I don't understand why that would be necessary? We're greasing the bearing surfaces, which means only the area the pinion makes contact with the the rack. That's inside the pinion housing. If the internal seals act as scrapers to keep grease from migrating out of the pinion housing, that seems like a benefit; we don't want grease escaping onto the rack and tie rod ends.

 

XS, what procedure did you use for this step? I assume it has to do with rotating the pistons under the t-bar?

 

Well, crap, I lied. McMaster-Carr.

 

I'm slowly discovering I can't trust myself to write from memory anymore.

 

After reading up a little on the way PS valves operate, I don't think 16valver's solution is any weaker than welding the spline shaft to the pinion.

The photos and discussion at the site you pointed to are a little misleading, the thin torsion bar doesn't transmit torque to the pinion from the steering wheel, it's just a spring that returns the valve body to neutral (A & B ports closed) when the driver releases the wheel. It allows the valve to travel a very small amount (the travel needed to open the A or B port by raising the associated spool against the spring pressure imposed by one of the two springs under the spool covers, aka "t-bar" in this conversation), then those pins that do the raising and lowering lock against the spools and transfer torque to the pinion. It's the spools that take the load, not the torsion bar.

It may be that welding the spline shaft to the pinion is stronger than using those two pins at the bottom of the spline shaft, but the torsion bar doesn't play a roll at all. The purpose of removing the springs and shims then replacing them with washers is to lock the spools in a closed position with the spline shaft and valve body in the neutral position. In that position, no torque is transferred to the torsion bar because the valves are locked closed (they can't move up or down so the pins can't move left or right). All torque is transferred directly to the pinion and since there's no valve play, there's no "slop" in the steering.

There's a very good discussion of how these systems work you can get online, search for "Hydraulic Power Steering System Design in Road Vehicles" by Marcus Ro ̈sth, it's a PDF file. Chapter 2 has an excellent overview. Here's one of the more important drawings:

 

BTW, if you're following 16Valver's procedure, don't remove the spools as suggested by The Machine Shop article, the rack won't work at all if you do that.

You can see in the photos on The Machine Shop site that the spools are ramps the pins on the spline shaft move against. Torque to the left pushes against the A port spool, displacing it and opening the A port while closing (or leaving closed, not sure) the B port. Torque to the right does the opposite unless I miss my guess. I'm working from photos, I haven't tried to remove the spools or the roll pins that attach the spline shaft to the torsion bar and valve body.

 

In light of my improved understanding of how this thing works, I think I can say what I called "pistons" are really spools and the procedure is to place the washers on top of them after removing the springs and shims, then torque down the retaining plate. The torsion bar and valves are already in a neutral position.

 

Another point; at first I was thinking the two covers over the spools should be kept free of grease, now I understand how this thing works I think the opposite is true, they should probably be packed with grease before they're closed up. they shouldn't move but they were originally bathed in ATF and if they dry out they're going to rust. Probably true of all the internal parts on the rack.

 

I think we have this mod on lockdown.

 

Not sure I'd say it is misleading, as it is a connection. But without locking those plungers/spools/whatever you want to call them in place, the narrow part of it will twist until the pins engage the body, at which point the unit will turn. I still think that for a track car, locking those pins in place is a bit less robust than making a more solid connection. It is pretty hard to fathom what kind of abuse a track car takes.......the most recent issue I had was that somehow the idle set screw (not the throttle stop on the bracket) came out of the throttle body. It has a rubber O-Ring, and is REALLY hard to turn, yet it somehow came out. Not saying that screw on the plate would, but I do feel more confident with what I have.

 

no, i did this years ago, before i was aware that type of mod was even possible.
but i am thinking of getting one of those weld-modified racks to swap in, there is a lot of slop.

 

But it it doesn't appear to be a load bearing connection. That's the part that troubled me; I couldn't understand why anyone would design such a weak link into what was otherwise such a substantial part. The pinion must be an inch thick cylinder of billet tool steel, it doesn't make a lot of sense to drive it with a quarter inch (if that) diameter piece of spring steel. I think the answer is, they didn't.

The pins in the photo don't look real thick either, but they're short. depending on how the two parts are welded together they may not be as strong. I'd like to see pictures of the welded part, but I also can't imagine it being real difficult to weld them together and since I have a tig welder I'll probably just bite the bullet and pull out that roll pin. I'd hesitate to call the mechanically locked design unsafe though.

If you look at the second photo on The Machine Shop's site you see the spools above and below the pinion. They have three channels cut in them and the collars between the channels have what look like about 45* sloped sides? Then look at the picture below, at the top of the spline shaft you see the two thick pins sticking out? It's those pins that engage the spools I believe, and the distance they move before bottoming out in those channels in the spool is the depth of that angled wall on the spool, a small fraction of an inch. As the pin moves into the angled wall, the spool is displaced in the valve body, opening or closing the associated hydraulic port. When the steering wheel is released the torsion bar returns the spline shaft to center and the small springs in the spool housing push the spool back, closing the port. It doesn't seem to me a person would really be able to feel that as slop. If the spools where removed though, as shown in the seventh photo, there would be lots of slop, possibly so much that the rack wouldn't work at all unless the pinion had been welded to the spline shaft. It's worth mentioning that if the springs are removed and the space between the top of the spool and the cover is filled with a couple washers, the spool can't move at all so the ports never open or close and the spline shaft and pinion are effectively pinned together. In this configuration, the torsion bar is mechanically isolated from any torque transferred between the spline shaft and the pinion.

I don't think the screw on the t-shapped plate will come out as long as the safety washer is installed, but I agree the two pins might be less robust than a weld. A lot depends on how the weld is done.

 

Both valid points, however a weld has to be done with TIG, in shop and it is irreversible.
The shim method with the lockwasher will not back out, and if for some strange ( aka Pcar law) it ever did, the driver would know instantly by the slack now back in the system and the car would still be on the road.
IMHO, a welded rack would wear out much quicker at the pinion rack gears due to the complete shock load now exerted on them.
The shimmed method allows for some of the immediate shock load to transfer into the torsion shaft, just like factory.
That's all I got to say 'bout that....FG

 

I ordered the shorter 12mm so I could drive them flush, otherwise the same parts.

 

My grease serts showed up yesterday, thanks Otto for the reference. If if rebuild these on exchage, Ill install the set sctrews and make them flush and add the 2 grease serts where they're needed. I'm rebuilding mine with new slightly stronger springs and thicker billet plates....