Help: Alignment Troubles

[crstibr]

Hello ! So recently i’ve been attempting to get my E30 aligned properly as the car does not self steer back to center and appears to be ‘crab walking’ when driving in a straight line.

modifications :

Front: E36 BC Coilovers / Knuckles, GAS (FAPO) E30/36/46 angle kit, E46 Purple Tag Rack

Rear: E30 BC Coilovers, Poly trailing arm bushings , Poly subframe riser bushings, Normal Poly Diff Bushing (not raised) 5 Lug Conversion, Toe / Camber Eccentrics

Method : A combination of Toe plates and string alignment. the string is measured from axel / wheel bearing nut on each corer and adjusted so that the string is equal distance from each front and rear independently. when these numbers match, i’m confident the string is parallel to the frame.

Questions:

1) when pulling back steering rack boots, i’ve verified that rack is centered among itself (inner shaft same distance from outer rack housing on both sides) , but when i zero out the toe i am clearly seeing a difference in exposed thread on the outer tie rods. a. is this okay ? and b. why is this? is the rack not centered among the chassis? i am fairly certain that caster (lollipops) and camber(tophats) is the same on each side.

1) in the rear, i have zero toe in reference to the strings (parallel to car) but i believe the rear thrust may be off. crude measurements show a slight difference in measurement from wheel to front of rear arch. i believe this to be the cause of my ‘crab walk’ issue. is there any way to correct this?

Car has a clean title and does not appear to have pickups damaged also.

Thank you in advance !

[TobyB]

Starting from the beginning-

1) it's going to be best to find the true centerline of the chassis.
In front, that's the position equidistant from the inner ball joints,
and in back centering between the trailing arm mounts is good.
You then use a string setup to indicate the centerline from behind
the rear bumper to in front of the front. (with aluminum bumpers,
you can drill small holes in the undersides and tap them for allthread.
Then you always have center when you need it.)

Then you align the rear trailing arms so that the wheels are exactly parallel
to centerline. That's the straightforward way to get started.
Adding a half- degree of toe in might help stability under braking...
https://www.amazon.com/How-Make-Your.../dp/0912656468
is where I learnt this, the first time.

1) as to the rack, have you confirmed that you're making the same exact
number of turns to each stop? I've always gone end- to- end, counted the
turns pretty carefully, divided by 2, and rechecked my work a couple
of times. It doesn't matter so much with a rack, but back in the steering
box days, the box was only truly tight right on center.

1) as to not self- centering, if nothing in front is binding, the first thing
to look at would be caster. Depending on what you're doing, a little is always
good, and sometimes, more is better. I like the positive camber gain I get with it
on the racetrack, but dislike the heavier steering on a street car.

1) good luck!

t
b) is by no means an expert at this stuffs.

[GasolineJess]

Been exactly here: swapped rack, angle kit, strings, and a car that will not tell you the truth. Two things in the current setup are hiding your answers, then the fixes.
WHY THE TIE ROD THREAD DIFFERS (Q1)
Centering the rack bar in its housing only centers the bar in the housing. Where the rack sits left to right in the chassis is set by the swap mounts, and E46-in-E30 brackets usually land wherever they land. If toe is genuinely zeroed on both sides against a true centerline and the bar is centered, leftover thread difference just means the rack body sits a few mm off-center in the car. Not a fault by itself.
Two checks before you call it fine:
1. Count lock to lock, divide by two, confirm center that way. Your housing measurement is probably fine, but turn counting does not lie.
2. Confirm both inner tie rods are the same part with the same effective length. E36 knuckles plus E46 rack plus an angle kit is a pile of mixed parts, and a mismatched inner on one side eats thread for breakfast.
If both check out, center the wheel with the bar centered and never touch it again. Thread exposure is cosmetic. Toe per side is not.
THE CRAB WALK (Q2): YOUR STRINGS ARE THE PROBLEM
Strings set equidistant from the front and rear hubs on each side are parallel to that side hub line. That line is only parallel to the chassis if both axles are square to it. If the rear axle is cocked, the left and right hub lines rotate in opposite directions, your two strings converge or diverge front to rear, and you can zero every wheel to its own string while the car still crabs. The reference moves with the error, so the method cannot see a thrust angle.
Thirty second test: measure the gap between your two strings at the front of the car and again at the rear. If they are not the same, the strings are not parallel, and that is the crab showing up in your setup.
Fix, in order:
1. Build a true centerline. Plumb bob the front subframe mounts and the rear subframe mounts, run a string down the middle of the car, then set your side strings at equal distance from it, front and rear. Now the reference is the chassis, not the wheels.
2. Set the rear first. Equal toe per side against those strings, a touch of total toe in. Your eccentrics are exactly the tool for this. That is your thrust adjustment: zero thrust means equal toe per side against a chassis centerline, not zero against a skewed string.
3. Only then do the front, and center the steering wheel with the rack centered.
Skip the wheel to arch measurements. Arches are sheet metal, not a datum, and ride height moves them. If you want a damage check, measure hub to hub diagonals instead.
NO SELF CENTER
Two usual suspects with this exact front end:
- Caster. Fairly certain it is the same both sides is not a measurement. Unequal caster pulls one way and kills return. You can read it in the driveway with a digital angle gauge on the hub face: sweep 20 degrees each way, camber delta times about 1.5, both sides, compare.
- Bind. Sweep lock to lock with the front unloaded and feel for notches. The column joint to the E46 pinion is a known bind spot if the coupler is wrong, overtightened rack or poly mounts bind, and every joint in the angle kit has to move free. A bind you can feel unloaded only gets worse loaded.
And set the expectation: a big angle kit gives back less self center than stock, full stop. It should still return from small angles. If it is dead from 5 to 10 degrees off center, find the bind or measure the caster first.
This is a fixable Saturday. Good luck.