Front/rear track width effects

[rpmag]

I have accumulated some basic data on sporting car suspension. All are RWD with front or mid-engines.
Approximately 70% have a wider rear track and the remainder wider or equal front track. Switch to a racecar list and the percentage is reversed to 70% wider front track.
This racecar % could be accounted for by the need to accomodate rear rim width's, but is there another reason why there is such a predominance of wider rear track sporting road cars?

 

[BillyShope]

All right, it's 4 equations and 4 unknowns. Anyway, you can solve it.

 

[GregLocock]

"Take a car with 50/50 weight distribution, equal roll stiffness each end, same tires etc etc in a steady state corner (no wonder vehicle dynamics is confusing, we regard continual acceleration as steady state)

Now decrease the rear track. This forces more load transfer onto the front axle, causing more understeer. Does it get any more complex than that?"

Well, no one seems to agree, but Car Sim behaves exactly as I'd expect, more understeer with a narrow rear track.

Look at the plots, top link in the gallery at

http://www.geocities.com/greglocock/

or this link might work

http://www.geocities.com/greglocock/gallery/track_us.jpg

Cheers

Greg Locock

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

 

[hemipanter]

Billy, thanks, no calculate problem. Sorry to say that I lend out my "Race Car Vehicle Dynamics" and newer get it back, strange...

First, as we see there are different opinions about the effect of track width. I know it usually are, that’s the reason I brought the question up in the beginning (n:5) of this tread.

I have been looking around the

http://www several

times about this issue. And there are a lot sites claiming wider track front as a cure for under steer. However, these sites use a larger "how to fix the balance" sheet that seem to emanate from the same source.

Here is a cut from a SAE paper that at least is on of its own.

"When selecting the track width, the front and rear track widths do not necessarily have to be the same. For example, track width is typically wider in the front for a rear wheel drive race car. This design concept is used to increase rear traction during corner exit by reducing the amount of body roll resisted by the rear tires relative to the front tires [4]. Based on the corner speeds and horsepower to weight ratio of FSAE cars, this concept should be considered by the designer".

Greg, I don’t say I don’t agree with you concerning the "wider front track understeer theory". In fact I was inclined to think so myself, but the doors are still open.
I have also performed a number of scale model experiments,
showing the same as my full scale test already mentioned.
When it comes to driving the car some inch wider front track is hard to detect and takes consistent repeated driving on the skidpad. Because of a surgery I have not been able to drive my car until now this summer, so I am in lack of driving evidence.

I may mention that my connection at Ohlins schocks is of the same opinion as we been discussing this issue before.

Based on my experiment I made my own Tw calculation formula for an excel sheet, very much like the one for the wheel rate I sent you. Dont know if I dare to show it now....

Regards
Goran Malmberg

 

[BillyShope]

So, Greg, are you going to break the bad news to the people who developed Car Sim? Obviously, there's an error in their code. The equations in RCVD are straightforward and easily derived.

Goran, I'm sorry that I dismissed your example so readily. In looking at the numbers more closely, I see that you meant 249-261 and not 349-261. I then realized the differences are just measurement error. And, of course, this is reasonable. You could push the left front 5 meters to the left and the right front 5 meters to the right and the loads must remain unchanged.

 

[BillyShope]

Goran, to combine a couple of American sayings, when I shoot from the hip, I end up with my foot in my mouth.

As the front track goes to infinity, the front tire loadings would become equal. As the front track goes to zero, the larger load would go to infinity and the smaller to zero.

So, it is a matter of 2 equations and 2 unknowns. In this case, the left front should have gone to 249.4 and the right front to 260.6 kg.

 

[BillyShope]

At my age, I'm beginning to wonder if I'll get this right before I'm gone.

What I meant was "the larger load would go to positive infinity and the smaller load to negative infinity."

There! I hope that's it.

 

[hemipanter]

Agreed, I am 63....

Goran Malmberg

 

[GregLocock]

Billy, Um, well you probably know Tom Gillespie. I don't!
CarSim is his baby.

I've just reclaimed RCVD - I think you are focussing too much on the rear axle equation, you then need to consider how it affects the front axle as well (because the total roll stiffness decreases). I'm also a bit leary of using the simplified equations, p 682 is much better.

Incidentally it is worth pointing out before anyone else does that I just used the generic car model in CarSim as a baseline, which is not 50/50 weight balance etc etc.


Cheers

Greg Locock

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

 

[BillyShope]

I'm probably one contact away from Tom. (Instead of the average of 4 which is claimed for any two people.)

I thought the idea was to isolate the effect of track width. If so, we can't change total roll stiffness. If all of the other parameters remain constant (using either set of equations), the results are obvious. I see no reason why KsubR, of necessity, should be a function of track.

So, if you have the time, please rerun the comparison, being careful to vary only the track. I would be very surprised if the software gave an answer different from the equations. But, if it does, Tom Gillespie has a problem and I'll see if my friend can contact him.

 

[GregLocock]

OK, done, same link as before. You're right, when I upped the rear sta bar to bring the total roll rate back to the baseline value, there was no longer an increase in load transfer across the front axle, but (obviously) it did increase at the rear, so the oversteer contribution from the rear increased.

So back on track... (sorry)



Cheers

Greg Locock

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

 

[hemipanter]

Talk about being old and getting things straight. First I wrote 349 in place of 249, and then I discovered I had another
Rc number for the rear axle in my excel sheet where I saved my testing numbers. Anyway, this did not affected the compareable numbers to much in that the rear axle wheel did get a more even load (less load transfer) and the front axle stayed the same.

Let us say we have a car that weight 1000kg, 40-60% f-r
weight distribution. Tw 1600-1600mm, Wb 2500mm, Cgh 400mm,
Zero Rch, wheel rate 10kg/mm all 4 wheels and 1g side
acceleration. Then Wt=1000*1*400/1600=250kg distributed 125 each axle, creating a tire load of 75-325 front and 175-425kg rear.

Now, raise the front axles Tw to 2000mm, what would the 4
numbers look like? Anyone that have a program, this would be an esy task.

It really should feel nice to come to an agreement here since I feel a little confused.

Cheers
Goran Malmberg

 

[hemipanter]

I should add, with 2000mm front track, still use the same 10kg/mm wheelrate.
Goran

 

[BillyShope]

Goran, now that Greg and I have overcome our misunderstanding (a misunderstanding which would have been resolved within 15 seconds if we had been close enough to see each other wave his arms, instead of halfway around the globe), we would agree that your problem can be solved either by recourse to the software or the equations.

I would love to have the software, but as I now don't have an employer to buy such things for me, I must resort to my books. (I'm tiring of retirement and would love to go back to work, but nobody hires septegenarians for engineering positions. Just when you get halfway decent at doing your job, you have to retire.) So, as you're near my age, you might consider purchasing "Race Car Vehicle Dynamics" by Doug and Bill Milliken. There are other books, but I've taught from this one and am familiar with it.

 

[GregLocock]

Billy

Here's the ejicational version of CarSim.

The license is pretty agreeable.

http://www2.ing.unipi.it/~d5973/meccveic.htm

Cheers

Greg Locock

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

 

[BillyShope]

Thanks for your good intentions, Greg (though I thought I'd have to stop and learn Italian before I could go any further), but the software evidently doesn't support WindowsME, which I happen to be stuck with. (Can't even install Microsoft Office.)

Actually, I do have the Millikens' "Program Suite," so I'm not totally without computer assistance.

 

[hemipanter]

I am not going to bore you out, just one more Q.
Greg said...

"Take a car with 50/50 weight distribution, equal roll stiffness each end, same tires etc etc in a steady state corner (no wonder vehicle dynamics is confusing, we regard continual acceleration as steady state)
Now decrease the rear track. This forces more load transfer onto the front axle, causing more understeer. Does it get any more complex than that?"

Then, if we want the roll stiffness back as stated here...

"OK, done, same link as before. You're right, when I upped the rear sta bar to bring the total roll rate back to the baseline value, there was no longer an increase in load transfer across the front axle, but (obviously) it did increase at the rear, so the oversteer contribution from the rear increased."

Does this mean (bringning the roll rate back to the baseline value) that if we reduce the rear axle to half the Tw we have to increase the wheelrate by 4?

Regards
Goran Malmberg

 

[BillyShope]

Not at all. The roll stiffness due to springs is unaffected by track. Similarly, you would think that the roll stiffness due to a sway bar would also be unaffected by track. After all, we're talking about torque per unit of angle, so, if the sway bar mounts are unaffected by the track change, there should be no change in the sway bar's effect. But (and this is why I don't like using spreadsheets when the original equations are available), those who put together the software packages are inclined to assume that a proportional change in sway bar mounting locations will accompany a change in track. (I notice the Milliken software does the same thing.) So, it was necessary for Greg to go back and fiddle with the sway bar in order to maintain the same roll stiffness.

 

[hemipanter]

Great,thanks!
Cheers
Goran Malmberg

 

[BillyShope]

One more comment: The roll stiffness due to the tire rate changes with track and will, in itself, necessitate a small change in sway bar stiffness. I don't know about Greg's software, but this does not seem to account for all the change indicated in the Milliken software.

 

[GregLocock]

I didn't work out the change in roll stiffness, just fiddled with the a/r bar until I got the right overall roll rate. Sorry, it's a bad habit analytically, it's just that in ADAMS I always end up fine tuning by iteration.



Cheers

Greg Locock

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