Question about dynamic toe change in back

[450knotOffice]

Guys, I searched but can't find the answer. When the rear suspension compresses due to a bump, do the rear wheels toe IN more or toe OUT more (in our torsion bar 911's)?

Thanks guys.

[rnln]

it is the same as lowering the suspension right? One of my questions in the past regards toeing when lowering the car, the response was toeing in.

Originally Posted by 450knotOffice

Guys, I searched but can't find the answer. When the rear suspension compresses due to a bump, do the rear wheels toe IN more or toe OUT more (in our torsion bar 911's)?

Thanks guys.

[Steve@Rennsport]

Scott,

911 semi-trailing arm suspensions toe IN under bump and toe OUT in droop, as when you lift off the gas onto the brakes,...

[450knotOffice]

Great! Thanks to both of you.

[burgermeister]

Here's a graph (dark blue arc)

Depending on the height of the inner pickup point relative to the outer, the dark blue "arc" rotates. Pictured arc is for inner pickup 10mm higher than outer pickup. If they're the same height, it's symmetric about the X-axis. I can only measure within +/- 10mm (my guess) with the arm in the car and a tape measure, so the calculated curve has some uncertainty as to its orientation.

Also note that if you have bushings (as opposed to bearings), thrust adds toe-in and braking adds toe-out from bushing deflection. No idea how much (obviously gear-dependent for thrust).

[450knotOffice]

Interesting graph. Thanks for the added detail.

[Flieger]

By "outer pick-up" do you mean the spring plate outer bushing or are you referring to the hub of the control arm?

If the latter, then a lowered car would have toe change symmetrical about the x-axis?

Thanks, I always enjoy your careful analyses, graphs, and experiments.

[Flieger]

As I visualize it, the semi-trailing arm should always inrease toe-out once the suspension is deflected to any angle away from level which is greater than the initial neutral setup geometry. Hypothetically, if the suspension was setup to have zero toe with the arm angled 10* downward towards the wheel, there would be toe-in during bump until the arm was level, then that would be over the bell curve and the wheel would start to toe-out back to zero. Once the arm was deflected greater than + or - 10* from level, the top view angle of the swing arm axis would start to dominate the toe rather than the geometrical shape of the control arm hub (bent towards center of car). This would cause toe-out at both extremes of suspension travel.

So does the chart apply to the more normal range of suspension movement? (Which I have no real idea of its magnitude)

What is the standard setup geometry for the standard steel trailing arms?- What angle are the trailing arms at when there is zero toe? The control arm hubs must have some angle built into their geometry for toe-in to compensate for the initial downward control arm angle towards the wheel which would cause toe-out. This angle would be greater than if the arm was meant to be level at static ride height with zero toe.

Sorry for the confusing questions.

I wish I had a model and a video camera to illustrate what I am saying....

Kind of like this picture: http://elephantracing.com/

[burgermeister]

Outer pickup = front outboard swing arm attachment
Inner pickup = inboard swing arm attachment

It's always an arc-shaped curve for a semi-trailing arm. The height of the inner pickup relative to the outer determines where the minimum toe-in is on the jounce - rebound scale. The rear suspension has about 100mm of jounce travel from its stock ride height to the jounce bumper. Something similar in rebound.

Bastow's suspension book has a pretty good explanation of semi trailing arm parameters.

[Flieger]

So you mean this where the control arm attaches to the spring plate?

[Walt Fricke]

No, where the spring plate pivots on the chassis. In your picture, where the torsion bar is. In my track car, where the Heim joint bolt is (which, in my car, is raised up a bit).

You can, as Burgermeister apparently has, with some welding raise the inner pickup by the tranny and still keep the torsion bar ('73 RS and 930s). The outer pivot point would be pretty tough to raise unless you have coilovers.

[burgermeister]

I didn't touch mine ... I just can't measure it accurately, so I took a best guess and came up with 10mm higher for the inner pickup. Also, it causes my cheesy XL model to match this curve, which I was sent by an old German friend, purportedly from an older German edition of Frere's book (and based on where in the wheel travel my car gets max neg camber up front, isn't quite the same car as mine. Also doesn't match a few other things I measured. So like everything, take the charts with a grain - or rock - of salt):

[Flieger]

Now that I have rear monoball bearings installed, I think I now realize how the toe change occurs. The rear control arm is free to rotate around its axis but also to swing fore and aft. That is why my monoball bearings need to be spherical.

The fore/aft pivoting is controled by the radius arm/spring plate, which can bend in and out. This complicates the calculations greatly.

Having the trailing arm axis above the spring plate axis causes the spring plate to pull the trailing arm forward on compression and causes toe-in. The opposite toe out occurs in rebound.

Do I have this correct?

[burgermeister]

The way I view it is that the semi-trailing arm only has one axis - it is defined by the elastic center of the spring plate bushings and the elastic center of the banana arm inboard bushing (or monoball). The axis has an angle in the top view, and possibly in the front view as well. Since this axis is not concentric with the torsion bar axis (which is parallel to "grid", ie perpendicular to the longitudinal axis of the car), something must deflect - otherwise the suspension would bind. Generally the spring plate bushings and to a lesser extent the spring plate do the deflecting.

You need spherical monoballs (or rubber) for the banana arm inner because:

1) for alignment purposes (meddling with the 2 eccentric bolts), the banana arm gets rotated about 2 axes (near vertical for the toe setting, maybe 45 deg from fore-aft by the camber setting) very different from the operating axis. Without a spherical bearing, you could not alter alignment settings by shifting the banana arm to spring plate connection around.
2) any structure deflects under operating loads. Axial bearings can be quite stiff, and thus can generate very large loads in response to applied off-axis moments. This can wreck the bearing, cause sticky operation (misaligned polybronze bearings on the front A-arm fall in this category), or cause structural damage to the vehicle.

[Flieger]

I think we are thinking of the same thing from the first part of your response. I had been thinking the control arm was the only thing effecting wheel hub orientation and that it simply rotated about its axis (45* from x and y viewed from above.) Now I realize the wheel path is also influenced by the spring plate geometry- it is not simply the way the springing force is applied. So, I now see the other link in your model.

You use the term "elastic center" and that is what took me a while to see. Even with the rigid monoball, there is still rotation allowed even though no (appreciable) radial movement.

[88911coupe]

This is a good thread but I'm still struggling with exactly what is driving the orientation of the rear wheel. I can see how the changing the plates length would change toe in/out but I don't see how it affects camber. Also, wouldn't there be a "natural" position for the in board trailing arm connection? I can't seem to mentally picture what controls what....if that makes any sense.

[5String43]

What Steve said. It's why you want to brake in a straight line and be on the gas early in a corner.

[Walt Fricke]

Buck

Look at Flieger's photo of a trailing arm minus the wheel hub stuff. As you have noted, easy enough to see that, with the inner banana arm mount mount acting as a fixed pivot, if you slide the banana forward on the spring plate, toe moves toward toe in, and vice versa if you slide it back. And the eccentric adjuster does just that once you loosen the two rear lock bolts - moves the outer end of the banana fore and aft.

It is much harder, for me at least, to get my arms around how the other eccentric causes a camber change. But the eccentric causes the whole banana to rotate (a bit), and because of the way the mating face of the banana is cut relative to some axis somewhere, this causes the top of the assembly to move in or out relative to the bottom, which changes the camber. The spring plate can twist enough to accomodate this easily enough, along with the fact that the plate is not at a right angle with the torsion bar (hence its name, in English anyway).

However, as the banana twists, the axle (which is not at the center of this rotation) also moves fore or aft some, which affects the toe.

I think the interaction is not as pronounced, and certainly is not as obvious in its geometry, as that between camber and toe up front.

Of course, this is for static toe/camber. Grasping why these two change the way they do during suspension travel is tougher for most of us.

There are some things a guy can study and at a point grasp somewhat. Then later realize he can't recreate the understanding.

[panzerfaust]

Originally Posted by 5String43

What Steve said. It's why you want to brake in a straight line and be on the gas early in a corner.

i recall there is a guy from the uk on this forum that trial brakes his aircool... carrying more speed before the apex, rotating the car quickly and possibly get on the gas later through the corner then you would braking in a straight line and getting on the gas early. he won the championship over there.

that is one area i think the old aircools loose a ton of time on the track to the moderns with ABS and shifting ability in mid corner under throttle. high speed trail braking can save time but you need ace driving ability.

[chris_seven]

I think everyone in the UK driving Appendix K Period F SWB cars uses trail braking and adjustable pedal boxes.

They also use LSD's that are set quite 'tight'.