Bought a rack and pinion (freshy reman from the looks of it!) from a 96 Impreza at PnP, about $60 after core and whatnot on half off day.
Then a $25 tap to cut more threads on it... and chopped about 2" off one side and about 1" off the other - had to put the rack in off-center in order to have the input shaft clear the frame rails. Other solutions, such as routing the shaft through the control arms or through the crossmember itself seemed like an asinine solution when I didn't even know how big an issue it would be.
Outer tie rods are from many Toyota things and were about $10 each. I had to bore out the steering arms about 1-2mm, but there's still loads of meat on the bone, so I think I'm solid there.
Made my self a fancy-dancy wooden crossmember to hold the rack while I mocked everything up. Worked out so well I may just fiberglass it to the frame rails and run it!
Since it's what the cool kids always talk about, and since it actually is probably a good idea with something this custom, I measured the bump steer. Very interested in any feedback on these numbers. Not sure I have any real solution, since the rack and tie rod length is fixed at this point.... moving the rack up a touch might be possible, but from postings in my OT musing threads suggest down is the way to go, and since that would create issues with the steering shaft, I think I'm just going to leave it.
My baseline ride height is 285mm (floor to tabs on the framerail - arbitrary point). Got that by setting the engine in and then standing on the frame to account for fluids, sheetmetal, etc.
Height - Pass - Driver (toe change all are toe-in)
230 - 0 - 3 (bottomed out, shocks, I think)
250 - 0 - 0
270 - 0 - 0
290 - 1 - 1
310 - 2 - 3
330 - 3 - 9
350 - 13 - 15
My thought is to lower the car (convenient solution) 25-30mm, putting static ride height in the sweet spot. That would give me 30mm compression and 30mm rebound with a max of 3mm bump toe.
Funny thing about all this, and the pics don't really show, but the long tie rod is on the driver's side, so the inner tie rod pivot is closer to the "ideal" line between inner upper and lower control arm pivots.
So, theoretically, that should be the side that has less bump steer. But it doesn't work out that way. Maybe there's some voodoo at work here, but since I never bothered to find the instant centers, probably not really worth getting excited about.
I *am* excited to be moving on this thing finally. The bump steer issue is one of the things that pops into my mind at 3-4am for some damnfool reason.
Kidding aside, I got a free.99 sprint car frame off CL (delivered, no less) that I'll use as a source of tubing. Will have to think things through about the actual mounting of the rack - feedback welcome on this as well:
The crossmember is a bolt-in item, not sure if that's worth preserving or not; but if I weld the rack mount to the frame to the crossmember, that would be an issue.
Not sure if I should worry about tying the rack crossmember to the suspension crossmember anyway....?
There are some captive nuts in the frame for swaybar mounts, so I could maybe run some tubes from the crossmember to the frame, use the swaybar captives for a bolt-in, and then mount the rack on top. Probably be a bitch if I ever have to take the rack out, haha. But this sure feels like an epiphany. My back is killing me (getting old is fun) otherwise I'd probably hustle out there and eyeball things.
Engineering a car from a stack of junk sure is fun! (actually quite enjoying the process!)