Poll: Weight forward or weight rearward for understeer?

[murpia]

Here's a question inspired by Pat Clarke's FSG column here (http://www.formulastudent.de/public-relations/fsg-news/news-details/article/pats-column-january-2/):


The bugbear of FS car handling is understeer, teams sometimes making great design compromises (like driveshaft angularity) to move some weight to the front wheels.

Who here has experienced this situation, increased front weight percentage reducing understeer?

'Classical' theory would suggest that increased front weight percentage would increase understeer, as the front tyres experience greater vertical loads. However, that doesn't take tyre temperature into account... For an FSAE car does anyone have data they're willing to share that suggests that increased front weight percentage increases front tyre temperature and reduces understeer?

Regards, Ian

[PatClarke]

Good one Ian =]
I haven't yet heard a compelling argument either way from teams, apart from the fact that most are chasing understeer.
The usual 'cure' is to try get both the front and rear wheels tucked under the drivers armpits.
Cheers
Pat

[ben]

We did our first FS tyre test last week and the conditions were shite plus the compound was pretty hard. The immediate comment relative to the baseline (Avon) was that the car had picked up understeer.

http://www.ubracing.co.uk/

The driver was very specific that he'd saturated the tyre by steering in the same way and then picked up grip as he wound lock back off.

I would imagine most vehicle dynamic theory goes out the window if the compound hasn't "come in" because it makes the tyre incredibly sensitive to steering rate.

Ben

[C.Zinke]

I think that a normal FSAE car icreases understeer with more weight at the front. This is because the effect of an anti-roll bar is quite the same. If you stiffen up the front of a FSAE car understeer increases.

Sometimes it is possible that more weight increases the tire temperature and therefore the grip. Personally i have never seen that. Our car is at 220kg.

BTW Pat's column is really interessting this month again. It shows that it is a good choice to build a front-heavy car with no differential. I'm the same opinion. Nevertheless i think you can build an even faster car if you make it noticable rear-heavy. But then you need a diff an wider tires at the rear to avoid overheating.

[murpia]

Originally posted by PatClarke:
Good one Ian =]
I haven't yet heard a compelling argument either way from teams, apart from the fact that most are chasing understeer.
The usual 'cure' is to try get both the front and rear wheels tucked under the drivers armpits.
Cheers
Pat
I think a lot will depend on the definition of the understeer that the teams are 'chasing'... Sounds to me like turn-in response is the problem if shorter wheelbase / lower inertia is working well.

Similarly, comments such as 'stiffening the front bar reduced understeer' seem to me to imply a turn-in response problem rather than a corner apex understeer problem.

Regards, Ian

[murpia]

Originally posted by ben:
The driver was very specific that he'd saturated the tyre by steering in the same way and then picked up grip as he wound lock back off.
Higher cornering stiffness construction, perhaps, with the driver going over the lateral grip peak?

Regards, Ian

[C.Zinke]

Would be typical for a radial tire. Is the dunplop radial? The Continental Formula Student tires are also radial. They have a very noticeable peak. Its really easy to feel it in the Steering wheel. Its a complete contrast to Hoosier i.e. (Continetal just did 4 new construvtions that we like to test soon.)

[ben]

Originally posted by murpia:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by ben:
The driver was very specific that he'd saturated the tyre by steering in the same way and then picked up grip as he wound lock back off.
Higher cornering stiffness construction, perhaps, with the driver going over the lateral grip peak?

Regards, Ian </div></BLOCKQUOTE>

Yeah, quite a stiff radial both vertically and in plane belt stiffness. We ended up running pretty low pressure. We're working on some new options.

Ben

[PBandJ]

Ben, It looks like, from the pictures, it was damp for the Dunlop testing. That would make any tire very 'peaky' and a handful to drive.

Also- what kind of wheels are those (the black ones) on the UB car?

[ben]

Originally posted by PBandJ:
Ben, It looks like, from the pictures, it was damp for the Dunlop testing. That would make any tire very 'peaky' and a handful to drive.

Also- what kind of wheels are those (the black ones) on the UB car?

Agreed about the conditions, but the Avon's were pulling 1.5G and driving ok. We just needed a reference to start further development as FSAE stiffness requirements are well below anything we make as standard.

I think the wheels are Compomotive CXR's:

http://www.comp.co.uk/wheels/wheels.asp?size=13%22+Motorsport+Wheels

Ben

[flavorPacket]

Understeer in a rear heavy car is almost always from the driver's inability to trail brake well rather than the race engineer's poor tuning abilities. I'm not saying that cornering stiffness, steering ratio, inertia, etc don't affect turn-in understeer, but the driver has been #1 in my time in FSAE.

Look at the past 50 years of 911 race cars, for example. Thay all understeer like pigs on turn-in if you don't transfer the weight properly.

If I were to design a FSAE car to understeer horribly, I'd put 80% of the weight on the front with skinny tires and a super stiff ARB. It'd be quick in the slalom, but in the sweepers it'd be unbearable when the fronts saturate so much earlier than the rears.

[B Hise]

anyone ever calculate the yaw moment from the diff on decel?

-Bryan

[Mark TMV]

ARB reduces overall grip on the end of the car it's mounted on. This makes me question the reason to have a front ARB for an autocross car, especially on a rear heavy car. Rear ARB is still needed, but only as a chassis tuning tool.

[ben]

Originally posted by Mark TMV:
ARB reduces overall grip on the end of the car it's mounted on. This makes me question the reason to have a front ARB for an autocross car, especially on a rear heavy car. Rear ARB is still needed, but only as a chassis tuning tool.

But if you use a RARB you run the risk of hurting corner exit traction because you're transferring the load at that end.

Not to say it's wrong but I've never seen many cars running a RARB without a FARB, and I've seen lots of cars with a FARB only.

Ben

[flavorPacket]

Originally posted by Mark TMV:
ARB reduces overall grip on the end of the car it's mounted on.

That's quite a statement.

[Pete Marsh]

I don't know about the drive shaft thing Pat? I don't think I've ever seen a car with the shafts angled backwards. Its pretty easy to put a bigger gap between the engine and diff if you want to get the wheels further from the weight. We swing them forward to meet our distribution and wheelbase targets.
Power on understeer acompanied by low front tyre temps is hard to fix with the chassis tuning without making the car slower on the watch and the drivers hate it. (requires absolute comitment and acuracy of line to be fast).
In this case I would move weight forward.......or lower CoG, or lengthen wheel base, or fit a smaller front tyre, or wider front track, or reduce weight/increase power, or some tricky diff, and more I'm sure.

Fun is'nt it,
Pete

[VinceL]

Great topic Ian. I noticed that line in Pat's Corner as well. I'm currious as to what teams he is reffering to when he mentions that they are trying to move weight forward.

We've never tested weight bias on track with ballast. I never bothered because the weight balance needs to match the anti-roll torque distribution well for good balance...at least I think so. This would mean that every time you change the weight bias you should change the anti-roll torque bias with it to rebalance the car in terms of understeer and oversteer. So I never bothered with that test because I thought that we would just be chasing car balance all day. I have however done a bunch of steady state sims changing weight balance and antiroll torque bias at the same time to try to get an idea of the 'ideal' weight balance for an FSAE car. I won't share the results of those tests, but I can say that the answer isn't simple. It depends on a lot of things, particularly corner size.

One thing that more front weight would help with is turn in response. So maybe what those teams Pat mentioned were experiencing is a better balace on turn in leading to an overall better corner. The entry is the most important part of the corner after all.

I kind of got off topic a bit. But to answer the original question in a steady state situation more weight on the front should increase understeer. But on turn in should decrease it. (Chuck Hallum had an interesting theory on that.) With the small sizes of FSAE corners the turn in is extremely important. So maybe you wouldn't even care about the steady state effects...now that I think about it, I should of tried this test on track after all.

Cheers

[flavorPacket]

Pete, the 1997 (I believe) McLaren had back angled shafts. ScarbsF1 has some great pics.

There are plenty of ways to improve turn in with adding weight to the front (and thereby reducing, relatively, the rear's longitudinal force capacity).

Dampers, tire pressure, driving style, Ackermann, ARB etc can all accomplish that task. AND they can do it without making the rear lighter and thus hampering longitudinal acceleration.

[Mark TMV]

Re: flavorPacked,

quote:
Originally posted by Mark TMV:
ARB reduces overall grip on the end of the car it's mounted on.

That's quite a statement.

I should've cited it, it's actually Claude's statement...

[Charlie]

We were always trying to get more weight rearward... quite the contrary to Pat's statement.

But I'm an old man now, in FSAE terms. http://fsae.com/groupee_common/emoticons/icon_wink.gif

[flavorPacket]

Originally posted by Mark TMV:
I should've cited it, it's actually Claude's statement...

In steady state, you are right. But FSAE is never steady state.

Yes, an ARB will increase normal force discrepancy across a wheel pair. BUT, if you use a stiffer ARB to get the car turned in faster (and hopefully settled faster), you may be able to have more grip sooner. So what you lose from load sensitivity will be offset by minimizing transients and improving normal load variation.

[Matt N]

I don't know how any of you F1 junkies missed this one -

http://img169.imageshack.us/img169/1710/upviewqd9.jpg

F1 cars are currently trying to get weight forward and off the highly-loaded (blistering) rears. Its hard to see but both the 2007 and 2008 Ferrari (linked above) have the driveshafts angled to move the transaxle, engine, and the whole mess forward.

Matt

[A Richards]

Originally posted by Matt N:
I don't know how any of you F1 junkies missed this one -

http://img169.imageshack.us/img169/1710/upviewqd9.jpg

F1 cars are currently trying to get weight forward and off the highly-loaded (blistering) rears. Its hard to see but both the 2007 and 2008 Ferrari (linked above) have the driveshafts angled to move the transaxle, engine, and the whole mess forward.

Matt


Im glad that you can see driveshafts in these pictures cause i sure as hell cant!, All i can see is a-arms.

[PatClarke]

Quote 'I don't know about the drive shaft thing Pat? I don't think I've ever seen a car with the shafts angled backwards'

Pete, there certainly was a car at FSAEA with the drive shafts angled backwards. It certainly wasn't a front runner. Ron T and I are going to visit that Uni soon to give them a little guidance.

And Ian, this is the most intelligent debate I have seen in here for some little time. Thanks
Cheers
Pat

[murpia]

Originally posted by VinceL:
We've never tested weight bias on track with ballast. I never bothered because the weight balance needs to match the anti-roll torque distribution well for good balance...at least I think so. This would mean that every time you change the weight bias you should change the anti-roll torque bias with it to rebalance the car in terms of understeer and oversteer.
I think you are right regarding matching the weight distribution and the roll stiffness distribution. But, with no minimum weight limit in FSAE I would expect adding ballast just to make you go slower... Once you lay out your car your weight distribution will be fixed unless you can slide your driver back and forward (not such a bad idea to try...) The exception would be if your tyres aren't generating temperature properly, in which case you need different tyres (width, compound, pressure, construction etc.) not a heavier car!


One thing that more front weight would help with is turn in response. So maybe what those teams Pat mentioned were experiencing is a better balace on turn in leading to an overall better corner. The entry is the most important part of the corner after all.

I kind of got off topic a bit. But to answer the original question in a steady state situation more weight on the front should increase understeer. But on turn in should decrease it. (Chuck Hallum had an interesting theory on that.)

I would be very interested to debate a theory of improved turn-in with forward weight... Any more details?

Regards, Ian

[VinceL]

You can test weight bias on an FSAE car if you set your base line senario as the case with the ballast on the drivers lap. Send the driver out with the ballast on his lap - or behind the seat if it can fit - and re-balance the car. Then have him do simple maneuvers like step steers or chicanes. Then strap the ballast to the nose of the car and repeat the tests.

Like I said I never tried increasing front weight percentage on track so I don't know first hand what it does. You can read SAE paper number 2002-01-3302 for Mr. Hallum's view on it. It's one thing I plan on testing soon on computer though.

Cheers

[flavorPacket]

Originally posted by murpia:
I would be very interested to debate a theory of improved turn-in with forward weight... Any more details?

Regards, Ian

More weight on the front = more force from the front tires (assuming proper tuning and no saturation). If you build that force sooner in the corner, you will create a larger yaw moment from the front and thus turn in quicker.

Am I thinking about this correctly?

[murpia]

Originally posted by flavorPacket:
More weight on the front = more force from the front tires (assuming proper tuning and no saturation). If you build that force sooner in the corner, you will create a larger yaw moment from the front and thus turn in quicker.

Am I thinking about this correctly?
But, more weight on the front = more mass (inertia) to accelerate to yaw the car...

If you had a perfectly linear tyre, then the two effects sould cancel each other exactly, so no change in understeer. Of course with real tyres with load sensitivity, you are supposed to get a little bit less lateral force and therefore you have the weight forward = understeer effect.

I think it must be another effect, perhaps related to tyre relaxation length / slip angle build up?

Regards, Ian

[flavorPacket]

I don't know, Ian. I can think of some situations where yaw inertia can remain constant while weight distribution moves forward (just flip a and b in a standard plan view). It will clearly depend on the exact numbers, but I'm willing to bet that you may be able to get away with it.

Regardless, from a FSAE-only perspective, more weight in the front will yield higher tire temps, which probably has the greatest effect out of everything.

[The Stigg]

Well after thinking about it quite a long time, I would definitely say that putting more weight to the front is the way to go in case of untersteer in turn-in.

In my opinion one of the most important factors in this situation is the yawing power comming from the drag force produced by the inside front tyre. (Which is highly dependent on the vertical load and of course your Ackermann setting)

There are several ways to put more weight to the inside front beside the static weight distrubution.
A good one of them is described in Pat's article:-) and another one is called trail-braking.

Regards Andy

[murpia]

Originally posted by flavorPacket:
I don't know, Ian. I can think of some situations where yaw inertia can remain constant while weight distribution moves forward (just flip a and b in a standard plan view).
Yes, but you're not just rotating the car, there has to be some lateral motion as well, else you're not cornering you're just spinning on the spot.

That lateral acceleration requires a lateral force at both axles in proportion to the weight distribution (for a given level of understeer). More weight forward requires more front axle lateral force, therefore for those mythical linear tyres understeer level is constant.

It's entirely possible for a given track and car that minimising yaw inertia is more powerful than optimum weight distribution. But, given the choice, I think I'd tune weight distibution to even out the tyre temps as best I could.

Regards, Ian

[flavorPacket]

Originally posted by murpia:
It's entirely possible for a given track and car that minimising yaw inertia is more powerful than optimum weight distribution. But, given the choice, I think I'd tune weight distibution to even out the tyre temps as best I could.

Regards, Ian

We have no min weight, so we don't play with it. BUT, if you take a look around design finals for the past 3-5 years, the avg wheelbase is roughly 61-62" (1550-1575mm). That says something.

Back to your earlier points:

Please keep in mind that I would also add weight to the front to cause steady-state understeer. I'm just exploring the idea that more front weight helps turn in.

When you turn into a corner, the yaw torque comes before the lateral acceleration (you can see this in data). So I guess the question becomes, how do you define turn in? Yaw torque development, or lateral force generation? If lateral force, how much? .4g? .8?


Originally posted by murpia:
That lateral acceleration requires a lateral force at both axles in proportion to the weight distribution (for a given level of understeer). More weight forward requires more front axle lateral force, therefore for those mythical linear tyres understeer level is constant.

You only need equal lateral force (in proportion to weight) on both ends if you want steady state cornering. This does not describe the state of a car in turn in, and not many FSAE cars fit this definition anyway.

[VinceL]

flavorPacket,

I agree with everything you just said, except for the first line. I don't think the average wheelbase in design finals has been 61-62". Here are some examples that I can think of that suggest otherwise. This is just from my memory and I could be off by 1-2". So if anyone from any of these teams is watching, please correct me if I'm wrong.

UWA: About 67"
Tu Graz: 64-65"
Penn State: 63-64"
Waterloo: About 67"
Kansas: Wasn't a short car. Looked like 63-66"

While we are on the topic of turn in response let's talk about wheelbase as well. I actualy don't see why people think that shorter wheelbases equals faster turn in. I would think it's the opposite.(And it's more than just me thinking. It's in Chassis Design and RCVD fron Milliken and Milliken.) With all else constant a longer wheelbase would give you more yaw moment, and thus more yaw acceleration. Of cource all else can never be constant because bringing your wheels in will decrease inertia. But the overall effect on yaw acceleration would be for a reduction.

I would think that in order to design your car for maximum response and nothing else, you should make your chassis as short as possible, then push your axles out as far as possible. Your thoughts...

Cheers,

[VFR750R]

my 2004 Colorado sport had better initial turn in then my vette...by what feels like a ton even if the mid corner grip wasn't as high. I attributed it to front weight bias and high Cg getting more weight transfer to front. Or as stated earlier, my skillz may be below the vette (very likely). The colorado did ridiculous lane change manuvours though.

i don't see the physics on how halfshaft angle could effect weight transfer.

[flavorPacket]

Vince, you are absolutely right about a longer a and b creating more yaw moment. BUT, sigma M =dH/dt=I * alpha - HOT. Thus, reducing inertia can also help. And those 4 corners are sure far from the cg...

[A Richards]

Originally posted by VinceL:
flavorPacket,

I actualy don't see why people think that shorter wheelbases equals faster turn in. I would think it's the opposite.(And it's more than just me thinking. It's in Chassis Design and RCVD fron Milliken and Milliken.) With all else constant a longer wheelbase would give you more yaw moment, and thus more yaw acceleration. Of cource all else can never be constant because bringing your wheels in will decrease inertia. But the overall effect on yaw acceleration would be for a reduction.

Cheers,


Lengthening the wheelbase will give you more "yaw moment" and the relationship is linear. The change in mass moment of inertia about the Yaw axis is quadratic therefore its increase 'MIGHT' have a greater effect and result in slower turn in.

Ash Richards

[VinceL]

Right, it might have an effect. It depends on how much of your total inertia is due to your unsprung weight, and also on the ratio of track to wheelbase - it sounds like we are agreeing on that.

I just did some quick hand calcs, and it ends up being pretty close.

[ben]

I haven't read RCVD for a while but I've just checked in Chapter 5 and yaw damping is proportional to Cf, Cr, a^2 and b^2.

The squared relationship with a^2 and b^2 suggests that a longer wheelbase is more directionally stable and therefore less responsive or have I missed something?

Ben

[Michael Hart]

Originally posted by PatClarke:
Quote 'I don't know about the drive shaft thing Pat? I don't think I've ever seen a car with the shafts angled backwards'

Pete, there certainly was a car at FSAEA with the drive shafts angled backwards. It certainly wasn't a front runner. Ron T and I are going to visit that Uni soon to give them a little guidance.

And Ian, this is the most intelligent debate I have seen in here for some little time. Thanks
Cheers
Pat We (accidentally) ran with angled half shafts for our first few testing sessions last year (miscommunication on upright points and sprocket size led to the rear wheels being ~1" forward of where they should have been relative to the diff) and any potential benefits that we may have seen from tweaking our weight distribution were greatly outweighed by the problems that the halfshaft angle caused...in our design at least, the halfshafts work best when they're closest to straight with the diff, so our drivetrain guys lost a fair bit of sleep scrambling to correct that little mistake

[flavorPacket]

Originally posted by ben:
I haven't read RCVD for a while but I've just checked in Chapter 5 and yaw damping is proportional to Cf, Cr, a^2 and b^2.

The squared relationship with a^2 and b^2 suggests that a longer wheelbase is more directionally stable and therefore less responsive or have I missed something?

Ben

That's true, but only for a linear tire. that Ch. 5 is more of a perturbation response model than any kind of serious dynamics model.

[Chuckster]

As an autocrosser since the early 1970's, I can tell you that plenty of people have relied upon shifting weight bias forward to improve front end "bite" or turn-in. There appears to be results to back the practice up.

Is it the only way or the best way?---dunno.

You will even find examples in books such as the old Fred Puhl handing book which I believe had a pic of a mid engined formula solo car with a concrete filled tube up front!

In the rain, it has always been popular to throw a sizeable chunk of ballast up front too.

I don't offer any physics explanation of my own to explain all of this, but apparently as the surface gets less grippy, moving mass forwards helps more. It also changes the roll couple distribution in favor of oversteer if one does nothing to the springs and bars after the mass forward shift.

So, I suggest that just grabbing ballast and trying it by itself is an incomplete experiment.

A more informative test might be:

Compare the baseline normal setup with:

(1) Rebalancing (roll couple-wise) the car with weight shifted forward
(2) No shifted weight with varying amounts of increased rear roll stiffness
(3) Just weight shifted forward-no other adjustments