Higher Speed Transitional Handling, 911
 

I've searched here and found very little about this question.

There are three places on our local tracks where I brake (generally in a straight line) to near 90mph and then enter a long sweeping turn. In each case I'm sure I could enter (and come out of) the turn faster if I weren't spooked by the abrupt shift of weight to the outside tires on entrance. I should say that I try to brake then let the car stabilize before turning. One additional turn is worse as the transition occurs after you leave a right hand sweeper and enter one to the left, meaning that the loads move from one side of the car to the other, not just from the center to the outside.

My 911's suspension has been tuned only a bit for the track. The car is a 74, with 22mm/30mm sway bars, a bigger stock front sway bar and a bigger rear Tarett sway bar. All bushings are stock and in good shape. Shocks are stock Bilsteins. Alignment is close to stock, except for 2° front negative camber and 2.5° at the back. I'm running BFG G Forces at the moment but the phenomenon I described happens no matter what tires I've tried. And alignment changes in the past haven't done much either.

Is the answer bigger torsion bars/coil overs to reduce lean and equalize tire loading or is there something more involved?

You could try lowering the car to reduce both the total lateral load transfer a bit as well as have a greater effect on the roll center height. Things that speed up weight transfer are #1 roll center height, #2 stiff anti-sway bars, #3 stiff torsion bars.

If you lower the car, the front roll center at least will get lower pretty quickly and so put the load into the anti-sway bar. In the rear, the roll center does not drop as quickly so you could go softer on the bar. If that does not help, back off on both bars. Don't raise the spindles unless the roll center goes below ground level.

This should help if the problem you feel has correctly been attributed to weight transfer occurring too quickly.

I think the real problem is a lack of rear downforce shifting the balance forward and contributing to oversteer.

I have a '74 as well with 21/28 t-bars and Tarett sway bars. My bushings are ER polybronze and monoballs. Stiffer bushings may help, but my suggestion is to be as smooth as possible w/your inputs. Deliberate, but smooth inputs help keep the car from becoming unsettled. You probably know this already, but it's something I tell myself every time I enter the track. . .

Have you had an instructor or pro drive the car with you as a passenger? I am always surprised and find big gains in my own driving when I ride passenger with a pro driving. I alway spay attention to his entry and exit speeds and try to get close myself.

Steady state cornering balance is managed by springs and sway bars. Steady state cornering is static balance and is not impacted significantly by shocks.

Transitional handling (weight shift) is managed by your shocks. Once your springs and sways are setup properly, I'd take a look at what you have for shocks. Bilsteins are just a manufacturer. Are they optimizied? By whom?

Unless you address the static and dynamic balance(in that order), the transitions will continue to be scary... Bushings, etc.. only help hold alignments and reduce friction while the suspension pivots.



During cornering, Shocks manage the speed at which weight is transferred around the car. Too slow and the car won't hook up soon enough, too fast and the tires exceed their grip circle.

For left to right (or vice-versa) transitions in corners, try and make a "mini straight" between the two to settle the suspension for a short time before changing directions.

it almost sounds like you could trail brake a little into the corner and start the turn with your car not pitching from nose down to nose up.

or finish your braking in a straight line and be back on the power before you turn. power on helps plant the rear tyres.

Just from your question, and the analysis you have done so far, you are exceeding your car's capability. Your CAR probably needs a bunch of things. But remember, we are working with an older 911, whose inherent chassis tendancy is to change direction ABRUPTLY. Part of the solution will be to determine what is simply unsettling to the driver, and what is actually unsettling to the car itself.


Here you are identifying chassis ROLL. Most of the preceeding posts, including yours, have good advice on this: higher-rate springs (torsion bars), stiffer (and slower reacting) dampers, lower ride height, lower weight placement in chassis, and wider track width work to limit how much weight is transferred to your outside tires. The anti-roll bars serve to regulate the RATE at which that weight is shifted.

As usual, Flieger has gone straight to the heart of the matter, with the simplest and least expensive solution with which to start.


Here you are identifying TRANSITIONAL chassis ROLL. As Chris so aptly points out, dampers play the primary role (not ROLL) here, combined with very smooth steering input, as others have said. In the absence of adjustable shocks, very STIFF dampers, ones that react slowly, will help here. At a minimum, the valving in your Bilsteins needs to be in top condition, and matched to your torsion bars.

As you probably already know, tires play a huge part in the confidence factor. Ultimately, who cares if the car leans, as long as the tires stick? Great tires can take away your "abrupt" factor.

And even with all this, in order to go quickly with an old 911, one needs a certain amount of throw-it-and-catch-it driving style. Initially, it takes a leap of faith, but that, in itself, is part of the thrill

Ed LoPresti

With an overdamped car, would not the weight transfer be more abrupt? I would think softer dampers would smooth out the load transfer to the tires, though the driver would feel a little sick and like a boat captain.

Not necessarily. For roll, increased bump reisitance acts to limit transfer to the outside. Conversely, increased rebound resistance acts to "keep" more weight on the inside. If nothing else, the limiting of the spring compression/extension, slows the process.

Ed

Wait a minute. If you take limiting the travel to the limit, then a car with no suspension would take an infinite amount of time to transfer weight. But the opposite is the case, the weight transfer speeds up as the suspension gets stiffer.

You want soft damping on rebound and hard on compression, but this keeps more weight on the tires at the expense of jacking up of the car as the left-right-left transition is made (not good).

The bottom line is that the more the body is allowed to roll and the smoother it comes to a static roll angle, the smoother the load transfers to the outside wheels.

Flieger,

I am with you. Jacking forces are pretty much immediate. The stiffer the suspension, the more quickly the forces reach the tires. Adding more compression damping will reduce the time it takes the forces to reach the tires.

Anything you do that reduces the roll has to reduce the time it takes the forces to reach the tires.

Scott

While what you say Flieger is true, the reality of it is somewhat different in the middle ground. A soft car is easy to drive but when we are talking about a car as soft as a minimally modified 911 that is under damped, it has enough suspension displacement to gain momentum as it rolls and then when it finally comes to its steady sate point and you have already asked it to turn the other way, by being under damped, it violently springs back the other way unsettling the rear. I would also imagine with dead shocks that the OP has enough body roll to get some significant toe and camber change in the rear which may also be a factor.

IMO you want very little compression valving on your shocks as in an ideal world this would be adjusted with spring rate, yet some substantial rebound valving to slow the weight transfer. How much of each on an nonadjustable shock is something you will have to research your self or trust one of the experts to get it close to right.

To the OP, I would recommend getting your shocks re-valved for your current suspension and weight as from my personal experience new off the shelf Bilstein Sports are marginal at best with 30mm rear T-bars and get worse with age.

I am unsure if you are paraphrasing my post here, but that is not what I wrote. By increasing the resistance to BOTH bump and rebound with dampers, one is limiting ROLL. In addition to limiting total weight transfer, it has to do with the magic of shocks - progression.

On a turn, increased resistance to bump effectively adds spring rate to the outside springs, progressively, limiting the tendancy of the chassis to tip outwards. Similarly, increased resistance to rebound effectively adds rate to the inside springs, progressively, "holding" the chassis from tipping outwards. Less "tip" = less weight transfer, and visa versa.

We all agree with the very rudimentary rule-of-thumb: a stiffer suspension transfers its weight more quickly. Adding damping certainly makes the suspension more stiff. So, Winders, when a turn is initiated, the outside (and inside) tires "see" the START of weight transfer more immediately. But what comes after that can be different, depending on anti-roll bar settings, and, yes - damper geometry and valving.

Ed

Evan,

You are swimming against the tide of popularity here! For (how many?) years now, I have contended that mono-tube, gas shocks, if they are kept fresh, are fine for the street. For anything except the most casual track applications, they are sadly inadequate. Yet, thousands of these things continue to live on Porsche and Corvette street-made-into-racecars.

If you make any headway with this argument, I certainly applaud you! Maybe it's an idea whose time has come?

[John: Sorry for the hijack. ]

Ed

The weight transfer due to the CG being above ground and shifting to the outside should be significantly less than the weight transfer due to lateral force on the ground plane acting on a mass above it and requiring the weight transfer to react the moment. Therefore, limiting roll will reduce the total lateral load transfer. Dampers cannot limit the static body roll, being fluid filled cylinders+ pistons.

I was not saying that stiff compressive and soft rebound damping was good. It causes the body to work its way up on the suspension with each turn input given when the car has not reached steady state. The higher CG means more weight transfer.

OK, so you are saying that the driving frequency is approaching the natural frequency of the system, so more damping will quell the body roll response at that frequency.

However, limiting the body roll speeds weight transfer.

I think what you are getting at is that the bumps stops come into play if the body roll amplitude is allowed to get too large, so by slowing it before the bump stops it limits the amplitude in a way that does not spike the load on the tires. It speeds up weight transfer on inputs that would not result in bottoming out but that trade off is acceptable if the track has a turn or two where the suspension would bottom out.

Where's James?

Yes, something like that. We will see how my theory works as I have now changed almost every single part on my car and semi tube framed it.

He's here, Max, and trying to absorb all this. I must say this has turned into a very useful thread, at least for me!
My take so far: a lower car would help (can't go any lower and still do the driveway.)
Wider tread would be good (not with these no flare fenders.)
Relocating the driver, motor, fuel tank or anything else lower to get the cofG down would be good.
Soft hands and trail braking are helpful (I'm not bad here though one can always be better.)
Stiffer torsion and/or sway bars are always good (This thing gets driven on the street, not sure how much further I want to go here.)
I need to stiffen up the shocks, though exactly how seems to be the subject of much debate.
I was particularly taken by Ed's statement that "... higher-rate springs (torsion bars), stiffer (and slower reacting) dampers, lower ride height, lower weight placement in chassis, and wider track width work to limit how much weight is transferred to your outside tires. The anti-roll bars serve to regulate the RATE at which that weight is shifted."
I don't understand (and there seem to be differences of opinion) what the relationship is between controlling the rate at which the weight is shifted (with sway bar size) and controlling that same weight shift with the shocks. I first read Ed's statement as suggesting bigger anti-sway bars would be the most important change I could make, but I think the consensus is that it is shocks. Comment?

Unless you are getting into, or do get into after making this change, the bottoming out area where the spring rate is increasing non-linearly and very abruptly, then you softening the anti-sway bars will slow/smooth the weight transfer.

You should look to get dampers that are critically damped for your spring rate in roll, which is different than one wheel bump or the pitch rate since you have anti-sway bars. Underdamped makes for softer/smoother weight transfer but the car will overshoot and go through a few cycles before settling down so that will be bad for your confidence level. - This I now see is what Evan was getting at. If the transition is made before steady state, the car will end up on the bump stops. - Overdamped will mean the car is just too stiff and it will not roll very fast but the tires will see abrupt load transfer.

I remember in the Milliken + Milliken book they mentioned that slightly underdamped can yield a shorter time to steady state than critical damping, at the cost of a little overshoot. The faster speed of the suspension movement makes up for the longer travel distance, though.

I recently experimented with the adjustable Konis on my car. I was really surprised what a (positive) difference stiffening the shocks made during turn-in. Stiffening the rear more than the front made an even bigger difference on my intentionally understeering set-up. I am running a very soft car (19/25 Torsion, 23/23 ARB almost full soft).

So I agree empirically (as well as theoretically) with the "stiffer shocks" suggestion.

Caveat: I am no chassis engineer, nor a hotshoe driver.

Personally, it sounds like you've got too much roll stiffness either from T-bars are Sway-bars. When you've got too much roll stiffness the car feels like it's on a knife edge. If you don't like the abrupt shift in load to the outside tires...

1) Go back to softer T-bars and/or Sway bars. I'd start with the sway bars. Reducing lean increases the shifting of load to the outside tires (all other things being equal). Lowering the car would help, but you'll need to do this by switching to RSR front struts. Just lowering the front end isn't going to help the situation and may cause the handling to get stranger.
2) Look at investing in adjustable shocks. Then you can tune how fast the load transfers load to the outside tires. Just be careful taking anyone's input verbatim in regards to stiffer or softer at one end or the other. This is because some shocks are adjustable in bump and others in rebound. They will affect the handling differently when adjusted. I'm not sure if either is better, you can most likely wind up at the right place using either, but the path to get there will be different.
3) In terms of adjustment, you might want to try increasing your camber a little bit to make sure that you aren't rolling over on the tires when the car settles over. I say a little bit because if you go too far you'll end up messing up your breaking under compression. This might be the place to start since it requires the smallest investment.

Note that Colin Chapman was "tongue-in-cheek" when he made the observation in my signature.

RSR struts raise the roll center and so will contribute to the fast (nearly instant) weight transfer. If his roll center is already above ground, don't use RSR struts. Lower the car and let the roll center be low. You want it at ground level. Too low (below ground) is bad as well.

The RSR struts were invented for a reason. They allow you to lower the CofG of the front end while still keeping the roll center where it needs to be. This will also reduce the roll-couple which will lower the roll. Just lowering the the front end on stock struts does some undesirable things to the camber curve and roll-center that can make the car's handling weird.

You need to know where the roll center is now before you go changing it. RSR struts aren't cheap.

Agreed. I wasn't advocating lowering the car. As far as knowing where the roll center is, an easy solution is that if the RSR struts have raised the spindle by 15 mm, than you can lower the car by 15 mm and you will have the front roll center back where Porsche intended it to be.

All that being said, as I said I really think that moneymanager should try playing with the camber, and/or backing off some on the roll stiffness. As long as he's not bottoming out hard, he'd most likely do better going a little softer. A softer suspension will make it easier to feel what the car is doing and reward smooth driving.

Yes, and lowering the car will both lower the roll center and reduce the total lateral weight transfer...

So long as the roll center doesn't get way below ground lowering is a good thing.

I think the RSR struts were to regain suspension travel not necessarily roll center height. They probably had a much higher roll center than that whichi is optimal for racing on the street car.

An interesting comparison comes to mind. A stock 911 is way better than most other showroom stock cars when it comes to getting around a track. In fact I'd say that a stock 911 is most likely comparable to a prepared car of most any other make. So how much difference does that make?

A showroom stock Honda Civic gets around Mid-Ohio (a classic handling circuit) in 1:48.737. An Improved Touring Civic (slightly older, noisy exhaust, aftermarket springs, shocks and sway bars) gets around Mid-Ohio 1:48.503. Both cars on on shaved street tires.

You can also take a look at how the lap times compare between the PCA classes at Mid-Ohio, and compare the lesser prepped versus the more highly prepped classes. A stock 911SC (Class D) gets around Mid-Ohio in 1:42.6. Jumping one or two classes and the lap time hasn't gotten that much shorter.

When you take into account the difference in experience level, I think that you'll see that the incremental difference as a percentage of lap time is not that much, especially since we're not talking about head-to-head racing. So my point is just that a stock 911 is relatively vice-free (acknowledging the rear weight bias issues) track car. Unless you are trying to fix a specific handling issue (for example: a push on the entry to turn 3), it's not clear to me what you're buying for yourself when you just stiffen the car's suspension.

The roll center in and of itself is most likely less important than the roll couple. If you lower the car and the roll center by the same amount, the roll couple remains the same and the car will roll by the same amount. The only difference is that you've now moved the car's suspension travel into a less than optimal zone -- especially when you're talking about strut suspensions. Under compression (as in when you lower a 911 by indexing the front T-bars) the front suspension starts to lose camber. If you start with the car in that setting while static, you'll find that when under load (with even more compression) you'll lose even more camber. Now throw some roll on the car and you'll wind up with the front suspension running out of travel (a very bad thing) and losing camber (a bad thing). What you'll wind up with is a car that pushes through a turn when you lean on it, until it runs out of travel, when the push becomes terminal. This push will come and go as the car traverses bumps -- which makes it hairy to drive.

Personally I think that it's more important to have the geometry right than to have the lowest possible car. Lowering a car will only help when the geometry is right. Having the geometry right means that there will be a fairly wide plateau in the car's cornering performance where it is happy, which allows the suspension to absorb bumps or sudden steering inputs without having the tires go into a sub-optimal attitude to the road. This allows the car to be predictable and the driver to have confidence in what it is doing. It's either that or you just have to lock-up the suspension so it doesn't move as Chapman suggested.

The entire reason for a stiffer suspension in roll is to control unfavorable wheel attitude (camber) changes.

The other reason for stiff suspension is to allow the car to run low without bottoming out. But that means that the stiffer suspension becomes almost just as bad as bottoming out. You want highly progressive springs with as little travel as you can get away with so the car can be as low as possible.

The roll couple isn't as important as the lesser total weight transfer. You don't want to have much of the roll couple reacted by jacking forces because those spike the load on the tires. We are suggesting that a softer car in roll would be better here since that slows and smooths the weight transfer. By having a low roll center you can have a greater window in which to soften the car by allowing it to roll and using the anti-sway bars to react the roll. This smooths out the weight transfer and ultimately gives more grip to the tires.

So you're saying to convert the 911 into a suspensionless go-kart. Why not just drop it down on some bump rubbers and run it like that?

I'd hate to try to race a car set-up like that on a track like Lime Rock that has bumps and geography in it.

Well, no suspension would save weight and the 911 chassis can be pretty flexy compared to modern cars, and the bigger profile tires are all the suspension F1 cars really need. And they go pretty dang fast with only a few millimeters of actual suspension deflection. Of course, downforce works wonders.

But more seriously,

I said to run the car as low as you can. I meant that if the track is bumpy you need to sacrifice speed in certain smooth turns to make up for the bumps to some extent. I did not mean to remove the suspension at all times, though go-karts are pretty darn fun and go very well around the track with just the frame as a spring.

Once you stiffen the supension a whole lot damping gets harder since the "rigid" parts now aren't so rigid and steel doesn't really have much internal damping.

Formula cars (especially those with ground effect) are completely different animals from a 911. But they actually have considerable suspension travel -- in the rear. Just watch the suspension from the rearward facing in-car cameras from the Singapore GP. The front is run stiff so as to not stall the ground effect air flow.

Anyhow, it's clear that we're advocating two different strategies for car setup. Moneymanager can read them both and make-up his own mind what he wants to do.

The funny thing is I was telling him softer.

Well thanks to John who has identified an issue here: we have two entirely different strategies for dealing with the transition handling issue: go stiffer or go softer. And I raised the question, but there must be a thousand of you out there who've had the same experiences I've had...big weight shift at turn-in or in a transition. So the question is important.
Based on what I think is a consensus that stiffer torsion and sway bars transmit load changes faster, and that stiffer shocks slow all that down, I'm pointed toward increased overall stiffness along with shocks that are a bit stiffer still. There seem to be a lot of votes for stiffer shocks so long as they aren't completely mismatched to the torsion and sway bars. Thoughts?
If this turns out to be wrong, we'll go back the other way. BUt my instinct at this point, based on a lot of torsion and sway bar changes in the last couple of years (which didn't dramatically alter transition handling) but no shock changes, is that the shock settings are hugely important.
Re lowering, I'm as low as I'm going (a couple of inches) given the desirability of continuing to enter my garage.

moneymanager,

I would talk to someone that sets up Spec 911 race cars or races a well setup and competitive Spec 911. These are the ultimate expression of 911 handling using torsion bars with 911SC sized fender flares.

Scott

Not picking on John here - but getting back to basics for just a moment --

We lower a car to get it's center of gravity closer to the ground. That way, it will not tip over when it goes around a turn a little too fast.

With that simple change, we now limit its side-to-side movement when it corners or swerves. As an added bonus, we simultaneously limit the car's natural tendancy to dive under hard braking, and limit its squat under hard acceleration.

HAPPILY, there is a very important side benefit in limiting how much air is allowed to flow under the car. Bonus time!

For those simple reasons, the car will now handle better.

UNHAPPILY, lowering a car has some undesirable side effects. The suspension no longer has as much "travel", and therefore might "bottom". This side effect is usually solved with higher rate springs.

HAPPILY, higher rate springs also limit side-to-side and front-to-back movements. Bonus time!

HAPPILY, higher rate springs keep the body from dipping down and rubbing the fenders on the tires. Double-bonus time.

In lowering, the suspension geometry may have changed, and now uneven track surfaces cause unwanted wheel/tire angles. This side effect is usually solved with bump steer techniques.

HAPPILY, bump steer techniques serve to correct (or improve) design flaws in the original suspension geometry, allowing the wheels/tires to move up and down in a more-than-less verticle attitude through their full range of motion.

UNHAPPILY, high rate springs tend to cycle too fast for bumpy tracks (like Lime Rock was for years.) This makes the car skip over the bumps. This side effect is usually solved by dampers (which I do not want to get into again!), or by using lower rate springs on excessively bumpy tracks, or by administering a sedative to the driver before he goes out.

In summation, lowering the car has so many benefits that we simply MUST do it. Likewise, making the suspension RESIST movement (ala Chapman) is the only known way to fully exploit the benefits of lowering.

Keeping it over-simple, I know, but that is the way we think about it.
Ed

By the way, it wasn't my intention to argue that lowering a 911 isn't good. Lowering the C-of-G is a good thing. My intention was to highlight the importance of geometry which is often overlooked in the rush to drop the C-of-G to ground level. All that you need to do is to watch one of those kids with a riced-out car which is scraping the pavement try to make it around a corner at speed.

There are certain situations (for example a limited prep racing class) where you cannot change the existing geometry. In that case as Ed pointed out you often need to take a more tortuous route to improving the cornering grip which often involves applying crutches to address the problems created by the previous "improvement". This can often turn into a succession kludges applied on top of each other.

Where such limitations don't exist, I think that getting the geometry right is important because it is the foundation of the car's suspension that defines everything else (after the tire). If the foundation isn't right the rest of the structure will not be robust. So for handling performance lowering a car is good, but lowering a car while maintaining an ideal geometry is better.