Spring and bump stop tuning for bumpy autocross

My s2000 and myself have been getting faster and faster at my local autocross events but I've been coming up on an issue that I feel is really holding the car back. The car feels amazing in it's current setup on smoother courses and out on the road, however on bumpy courses I feel the rear is bottoming out which obviously makes the car very hard to drive.

The car is currently set up the STR class, it has Fortune Auto 500 series coilovers (definitely not penske level but much better than chinese garbage), front spring rates are 900 lbs/in, rears are 560, Karcepts FSB with the .250" center section and usually set on the stiffer end, rear sway bar is also Karcepts and adjustable but usually disconnected. 

The dampers have a 3-7/16" stroke including bump stop area. With the rear's motion ratio that gives a total possible wheel travel of just under 6". The standard fortune auto bump stops are really terrible 5/8" thick hard rubber pucks. I've already fixed this with conventional foam stops on the front pair. But the rear still has them. With the amount of preload I currently have that leaves about 3-3/8" of bump travel from ride height until hitting the bump stop. I am under the impression that my current spring rates are not stiff enough to keep the rear shocks from bottoming out hard in the bumps.

I am very happy with the balance of the car right now though so I really don't want to experiment with stiffer rear springs at this time. My thoughts are that I could use longer bump stops in the rear that are short enough to not come into play during normal suspension movements like roll, but long enough to give a softer deceleration on large bumps.

Will this work out like I'm thinking, or is this a waste of time? Obviously the real solution would be to get dampers with longer travel but I'm kind of married to these right now and they ride really nicely on the street. And I'm really not sure if the fenders could even accomodate more travel than it has.

In reply to freetors :

I’m not familiar with S2000 specifics, but that concept works well on a Miata, so it should be possible to make it work on your car too. Your spring rates sound like they might be a bit light though. I’m running 800/650 on my ~2150lb car and could almost use more rear spring. 

You might also see if anyone makes a shock mount for your car that allows more travel. 

Wrap a ziptie around each of your shock shafts opposite the bumpstops.  Drive car in representative fashion on representative surface.  Check zipties.  If they are now against the bumpstops the car is in fact bottoming out that corner.

Raise the ride height is the simplest fix.

In reply to Pete Gossett :

Wow. I had no idea Miata people ran that much spring. Are you autocross or track use? Some 30 second research shows that the miata motion ratios are a bit more direct than the s2000, so that would be really, really stiff!

I have a pair of 670 lb/in springs I could try out on the rear. Although I'm very happy with the current balance, I wouldn't mind slightly more rotation.

My particular coilovers have independent preload and shock length so I should be able to (theoretically) use all of the of the available travel. I am looking really hard at getting some karcepts spherical upper mounts though. The fortune mounts are very basic. You can see them and the new vs old front bump stop in this pic:

In reply to freetors :

Just autox, though I still have exhaust & wheels/tires before it’s ready to race. I just got it on the road Friday night & have our about 40mi on it so far. It’s definitely not too stiff, though you do feel the big bumps. 

How much preload do you currently have? What is the motion ratio? Do you know your corner weights?

Driven5 said:

How much preload do you currently have? What is the motion ratio? Do you know your corner weights?

The motion ratios are .589 in the front, .578 in the rear. Front preload with 900 lb springs is close to zero, maybe just a couple of turns, it droops 1-3/8". Rear preload with the current 560 springs is about .25", with 1.5" of droop. Last time I has it on the scales it was LF: 739, RF: 699, LR: 747, RR: 706. That's with weight in the driver seat.

Running a sanity check on the numbers, something doesn't add up here. Aside from your 16k/10k spring rates being rather unconventionally split for an S2k, if those are indeed the spring rates you have, then particularly in the rear either your stated preload is considerably off or your stated droop is considerably off.  Just based on the math, you should have an natural droop of LF: 2.38, RF: 2.25, LR: 4.00, RR: 3.78. Those rears numbers aren't even in the right ballpark.

If your rear preload is stated correctly, you should have considerably more droop than stated. This combined with the bumpstop will leave you with likely not enough available compression travel and have you hitting the harsh bumpstop. Increasing the preload and decreasing the unloaded shock length to keep the same ride height should help. 

If your rear droop (and available bump before bumpstop ) travel is stated correctly, you should have considerably more preload than stated. This may actually leave you with not enough available droop travel and have you harshly hitting full extension of ('topping out') the shocks.  If this is the case, decreasing the preload and increasing the unloaded shock length to keep the same ride height could help...Unless perhaps you are already (or would begin) hitting the bump stops too.

Regardless, I would definitely try the 12k springs on the rear, as it should help on all fronts...Including rotation.

Even if they come into play more frequently, I would expect longer and (initially much) softer progressive bump stops to also likely be beneficial. Eibach makes ones that might be good for your situation in 1.81 and 2.13 lengths.

In reply to Driven5 :

Interesting take on the numbers. One thing that's probably muddyiing up the numbers is the bushing resistance. The oem bushings do not rotate, they twist. I have them all set to have no preload on them at ride height. But I'm sure they add a lot of resistance. Just as an anecdotal note, when I was installing the 900 front springs I had to loosen the upper control arm. With a 900 spring in one side and lose UCA bushings, the other side untouched with a 670 spring, and the sway bar disconnected, bouncing the car felt identical in stiffness between the left and right sides!

Regardless of all that, I do indeed have a measured amount of droop as previously stated. I may have my preload amounts misremembered though. Also the spring rate I'm using are in the same ballpark as what Nationals people are using for their cars in str. I will take the rest of that into consideration. Thanks!

I'll agree with longer and more progressive bumpstops.  The ramp-up in rate at the end of the travel will be less sudden which should keep the car better behaved.  

As said the simple solution is ride height increase which it sounds like should be easy to do. I have no direct S2000 experience but given the motion ratios and what should be a reasonable weight distribution I would expect the spring rates to be closer together. Miatas run a pretty decent split, but the rear motion ratio is significantly better than the front. Adding bump stops can work but you need to be careful. It's effectively just an extra spring so if your dampers can't handle it, it can create some really odd problems. I did this on the rear of my undersprung BMW Z3 to keep it from bottoming out as much as an experiment. It does do that and limits body roll a little more, but over large bumps it slams into the bump stops then launches back up hard because it's wildly underdamped at that point. It's worth experimenting with, just make sure you know what you're getting into. If you do go that route, I'd actually probably start with the stock bump stops if you still have them, or some newer OE part since they tend to be well engineered for progressive engagement and durability. 6" of wheel stroke on an autocross car is a decent amount, I don't think that dampers with a longer stroke is what you want, if they even exist.

In reply to freetors :

Whatever effect the bushings do have on suspension rate, should affect the front and rear similarly.  You definitely have something else skewing those rear numbers. 

I'll take your word for it that your setup is similar to other national level STR S2000, but it does make me curious to better understand how/why such a heavily front (understeer) biased LLTD (which is far from the be-all-end-all of suspension tuning) should produce such desirable results.  Nearly equal weight distribution, nearly equal motions ratios, but heavily front biased springs and heavily (entirely) front biased bar(s). Most standard aftermarket spring rates for the S2k appear to be equal all around, with some having only a 1k-2k bias towards either the front, or more commonly the rear. Which is what I would expect. Even if running a square tire setup this screams of excessive understeer, and even more so if staggered, but if the stopwatch says it works....Then it works.

Your plan with the bump stops sounds good, and I agree that it sounds like you could use more rear spring rate (for an autocross-oriented setup), especially with a near 2:1 motion ratio on the rear suspension. I'm running 650lb/in all-around on a rather front-heavy 2300lb car with a motion ratio close to 1:1 (macstruts where the coilovers bolt to the back of the uprights).

Driven5 said:

In reply to freetors :

Whatever effect the bushings do have on suspension rate, should affect the front and rear similarly.  You definitely have something else skewing those rear numbers. 

I'll take your word for it that your setup is similar to other national level STR S2000, but it does make me curious to better understand how/why such a heavily front (understeer) biased LLTD (which is far from the be-all-end-all of suspension tuning) should produce such desirable results.  Nearly equal weight distribution, nearly equal motions ratios, but heavily front biased springs and heavily (entirely) front biased bar(s). Most standard aftermarket spring rates for the S2k appear to be equal all around, with some having only a 1k-2k bias towards either the front, or more commonly the rear. Which is what I would expect. Even if running a square tire setup this screams of excessive understeer, and even more so if staggered, but if the stopwatch says it works....Then it works.

I can't tell you exactly why it works, only my theory on the subject. I'm merely following in the footsteps of people who figured this out years. My car will likely never go to nationals anyway, but I do want to maximize what it can do without spending a bunch of money.

In regards to aftermarket spring sets, I feel most of those are just oriented towards lowering for looks.

You would expect the s2000 to have similar spring rates front and rear and indeed that's how they come from the factory, however for autocross with wider, stickier tires, and higher spring rates, the car becomes harder and harder to control at the limit. I think most people in STR would find that similar rates front and rear would be darn near undriveable. Part of the need for stiff springs comes from heavy lowering. AP1 chassis drivers need to lower the rear of the car quite a bit because the built in passive rear steer. The lowering puts it into a curve where there is no dynamic toe out, only toe in. Stiffer springs also further restricts the amount of toe change. In the front, lots of lowering seems to improve the dynamic camber change and puts the roll center at ground level or lower. I was personally very surprised after going from 670 to 900 lb/in front springs and had a reduction in understeer! An additional benefit of the soft rear with a weak or no swaybar is that it helps keep the inside rear on the ground so the torsen diff can keep up.

In summary, I (and probably many other str drivers) believe that the s2000 just works best, in an autocross scenario, with a front that is stiff and planted, keeping the tires with a favorable amount of camber to the ground, while keeping the rear moving freely enough to keep putting power down and just soft enough to keep it from spinning. I just don't think a softly sprung s2000 can respond fast enough to take full advantage of modern tire grip and the fast transitions of an autocross course. Another interesting note is that Honda changed the spring and swaybar rates basically every other model year throughout its life span. The original cars had very stiffly sprung and barred rear ends in comparison to the front and that partly led for their legendary propensity for oversteer antics. Every model refresh after that saw honda incrementally stiffening the front and softening the rear, pretty much what we do in STR, although we make it a little more extreme. If I was setting the car up for track work I probably would use a more balanced setup.

In reply to Driven5 :

I'm in the process of changing out the rear springs and installing the longer, softer bumpstops. I remeasured some stuff to verify. The rear had 2 inches of droop with the 560 springs and 1/2" preload. I eventually figured out that spreadsheet that I was using created by a Steve Edwards and downloaded from the internet was riddled with mathematical errors. It has good bones but some of the details aren't right. The new springs should have .45" of preload for the same droop. My numbers still don't totally match yours. Does your spreadsheet separate the unsprung weight from the sprung weight for its calculations?

I'm just using some simple formulas for this, no spreadsheet right now.

Wheel Rate = Spring Rate * ((Motion Ratio)^2)
Static Droop = Sprung Weight / Wheel Rate

For some reason I thought I put in 50-75 pounds of unsprung per corner, but reversing my calculations it looks like I just did it all as sprung since it was just a rough check. Revising with say 65lb of unsprung weight per corner, the numbers go to LF:2.17, RF:2.04, LR:3.65, and RR: 3.43. Note that I'm also using 896 rather than 900 on the front springs, so that would create a minor difference there as well. In a purely mathematical sense, 2 inches droop and .5" preload should now be roughly equivalent to 2.87 inches of no-preload droop. So between your numbers (preload and droop) both coming up, and my numbers (theoretical droop) going down, it does bring them noticeably closer together and more consistent with the difference to the fronts.  So you might be more 'ok' than I first thought on droop travel.

Oddly enough, the same math would predict only a .0013 difference between the back calculation from your measurements and what a 670 pound spring would theoretically be on the RR. Not really relevant to anything, just a coincidence.

I'm not understanding what your .45" preload on the 670 springs is for though. It seems to me that for the same 2 inches of droop, assuming the other factors contributing to the difference between theory and reality all stay the same, you would only appear to need about .25" preload. That's looking at it with either the same 110 lb/in added to the 670 as noted in my proceeding observation, or altering the motion ratio to get the 560 lb/in springs to yield a 2.86-2.87 static (no-preload) droop and then recalculating that with 670 lb/in springs.

Ehh... At a certain point you just have to say the math is close enough and focus on how it behaves in the real world. I try to just use the math to get a decent prediction of how changes will affect the cars behavior. By no means do I take is as an end all be all. If that were the case I probably wouldn't have put 900 lb/in springs on the front of my car!

Agreed...Although with motion ratios like that, using 900 lb/in springs isn't nearly as crazy as it sounds.