Here We Go - Trans Shimming Doubts

[Ray from Indiana]

This post is about trans shimming. Yes, I did searches but I’m not satisfied with what I found.

I lifted a photo of the BMW manual shimming example from another site. Its shown below.
http://www.flickr.com/photos/98366272@N ... hotostream

Now just take a peek at the input shaft shimming.
I would think a better estimate of the assembled input shaft endplay (assuming zero gasket crush) would be:

Endplay = (measured bearing pocket depth in cover + gasket thickness) – (shimming plate thickness) – ( measured bearing height from plate) – (shim thickness used )

If you look at the example you’ll notice that BMW negates the thickness of the gasket when calculating endplay. The assembled endplay that would result for the input shaft using the BMW example dimensions would be

15.35 + 0.20 – 7.25 – 7.5 – 0.35 = 0.45…way outside the 0.0 to 0.1mm spec

Did BMW show this wrong in their example or am I seeing this wrong?

Then there is the gasket crush question. If I have a 0.31mm thick gasket and it crushes to 90% of original thickness upon assembly, that would reduce endplay by 0.03mm. If I had a target of 0.05mm endplay the assembled endplay would be reduced to 0.02mm – a little tight. That would mean a lot rides on the amount of gasket crush!

Can anybody with trans experience shed some light on this?

Regards,

Ray

[Kurt in S.A.]

Never seen the inside of a transmission...saw Matt Parkhouse go through the shimming process last year at the MOA National. Seemed simple enough...I don't remember anything about gaskets.

Here's a decent website on gearbox overhaul...shimming stuff is at the end.

http://jhau.maliwi.de/mot/gearbox.html

Kurt in S.A.

[esman100]

I have been very successful rebuilding two boxes and factor in the gasket to allow .002" or .05mm clearance. I am a moldmaker and know how to measure precisely which is important when trying to achieve tight clearances. A measuring plate is very helpful and in my opinion necessary for tight precision without guessing or refitting. If I used parallels to measure I would open up the gap to .1mm (.004") to allow for a potentially less positive reading. The more the clearance the more play in the system, too tight and it will run hot and burn bearings and destroy the case.

[Ray from Indiana]

Jeff

I wasn't questioning whether gaskets should be factored in - just how they should be factored in. Did you shim according to BMW's example or was your shimming calc more like the one I showed in blue?

BMW factored in the gasket but they subtracted out its thickness. The way their example was shown they calculated endplay assuming things were assembled without a gasket. This , at least to me, seems like it would end up outside their own 0.0 to 0.10mm spec.

I also really would like to know what people are assuming on the % gasket crush

[Ray from Indiana]

Here's a decent website on gearbox overhaul...shimming stuff is at the end.

http://jhau.maliwi.de/mot/gearbox.html

Kurt
I saw this site earlier too. It would appear, using Jurgs instructions that didn't add gasket thickness or address gasket crush.....which was why I wasn't satisfied with the instructions. Note the wide clearance he found in the box's original shimming

[esman100]

My gasket thickness was .008" and you do have to allow for it. I don't think that the gasket would compress more than .001". I do know that once the box is up to operating temp it grows in size and translates to even more clearance because Aluminum expands at a higher rate than steel.

[Ray from Indiana]

So you shimmed by adding gasket thickness to the bearing pocket depth measurement then subtracted bearing height above the bare case for the gapped to be shimmed to 0.05mm?

If gasket compresses .001" then there would be a reduction in endplay of .025mm, right? So wouldn't that have to be accounted for if shimming to a 0.05mm target?

[esman100]

So you shimmed by adding gasket thickness to the bearing pocket depth measurement then subtracted bearing height above the bare case for the gapped to be shimmed to 0.05mm?

If gasket compresses .001" then there would be a reduction in endplay of .025mm, right? So wouldn't that have to be accounted for if shimming to a 0.05mm target?

I don't know what the gasket compresses, but if it compressed .001" you still have clearance and shouldn't burn the bearings. The BMW gasket is very firm and the thickness is about .008" and the torque applied to the 6mm bolts is minimal, so it may very well be in the tenths of a thousandths. I didn't worry about that, but I worried about it turning freely by hand after it was assembled. I found you have to do some averaging because of the warp or inaccuracies of the bearing pockets and the cover. So the minimum clearance that I shot for was .002" and as much as .003 to .004" in some spots. If you can turn the input shaft by hand thru all the gears it will have a lot more clearance when at operating temp. If it doesn't turn freely then I think it's too tight.

[fig]

Why not install the plate with the gasket and torque to spec. Then take your measurement? I recently got a shim plate from cycle works and I think it says to do this in the instructions.

[Duane Ausherman]

I would guess that we did at least a few hundred transmissions in my shop. All employees also attended the official BMW workshop training, even the secretary and parts person.

The BMW instructions were to shim for zero end play and get the desired end play by adding the gasket. The details about the gasket came up and took a few minutes of discussion. The factory gasket of that era for the /5 was .1 mm thick, or .004" thick. It was red in color. At least one after market firm in Germany sold a similar gasket that was white and very slightly thinner. It didn't have the self sealing properties and probably didn't compress as much.

Too much end play on the input shaft and you will get the "grabby clutch" affect. Too little and you will have a quick failure.

Since the end play grows with temperature, it is necessary to have the transmission cool when doing this measurement. We didn't do any scientific measurements for how much one loses at freezing. There is some small end play that at some very low temperature will cause binding.

We aimed for just under the .004" end play at room temperature. We didn't accept .002" as "large enough" as we didn't know the lower limit. After all, we were in California. The German techs couldn't tell us either.

Do not dream that you can effectively use parallel bars to do the measuring. You really need a plate. We made our plates from old damaged covers. I think that I still have one or two of them.

Over time, I have read some of the experiences on the Internet written by someone who did ONE transmission. After I personally rebuilt more than a dozen, I thought I knew what I was doing. Nope, I was still finding glitches after at least 50 of them.

I suggest everyone stash a working transmission to use while you do your rebuild. The main reason to do your own is for the satisfaction of doing it.

I suggest reading every Internet site about this subject before you start. Put them all together and you still won't have all of the hints that you may need. Are you a person of immense patience?

Make a lot of notes and take a lot of photos. Do not take short cuts. Tell us about it when you finish. Good luck.