there will be an improvement in handling, control and steering effort." According to the August 2003 Sport Compact Car magazine wheel guide article, positive offset is where the hub mounting surface is outside the wheel's centerline and negative is where the hub mounting surface is inside "like on the outer wheel of a duallie truck". In that case negative offset would increase scrub radius rather than positive. In the next chapter of "Chassis Engineering", Adams says "If you had a wheel with 1.00 inch of negative offset (mounting surface outboarad of the wheel centerline) the backspacing would be [blah, blah, blah]." Somebody needs to get their story straight (time to write Josh Jaquot a little email). Nevertheless, the closer you can get the ball joints to the wheel's centerline, the better. I'm going to leave out most of the conversation on control arm length since front lower are the only ones on the EL chassis. Let's leave it at, "In general, it is best to make the lower control arm as long as possible. This reduces the angularity the ball joint must accommodate as well as slowing down angular change of the suspension members as they go through their travel." Here I'm going to skip the rest of the chapter which deals with computer analyzing suspension design and a discussion on bump steer (change in toe as the suspension goes through its travel). On second thought, I'm going to skip the next chapter on front suspension building because it basically tells what we already know and go straight to chapter 9...
...Which is on live axle rear suspension design. This design is extremely important to us because the EL chassis uses this design with associated panhard bar. "A live axle rear suspension is one where both rear wheels are mounted on a rigid axl. Because the whole axle moves as a unit, and because it moves whenever either wheel hits a bump, it is called a live axle. The advantages of a live axle are simplicity and rigidity." Ergo, no lower frame brace needed since the axle is the lower frame brace. "A well-designed and well-developed live axle will beat a poorly designed independent rear suspension, even on rough roads. On smooth roads, it is usually difficult to see any advantage for an independent rear suspension. [...] The major disadvantage of a live axle is its inability to allow each rear wheel to follow the contours of a rough road. Most of this inability comes from the unsprung weight of the differential." As the EL chassis is of course fwd, Tercels and Paseos don't have anywhere near the unsprung weight in the rear suspension as a rwd car would. On the other hand, there are suspension designs (covered in the book even) that have a live axle, but with the differential mounted to the frame. Like I touched on before, our chassis uses a panhard bar which connects to the chassis on the driver's side and to the axle on the passenger side. "A Panhard bar is simply a link between the axle and the frame which controls the side-to-side location of the rear axle. The advantages of a Panhard bar are that it is simple, effective and lightweight. The disadvantages are that it must be as long as possible to minimize the slight side-to-side variations that result from the arc scribed by the bar. This slight variation has no real adverse effect on the car's cornering capability, but the driver does feel it so it can affect his performance." This lateral movement of the rear axle is the single most common cause of fender rubbing with our cars. The passenger side rear tire has a tendency to rub the outer fender lip because the axle and both wheels move to the right as the suspension goes through its travel. Though this really doesn't happen too much with stock tires and stock ride height, once you get up to and past 195mm width tires and the car has shorter springs, rubbing might become an issue with your particular car and set-up. There's another possibility that might end up being a potential modification for EL chassis which is called a Watt's link design. "A Watt's link eliminates the slight side-to-side variations that happen with a Panhard bar." Imagine cutting the panhard bar in half and attatching the cut ends to the center of the axle with connections to the chassis on either side, one chassis connection where it was before and one directly above where the bar formerly attached to the axle. Adams says this is a better design, but the complexity makes it more fragile and therefore, the advantages are pretty much outweighed by the disadvantages. One thing I really need to check on is how the axle is located front to back. Trailing links, maybe? As soon as I get a look, I'll add in more commentary. "The best rear suspension design is only effective if all the parts are strong and rigid enough. Any deflections, or looseness of the parts, will have a greater effect on the car's handling than will its design geometry." Sounds like another call for urethane bushings if you ask me.
The next chapter is on independent rear suspensions. Interestingly, the majority of the chapter spends time knocking the Corvette's suspension design up through the C4. Apparently, an independent rear suspension is smoother in ride quality than live axle, but can't get the power to the ground as well (for rwd at least), and any looseness in drivetrain or suspension pieces completely throws camber and toe settings out the window (at least with the Vette since the axle acts as upper control arm). Also, "without a rigid frame, the precision of the suspension is obliterated by chassis flex." Outside of all that, the independent rear suspension is pretty similar to the double A-arm front suspension in most cases. "The major advantages of an independent rear suspension are the potential for a smoother ride and better adhesion over rough pavement. Under these conditions, the ndependent rear suspension can be better if the system is correctly designed. This potential advantage must also be evaluated with regard to the added cost required." Any embarassment I used to feel about not having an independent suspension in the back has now completely disappeared through reading this chapter and the previous. I'm going to skip chapter 11 on rear suspension building.
Chapter 12 then, is on frame design, the part everyone's been waiting for. "Before actually designing a frame and/or roll cage, it is necessary to recognize what shapes and arrangements are rigid and which are not. The basic shape for constructing rigid structures is the triangle. Its shape and dimensions will not change much unless one of its three legs is broken. In contrast, a square-shaped set of tubes has very little structural rigidity, in that it will bend diagonally when even a small load is applied to it." One solution is to brace the rectangular shape with a diagonal member. "This divides the rectangle into two triangles, so that a shape which was very weak becomes rigid. Double diagonals can be used (which creates four triangles) for still more rigidity, but these additional members are usually unnecessary unless very high loads are anticipated. Figure 12-4 (again, not included) shows the use of a panel of thin metal to give the rectangle diagonal rigidity. This is called a shear plate and its effect is the same as a diagonal brace. Shear plates can be used to advantage in race cars, because they can function as firewalls, floorboards and bulkheads, thus eliminating the weight and complexity of diagonal tubes. Applying the rigidity analysis from the two-dimensional examples to a three-dimensional box reveals how the basic structure of a car can be improved. The most difficult forces to resist in a chassis are the loads that put the frame in torsion. Twisting an open box clearly shows how poorly it absorbs torsional loads. This is what happens to areas of an automobile that are not triangulated. Even though five of its six sides are rigid, the one open side makes the whole assembly very weak. Just as a single diagonal brace strengthens a two-dimensional rectangle, a diagonal across the open box makes the total package torsionally rigid." That takes care of the basics. Beaming stiffness, "how much a frame will flex as it is loaded in the center and supported at both ends," is really a moot point, while torsional stiffness, "how much a frame will flex as it is loaded when one front wheel is up and the other front wheel is down while the rear of the car is held level," is all-important. Adams builds models out of balsa wood to demonstrate the characteristics of different frame designs, how effective or ineffective the minimum SCCA rollcage and standard NASCAR rollcage are for increasing torsional stiffness, and how to apply shear panels and strategically place extra bars to minimize weight and maximize torsional stiffness. The best thing we can do (besides a front strut bar) is to build our own models of an EL chassis and play around seeing what combinations and set-ups make the most difference.
The frame building chapter will be passed over in favor of jumping right into aerodynamic downforce. "Aerodynamic downforce is developed on a car when the air pressure on the top of the car is greater than the air pressure on the bottom of the car. If the air pressure on the top is less than on the bottom, the car will generate lift just like an airplane. [...] The amount of downforce (or lift) a car has will change in magnitude in proportion to the speed squared. An example of this relationship shows that the downforce will be four times as great when the speed is doubled (Chart 14-1)."
| Chart 14-1 | |
| Speed | Downforce |
| 50 mph | 100 lbs. |
| 100 mph | 400 lbs. |
| 200 mph | 1600 lbs. |
Using the example from Chart 2-2, we'll "add 400lbs of downforce (100lbs at each wheel), you'll have results as seen in Chart 14-3 (not included), which says that adding 400lbs of downforce can increase the cornering power by 10 percent to 1.15 g's. [...] It should be noted that many cars do not have aerodynamic downforce, they have lift. Older model production cars had considerable lift, especially at the front. It was typical to see 300lbs of lift on the front of cars like 1970 Camaros. This was what made them feel 'light' at 100 mph. [...] If you accept that causing more pressure on the top of the car will cause more downforce and improve lap times, the next question is how do you get more pressure on teh top of the car and less pressure on the bottom of the car?" Adams' number one suggestion is to adjust body rake. "When the front of a car is lower than the rear, less air goes under the car. When there is a lack of air to fill the void under the car, it causes a low-pressure area. This low-pressure area under the car increases downforce." The best way to adjust the body rake is to get some springs that lower the front more than the back. "Some ways to optimize the low pressure under the care are: 1. Make the bottom of the car smooth and flat to keep the airflow smooth." Not mentioned in the book, from Bernoulli's law, we know that faster moving airflow has lower pressure than slower moving airflow. That's what generates lift on an airplane wing. Air flow on top has to go a longer distance to reach the back of the wing at the same time as the air flow on the bottom of the wing. On the underside of cars, there is plenty to slow down the air producing a high pressure area under the car. This is why smoothing out the underbody is so important. Adams' second suggestion for optimizing the low pressure under the car is to "build a fence between the low-pressure area and the airflow along the sides of the car." Although Adams suggests doing this by "slightly recessing the floor inside the frame rails," side skirts of a body kit make more sense on our application. After body rake, front and rear spoilers are the next biggie. "A front or rear spoiler is really an air dam that causes a higher pressure in front of it and a lower pressure behind it. When the high pressure ahead of a spoiler can be directed to a horizontal surface on top of the car body, the downforce factor will increase. And, when the low pressure behind a spoiler can act on a horizontal surface on the bottom of the body, it will also increase the downforce factor." To go along with a big ricey rear spoiler (or a smaller one that's still functional, but not so gaudy), those carbon fiber canards look good to me, but neither of those affect the underbody. If you can use the car's underbody to create downforce, then you won't need canards or a big rear spoiler that while effective at creating downforce, also create aerodynamic drag which negatively affects top speed, fuel economy, and in part, acceleration. Incidently all the irregularities of the car's underbody also create drag, so if you can smooth the underbody out or just keep air from going down there at all, then you decrease drag and increase downforce at the same time. Better performance, better fuel economy. So how do we manage the underbody? First step is a lower lip or chin spoiler. The front bumper and company are the first line of defense. You do need air to get past and into the radiator, so there has to be a hole in the center, but what happens to the air between the front body panel and the radiator? It's all about messing up your nice low pressure zone right behind the chin spoiler. The best thing to do is make a shroud that goes from around the hole in the bumper back to the edges of the radiator so that all the air going through that hole takes a non-stop flight to the radiator without screwing with your aerodynamics. Shoving more air through the radiator helps cooling too, so it's a win-win situation. Ok great, but what happens on the other side of the radiator? Most of us have a huge gaping hole in the bottom of the engine bay. Right. Adams has a solution for that which involves another radiator shroud on the back side of the radiator with dryer ventilation ducting to the wheel wells to let the air out. That's a solution, but I think one that'd be a bit more practical for us, if a little less effective, would be a carbon fiber or fiberglass hood with a "reverse" hood scoop so that at least some of the hot air coming out of the back of the radiator gets routed above the car helping to create that high pressure on top. Don't know what all would be involved in functional Z3 style fenders, but that's an idea too. That pretty much wraps up chapter 14 then.
Since I'm skipping chapter 16 on testing and tuning, chapter 15 on rotating inertia will be the last. "Most car enthusiasts know that reducing weight will increase acceleration with the same horsepower. What is less well-known is that iff the weight, and its distribution, of the driveline components are reduced, the improvement in acceleration can be much greater than that which would be realized for just reducing the car weight. The weight and its distribution of a driveline component about its center of rotation is called its rotating intertia. This resistance to rotational acceleration is called rotating inertia, because its effect is seen when the parts are accelerated rotationally. [...] If your car had a flywheel that had most of its weight concentrated around the rim, it would have greater resistance to revving quickly than a flywheel that had its weight concentrated around its center. Obviously, a lighter flywheel would be even better because it would not only weigh less, but it would also have less rotating inertia. [...] All the rotating parts in the chassis have inertia which resists angular acceleration. These parts include the tires, the wheels, the brake rotors, the hubs and the ring gear and the differential. To show the effect of the rotating inertia on these parts that turn at wheel speed, we used the same example car (as the one I neglected to mention previously) but with a 15lb reduction in these parts. Under the same test conditions, [yadda, yadda, yadda]. This shows that a 15lb reduction in rotating inertia on the chassis rotating parts will have three times the benefit of a 15lb weight reduction on the rest of the car. [...] The rotating parts of the driveline include the crankshaft, the flywheel, the clutch, the transmission gears and the driveshaft. Since these parts operate at a much higher rpm, the effect their rotating inertia has on acceleration is much greater. If we reduced the rotating inertia of the flywheel on the example car by 15lbs, [yadda, yadda, yadda] the effect of reducing rotational inertia on driveline parts has 15 times the benefit of just reducing the weight of the car." Impressive. Just wish that Aasco flywheel was cheaper. Shows a benefit of knife-edging the crank too. If anyone has the weights of ACT clutches as compared to stock, that'd be good to know. Last but not least, you take 10lbs off the wheel/tire combo at each corner (not hard to do by just getting new rims and tires) and that'd be like taking 120lbs (10lbs x 3 x 4 corners) off the car! Easy and stylish. Plus it'd help steering response due to lighter moving parts, decrease unsprung suspension weight giving a smoother ride and better handling, and the new tires would (or at least should) help your handling more anyway. For $650 or a little more (approx. price of 15" Rota wheels with Kumho tires), that's a pretty good deal if you ask me.