Steering Boxes: VW vs. TRW; VW Teardown & Rebuild

[Zundfolge1432]

Looks good

[OKType3Tim]

Thanks Zundfolge1432.

I've been holiday traveling for the past few weeks, but will be able to put up a complete rebuild post next week.

In the mean time, the repop shims sets have arrived and I'm pleased with the quality:

In an effort to keep the cost lower:

I've had 4 of the 6 roller shaft shims part#s reproduced. My thought is that if the top of the roller shaft is worn to where you need the 2.20 or 2.25 shim; then you likely have other severe wear on the roller anyway. Time and experience will tell if this was a good or bad choice.

For the Steering worm shims, I've had 5 of the 7 shim part#s reproduced. If you need the 2 thickest ones; you can get there by stacking two of the thinner ones. E.G. .20 shim + .25 shim gets you the .45 shim thickness.

And the repop gaskets are also here:

I'll be able to get the details on the rebuild kits posted next week.
_________________
'69 Squareback restoration
'69 Fastback restoration
'66 Type34

[OKType3Tim]

I'm going to use the next few posts to document the process I've used to rebuild several units. Thanks to Josh and my own needs, I've got a half dozen under my belt now. The pictures will be a mix from different individual rebuilds, with my intent to share the things I have learned along the way.

This first post concentrates on the housing.

The housing and cover have been completely degreased, and then blasted clean of all residue, rust, and paint using glass beads in my blast cabinet. I tape off the brass bushings with painter's masking tape to keep from blasting the bushings.

At this point I make sure that the surfaces where the small and large seals are located, are clean and free of burrs. (these pictures from after the fact.)

You might find a ding from a previous removal of a seal. This particular picture shows a ding but it will be ok. I.E. Use fine emery cloth to clean out the paint and make sure that there are no burrs; and the new seal will work fine.

For the first few units, I then primed and painted the housing and covers using OMNI MP170 epoxy primer with L41 (VW glossy Black) PPG Acrylic Urethane.

The ground wire cover bolt and washers, the large nut, the adjuster nut, and the marker rings were cleaned, and then zinc plated using the kit that Eastwood has.
http://search.eastwood.com/search?w=zinc%20plating%20kit

The top of the housing should be checked to make sure that the surface that the cover seals against is flat/true. I had one that was very much out. So I used my granite surface plate, with wet 1000 grit paper.

This got it down to an acceptable point. The arrows point to where the low spots were. You can still make out some of the original factory machining marks. The rest of the surface is bright and smooth from the work on the surface plate.

Finally, 10x1.50 and 8x1.25 taps are used to make sure that the threads are cleaned up of any debris and/or paint.

_________________
'69 Squareback restoration
'69 Fastback restoration
'66 Type34

[OKType3Tim]

Next, I build up the roller shaft.

The roller shafts are degreased, all the paint and gunk is removed from the bottom of the shaft and the splines where the steering drop arm attaches.
Then checked for wear on the surfaces that ride on the brass bushings. This one is pretty suspect at the top of the shaft. I rejected it:

Then check for wear on the roller surface:

With my limited experience I think this one has only moderate wear, and will still have some life in the roller surface. I.E. the shinny wear is not all the way to the bottom of the groove, and there are no FLAT spots anywhere on the roller surfaces.

Then the roller is checked for axial (side) play. The Bentley spec is that if a 0.05-mm (.002-in) feeler gauge can be pushed between the roller and washer, then the roller shaft must be replaced.

This one is bad (double the allowed play):

This one is just barely ok:

Note, that every bit of side play at the roller translates directly into free play when you move the steering wheel back and forth.

The roller rides on a cage-less needle bearing. The movement should be smooth without excessive play. Here are some pictures of one that was NOT smooth. (but it still turned.)

Yep, those needle bearings are worn out.

The adjuster screw bearing surface at top of the roller shaft must be clean and smooth:

And the circlip WILL BE worn, so replace it.

Note: The adjuster screw is a bearing. The adjuster screw is fixed/locked in the cover plate. And applies continuous pressure to the top of the roller shaft as the shaft rotates back and forth.

I found a unit that had been assembled "DRY" and even though it had never seen service, the adjuster screw already shows scaring.

Since the adjuster screw is a bearing, I make sure that the surface is nice and smooth,

by doing a few figure 8 motions on the surface plate:

With this result:

So now I have all the parts.

Acceptable shaft
Polished adjuster screw
Selection of Roller shaft shims
New circlip
Adjuster screw lock nut

Note that the shaft bearing surfaces are clean, bright, free of damage. On this unit, the bottom section of the shaft has been "blackened" to protect from corrosion.

I test fit the adjuster screw, shim and circlip dry.

And the adjuster screw turns nicely, just right.

BUT I disassemble and fill the cavity with corn head grease,

and the circlip has to be forced into the slot and the adjuster screw is so tight that it will not turn. So I back off one size on the shim:

Note: both the shims shown above are VW shims. So you get an idea of what the manufacturing tolerance on those shims were.

And all is well. No sloppiness, adjuster screw turns nicely, but firmly:

Next, I'll examine the worm shaft.
_________________
'69 Squareback restoration
'69 Fastback restoration
'66 Type34

[raygreenwood]

Niiiiiiiiccccce! Ray

[glutenfree]

Get some Ray! Thanks for sharing the process...adding to my bookmarks and I know I will be back here reading through every post when it's my turn to rebuild a box.

Why is this thread not a sticky yet?
_________________
1971 Fastback Build
VW/Type 3 Parts for sale
Instagram
my Shapeways store

[raygreenwood]

So...how well did the electroplating kit work?

I am not far off from trying to do some electroplate at home. Maybe even CAD plate with yellow zinc top coat. Ray

[ataraxia]

Tim,

Did that roller assembly with the shredded needle bearings come from one of the boxes I sent to you? Just curious.

Good to see you getting back to this...I haven't had as much time behind the wheel of my car as I'd hoped by this point.

[OKType3Tim]

[OKType3Tim]

[raygreenwood]

I found an,excellent series of videos on YouTube from a,guy who has large home tanks who does zinc and CAD with yellow zinc over the top...just like factory.

Yes....muriatic acid will strip it all. I have lots of little parts that need to be acid stripped just to get them clean.....but really need a protective plating after that.

I saw your fuel injection parts in the review comments list on Eastwoods site.....I am assuming ......the zinc/Tin is an interesting concept.

When I strip and rebuild calipera.....the acid takes the plating off the piston. It still works fine with polishing.m..but you lose about .0002"- .0003". A plating....and then dropping it into a vibratory polisher would be ideal.

I will find the links to tue ppating videos and post them. Ray

[Brent]

Your parts look great.

Note the Eastwood kit is tin-zinc. I had one I did some small parts in. For my carb parts, I ended up sending them out for yellow zinc as wanted a brighter finish. Caswell has kits that are pure zinc with the yellow/brite/black dips.

I'm not aware of any home cad kit. I think cadmium plating gets into a realm of much more toxicity.
_________________
69 Fastback Build
Berg5 Build

[raygreenwood]

[OKType3Tim]

So, on to the Worm Shaft.

I had to think through a way to explain the importance of the bearing and bearing race wear on the operation of the steering box. You may have to put your thinking cap on to follow me on this one. So here goes.

First, here is a well worn but nice looking worm shaft. It is the one that I used in an earlier post to find out the rockwell hardness of a VW worm shaft. There are no flat spots, but in the middle section of the worm it is nice and shiny all the way to the bottom of the worm groove. The bearing races at either end are shiny bright, but no pits or grievous damage.

Here is a set of used bearings which show no pits or obvious damage. However, they just "feel loose", compared to new bearings. They are from one of Josh's units and I just didn't feel comfortable with them. I.E. the bearing balls just seem to have a lot of clearance from each other, which to me is a sign of wear. When you measure the diameter of an individual bearing ball, it is only .0005" (1/2 thousandth) different from a new bearing ball. BUT, with 13 balls in the race, that is .0065" throughout the entire cage. In addition, the holes in the cage are worn enough that I might just be able to pop a ball out through the cage hole. However, I could see someone being tempted to reuse them.

For comparison, I have a new worm shaft that came from the TRW tear down. It is the one that I used in an earlier post to find out the rockwell hardness of the TRW worm shaft. It has some marking from the fact that the shaft was rotated many times prior to disassembly; but this shaft was never in service with any load on it.

And then a new set of original FAG bearings:

After a few attempts I came up with the method shown in this picture to measure the difference in the two assemblies.
[The machinists more experienced than I, may be slapping their forehead saying, "I can't believe he did that". But after trying a couple other methods, this provided repeatable results. And I'd be glad to know if there is a better method. The indicator is hanging out there pretty far, and it would have been better to shorten up the arm. I did have it locked pretty firm, and I tipped the parts stack in and out carefully to not disturb the indicator.]

When I stacked up the used stack of bearings and shaft, the dial indicator showed this reading:
(note that I'm going to be after "difference" reading, not an absolute reading.

Then I stacked up a new stack of bearings and shaft, with the dial indicator showing this reading:

(difference = 15.25 - 4.25) Working out the conversion, the stack of new parts is .14-mm longer (taller).

[Disclaimer here: I'm ignoring the potential issues of manufacturing variances, in order to make the explanation.]

Let's suppose that you are going to rebuild using the original parts. It would be reasonable to assume that the wear difference is equal on each end. So adjusting the big nut you take out all the slack. (Per Bob Hoover, adjust the big nut first!) So, you have moved the worm shaft .07-mm up toward the steering column. (The other .07-mm being taken up at the big nut end.) If you were keeping the original parts, and just adjusting the settings of both the Big nut and the adjuster screw, then you would need an additional .05-mm of shim to get the "no-play" zone of the worm back to the middle. E.G. if you originally had a .25-mm shim, then you would go to a .30-mm shim.

[in the upcoming post on final assembly, I'll detail more about the "no-play" zone. And discuss issues about more shim vs. less shim.]

So, lets keep the old shaft, but put new bearings on it:

Working out the conversion, we are now just .048-mm short of the full new stack. We got 2/3rd of the wear back just from the bearings! So we have gotten the worm shaft back pretty close to where it should be (or originally was). Final assembly and test will determine if we will still need to change the shim.

SO...What's the point? If you can't measure it the way shown above, and you don't have a new worm shaft!
My point is to share, that just because the bearings and races aren't blown out doesn't mean they don't have wear. And that wear is going to affect you as you try to adjust your existing unit. Adjusting the big nut has to be done, and it is going to take up the wear; but taking up that wear shifts the worm shaft up toward the the steering column. That shifts the center point of the worm shaft, which shifts the "no-play' zone. But you want that "no-play" zone to be located so that the car tracks nice when you are driving straight, with the steering wheel straight, motoring along the freeway at speed. To shift that "No-Play" zone without changing the shims means you start messing with steering wheel orientation and tie rod adjustments.

As always, comments and discussion welcome.
[If this post is not understandable, don't hesitate to speak up and I'll try to make it better. I had to put my own thinking cap on just to write it. ]

It also will help you to review this:
http://www.thesamba.com/vw/archives/manuals/practicalwork_steering.php
_________________
'69 Squareback restoration
'69 Fastback restoration
'66 Type34

[OKType3Tim]

The Cover:

I apologize for not finishing these posts up sooner. I was prepared to finish the entries, but as the holidays arrived, some advances came regarding the cover and it's bushing.

Evaluating the cover is pretty straight forward. Check the basic items:

1. That the mounting holes have not been crushed or cracked from over tightened bolts. The cover screws only need 15 ft-lbs; that's not much. So it is easy for someone to have over-tightened in hopes of stopping leakage. {Remember the information proceeding about getting the housing surface flat.}
2. No excess corrosion, or damage to the sealing surface where the cover meets the main housing.
3. make sure the threads for the roller shaft adjustment screw are clean and undamaged. Run a thread chaser through them if necessary. (M10x1.0)

Then check the bushing for wear. Because of the way that the adjusting screw applies pressure of the roller onto the worm; the wear pattern will be for wear at the top of the right side:

Notice the wear groove in the picture above.

And at the bottom of the left side:

Thus, it is hard to get a true wear indication with a simple inside diameter measurement. Measure the ID front-to-back, which will have the least force and least wear. Measure the ID right-to-left at the top of the bushing; and right-to-left at the bottom of the bushing. Comparison of those three measurement gives you a technical comparison. A practical comparison is to remove the adjusting screw from the roller shaft and then insert the top of the roller shaft up into the bushing. Compare the play (wobble) front-to-back vs. right-to-left. Excess right-to-left play is an indication of more severe wear on the bushing (which you probably already saw with a groove in the top of the bushing). And perhaps some wear on the top of the roller shaft.

Through a handful of rebuilds, I was able to find and use covers that had minimal wear.
But what to do with covers that had too much wear? How to replace the bushing? I discussed with a couple machinist friends and with the owner of a local machining shop that is also a VW enthusiast. Commercial machine shops would have to charge way to much to do a bushing replacement. But from from the discussions, I got the clues I needed to proceed.

First you need a fixture that you can use to hold the cover for accurate machining. Cutting off the top of a less useful housing provided the start.

Then milling the top to a consistent thickness.

Threaded rod to create a mounting and adjustment mechanism.

The first task is to be able to remove the existing brass bushing. By mounting the cover "square" in the fixture, the bushing can be cut out on the mill without damaging the cover bushing cavity.

Here are two. Note that the one of the left has severe corrosion. It is my "test" cover for this bushing replacement effort. If you tried to reuse that one, it would be pretty hard to get a seal between the cover and the housing.

Also in the picture is the source for the replacement bushing. It is an oilite cast bronze bushing (ISO EP-1418-24_SAE841). These can be obtained for around $3 each.

So then you need a way to cut the bronze bushing to size. One of the machinists found me a small legacy lathe. After some basic refurbishment:

I was able to learn to create the first successful part after only a few ruined bronze castings.

Based on discussions, internet research and some other VW experiences, I though I might be able to use heat vs. cold to easily slip the new bushing in place.

But, no that won't work. The mass of the bronze is so small that it almost immediately comes up to the same temperature as the cover.

But, mounting the cover in the fixture, the new bushing was pressed into place.

The ID had been turned close to finished size on the lathe. But pressing the bushing into the cover will change that ID. The bushing needs to be an interference fit; it can NOT be loose or spinning. So as the bushing goes into place, the ID is going to get smaller. So the last step becomes remounting the fixture/cover on the mill and boring the ID to final size.

I then lubricated the roller shaft and new bushing for test fitment;

and let it spend a night in the freezer to make sure all would be well when it is really cold out.

Now, on to final assembly.

As always, comments and discussion welcome. Including any tips or advice from the experienced machinists out there.
_________________
'69 Squareback restoration
'69 Fastback restoration
'66 Type34

[raygreenwood]

Really nice stuff!. In a month or two I will post the pictorial for my type 4 steering box rebuild.
While the type 4 box is very different than the type 3....more complex in that its a recircupating ball unit.....its actually simpler because there are less primary parts and only two bearings and two seals to source (same as one of the type 3 FAG ball bearings)....and no machine work.
I rebuilt two and have a thousand pictures to sort through. I have a third one to do and will get to use my new inch pound torque gauge for setting the worm preload. You can see some pics of that tool in the 411/412 forum.

But I will drop in a link when its done as it has some similar problems and interests to your type 3 build saga. Ray

[OKType3Tim]

Final assembly: (finally)

The unit I'll be showing is courtesy of Josh and the effort he went through to obtain true original VW shims.

The "Bentley Workshop Manual" and "Official Service Manual, Fastback and Squareback" have good intructions on this process. Further illuminated by the "Notes on the Practical Work for Subject No. 4, Steering". Although there is one point of confusion which I will point out below.

So here we are laid out ready to start. With (hopefully) all the tools we will need,

And all the parts, including new shims, new double lipped seals, new gasket, new circlip, and new bearings.

The housing, and roller shaft will be reworked as documented previously, and this cover is one that was measured and determined to be satisfactory.

With the housing, roller shaft and cover ready, the housing is mounted to the core of VW tool 280.

And the parts laid out for the first trial installation of the worm shaft.

To begin, a worm shaft shim of .035 is selected.

and inserted into the housing:

The inner bearing is placed on the worm shaft,

and it's race is placed in the housing, being careful to keep it oriented square to the opening.

The outer bearing and race are placed on the other end of the worm shaft and the full assembly is started into position.

Note that the race at the red arrow in the picture needs to be placed square to the race housing. You don't want to get started in a 'cocked' manner. Also, at the yellow arrow, I am using a worm shaft adjusting screw that is part of my toolkit. I don't want to damage the finish on the actual worm shaft adjusting screw that I will use in the final sequence. The worm shaft adjusting screw has only been engaged for about two threads at this point.

Using the worm shaft adjusting screw wrench, the shaft assembly is tightened into the housing

to set the inner race firmly against the shim.

The tension on the worm shaft adjusting screw is backed off and then set "until the worm feels rough when turned". I've interpreted that to mean, just feels a little rough when turned. The lock nut is added to the adjusting screw. I don't (yet) have an inch-pound torque wrench, but the spec of 1.3-2.1 inch pounds is very very light.

The bottom of the roller shaft is lightly greased

as is the associated housing bushing

The top of the roller shaft and associated cover bushing are lightly greased and assembled with the adjusting screw all the way out (roller shaft all the way up). And then inserted into the housing with the roller meshing with the worm. Bolts are installed on the cover and snugged.

The Steering Drop Arm is installed. (Again this is one that is part of my toolkit.)

Note that there are no gaskets or seals installed at this time.

The Measuring Head of VW Tool 280 is installed. At the red arrow, the arm should be as close to the worm shaft as it can without touching it. The screw is then tighten at the yellow arrow. Finally the setting plate is slid to the appropriate marked location for steering boxes "311.415.371" at the green arrow.

The adjusting screw on the roller shaft is turned in to provide adequate engagement with the worm shaft so that the Drop Arm can be rotated such that the alignment pin will drop into place. There are a set of alignment pin holes for left hand drive and a separate set of alignment pin holes for right hand drive. The drop arms are different between left and right hand drive vehicles.

Now you have the Drop Arm, on the roller shaft, set in the correct "Center" position.

Now I have added the steering column flange to make it easy to turn the worm shaft. And I've removed the alignment pin.

I'm going to turn the adjusting screw in until I have no wiggle or play, at this center spot.

At this point I want to be able to turn the Drop Arm 11 degrees (+/- 2 degrees) left or right with no play. So, that could be 13 degrees left and 9 degrees right as an example. The correct tool for this is VW 279, but I haven't located one; so I will have to improvise.

A simple paper drawing,

adequately aligned will do the trick.

I will rotate back and forth to see where the play is, and where the play is not.

Then, I turn the adjusting screw so that I have no play at 11 degrees left, and see if I can turn back through the center point without binding, and get to 11 degrees right with no play. And I keep fine tuning the adjustment screw to arrive at the best I can.

Now, here is where I got confused with the Bentley insturctions
The Bentley says:

[raygreenwood]

:-

Very Nice!.
One note on the cornhead grease that may make it easier to install and make it easier to keep your gaskets clean while sealing it up. The whole point or corn head grease is that its thixotropic.

It shear thins rapidly. Heat is a leverage/multiplier to this. You can heat corn head grease just a bit.....work it a bit....and it becomes a syrup.

You can then heat the steering box....maybe to 110...120F......and pour the cornhead grease in....turn it a rev or two...top it off....and it fills every void.

You are one or the few people I would loan my new inch pound torque gauge to if you need it. The memory needle will be in thursday and installed. I can put the spinner handle on it...and it has 0.5 inch pound increments.....0-30 inch pounds total. Ray

[OKType3Tim]

Thank you Ray.

Regarding the Corn Head Grease: I stirred it vigorously with the end of my screw driver. That made it liquid enough that I was able to pour it in. I was prepared to heat it, but turned out to not be necessary in my case. (note: My shop is heated so everything was regular 68-70 degree temperature.)
_________________
'69 Squareback restoration
'69 Fastback restoration
'66 Type34

[jgerock]

Very impressive thread! Bet you could rebuild the steering box that is in the BMW's. I'll bookmark this thread if/when I get another VW.

ZF brand 1974 gearbox. It had plastic/nylon bearing cages.