D2 Coilovers for 2G

[asian312]

Been reading through the posts regarding many of the suspension setups available, but most seem to pertain to the 1G setups. I've looked at a few sites with information on the D2 setups for the 2G AWD and notice that they do not have camber plates on the fronts like the 1Gs do. Despite that drawback, do they still come with upper pillowball mounts? I read somewhere in these posts that they were nessessary on the fronts and not the rears, but was not able to differentiate if that pertained to the 1G, 2G, or both.

And I know this is a most hated question, but what about spring rates? Once again I've heard 7 front / 5 rear for the 1Gs, but what about the 2gs? D2s are suppose to come stock with 11 /9 split, but is this too harsh on the street? My car is a daily driver, and by that I mean it's my livlihood. No intention of turning it into a monster or even a weekend worrior, but need a little performance when I start going out to a few track days. Ride quality is a must and fore most since I commute alot.

If there are any other suggestions plz feel free to shout about other setups. I'm in the $900-$1K range, but if a koni and GC setup is more suitable than I'm all up for saving some $$.

 

[DG-FNR]

Because the 2G is a double A-arm, you don't need a "caster/camber" plate on it. In fact, it would do nothing other than perhaps change the motion ratio a little bit.

The spherical bearings themselves aren't really the big deal - the big deal is keeping the spring axis coaxial through the range of motion - that's the "coaxial hat" thing you may have read about. In order to do a coaxial hat properly, it needs to be located on a spherical bearing (they aren't "pillowballs").

The added side benefit is that you eliminate a rubber bushing, which is a spring in series wth the shock and which robs the shock of some sensitivity. You paid for that shock, why not use it?

The coaxial hat is less important in the rear than in the front, but that doesn't mean "not important". With a full set of coaxial hats mounted in spherical bearings, you give the shock and spring their best chance to do their job. And you can reasonably expect the spherical bearing mounts to last the life of the car.

As far as springs go, we are recommending a 600/300 setup. That puts the car at roughly the same stiffness in ride as a Z06 Corvette, which is a good compromise between street and track use. Happily, springs can be upgraded at any time, and Hypercoils keep their value very well, so changing spring rates is not very expensive.

Be very careful of any setup over about 400lbs of rear spring.

DG

 

[asian312]

Thank you for the clarification and insight on the spring rates. Other than just a harsher ride, would increasing the rear spring rates cause the car to become less stable or induce more of a snap spin scenario? Just curious.

Since I have no experience with suspention upgrades, it's wonderful to have knowledgable people give their advice saving me a lot of headache in the long run. Thanks!

 

[DG-FNR]

No problem; it's what I'm here for.

There's no reason to expect stiffer springs to increase *harshness*. If the car is properly damped, you can run very stiff springs without shaking your kidneys to death. IT'S SO BOUNCY!!! is usually a DAMPING problem, not a spring problem (although sometimes the reason why it's a damping problem is that the car is riding on the bumpstops and the poor damper can't damp that - so *that* is a "spring problem" as in "too little spring")

What stiffening springs *will* do is increase the natural frequency at which the suspension wants to vibrate. Think of a big ol' plush Cadillac. Hit a bump, and you get this slow up-and-then-down motion with a period of about a second. Stiffen the springs, and you shorten the time that it takes for the suspension to cycle. That increases the "immediacy" of the car, because it reacts and recovers quicker. This, in turn, tends to make the feel in the car more sensitive - you start to notice road irregularities that would have been isolated away before with the softer suspension.

This sensitivity gives you, as the driver, more information, and that's never a bad thing. Your passenger may or may not agree. But there's no reason why you can't have a nice taut suspension without shaking your passengers to death.

A nice compromise is a natural frequency of about 1.8-2.0 Hz. That's Z06 Corvette - sports car, but still streetable.

All else being equal, we expect that raising the spring rate on that end of the car to decrease the grip on that end. Sadly, all else is rarely equal. On a DSM, all else is almost NEVER equal, especially an AWD where the diffs crosstalk.

If the springs are stiffened both front and rear at the same delta WHEEL rate (the spring force felt at the wheel, not the spring force felt at the spring) then we expect that the overall balance of the car won't change. Again, this is one of those "all else equal" deals - if the reduced roll that comes with the stiffer springs improves one end more than the other (which can happen due to a whole slew of different reasons) then the handling balance can change. And a lot of what determines THAT is in the tire.

As a very VERY rough, general rule of thumb for an AWD 2G DSM, stiffening rear spring or rear bar relative to the front increases corner-entry oversteer, increases corner-exit understeer, and reduces the maximum lateral grip level. Your milage may vary. Void where prohibited.

Because the motion ratio (the effectiveness of the springs) is higher in the rear than in the front, the rear is more sensitive to spring changes. Again, as a general rule of thumb, 100lbs of rear spring equals 300lbs of front spring in terms of effect on natural frequency (the car is also nose-heavy which further contributes)

Going too stiff in the rear, both relative to the fronts, and overall (the harder the tire, the less spring it can stand, especially in the wet) can cause sudden snap-spins on corner entry with little chance to recover. This is something to be avoided. I recommend people on street tires (or on race tires running at high speeds) start with a 300lb rear spring, and then play with the front springs to tune the balance.

DG

 

[wret]

There's no reason to expect stiffer springs to increase *harshness*. DG

I hear similar statements made often. As a suspension tuning novice, but a logical thinker, I cannot agree. The function of the suspension is to mitigate impact from terrain over distance and time. Distance being the travel of the suspension and time being the frequency of suspension response. While I agree that properly dampened springs will present a more comfortable and controlled ride than those that are improperly dampened, how can it be said that a suspension with firmer springs, shorter travel, and theoretically more dampening is not more “harsh?”

 

[DG-FNR]

To really do this justice requires a thorough discussion of vibration theory, and a couple of dozen SAE papers. Let's just say that Detroit, which has been historically more concerned with ride quality over ultimate performance, has done a lot of work on determining what aspects of suspension design affect what aspects of ride quality, and by how much.

When we talk *harshness*, we usually are talking about *transmissability*, which is a measure of how much of the impact force of a bump is transmitted through the suspension to the sprung mass, and ultimately, the passengers. Transmissability is a function of damping ratio, not spring force.

If you think about it, if we assume that hitting a bump of X amplitude at Y speed results in a wheel force of Z pounds, all that really changes if the spring rate changes is how much the wheel deflects in response to that force input. The amount of force itself doesn't change.

If you further think about it, a harsh suspension, one that bangs the passengers around a lot, must be doing that to the sprung mass as well. And if *that* is happening, then the car is getting upset and not gripping well. So it is in the best interests of overall race performance that the car *not* be harsh - both in terms of ultimate grip, and in terms of the driver being able to drive the car in comfort.

Every race car that I have driven that was properly set up has been a sweetheart to drive. A Z06 Corvette is the easiest car to drive that I have ever been in, save a Formula Ford (that was also a total sweetheart) My own race car, with 900lb front springs, can be driven around town in perfect comfort. It also took 5 separte valving iterations and suspension position sensors to get there.

Where you sometimes see tradeoffs in ride quality on racecars is when the car has signifigant aero downforce, and where the amount and balance of that downforce is VERY dependant on the pitch angle of the sprung mass. If that is the case, it becomes vitally important to keep the attitude of the sprung mass relative to the airflow under complete control, and it may take very stiff springs to do so. That, in turn, can result in very high natural frequencies, which aren't HARSH but have their own problems with regards to driver comfort (and native road-holding ability)

Like everything else, it's possible to go too far. But in order to get natural frequencies that get really uncomfortable, you have to go to springs that are SO stiff that the car lost hold of the road a long time ago - or you're on the bumpstops.

DG

 

[jtmcinder]

When we talk *harshness*, we usually are talking about *transmissability*, which is a measure of how much of the impact force of a bump is transmitted through the suspension to the sprung mass, and ultimately, the passengers. Transmissability is a function of damping ratio, not spring force.

Nope. Bump transmissibility is a function of the spring and shock combination, since both the spring and the shock resist the upward movement of the unsprung parts. On most cars, the spring is playing the larger role, not the shocks. (Search for such things as Hooke's Law for details.)

If all you care about is bump transmissibility, then you should run extremely low springrates and underdamping shocks. The problem is this sort of set-up will cause the car to be overly sensitive to swales (i.e., low-frequency "bumps") as well as lateral and longitudinal acceleration (i.e., turning, braking, and hitting the gas). So you find some compromise, often in the area of 1.7 to 2.2 Hz.

The ride quality of a Caddy has a number of reasons. One that is often ignored is the incredible ratio of sprung to unsprung weight on those cars. To match this ratio in a DSM, you would have to delete the brakes and run four donuts.

- Jtoby

 

[DG-FNR]

"The Shock Absorber Handbook" John C Dixon, Society of Automotive Engineers Press, 1999

In particular, page 61, Figure 2.5.1(b), showing transmissability as a function of damping ratio, and pge 62, Figure 2.5.1(c), showing peak transmissibility as a function of damping ratio.

Page 64 "From this figure it is apparent that as the damping ration increases from 0.1 to 0.4 the peak transmissibility reduces considerably, but beyong this value the rate of improvement is relatively modest. This provides some logic to support the use of practical vehicle damping ratios fpund by experience"

And from MY experience, valving is optimised between 0.65 and 1.0, which fits.

Changing the spring rates will change the natural frequency of the suspension and change the forces that the shock needs to provide in order to maintain any particular desired damping ratio. But the actual transmissibility is a function of damping ratio alone. For any spring rate (or sprung mass) giving a reasonable natural frequency, the appropriate damping ratio WILL keep transmissibility in check.

DG

 

[jtmcinder]

"that as the damping ration increases from 0.1 to 0.4 the peak transmissibility reduces considerably, but beyong this value the rate of improvement is relatively modest. This provides some logic to support the use of practical vehicle damping ratios fpund by experience"

And from MY experience, valving is optimised between 0.65 and 1.0, which fits.

Note that the peak in the curve is always down at a rather low frequency. If all you care about is reducing the transmissibility of low frequencies, then go with some outrageously overdamped set-up. If you keep in mind, however, that higher damping ratios means more transmissibility of the high frequencies (i.e., all frequencies above the cross-over for the set-up), and you note that bumps are the issue, not swales and such, then you'll see why Dixon suggests what he did.

Isn't Topeka rather bumpy? Could your belief in overdamping be part of the problem?

- Jtoby

edit: I will now attach an image of the transmissibility curves. There are three levels of damping ratio (expressed here as percents). Note how higher damping ratios reduce the tranmission of low-frequency inputs (left end of plot), but increase the transmission of high-frequency inputs (right end of plot). Humans are particularly sensitive to vibrations in the 4-8 Hz range, so assuming a natural frequency around 1.8 Hz, the part of the curve that matters to ride quality are in the range of 2.2 to 4.4 relative frequency. These are all above the cross-over, which is why higher damping ratios make the ride seem worse.

 

[DG-FNR]

I do NOT believe in overdamping. Quite the contrary. I believe that most people are WAY overdamped, and that the car will be faster with much less damping.

Experience has borne out this observation. When the rules permit, use more spring and less shock, and go faster.

Besides, I don't have to guess at this stuff; I measure it directly off the car.

http://farnorthracing.com/newimages/suspension_velocity.jpg

Those are the live suspension velocities taken off the right front corner at three different National events - and one of them is Topeka. There are techniques that analyze this data and directly report on the suitability of the shock valving.

Dixon again (Page 54) "A racing car will have an effective damping ratio approaching 1.0... ... the optimum handling occurs for a ratio approaching 1.0"

I think he's too high, and the data (and a number of race teams) agree - optimum is usually closer to 0.65. I have personally tried damping ratios from about 0.3 to close to 3 (by mistake) and 0.65 - 1.0 is where the car is fastest. There is room for personal taste; I like less damping, but some of my fast friends prefer more.

I'm not suprised that you've mischaracterized my damping setup and the nature of the Topeka surface. After all, it's not like you've ever been there to see my car run on it....

DG

 

[jtmcinder]

Let me try to express myself in a more straight-forward manner. I will also try to slow down a bit, in case some readers skipped physics in high school (as I did).

The cross-over point in the transmissibility curve is always at 1.414 times the natural frequency of the suspension. Assuming a passive suspension (like ours), any input to the system (i.e., a bump or swale) that is below the cross-over point is magnified; any input above the cross-over point is shrunk. (You will often hear people talk of the second half of this -- viz., the shrinkage of inputs above the cross-over point) -- in terms of isolation. The idea is that the suspension isolates the unibody from any and all inputs above the cross-over point.)

None of the above has anything to do with the shocks. The only variables that influence the natural frequency of the suspension are the sprung mass and the wheel rate (i.e., the springrate multiplied by the square of the motion ratio). The only variables that influence the frequency of the input to the system are the size of the bump (or pothole) and how fast you hit it.

Moving on, the damping ratio is the ratio of the damping forces produced by the shock to the damping forces required for critical (i.e., one-oscillation-and-then-stop) damping. One way to select the desired damping ratio is to look at a plot that gives the transmissibility curves for a variety of damping ratios, which are available on the web if you search. What you will see is that there is a trade-off. Higher damping ratios reduce the magnification of inputs below the cross-over point, but they also reduce the shrinkage above the cross-over point. In short: higher damping ratios make the car wallow less (by decreasing low-frequency transmission), but also make the ride more harsh (by increasing high-frequency transmission). This is what most people already know, only expressed in fancy words.

At this point, I hope that you can see why people often say that it's the shocks that matter. Stiffer shocks have higher damping ratios (for a given springrate), so they increase the transmission of high-frequency inputs, which makes the ride nasty. A slightly higher damping ratio is usually OK since the nastiness isn't too bad and the reduction in wallow (i.e., low-frequency shifts) more than makes up for it. You crank up your adjustable shocks for racing and then turn them back down to fetch groceries.

If you stop here, you're ahead of the game, but one more step will really help.

Assume that the damping ratio is fixed at some value (e.g., 0.6) and the springrates are changed. (Yes, I know that you have to run stiffer shocks to maintain the same damping ratio when the springs are stiffened, but assume that this is done so the damping ratio really is constant.) What happens when the springrate is increased is that the natural frequency of the suspension is increased. Now add in the fact that the cross-over point is always at 1.414 times the natural frequency. This implies that increasing the springrate will expand the region of the transmissibility curve where inputs are magnified, because magnification occurs below the cross-over point and you just moved the cross-over point upwards. Increasing the springrates also shrinks the region where inputs are shrunk, again because you moved the cross-over point upwards. So, as a rule, upping the springrates will always increase the amount of the input that is transmitted to the unibody.

That is why springrates also matter. A higher springrate will cause the ride to be nastier, assuming that the damping ratio is the same.

If you are near overload, stop now. If you often cruise BDSM sites, carry on.

All of the basic physics texts seem to assume a damper that is symmetrical -- i.e., a damper that produces the same amount of compression damping as rebound damping for a given shaft speed. Modern shocks are way beyond this. Good ones produce much more rebound damping. In other words, modern shocks seem to be designed on the assumption that there will be many more half-sine inputs in the upward direction (i.e., bumps) than half-sine inputs in the downward direction (i.e., pot-holes). This has allowed them to design shocks that have higher damping ratios in one direction than the other. Therefore, you can actually try to get the best of both worlds. You can run a lowish damping ratio for bump to increase the amount of shrinkage of the transmission of actual bumps, while simultaneously running a highish damping ratio for jounce (i.e., extension of the inside shocks due to body roll in a corner) to decrease the amount of magnification.

Now you not only know why stiffer shocks produce a nastier ride and stiffer springs produce a nastier ride, but you also know why some of us go ape over Konis, which have a very high rebound-to-compression ratio.

Or you think I have too much spare time.

- Jtoby

 

[DG-FNR]

*sigh*

Asian312, I apologise. He follows me around and there's nothing I can do about it.

The simple bottom line is this:

1) The natural frequency of the car is a function of the amount of sprung mass, and the effective spring rate (normally called "wheel rate")

2) The higher the frequency, the "stiffer" the car is.

3) It has been determined that, for a car without signifigant aero downforce, the natural frequency of the car should lie between about 1.5Hz and 2.5Hz. A Z06 Corvette is in the 1.8Hz-2.0Hz territory

4) It has also been determined that ideal damping for performance should lie in a damping ratio somewhere between 0.65 and 1. The exact *forces* required to produce this damping ratio depend on the natural frequency of the car, which in turn depends on spring rate and sprung mass. But independant of what the forces are, for any given natural frequency (and thus for any given wheel rate) the damping ratio can be tuned to lie in the "happy zone" of 0.65 to 1.0

5) Happily, the function of "harshness" is primarily that of transmissibility, and transmissibility of road impacts is minimized starting at around a damping ratio of 0.4. A properly damped suspension, set up in the proper zone, is not particularly harsh.

6) There are other aspects of ride quality besides transmissibility, and ride quality itself is very subjective and varies from person to person. But a properly set up racing suspension should not and does not abuse the driver or passengers.

If your car is shaking your fillings loose, the setup is wrong.

DG

 

[jtmcinder]

You seem to have a decent handle on most of this. Then we get to:

5) Happily, the function of "harshness" is primarily that of transmissibility, and transmissibility of road impacts is minimized starting at around a damping ratio of 0.4. A properly damped suspension, set up in the proper zone, is not particularly harsh.

Nope. Transmissibility is not a single value; it is very misleading to talk about it terms of minimization. It is a curve with input frequency as the independent variable (or, if you like, the frequency ratio [input divided by natural]). Furthermore, because the curve must pass through a common point defined when the input frequency is 1.414 times the natural frequency, the left-side part of the curve must be high and the right-side part of the curve must be low. Therefore, it is always a tradeoff. Read that sentence several times, if necessary. The more you reduce the transmissibility of the lower frequencies by upping the damping ratio, the more you increase the transmissibility of the higher frequencies. You cannot get around this. It is basic physics.

- Jtoby

 

[DG-FNR]

I will also try to slow down a bit, in case some readers skipped physics in high school (as I did).... ....You cannot get around this. It is basic physics.

A bump is a low frequency input.

It may be high-amplitude, but it is low frequency.

A steering input is an even lower-frequency input.

High frequency inputs are vibrations, and the best thing in the world for damping them out is SEAT FOAM.

So if we are talking about the harshness of the suspension in response to the bumps one sees in everyday driving, we are talking about low frequency transmissibility.

Yeesh! This is what happens when your only interface to the subject at hand is the Internet and textbooks.

DG

 

[jtmcinder]

A bump is a low frequency input.

It may be high-amplitude, but it is low frequency.

Nope. Not even close, big guy. At 45 mph (which is a useful speed to use for autocrossers), a bump is about 7.2 Hz. So if you aim your natural frequency to be around 2.0 Hz, as you have suggested, then anything above 2.828 Hz is above the cross-over. And anything above the cross-over enjoys less isolation as the damping frequency is increased.

Roll your eyes and sigh all you wish. Keep posting utter nonsense and expect more replies. The only positive that I can see to this exchange is that some lurkers might be learning something. I sure as heck don't plan to be your physics tutor.

- Jtoby

 

[Jon Lane]

A bump is a low frequency input.

It may be high-amplitude, but it is low frequency.

A steering input is an even lower-frequency input.

This isn't my field and pardon the intrusion but you just implied velocity. Varying amplitude, as in high-amplitude vs low amplitude, means high velocity vs low velocity, and are not dampers velocity devices? Doesn't a damper function against velocity instead of position or amplitude?

Meanwhile, high suspension amplitude is a large signal input plotted against time. Since the opposing body (the car) has finite mass, it seems to me that the spring is a position rate device, increasing resistance with both increasing deflection and rate, but tied to the period of the event.

Therefore it seems that both devices increase input to the opposing mass with increasing damping and/or spring rate. Which is also intuitive.

I went from stock to Koni/GC with only 450/200 springs. On the street the car rides much harder on all surfaces and runs out of travel sooner thru big vertical highway oscillations. The suspension simply telegraphs a hell of a lot more to the car, as I would expect.

What is my field to a degree is mechanical circuits. When you decrease compliance (higher spring rate) you must increase damping (higher shock rate) to maintain a critical damped Q, or how the system settles to zero after deflection. This is obvious. With masses unchanged, when you decrease compliance and increase damping you also raise frequency, which means you've lowered amplitude. Again, this is intuitive: Stiff stuff moves less than soft stuff under the same input and during the same period. Less movement per period means more energy transmitted to the car, no?

I'm taking a throw at this by trying to analogously translate another field that deals with masses, compliance, frequency, and damping, so apologies in advance if this is whacked...

 

[Jon Lane]

Incidentally, back in the day Herb Adams did quite well in Trans Am with, well, modified Pontiac Trans Ams. His approach, if I recall, was low unsprung weight, soft, long travel springs, and a ton of roll bar and damper. Even with about 500 lbs of big block iron in the front of the car, he kept the springs relatively long and soft, which seems to me also linearizes them a lot. Front TA coils are probably nearly a foot long when compressed.

Made sense to me at the time: Let the car respond to the input through the springs so as to keep all four down but keep it controlled with roll stiffness and planted with damping.

While a TA almost couldn't be more different than a DSM, and while I wouldn't know enough about today's autocross wisdom to save my bump steer, something about not sending the car toward the rigid suspension end of the scale with very high rate springs and high damping still seems intuitive.

Yeah, I know DSM's have shite for front travel. But I'm curious if jamming them way down onto 900lb springs is the only way to run fast.

 

[jtmcinder]

I agree with just about everything in your first post. I also started with the intuitive approach and description. In this thread, I added the technical side. This is one of those rare situations where intuition and physics agree, which is why I was surprised by the misconceptions expressed earlier.

Incidentally, back in the day Herb Adams did quite well in Trans Am with, well, modified Pontiac Trans Ams. His approach, if I recall, was low unsprung weight, soft, long travel springs, and a ton of roll bar and damper. Even with about 500 lbs of big block iron in the front of the car, he kept the springs relatively long and soft, which seems to me also linearizes them a lot. Front TA coils are probably nearly a foot long when compressed.

He is not the only one to take this approach with F-bodies. Sam Strano, two-time national champion autocrosser in F-bodies, also runs soft springs and low damping ratios. If you are a good enough driver to make the weight move to where you need it, this approach has several advantages. Watching him make a pig dance is a joy and my motto of "Pray for Rain" comes from trying to keep up in my Talon.

- Jtoby

 

[jtmcinder]

One other point with regard to transmissibility. What we have been talking about is how much of the input to the wheels is transmitted to the unibody, since we have been talking about springs and shocks. There is also the question of how much of the input from the road is transmitted to the wheels. (If you are really into suspension modelling, this shows up in the form of 2 df models vs 4 df models.) At low springrates, adding in the behavior of the tires isn't very important, but at high springrates, the action of the tires become very important. Tires are seriously underdamped, which means that they isolate the car from high-frequency inputs. Again, this is intuitive: underinflated tires with soft sidewalls give a nicer ride than Azenis Sports pumped up to 40 psi.

As to the frequency of bumps and such, I was using the simple case of a crack in the pavement -- as in: the thinnest possible bump you could hit. Then you apply some math making a bunch of assumptions about tire deflection etc as you have the car hit the bump at various speeds. When I model stuff, I always use accelerations of 1g and a speed of 45 mph. It makes a huge difference what speed you model, so road-racers should take my assertions with large doses of Morton's. The deal with SWMBO is no road-racing until our younger is 18, so I haven't spent much time on it.

I have no idea if discussing this in detail would be useful or just a way for some people to show off and/or pick fights. I leave it up to you (plural) to decide.

- Jtoby

 

[jtmcinder]

Besides, I don't have to guess at this stuff; I measure it directly off the car.

http://farnorthracing.com/newimages/suspension_velocity.jpg

Those are the live suspension velocities taken off the right front corner at three different National events - and one of them is Topeka. There are techniques that analyze this data and directly report on the suitability of the shock valving.

If you post these in ascii with a time index (or course), I will conduct the analysis and post the results.

- Jtoby