I got Silver Ghost back from the body shop on Friday, May 6th, installed the Penske 8300DA shocks and poly upper mounts on Saturday and drove to Lebanon, TN for the TRSCCA autocross on Sunday at the Nashville Superspeedway.
The car was great. I really missed it! I felt like I drove pretty well and paxed 4th of 118 drivers on a short, tight, low-speed course in the newly-paved infield lot. It was especially cool because the One Lap Of America crowd were racing around parts of the tri-oval and the infield road-course track simultaneously.
I set the knobs to match the targets I’d given Penske. Big thanks to Penske’s Steve Horn for guiding me through the process and helping me make the many decisions necessary.
And big thanks to Dennis Grant (DG) and his Dynamics Calculator that allowed me to figure out the shock forces to specify.
I’m going to explain a little about how I went about this whole shock specification thing for those who may be interested. (Some of this may have been discussed in previous posts.) I’ve now done this for several friend’s cars as well as three of my own. I wrote Penske a spec describing what I wanted with targets (force at velocity) and piston type (double digressive) and tolerances and gas pressure. Also required was shock body length (to be legal in Street) and reservoir hose length. Steve and I discussed my spec and he declared it doable as long as I was realistic, which I took to mean not to be too much of a corporate engineer who would throw a tantrum if the numbers were not exact. The impression I got was that the forces I was asking for were low relative to the design range of that shock and piston combo.
Those familiar with DG’s on-line “book” will know that he recommends a starting point for shock forces at the digressive knee of the curve to be those that produce a total of 65% Critical Damping*. His calculator allows you to figure out how much force that would be.
Except that DG isn’t talking about Street class autocross cars. He’s talking about cars that have optimum, autocross-racecar-stiff springs, which he defines as ~2.2Hz** in the front and 10% higher in the rear. My Corvette is relative stiff for a production sports car but is only ~1.7Hz. Should we still damp at 65% Critical?
In Street class we are stuck with two key things: 1) the stock weight, and 2) the stock springs.
Way back in autocross pre-history, like maybe the 1980s(?), someone figured out that we can do a couple of legal things with shocks in Street class to partially make up for the deficiency in spring rate. Those two things are 1) use additional shock damping to mimic stiffer springs in dynamic conditions, and 2) use high gas pressure to mimic a statically stiffer spring.
DG states that he’s not sure of the practical limit for shock damping of a Street-class car in order to mimic stiffer springs but thinks that it may be somewhere above 100%Critical. In my old C5 I tried readjusting my shocks for 90%Critical as a guess. This required significant reductions in forces compared to what I’d been running, so I’d probably been at 100% or maybe even more. In fact, I couldn’t even get down to 90% in all cases they were valved so stiff. (I had 4 numbers to set: Front and rear bump, front and rear rebound.) The car gained grip, was easier to drive and got faster. (Just in time for it to be obsoleted in B-Street by the M2 and Supra.) So, that’s what I specified for the Penskes: 90%, with, hopefully, some adjustability above and below.
DG’s calculator seems to calculate the total damping and then allocate the damping at about a 2 to 1 ratio between Rebound and Bump. I decided I wanted a 60/40 split, so I did a little trick. I totaled bump and rebound forces from the Calculator and reallocated the total to produce that split. (Remind me to ask DG if he thinks that makes sense.)
I’ve discussed the allocation of damping between the rebound and bump directions in a previous post here where I reviewed the historical 4 to 1 ratio vs. more modern ideas of how to split up the damping. I don’t have a scientific reason for choosing a 60/40 split instead of DG’s 2/1 split, just experience that tells me that more compression damping is better for transient response, though possibly not good for grip, especially on a bumpy site. I also had Penske put in all the gas pressure these shocks can take and still be long-term reliable, i.e. not stress the seals too much.
The weekend after Nashville I trailered Silver Ghost to Charlotte for an SCCA Pro-Solo event. So far I’ve never done particularly well at Pros, I have a single trophy and no class wins, but I needed the seat time with the new setup and Pro-Solos are lots of fun. The one other car registered in B-Street was a no-show. I was put into Bump class against a Tesla Model 3, also by himself in the new Electric Vehicle Experimental (EVX) class. I was able to prevail on corrected time, and I’ll take the win, but was not proud of how I drove. I don’t have any idea how you can come up with an accurate index for a new class like EVX. Electric cars with Street Touring suspension allowances and no tire/rim limitations except whatever you can get under rolled fenders. Who knows how fast those cars should be?
I drove the course okay, but my starts were terrible. Reaction times were slow and 60 foot times worse. In the total of 12 starts over two days I managed to massively and excessively spin the rear tires 12 times. Giving away 0.4s in the first 60 feet made it interesting to watch the Tesla rocket away into the distance beyond the light tree while my tires churned away!
I started the event with 97 runs on the tires and about 4/32nds of tread. Grip was only okay on this surface that’s known for high grip. I suspect the tires were heat-cycled out. The next set has arrived and will go on before a big event next month, the Peru National Tour, where we have seven drivers registered in three Supras, an M2, an RS3 and Silver Ghost. That will be a good test.
*A spring/mass/damper system, like the ones each corner of your car (if you divide the total sprung weight into components at each corner) is critically damped if, when disturbed, the system returns to the static position in the least amount of time and with no overshoot, i.e. no bounce. So, %Critical is a measure of the rate the damper takes energy out of the system as it resists the motion. No real-world system is 0%Critical because there are always energy losses with any mechanical motion. Shock absorbers on cars can be configured to produce a wide range of %Critical, even more than 100%. The more damping above 100% Critical the longer it takes to get back to the static position after being disturbed, like a tire being smacked by a bump.
**The natural frequency in Hertz, or cycles per second, of a spring-mass-damper system is the correct way to talk about suspension stiffness. The spring rate is one input. Another is the sprung mass on that corner of the car. Motions ratios (mechanical advantages of springs and shocks given the specific geometry) must be accurately known. Damping has an effect on natural frequency as well, especially at the high damping values sometimes run with racing shocks. The natural frequencies for front and rear are also produced by DG’s Calculator. The stiffness of the suspension largely controls the transition rate capability of the car, that is, how fast it can change direction.
4 thoughts on “An Autocross Season- Part 12: Back In The Mix”
What did you use for tire stiffness? Your own measurements? DG’s sample number? Or found info online for the two different brands?
After some internet searching and finding some tested numbers, I use 1200 as a generic number for 200TW autocross tires. I haven’t found any actual as-tested numbers for them specifically, but the difference between 1000, 1200 and 1500 in the final result is very small given how low the steel spring rates are in Street-class cars.
> DG’s calculator seems to calculate the total damping and then allocate the damping at about a 2 to 1 ratio between Rebound and Bump.
DG’s calculator works out 65% critical in both directions, bump and rebound, by working out the critical damping for rebound as if the unsprung mass is fixed to the ground, and then for bump as if the sprung mass was fixed in space – effectively “flipping the car over”.
This is an elegant and immediately comprehensible model, which, unfortunately, doesn’t work. It gives the correct values for rebound forces, but tends to generate bump force values that are a little too soft in practice.
You can consider the bump forces a “floor” value, as in “go no softer than this”.
Adding 10-15% to the calculated values seems to work out about right.
Another way to look at it is to select the shimstack that produces the calculated compression values, then use the adjusters in the canister (which are additive to the shimstack) to tune in the rest.
Testing and suspension speed histograms will provide the final answer, but I expect ideal damping to be 65%-70% critical in rebound, and then what my calculator reports +10%-ish in compression.
Given this information I’ve got to go back and figure out where I stand with the settings we’ve settled on. Thanks.