Lateral grip vs weight transfer

[CornerHard]

Here's something that's always bemused me. I'll try to explain my current understanding, and hopefully JToby, DG, or someone else will come in and tell me where I'm wrong.

Thicker anti-roll bars increase weight transfer. If you put a jumbo anti-roll bar in the back of your car, it will transfer more weight to the outside tire and transfer it faster than stock. If you put a jumbo anti-roll bar in back and a stock anti-roll bar in front, the back end of the car will lose grip before the front, and you will increase the tendency to oversteer in the car. In this way, the relative sizes of the front and rear anti-roll bars can be used to adjust the understeer/oversteer balance of the car.

However, as I understand, if I make both bars thicker, I increase the lateral grip of my car in corners. Why is this? My understanding is it improves the tire contact patch by preventing the car from pitching excessively in the corner. Thus the benefit of the larger contact patch outweighs the penalty of transferring more weight to the outside tire. How do you detemine which roll resistance rates will yield the maximum lateral grip, assuming stock spring rates? Am I totally wrong?

 

[ACM]

You are talking about two different things here.

The first is weight transfer 'trimming' - you are using the a/r/ bars to fine-tune the roll stiffness (don't forget however that the effect of an a/r bar change is not limited to the end the change is applied to) and thus the weight transfer.

Your second comment is in regard to using a/r bars as a Band-Aid for improper geometry control. This applies mostly, but not exclusively, to McStrut front ends. This isn't about weight transfer directly, it's about not falling off the tyre. A/R bars should not be used for this purpose, correct geometry should be; unfortunately - just like the commercial - we don't live in Perfectville, and until we do there's a/r bars to help us out.


Charles

 

[jtmcinder]

Your second comment is in regard to using a/r bars as a Band-Aid for improper geometry control. This applies mostly, but not exclusively, to McStrut front ends.

I think that there's a bit more to this, especially when you're talking about a near-stock car. For example, the amount of lateral force going through the front suspension is a lot higher than that going through the rear. When you add in the relative sizes of the bushings on the inboard end of the control arms, you are losing a lot more camber to bushing squish in the front than in the rear. When you next add in that the front started with less camber (on a stock car) and has a slightly flatter bump-camber curve (again, assuming stock or near-stock suspension), you find that the front tires are nowhere near where they need to be camber-wise in a hard corner.

Baidaids like A/R bars won't do anything about the bushing squish and the imperfect geometry, but they can at least reduce body roll and stop you from losing camber due to bodyroll coupled with an incomplete bump-camber curve. This is why it doesn't matter very much which end of the car that you add more bar. As long as you are adding to the total roll resistance, you are helping the front tires to keep their camber.

Relatedly (tee hee), if you do the math on the stock suspension and Dennis' current set-up, you find that the weight-transfer balances are about the same. So why does the stock set-up understeer at the limit while Dennis tells us that his does not? The answer would seem to come from issues like the above. Dennis has spherical bearings in his lower front control arms. He is not giving up camber to bushing squish. He also has a heck of lot more total roll resistance than a stock 2G. Therefore, he has good control over his front camber and can keep the fronts where they need to be so he is getting a lot more from his front tires than someone on a stock or near-stock suspension. That is why he doesn't suffer the understeer of another car with the same weight-transfer balance.

In other words, the logic of moving weight transfer to the rear when the car is understeering only really works when all other things remain equal. But because there are interactions between the various issues that determine lateral grip, all other things do not remain equal. Expressed differently, adding more rear bar to reduce understeer only makes sense when the cause of the understeer is a lack of rear roll resistance. If the problem is actually poor front camber in hard corners, instead, then that is what you need to be working on.

Put a third way, when you are trying to increase the handling performance of the car, you would be well-served to start by figuring out why the end of the car that lacks grip is lacking grip. If you are sure it's because the tires are being overworked and, therefore, are either at too high a slip angle or are being asked to do more than they possibly can, then, fine, move some weight transfer to the opposite end. If the reason is because you are not getting all that you can from the tires at the grip-lacking end, then work on whatever is making the tires unhappy. Hint: it's the camber and/or the ridiculous front air pressure that you're running to keep them from rolling over.

Stock 2Gs (as well as many 2Gs that have shifted a ton of weight transfer to the rear but haven't done much about keeping the front tires happy) behave in a way that tells you that something is very wrong: they push and push and then suddenly snap-spin. If you keep throwing more rear roll resistance at this, all you will do is make the snap-spins come sooner at lower speeds, while making the car very loose at high speeds. (Ask me how I know this.) Anyone with a 2G that pushes would be well-served to focus on making the fronts happier and not on making the rears unhappy.

- Jtoby

 

[T2Small]

So it sounds like we should be increasing front stiffness and decreasing rear roll stiffness over the older style autocross setups if we are fixing camber robbing busing flex problems? (Something like 550/400 springs with RM bars)

I will be trying different spring rates for the beginning of the race season as I move from 550/450 to a 750/375 setup with spherical bearing lower control arms and polyurethane bushing everywhere else appropriate.

I am sick of wet track snap spins and some hairiness exiting corners at 100 km/h with the tail wagging in an attempt to make slow speed turn in managable. Of course a big problem with my results will be that I added a front Quaife and polyurethane bushing while changing the spring rates. So it will be hard to attribute any noticable changes to one factor.

 

[jasonlee]

where do you pick up the front bearings? ive never heard of those before?

j

 

[ACM]

where do you pick up the front bearings? ive never heard of those before?

j

Two alternatives; if you don't mind modifying the arms (ie honing out the hole slightly) you can use anu off-the-shelf HD spherical bearing that's a shade over 40mm OD, a #14 is 1.625".

If, like me, you're constrained by rules (Street Prepared) you need either a metric 40mm bearing or use a #12-ish bearing and sleeve it (load capacity drops though). A suitable Teflon-lined metric bearing is produced by Fluro and can be ordered through Hoerr Racing Products - this is the only 40mm bearing I have been able to source; it is ~$30 in steel and ~$50/ea in heat treated SS. The SS has to be special-ordered from Germany and mine took about 6 weeks to show up. Go to Fluro.de to check the bearings out. They have the perfect studded rod ends for a/r bar end links too

I believe the inboard pickup is the same OD as the lower damper mount, as iirc only the width differentiates the two bushings - DG should be able to confirm this.

Charles

 

[DG-FNR]

I don't think bushing squish is a particular smoking gun, at least as far as camber is concerned. It's there, but it is a small-order change; not a large one.

The stickier the tires, the less true this is.

The bushings are more problematic from a shock sensitivity standpoint. An isolator in series with a control device is self-defeating.

The large-order contributor to understeer in a stock car is that the spring rates are so soft that there's not much roll resistance. Put good tyres on, and the roll angle gets large enough to drag the suspension pickup points farther than the car's camber compensation in roll can compensate for. The tyre's dynamic camber goes increasingly positive and the front loses grip.

This is why you can increase the rear roll resistance (on a stock 2G) and gain FRONT grip. Making the rear bar thicker not only shifts the roll resistance bias rearward (normally, a tendancy towards oversteer by unsticking the rears) but it also increases TOTAL roll resistance, which will reduce roll-induced front positive camber, increasing front grip.

At some point though, with increased rear roll stiffness only, weight transfer effects will take precidence over geometry effects, and you'll start unsticking the rears. It doesn't help that the situation varies as the car enters and then leaves the corner.

The other tradeoff is that if we keep tire grip constant but increase roll resistance, we make the car more "abrupt" - the line between "I've got it" and "it's got me" gets narrower. This is OK for a race car, where grip levels tend to be elevated fairly constant at a given event, but is less acceptable for a car that may see greatly reduced grip levels that vary wildly. The ultimate setup for tarmac is not going to be very forgiving in snow or mud.

I have a hard time recommending a heavily rear-biased setup for a car that sees street use. You want the car to understeer in reduced-grip situations. A car snap-spinning on a freeway onramp is no fun at all.

DG

Learn: http://autocross.dsm.org/books.html

 

[jtmcinder]

Right. Bushing squish probably costs you something like half to three-quarters of a degree of camber, but never any more than that. The fact that the bump-camber curve only makes up for about half to two-thirds of the effect of body roll never stops, but just keeps getting worse and worse. That is why I said that, when starting from a stock car, it doesn't really matter very much whether you put the stiffer bar on the front or the rear; increasing total roll stiffness is the key to keeping the fronts happy.

- Jtoby

 

[CornerHard]

Okay, interesting points by all, thanks for the input. I didn't think I'd DG, JToby, and Charles all replying to my thread

Follow up question:

Assuming 2G AWD DSM, a specific street performance tire, perfectly flat tarmac, and dampers capable of critcally dampening any spring rate, what combination of springs and anti-roll bars will yield the highest lateral grid around a skidpad? Very high spring rates and no A/R bars?

 

[NDgsx]

Assuming 2G AWD DSM, a specific street performance tire, perfectly flat tarmac, and dampers capable of critcally dampening any spring rate, what combination of springs and anti-roll bars will yield the highest lateral grid around a skidpad? Very high spring rates and no A/R bars?

The one that you find in testing

No, they'd be overly stiff. Numbers on glass mean nothing, the spring rate has to allow for enough compliance to let the wheel absorb bumps in the real world.

 

[ACM]

Okay, interesting points by all, thanks for the input. I didn't think I'd DG, JToby, and Charles all replying to my thread

Follow up question:

Assuming 2G AWD DSM, a specific street performance tire, perfectly flat tarmac, and dampers capable of critcally dampening any spring rate, what combination of springs and anti-roll bars will yield the highest lateral grid around a skidpad? Very high spring rates and no A/R bars?

Doesn't really matter - you're talking about what in essence is a static condition, going around in a circle at max lateral G. All that matters is that you have the "optimum" roll stiffness for the setup.

Were you to then brake hard, turn around and try to get as quickly as possible to the same point in the opposite direction, then the method you used to achieve the roll stiffness, the camber curves, and all the usual things we care about, would all come back into play.

Charles

 

[jtmcinder]

Were you to then brake hard, turn around and try to get as quickly as possible to the same point in the opposite direction, then the method you used to achieve the roll stiffness, the camber curves, and all the usual things we care about, would all come back into play.

The deep implications of the above are often missed by autocrossers, even those with a lot of experience. 99% of the texts out there are written for road-racing where steady-state cornering ability is much more important than transient behavior. These texts don't really have to worry about issues like instantaneous vs controlled/delayed weight transfer because nothing happens fast enough to make these very different. Therefore, they often argue in favor of setting the weight transfer ratio equal to the weight distribution ratio.

Autocrossers are much more interested in transient behavior, so we really must pay attention to the difference between instantaneous vs controlled/delayed weight transfer. The co-designer of the fastest autocrossing vehicle ever made really focuses on this and often points out that you can get the car to transfer weight to the inside [not a typo] by having a roll center that is below ground level.

- Jtoby

 

[MrAWD]

Assuming 2G AWD DSM, a specific street performance tire, perfectly flat tarmac, and dampers capable of critcally dampening any spring rate, what combination of springs and anti-roll bars will yield the highest lateral grid around a skidpad? Very high spring rates and no A/R bars?

Hmmm, this is almost like asking which karate style is better!
There are at least two different schools about this. Once I was told (or maybe read somewhere) that European way is going with softer bars and stiffer springs, while here we do it more so in the other way. I don’t know whether that is true, but those two ways are out there and properly done, they would both be very fast. Just like those karate styles, once learned and implemented in the correct way, are equally successful!!

There is only one thing that I could think of right now as important difference between those two styles. The fact is that it is much easier to change the stiffness of the roll bar (assuming there is adjustability there already), then to change the spring and most of cars out there. So, what I believe that is better way to go is to figure out the best spring rates for your car. Once when you did that kind of tuning use you’re A/R bars for the fine tuning if you wish! The best thing with this setup is that you could run a stiff bar for the grippy surface and quickly switch to much less bar when things get wet. Of course there is that option of disconnecting that bar to start with in the case of the slippery track, but that is just one way adjustment.

Good luck either way!


Fedja

 

[c0degsx]

I have a hard time recommending a heavily rear-biased setup for a car that sees street use. You want the car to understeer in reduced-grip situations. A car snap-spinning on a freeway onramp is no fun at all.[/url]

Why would you ever want to understeer in a situation. I think I would rather lose my rear in the turn and help recover with a countersteer, then lose front traction and go straight into a pole/etc. head on. Yes I am aware that if you lose front grip you just have to mash the brakes, but sometimes its too late.

Heres a scenario....

Understeer: your exiting an offramp, your turning fast and hard. suddenly just as your finished exiting the apex and accelerate hard, your front loses weight. Now you just didnt make the turn because you went off the road and didnt have time to brake after the apex.

 

[T2Small]

Why would you ever want to understeer in a situation?

(I added the question mark.)

Because your car will spin so fast you will be turned over 90 degrees before you even begin to react. If you think I am joking, I was running 550/450 with RM bars and on race tires I was well over 90 degrees of yaw before I had even a chance to react. A little bit too much trail braking into a corner on these "old style" tail happy setups are a disaster waiting to happen.

I thought long and hard after I pulled my car out of the 6" swamp I spun into (over a curb and some tires on the way in).

Your understeering car in the wet is far, far safer.

Granted on street tires the wash out is much more progressive, but not by much. It is nothing like correcting a rear slide in a RWD with decent balance.

 

[c0degsx]

Yea I guess I cant really say to much about oversteering in a 2g. I have however oversteered a 240sx, m3, and mr2. The mr2 was the worst due to its infamous "snap oversteer". That was probably the only car I usually couldn't get to correct itself. The other RWD cars however easily came back into place even after a 90 degree yaw.

I hadn't realized that a 2g would be so bad on the oversteering. Why dont I ever see 2g drifters then?

So basically while trailbraking the weight gets too far in front of the front axels that it oversteers? I never realized how horribly places our engines are.

BTW does anyone happend to have any videos of DSM's roadracing?

 

[jtmcinder]

I had the seventh MR2 in North America. That car was almost perfectly balanced and didn't "snap"-anything. It would do whatever you asked it to do and could be steered with your hands or feet. It sure as heck wasn't infamous for bad handling of any sort. The only way for it to snap oversteer was for the driver to be hard on the brakes at turn-in. In other words, all snap-whatevers would be driver error.

- Jtoby

 

[MrAWD]

Why would you ever want to understeer in a situation. I think I would rather lose my rear in the turn and help recover with a countersteer, then lose front traction and go straight into a pole/etc. head on. Yes I am aware that if you lose front grip you just have to mash the brakes, but sometimes its too late.

I guess you would rather hit that pole sideways instead of hithing it straight on!!

Fedja

 

[T2Small]

Why dont I ever see 2g drifters then?

Mostly because the DSM drivetrain cuases understeer under power application. The EVO has a selectable center diff that allows more than 50% rear bias.

Just to clarify what I said earlier. Excessive trailbraking can unsettle and even spin almost any car if applied too much. Most I have driven either communicate their intention to spin or have a high threshold. A stock suspenion 2G seemed quite easy to control under heavy trailbraking in wet conditions. An impending spin could be controlled or avoided entirely. My high rear spring rates made it very unsafe for me to even think about trailbraking in the wet.

Jtoby : I have seen several MR2s at the track and I consider them in the same league as 911s. Normally there are no issues, but when things get out of hand, they get get out of hand fast. I have yet to see either a 911 or a MR2 recover from an impending spin. (But I realize it probably is possible.) They either have to avoid them or be excellent drivers. And as you mentioned all of these cars started spinning under braking.

 

[zephyr_8]

Just out of curiousity, How would chassis rigidity(seam welding, roll bars braced to the body, welded in strut tower bars) affect roll stiffness and the effectiveness of the springs? And how does all that has been said apply to 1gs?(the drivetrain making the car tend to understeer, the camber curve, and the roll bar strategies)

My car never understeers(even in snow) unless I lift a wheel over a bump or if I lock the tires up when hard on the brakes. I have fwd springs on the front though, which I have heard are softer, but it seems like it balances the car out. I plan on upgrading to a more accurate suspension setup(coilovers) But I am poor and the fwd I was ripping for parts had those springs and good shocks. How do you guys think these springs will effect the way my 90 awd acts? It may only act this way because of driver inputs. I tend to balance it on the brakes/throttle. I have played with the alignment a little too, mostly rear camber.

Sorry about grammer, and the way i worded things, I didn't know quite how to put most of it.

 

[Poloturbo]

Mostly because the DSM drivetrain cuases understeer under power application. The EVO has a selectable center diff that allows more than 50% rear bias.

Just to clarify what I said earlier. Excessive trailbraking can unsettle and even spin almost any car if applied too much. Most I have driven either communicate their intention to spin or have a high threshold. A stock suspenion 2G seemed quite easy to control under heavy trailbraking in wet conditions. An impending spin could be controlled or avoided entirely. My high rear spring rates made it very unsafe for me to even think about trailbraking in the wet.

Jtoby : I have seen several MR2s at the track and I consider them in the same league as 911s. Normally there are no issues, but when things get out of hand, they get get out of hand fast. I have yet to see either a 911 or a MR2 recover from an impending spin. (But I realize it probably is possible.) They either have to avoid them or be excellent drivers. And as you mentioned all of these cars started spinning under braking.

I found out also that for a conservative spring rate setup as mine ERS 450/350 with konis to 3/4 of a turn from full stiff, with the stock sway bars that the car with some trail braking on a very wet track was very stable. Following a modified STI in the wet , he could'nt beleive the grip I had in a lot of corners were the STI usually gets their advantage. Very progressive understeering that as a lot to do with my ancient not that good but fair in the wet tires. Never snapped or even ad any issues.

 

[logic]

Two alternatives; if you don't mind modifying the arms (ie honing out the hole slightly) you can use anu off-the-shelf HD spherical bearing that's a shade over 40mm OD, a #14 is 1.625".

If, like me, you're constrained by rules (Street Prepared) you need either a metric 40mm bearing or use a #12-ish bearing and sleeve it (load capacity drops though).

Rather than start a new thread, I figured I'd resurrect this one; there didn't seem to be another thread talking about modifying the arms that got into this level of detail.

I read DG's write-up on doing this, and he basically went the "standard bearing" route: hone out the arm a bit (leaving a shoulder on one side), cut a retaining notch, press the bearing in, and install a retaining clip to hold the bearing in place. Then, machine standoffs to take up the space the bushing used to consume. Exactly as Charles described in his first case.

I can probably borrow a friend's lathe to make the standoffs (although with his first child on the way next week, that may take a while ), so my question is: would a press-fit be sufficient for the 40mm OD bearings, or would retainers be necessary? Charles, if you're still around, I'd love to know how you configured this (and jtoby as well, although I don't recall if you made this change or not). Once you get into notching a retaining ring, we're out of the equipment that I have direct access to, and I might as well have someone else hone it out for a more easily-obtained bearing.

(I've always been surprised that nobody has made a small side business out of doing this modification for people, given the number of other specialty milled items you see floating around in the DSM/Evo world.)

 

[ACM]

I pressed the 40mm bearing into the arm, I did not cut any retaining clip grooves as the rules are unclear whether that would be legal. I decided to err on the side of legality rather than safety (this is autocross, not road racing) and monitor the bearings carefully.

Neither side has moved at all, or if they have it's too small to measure. The fit is very tight, causing the bearing to stiffen up a touch, but I was reassured that the fit was within tolerance for the bearing. all of this only applies to the Fluro brand GXSW bearings I purchased from HRP World, I have not used, tested, or even spec'd any other bearing.

The same bearings are still in the car, 3 seasons now ? Obviously the car isn't used on the road at all now, but the mileage it does see qualifies as "severe duty". Were I to use this setup on my street DSM (and I am seriously considering it), I think I would cut retaining grooves and fit clips - small insurance against a very nasty and dangerous possibility.