Optimum Oval Track Anti Squat? 4

[GaryJohansen]

Hi, I'm new to this forum and have really enjoyed reading the threads. Hopefully someone can help me with a couple of questions...

I'm involved with a pavement oval track car (2006 Lefthander Chassis) running on a 3/8 mile medium banked oval track with approximately 500bhp. Car weighs 2900 lbs with slightly over 50% rear wgt percentage.

The rear suspension is a 3-link with top link at 7 deg down toward the front (27" long). The lower trailing arms are both 28" long. We've been looking at trying to introduce a little roll steer to help the car turn in the center, but are also trying to avoid a loose condition off the corners. The right lower trailing arm is 2 deg up to the front and the left lower trailing up is 2.5 deg up to the front.

Currently the car has about 40 percent anti squat. This seems like too much as the car is very unpredictable regarding tight and loose each week and the rear sustpension really never seems to take a set during acceleration. Rather, the rear of the car seems to be bouncing with the track as if there were no suspension. Seems like too much AS!?

Does anyone have any experience with this type of car regarding optimum antisquat? Also, is there value in looking at antisquat in the Left Rear independently of the Right Rear? Currently the LR has more antisquat.

I really appreciate any help and insights anyone can provide.

Thanks

 

[BillyShope]

If you are achieving the anti-squat with an instant center position that is relatively high, I can understand the conditions you describe. What I'm getting at is that the 40% anti-squat line passes through the rear tire patch with an upward angle, so that 40% anti-squat could be achieved with the instant center at virtually any height. You could have 100% anti-squat and no problems IF you keep the instant center height down at approximately the tire radius. With a higher instant center, the links are "available" for vertical loads and you can get the kind of "bouncing" you describe.

 

[GaryJohansen]

Thank you very much for your insights. The IC is at around 11" above ground and the rolling radius is around 13" above ground. IC is about 80" forward of rear axle.

I should have mentioned earlier that the top link also has a spring included, with several rates available. Of course, it is intended to reduce the impact of rapid application of torque when the throttle is picked up. I apologize for not mentioning this in my first post. At any rate, my thinking is that with antisquat levels above 40%, even 45%, doesn't the spring in the top link start to behave as a factor in the overall wheel rates? And, with no damping in that spring, ...?

You can probably tell that I am not a suspension engineer, but I find it fascinating.

Thanks again for your time and insights.
Gary

 

[BillyShope]

Gary, things aren't working out quite right. If the IC is 80 inches forward and 11 inches off the ground and you have 40% anti, this means that, with a wheelbase of, say, 112 inches, your CG is at a height of 38.5 inches. It looks like you're closer to about 90% anti-squat.

Looks like your upper link is mounted almost 8 inches above the axle centerline and the lower links are a bit over 5 inches below. Right?

I don't see any direct relationship between anti-squat and the upper link load seen by the link spring. That is, you can change anti percentage by leaving the upper link alone and moving the lower links or you could accomplish the same thing by leaving the lower links alone and moving the upper.

Damping is at the wheel and is unaffected by the number of springs "upstream," but that preloaded spring in the upper link would have little effect on wheel rates, anyway. Personally, I dislike the whole concept of adding springs to suspension links.

 

[GaryJohansen]

Billy,
I'm trying to find the best way to show you a picture of the excel spreadsheet I've developed. Perhaps my understanding of antisquat is wrong. IC is the intersection of trailing arms, correct? anti squat is the height of a line from tire contact patch to IC as it passes under CG, diveded by the CG height? Car is 105" wheel base.

I have to agree with you regarding springs in the suspension links. I see all kinds of things on cars at the track. Most of them look pretty crazy.
Thanks again for your time.
Gary

 

[GaryJohansen]

I don't actually know CG height, but many say for these cars that it is around camshaft height...15" above ground.
Gary

 

[BillyShope]

The percent anti-squat would be:

100 (11/80) / (15/105) = 96.25%

 

[GaryJohansen]

Billy,
It would appear that the description I provided at 22:07 for calculating anti-squat was incorrect. Thanks for correcting me.
Gary

 

[BillyShope]

As for suspension asymmetry: It would be better if the lower RIGHT side link had the steeper angle. This would tend to cancel driveshaft torque and provide more equal rear tire loading during forward acceleration.

I discuss this on Page 15 in my blog:

http://home.earthlink.net/~whshope

While this is directed to modification of the Mustang 3link for dragracing, the spreadsheet, when filled with your car's specifications, will provide the proper answers. Effects on braking performance are minimal. (Jaguar used a similar asymmetry on their C-Types.)

 

[BillyShope]

I'm sorry, Gary. I shouldn't post until after breakfast. Disregard my last post, for, while such a setup would be fine for a car on a road course, you wouldn't want it on an oval.

You might, however, keep such a trick in mind as a means of correcting a handling problem on the oval. If, for instance, the car is loose on exit, you'd want a steeper angle on the LEFT lower link.

On an oval car, you're certainly not worrying about the driveshaft torque unloading the right rear! But, as I indicated, asymmetry can be used to correct problems at the turn exit.

 

[GaryJohansen]

Thanks again for all your help. Since yesterday I did more searching on the internet and found your article about counteracting the effects of driveline torque on drag race cars.

Your comments about the differential angles of the lower trailing arms really resonate with me. In fact, that is where I got interested. We have been doing a lot of fine tuning of these differential angles as we try to optimize the condition after the apex and off of the corners. I'm just struggling with the balance of that vs. the antisquat. I mentioned to the team I'm associated with that we might want to introduce a very small amount of static "thrust angle" into the rear end while minimizing the roll steer. Sorry for bad terminology. By thrust angle I mean the angle that the rear end is off from perpendicular to the longitudinal axis of the car.

Last week, they raised the heim joint at the front of the top link to reduce the antisquat and the driver liked that and the car was much better. I'd like to run some sort of DOE on this when we get time. (I've run a number of DOE's in the past as a Diesel combustion/performance engineer, but have never applied it to race car testing).

Our car is not too far off. This is our first year in late models. We're 4th in points with 1 feature win and 1 fast time (and a bunch of near misses on fast time). Last year we ran in a class based on late 80's GM metric frame and were very dominant. Lots of learning this year.

Thanks again
Gary

 

[GregLocock]

Gary, your idea of using a DOE is excellent. Make sure you use a quadratic experiment (3 levels) as most geometry effects are arcs of circles.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[GaryJohansen]

Thanks very much! I've been wondering about the 3 level DOE and trading off the risk of missing an optimum or diminishing impact, vs. complexity in our first try.

Seems we are pretty close to the best combination of static rear steer and roll-steer for our current setup.

Perhaps since you and Bill were kind enough to respond to my questions, I could put another in front of you. Is there any rule of thumb or best thought on how to interpret shock data from the car? I'm thinking we can use shock travel to get a lot of "virtual channels" such as dive/squat and roll, roll centers, shock velocity/acceleration/jerk, etc, etc. It seems the possibilities are numerous. Is there value in using shock dyno sheets, shock gas pressure, and spring rates to find a way to infer wheel loads around the track (assume of course the motion ratios are understood)? I recall reading in another thread on this site that it is possible to build 4 shocks that all feel completely different to the driver but look the same on the dyno. Is that due to the relatively slow velocities on the shock dyno?

Thanks again,
Gary

P.S. Reduced anti-squat helped us on Friday night. Second in fast heat, 3rd in feature...Might have been better.

Woulda, coulda, shoulda...Racers' 3 favorite words!

 

[patprimmer]

You forgot almos, and except for.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[GaryJohansen]

Ha! How could I forget the 2 most important ones...!

 

[GregLocock]

If you are close to optimum you'd probably get away with a two level experiment, which would be a huge reduction in complexity.

What can you infer from shock travel?

Load transfer in its various forms, is your best bet. What instrumentation are you using to measure the travel?

If you are using stringpots or LVDTs then I've found that differentiating once, to get velocity is OK, but acceleration is not much good.

You'll find that the actual force you get from your shock absorber often bears little resemblance to the curve you measured on a shock dyno. With circuit racing that might not be true so much - you use lower velocities than we do. (I get up to 5 m/s)



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[GaryJohansen]

We're going to be using LVDTs. Thank you for your insights. Load transfer was the main thing I was looking for. And, perhaps, to get some idea of front and rear geometry by analyzing the kinematics with the shock displacements. My simplified model doesen't take into account friction or deflections but it might give us some more insight.

Thanks again
Gary

 

[GaryJohansen]

I should have said:

I think we are getting close to optimum anti-squat vs. roll steer.

Probably nowhere near the optimum setup overall for the track and driver....

thanks
Gary

 

[GaryJohansen]

Anyone have any experience with 3 link suspensions where the rear attachement point is behind the rear axle? Currently our car has the rear attachment of the top link approximately 6 inches rear of the axle CL.

In comparing the kinematics of this setup with a shorter top link with rear mount directly above axle CL, there does not appear to be a significant rate of change for the antisquat as the rear suspension goes into squat.

I've heard that some oval track drivers don't like the feel of the car with the rear attachment of the top link behind the rear axle. I have no idea what the other details of their chassis setup was, so I'm taking this input with a grain of salt.

Again, any input is appreciated.
Gary

 

[GaryJohansen]

Wondering if someone could help me carry the anti-squat thinking some more. I mentioned earlier in this thread that we had problems with forward bite and that the rear would not take a set. One item that we found was that the pitch center was located very far to the rear of the car. This was due to the geometry of the front suspension, primarily in the lower control arm. Anyone have any recommendations for pitch center for a pavement oval track or road race car? Should the pitch center start below the CG and move forward as the rear squats? Also, is there a limit as to how LOW the anti-squat percentage should go when the rear does actually squat?
Thanks very much,
Gary