Bump Steer check

[dannobee]

I followed this thread https://forums.pelicanparts.com/porsche-911-technical-forum/1116600-19mm-raised-front-spindles-adjustable-bump-steer-no.html when it came out, but held off on commenting until I had all of the parts to overhaul the front suspension in my car.
Some of you know that I was a nascar crew chief for a day or three. I'll just say that no self respecting crew chief above the Late Model level would ever put a car on the track without checking (and correcting) the bump steer. Based on the mystery and misinformation on the aforementioned thread, I'm hoping to dispel some myths and show how it's done. Instead of just buying a bump steer gauge, I tried to keep everything cheap and available.
You'll need a FLAT 2' x 4' piece of plywood. Baltic Birch plywood works well and is dimensionally stable. Check it with a framing square for flatness before purchasing. You'll also need a pair of hinges. You'll also need a dial indicator and something with a roundish tip. While the car is still on the ground, measure the angle of the lower control arm. Write this down. Then jack the car up and safely support it so that the control arms can move freely. Give yourself some room to move around underneath, as you'll also need to remove the sway bar (or the links if it's through-the-body style.

Make a wheel plate, something like this. Wax it, mark it every 1/4 inch. Put a sharpie line in the middle to indicate your ride height. Porsches use 5 x 130mm bolt pattern. You'll want to check at least "3" up, 3" down." You might have to remove the brake caliper to get the wheel plate to bolt up. You can usually carefully remove it without loosening the brake lines. Put the lug nuts on finger tight.


Next you'll need to remove the torsion bar adjusters. Carefully measure the distance between the stop and the adjuster. Write it down so that you can return it to that spot when you reassemble.



Next you'll need to cut two more pieces from your plywood. One will need to be tall enough to reach the sharpie mark on your wheel plate. 22" and 16" seems to work, and gives yourself enough room to work under the car. Hinge your remaining two pieces of plywood. You'll need to affix your dial indicator to one end. I used a magnetic base dial indicator and just found a piece of scrap steel and screwed that to the vertical section. For the other end, you'll need to affix your rounded tip tool. Make them solid so that they don't move. There's no real pressure involved, but you don't want anything to flex. Now put the vertical plywood against the wheel plate. The solid point will touch at one end, and the dial indicator point will touch at the other end. Verify that the two points touch and clear the lug nuts throughout the range of travel.



Now using your floor jack on the lower ball joint, you can match the lower control arm angle to what you had in the first step. Since the torsion bars are not supporting the car, the only things restricting the movement of the spindle are bushing stiction and shock valving. Move stuff around (or mark your plate) so that at the correct control arm angle, the two measuring points are in the middle range of the plate and your dial indicator is at 0.500". At the middle point (ride height), you want to measure how much the toe changes throughout the range of motion. Write these numbers down. On my car, with stock tie rod ends at euro ride height, I got:

3" jounce -0.575"
2" -0.410"
1" -0.240"
0
1" rebound 0.185"
2" 0.400"
3" 0.590"

Yes, it toes out on jounce and toes in on rebound. Over 6" of travel, the toe changed over 1". Over 4" of travel, "2 up, 2 down," it changed over 0.800" Most crew chiefs would consider this...a..."wicked" amount of toe change.

Cont. Pt. 2

[Nick Triesch]

Or, if just regular driving, stay away from railroad tracks!

[dannobee]

Part 2.

Now that we know what we have, what do we do?

I cut a couple of rack spacers on the lathe. A couple of 12mm spacers changed my readings to:

3" Jounce -0.240"
2" -0.170"
1" -0.075"
0
1" rebound 0.070"
2" 0.150"
3" 0.210"

Clearly, a new plan of attack is in order.

Not seeing any commercially available kits that piqued my interest (unconventional design, too costly, bolt size too small), I looked elsewhere. I did find a bump steer kit for a '94 Mustang that looked promising for $100. Heim joints and shaft were 5/8", plenty strong for something as light as a 911. The taper wasn't quite the same as the Porsche tie rod end, but at least I could easily correct it on the lathe. And the top side shaft size was 1/2". I matched the tie rod taper on the lathe, and ran a 1/2-20 tap through the spindle, to catch a thread or two without reaming the taper (in case I ever wanted to return it to stock). On the inner tie rod side, I wanted to extend the threads in a bit to clear the adjuster that was included in the kit. Off to the lathe to turn down the inner tie rod shaft to match the rest of the threads (14mm-1.5). The result was this:



Big ole rigid shaft the same size I'm used to. Now we're getting somewhere. After fiddling with the shims, I ended up here.

https://www.youtube.com/shorts/g-Rs2NU9ZqM

Now the toe changes less than 0.010" through 2" of jounce, and about 0.020" per inch in rebound. Good enough for the girls I go with.

Now, if I was racing this thing, I would also check bump steer after turning the wheels each way 10 degrees. If it was off, the rack would need to be moved backward or forward until the readings were acceptable.

Hope this helps anyone who is interested in checking their bump steer.

Total time for initial checking was less than an hour. Adjusting after fitting the bump steer kit took another hour and a half. YMMV

[stownsen914]

Thanks for sharing. From what I gather, most don't actually check bump steer.

[famoroso]

Yes, indeed! Thanks for sharing.

Jae Lee / Mirage International offers this service (both the actual measurements and the proper outer tie-rod to then minimize bump steer).

He did (the verb version of "bump steering") on my '88 M491, in conjunction with new Bilstein HDs, front bushings, corner balance & alignment, and fixed a hellacious case of steering wheel kickback / "bump steer" (the noun version). He also did my '87 in connection with raised spindle KW V3s, new bushings, corner balance and alignment.

Totally worth the investment in both cases given how sublimely the cars handled / performed after the work.

Once my '89 project is ready, it'll go to Jae for the same procedure.

[Mixed76]

Hi Dannobee,

I like your method, thanks for sharing! How do you know you are measuring toe change only and not camber or the arc of the lower control arm swing? Are those just much smaller than the toe change?

Sent from my Nokia 7.1 using Tapatalk

[dannobee]

Camber change will move the vertical board in a different plane. If the dial indicator and point were both vertical, we would be measuring camber change. Since they're horizontal, all we are measuring is the change in toe.

[Mixed76]

Ah, thank you so much! I didn't see that the two points were horizontal to one another!

Sent from my Nokia 7.1 using Tapatalk

[Tremelune]

Isn't bump steer correction just a matter of lowering where the tie rods connect to the spindle by an amount that has them level at rest...?

[dannobee]

Nope. If you want to just guess, a closer approximation might be parallel to the control arm.

[tperazzo]

Thanks for sharing this method. Do you position the solid tip and indicator at the tire diameter? Or is your distance arbitrary to only improve your bump steer before and after changes.

[dannobee]

Tire diameter or the width of the plywood will give a more accurate final result. Since the plywood I used is 24" wide, it approximates the tire diameter. If we extrapolate the 24" readings to a 16" diameter wheel, obviously knock off 1/3 of the reading. In this case, 0.030" toe change over 4 inches of travel would be 0.020" movement at the wheel.

[Jonny042]

Dannobee, thanks for posting this! Good to have an expert example of the measurement process done the conventional way.

Wicked toe change is right!!!! I never tried to quantify the amount of bump steer with a stock setup at Euro ride height, it's good to see the numbers in black and white.

I understand that measuring with the hub mounted board and a dial indicator is more precise and the only way to observe the actual curve, but do you take exception to my simplified method of simply measuring toe change vs ride height change using toe plates, then changing shims to minimize bump steer over a set range of jounce and droop?

[dannobee]

I saw your way and although it isn't quite how we'd do it in racing, it's better than what the factory kicked out.

In the big scheme of things, the wheel plate and dial indicator method doesn't really take that long to perform (especially once you've done it a few times), and is accurate down to a few thousandths of an inch.

Speaking of factory compromises, I'm not going to fault them for their bump steer curve. Between the bean counters and the time constraints of mass producing a product, it is what it is. And lets not forget that the vast majority of their customers are not race car drivers vying for the very last tenths of seconds. The "wicked" toe change is at least in the "safe" direction and will give the driver plenty of tire squealing as feedback as the car approaches the limit of adhesion before it flies off the track.

[Jonny042]

Originally Posted by dannobee

Speaking of factory compromises, I'm not going to fault them for their bump steer curve. Between the bean counters and the time constraints of mass producing a product, it is what it is. And lets not forget that the vast majority of their customers are not race car drivers vying for the very last tenths of seconds. The "wicked" toe change is at least in the "safe" direction and will give the driver plenty of tire squealing as feedback as the car approaches the limit of adhesion before it flies off the track.

The factory also designed around a 14" wheel! We've come a long way from this:



to this... some adjustments are going to be needed!

[dannobee]

Exactly! I just wished that over time they updated the front end to Short Long Arm instead of sticking with struts for as long as they did. Upper and Lower control arms, like the 959, 956, 962, 919, etc. Oh, and like the new 911's!

[David Inc.]

Man I knew going from the normal rack to the rack spacers at lower ride height felt better but I had no idea it was that radical. Taking it all the way down sure sounds better...

[Jonny042]

Originally Posted by dannobee

Exactly! I just wished that over time they updated the front end to Short Long Arm instead of sticking with struts for as long as they did. Upper and Lower control arms, like the 959, 956, 962, 919, etc. Oh, and like the new 911's!

So now that we've corrected the front, what to do about the rear? I've been thinking (dangerous I know) more and more about what to do with the bump steer curve in the rear.

I know improvements can be made when you go to coil overs and relocate trailing arm pivots upwards in relation to the chassis, but assuming we are going to stick with torsion bars, the only conclusion I can come to is to fit a stiffer torsion bar in the rear and try to limit pitch and roll to some degree, thereby reducing rear toe change.

I've tried upping the rebound and compression damping to better control pitch and roll or at least slow down the transitions so the rear steering isn't as odd feeling, but I pretty quickly got to the point where the car felt overdamped and uncomfortable.

To that end, I've ordered a set of slightly upgraded 20 front/27 rear bars (up from stock 19/24) which will help dial out a bit of understeer and do a better job of controlling the rear ride height.

A lot of the conventional wisdom of 911 tuning (get yourself some 22/29's and the problems go away) have merit but I'd rather maintain some compliance, especially in a car without a full cage to stiffen the structure.

[dannobee]

Regarding stiffer springs, and I'll likely get flak in this forum for saying it, is that we moved away from "stiff springs, soft bar" decades ago. Once we started "doing the math," and switched to soft springs and big bars, the cars went faster. (It's kinda tough to argue tuning philosophies when your competitor is consistently faster than you)

If you have ever checked tire temperatures before and after any changes, you've probably noticed that "springs build heat." If we could somehow keep soft springs in but still limit body roll, that would be good. And from an aerodynamics point of view, the soft springs will compress easier at speed and reduce the overall frontal area and therefore drag. That's also a good thing. You just need bigger sway bars to make it work.

[David Inc.]

Originally Posted by dannobee

Regarding stiffer springs, and I'll likely get flak in this forum for saying it, is that we moved away from "stiff springs, soft bar" decades ago. Once we started "doing the math," and switched to soft springs and big bars, the cars went faster. (It's kinda tough to argue tuning philosophies when your competitor is consistently faster than you)

If you have ever checked tire temperatures before and after any changes, you've probably noticed that "springs build heat." If we could somehow keep soft springs in but still limit body roll, that would be good. And from an aerodynamics point of view, the soft springs will compress easier at speed and reduce the overall frontal area and therefore drag. That's also a good thing. You just need bigger sway bars to make it work.

How soft is soft?

This reminds me of the F1 design philosophy which tends to be: just enough spring to keep it from bottoming out at speed, then add bar to keep body roll under control.

Edit: Then again we're not quite dealing with F1 aerodynamic loads on 911s are we?