A locked diff - to much limited, not enough slip - can cause understeer. The OE diff is not adjustable on the RX-8, and no one would accuse the OE diff of having to much lock up.

A bad diff is evident by a significant increase in one-legging, and is typically accompanied by oversteer, when you light up the inside tire the back end sees a loss of traction.

 

Heh, thread resurrection.

I have the time before our next event...I might just increase the negative camber in the front and see what happens...

 

What about too soon?

 

I'm not sure. It seems to me it even happens when I have the throttle neutral or closed. I'll try to pay more attention this weekend and see if I'm right.

[edit] Oops, this was in repsonse to Mikes' observations about weight transfer.

Jason, that's what I thought.

Charles, pushing on exit under power is a different problem, and that can happen at most any speed with the RX-8 (in BS trim, anyway) if you have too much rear negative camber relative to the front.

My understanding is that can be lessened with a different diff that engages more progressively, like the OS Giken, but I have not personally experienced that.

 

That's my understanding as well. That's why changing the rear lsd is on my list of things to do for STX.

 

Now THIS is a great thread.

OK....Here's my take.

Myself and 3 other very fast and talented drivers have all concluded that the car has issues when it's under roll for longer periods of time in slower corners. We can make the cars handle wonderfully at just about everywhere we go except I'm still haunted for a particular style of corner---longer, slower corners where the car is in roll for a longer period of time. Our cars exhibit strong tendancies to U-steer at mid corner with ANY throttle inputs and can/will O-steer at track exit on these same corners.

We found this when we constantly battled the longer right hand International Horseshoe at Daytona and then the 2nd right hander horseshoe that follows (West Horseshoe??). The first horseshoe is a full 180 degree rt hander run in 3rd gear with our 5.12 ring and pin (most likely 2nd g in a stock 4.44 gear). We would have to wait soooo long to get on the gas that you'd get totally smoked on exit. We tried everything. Finally got it to be so-so with 3 mm's toe-in rear (and it didn't hurt our straight away speed because we were getting off the corners faster). The car was wicked quick in the faster stuff everywhere except those two LOONGER, SLOWWER turns. Honestly I was wo rqw5 through the bus stop and you could KILL anyone there (I offer this because it appears that other experienced drivers of our cars offer this as a major strength of these cars----yes? correct?). The bus stop is a high speed chicane type of turn with predictable oversteer at entry and mid and goes to rock solid when power is applied. These are the type of corners that I believe our cars are very strong.

Typical suspension setup that we run just about everywhere.
-2.7 FL, -2.5 FR, 1mm Frnt total toe out, 1 mm Rr total toe in. -1.9 LftRr, -1.7 RtRr. Big front bar. No rear bar. 550 fronts. 450 rears. Obviously we change lots of this setup once we get there but for the most part the car hovers around the numbers above (give or take).

There are only a few other corners similar to this where the car has revealed these same tendancies (strong sensitivity to u-steer with tiny throttle application at mid-corner) and they are few. Since I've driven this car on a zillion tracks we were surprised and perplexed at Daytona. Turns out that this type of turn is unique to the many tracks we run and we never experienced anything like this before.

A suspension engineer suggested that the roll center might be in the depths of Hades and that is what we're experiencing. Currently I have a suspension analyzer program on order (Performance Trends) and plan to dabble in the black arts of pickup point locations and find out what's technically going on. I know when we run the car's ride height low the angle of the upper control arm is all whacked out funky and this must have something to do with this issue.

Has anyone gone down this technical road? This is new ground for me and the team.

Cheers,

 

We use the stock diff and it has proved itself rock solid. Used an aftermarket diff before which went to full lock on accel and we chased our tails on that one for a long while. Cusco diff. Terrible for us. Went to stock diff and the data for inside vs. outside rear wheel looks normal. We also run the car as low as possible in the rear to reduce the rear roll center to eliminate any inner rear wheel spin. We pretty much never experience this anymore.

 

He lives!!

 

That was a really good post up there^^ Im scary near those alignments already and the rest im probably going to steal and test next season.

I wish this site had more users like Eric Meyer!

 

Thanks for the posts, Eric.

Your post does serve as an good reminder about the difference between "low speed" for track use vs. autocross. For Solo duty, my car (in STX trim) does just fine in second gear corners - I have understeer problems in first gear corners. Of course, this is probably due to our setups being optimized (hopefully ) for different contexts, but we both have the same problem - get far enough below the most common cornering speeds and we both get understeer.

I did some experimentation with softening low-speed compression damping on the front to lessen low-speed understeer, and it seemed like it helped, but it's nothing definitive yet.

As for the roll center question, I think Mark has been investigating this so perhaps he will chime in. On my last car, a 94 Miata, I made a 2D CAD model of the front suspension that helped me understand the relationship between ride height & roll center height, and could of also been used to investigate alternative pick-up points. I've been meaning to do the same for the RX-8 but have not done that yet. Of course, for the model to be usefull we need accurate measurements of control arms, uprights, existing pick-up points, etc. which I don't have.

I would really like to know "how low is too low" on the RX-8. Particularly for the stock pick-up points, since I can't change those under STX rules. I'm happy to collaborate any way I can, and/or am interested in hearing what you discover in this area.

 

UGH. When is someone going to check for excessive toe in at mid and corner exit with 2 video cameras?

I have always suspected this is happening but have never seem proof.

Maybe bushings would solve it?

FM

 

I remember my guys bump steering the car and not seeing any crazy toe issues. Are you suggesting there is a migration to front compression Toe-In during body roll?

We run a wickedly huge front bar and some 450+ spring rates. I've used up to 750 front spring rates and have been slowly reducing them as I added more front bar.

Momma didn't raise no engineer so this suspension geometry stuff is not second nature. Something I did find interesting is this: With a big front bar and a pretty good size spring (and looking at shock pot data travel) it didn't appear that the low speed, longer corner, body roll u-steer issue was because of toe-in (or out) bump steer. The suspension simply doesn't have that much travel. My guy says it's those crazy forces at work that I've yet to discover.

Will share more data when available.

BTW, the game plan is to run the car as low as possible and change the pickup points to something more optimum.

 

If the car does have toe issues in cornering, it might not show up in a bump steer measurement if you're just cycling through the suspension in a garage. Pretty much every modern car has tuned bushings to dial in tiny amounts of compliance toe-out up front and toe-in out back to make the car understeer. Additionally, the arm geometry is usually chosen to isolate the compliance effects between vertical ride motions and lateral cornering motion, so it might only become significant in small-radius low-speed corners when small changes in steer angle have a large effect.

I guess the only way to figure out if that is the case is with 2 cameras or a reasonably good model of the suspension with bushing stiffness numbers. 2 cameras might be easier.

 

You know, I just spent a full weekend on the Summit Point Shenandoah Circuit embarrassing many cars including a supercharged Lotus Exige S 240 (that's a point-by, video to come... ), and my really only complaint is the 2nd-gear corner understeer. (Well, OK, I'd like 500 more RPM too. )

There are three corners there -- Hammer, Hook, and Corkscrew -- that are slow and longish. Entry is fine, but the car pushes badly through mid-corner. I found myself dumping the throttle to get the tail to rotate. Left-foot braking is no help. Tire pressures were just-right (BFG R1s), though I might go up a PSI or two. On corner exit the car was glorious, getting the power down and steering easily with the throttle and just a little bit of counter-steer when needed.

I definitely have much more suspension work to do. I have Konis, but still run stock springs. Based on what I've learned here and in the camber thread, I need a lot more camber and may want more rear bar. I put in a MS front bar and drilled the rear bar for more stiffness, but I think more rear bar might be called-for to decrease understeer. I have Progressive street springs in a box waiting for installation.

I think if I can get the springs in, then get -2+ degrees of camber front and slightly less in the rear, and stiffen the rear bar, I can mostly defeat the understeer and make this thing even more of a g-monster. I wonder if I should ditch the MS and drilled stock bars and go with RB at both ends?

 

As said earlier in the thread, you might want to mount a camera pointing at the rear wheels.
I think there is a toe issue that will show up from mid corner on out while under power.
If so, you might just be able to replace one or two bushings and you are good to go. No more toe change and you could leave the rest of the car alone.
Think of the outside rear tire trying to push the car forward and see which bushing would be good to change.
I wish someone would do this and report back if it works or not.
FM

 

I just went and looked under my car to see where the squishiest looking bushings are. As far as I can tell, the upper camber control rod has a huge bushing carrier in it at the wheel end (at least quite a bit larger than the lower camber link and toe link), so that combined with the high and forward mounted toe-link causes compliance toe-in under lateral forces. So if I was going to recommend a bushing change, I'd recommend that one. Does anybody have pictures of the individual links and bushings off the car?

 

I refuse to beleive that this is a unique problem only to the RX8. This must be an issue that other cars have and has been confirmed and fixed. Where do we find this info guys?

Idea.....let me get back to you. HINT: I believe the MX-5 has the same rear suspension as the RX8.

 

Eric, any idea what ride heights the koni challenge MX-5's are using?

 

Same. Minimum of 3.5" to the lowest part of the car. We found this was the front of the nose or depending on how much rake you had in the car our exhaust just under the diff. The lower you run the car in the rear the lower in height you decrease your rear roll center and the less rear wheel spin you'll get under power on the inside wheel during lower gear turns. Remember, we run a 5.12 ring and pin so 2nd gear (which you almost never use unless something has checked up pretty bad in front of you and traffic goes to a near stop)....so 2nd gear can be a spin master.

I talked to my good friend Tom Long about this. Tom is VERY, VERY fast and experienced. Tom offered that the MX-5 has the same tendancy and that ANY throttle input in these types of corners goes to understeer. ANY. This would echo my experience with the car and with looking at the TPS data.

But I still want to figure out how to change the suspension mounting points so I can add throttle earlier and corner exit at 1, 2 or 3 more mph.

 

That's my next question -- how much rake did you typically run in your RX-8's? What about the MX-5 guys?

 

Here is how I understand roll center as it related to suspension geometry. Its not entirely my words since most of it is coming from a text book but I paraphrased most of it to fit neatly in this thread.

As cornering force is applied to the Center of Gravity (CG) it causes the car to rotate around the Roll Center (RC). The greater distance between te CG and the RC, the more leverage the CG will have against the RC and thus with the same cornering force cause the car to rotate even more around the RC. If the CG and RC have no distance between them then the car will not rotate at all around the RC. In most cars, the RC is below the CG and thus causing the chassis to roll away from the cornering force. If the roll center was above the CG then the Chassis would actually roll towards the corner force or in other words lean into the corner like a motorcycle.

The position of the roll center as already explained is controled by the position of the upper arm on our cars. If we can move the position of the lower A-arms then we could also use those as a way of adjusting the roll center aswell. The closer the upper arm becomes to the lower arm, the higher the roll center becomes thus causing less roll (think back to the comparison of Roll Center in a motorcycle).
When raising the upper arm thus moving it away from the lower arm, the roll center becomes lower thus creating more roll.

Hope this helps people's understanding of roll center. Any questions or disagreements I'd be glad to continue contributing my knowledge to this very constructive thread! I'm going to make a few diagrams once I get some time.

Steve

 

just got back. great question!

beers

 

Or also how much scrub inducing wheel offset are you running as well?