BMW Luxury Touring Community banner

1 - 20 of 35 Posts

·
Premium Member
Joined
·
2,086 Posts
Discussion Starter #1
Tried a different way of checking final drive bearing shim thickness.
Have used Dman's dial micrometer method and it works very well.
This method uses epoxy between the bearing and cover.

Steps:
1. Remove cover and shim.
2. Dry top outside race of bearing.
3. Put small amounts of solid 2-part epoxy on bearing.
4. Oil inside of cover
5. Install cover and tighten in place.
6. Remove cover and let epoxy cure overnight.
7. Peel off epoxy and measure thickness with micrometer.
 

Attachments

·
Registered
Joined
·
2,176 Posts
I assume that you made some sort of comparison between the two methods? If so, how close were the two. And if they were close, again I assume that the epoxy method would be used by those without a dial mic?
Very creative.
 

·
Premium Member
Joined
·
2,086 Posts
Discussion Starter #3
fpmlt said:
I assume that you made some sort of comparison between the two methods? If so, how close were the two. And if they were close, again I assume that the epoxy method would be used by those without a dial mic?
Very creative.
You have to be careful to only measure the flat part of the epoxy. Trimming off the sides works best. I did one this morning. Waiting for the second to harden. I did the first one three times with 6-8 samples each time. If you put the max pressure on the micrometer I.E. don't use the friction knob,you get the exact same reading as the dial indicator method.

dan
 

·
Registered
Joined
·
3,402 Posts
gbob said:
Here it is and it comes with "scale"
www.plastigaugeusa.com/
Used to get it at any automotive machine shop supply house / old time auto parts store

Bob G
I used that stuff in engine rebuilding many years ago. It works.
 

·
Registered
Joined
·
4,404 Posts
Use of PlasticGauge has been discussed here previously.
I have some and tried using it for the FD application.
It really isn't intended for this type of application and I couldn't make it work.

The epoxy method is interesting. Really neat idea. Lots of variables to wonder about it seems to me. Dimensional stability with temperature changes, and variation between epoxy formulations come to mind. You would have to trim the piece of epoxy down to get just that section which represents the gap between the crownwheel bearing outer race and the FD cover. But if it works repeatably, it's a gem of a method.

It would be worth repeating the method a number of times and comparing the results with another known method to demonstrate consistency of results.

But if it works with good reliablily it certianly would simplify the measurement instrument set required down to just a good caliper micrometer.

Keep us posted.
 

·
Premium Member
Joined
·
2,086 Posts
Discussion Starter #7
I examined 2 rear drives this weekend.

First: Dman dial indicator .020" measured at least 12 times
epoxy also showed .020" measured 8-10 samples
This converts to .51mm
add .05 or .10mm preload
Shim should be .55mm or .60mm, existing shim was .90mm


Second: Dman dial indicator .008" measured at least 10 times
epoxy also showed .008" measued 12 samples
This converts to .20mm
add .05 or .10mm preload
Shim should be .25mm or .30mm, existing shim was .60mm
 

·
Registered
Joined
·
322 Posts
It would be interesting to try using a piece of solder, it should flatten nicely and hold its dimension nicely.
 

·
Registered
Joined
·
4,404 Posts
When removing the crownwheel assembly from the cover after checking the gap, I heat the cover to ease removal. On some drives quite a bit of force is needed to remove the bearing from the cover. Using a press or hammering the bearing out of the cover could be damaging to it.

Solder might melt.

Dan, I wonder if you are heating to remove the cover and what temps you are using?
 

·
Premium Member
Joined
·
2,086 Posts
Discussion Starter #10
CharlieVT said:
When removing the crownwheel assembly from the cover after checking the gap, I heat the cover to ease removal. On some drives quite a bit of force is needed to remove the bearing from the cover. Using a press or hammering the bearing out of the cover could be damaging to it.

Solder might melt.

Dan, I wonder if you are heating to remove the cover and what temps you are using?
I use a heat gun to heat up the cover prior to installing. Or just stick it in the Texas sun for a while. 150-200 deg is all I want. When using epoxy the cover is only in place long enough to tighten the bolts,take the bolts out,remove the cover and crownwheel assembly and tap the bearing loose with a rubber hammer. Never considered the harm to the bearing because up to this point the bearing is going to be replaced anyway.
 

·
Registered
Joined
·
4,404 Posts
DanMartin said:
I use a heat gun to heat up the cover prior to installing. Or just stick it in the Texas sun for a while. 150-200 deg is all I want. When using epoxy the cover is only in place long enough to tighten the bolts,take the bolts out,remove the cover and crownwheel assembly and tap the bearing loose with a rubber hammer. Never considered the harm to the bearing because up to this point the bearing is going to be replaced anyway.
One of the possible reasons for bearing failure I have wondered about, in addition to improper preload, is trauma to the bearing during installation. Mechanical engineers I have spoken to on the subject have indicated that it would certainly be possible to cause micro-fractures in the balls or races that would lead to eventual failure. I don't know how much pressure would be too much.

I remove the bearing and crownwheel assembly from the cover using both hands, holding the cover with fingers and pushing the hub with both thumbs. If I can't get the bearing out doing that, I put the whole assembly back into the oven at 250 degrees F for a few minutes. The bearing will drop right out after that.

Not sure what that heat would do to you epoxy. You could try lesser heat. I've had some drives where I could press out the bearing from the cover when all components were equal temperature, but that is the exception. Most of 'em require some amount of heat to get the bearing out without pounding, tapping, or using a press.

(It is kind of surprising to me how much variation there seems to be between drives in terms of machined tolerances.)

Thanks for posting your experiences with the epoxy method. It seems to be a method most easily done by do-it-yourselfers.
 

·
Premium Member
Joined
·
7,546 Posts
DanMartin said:
I use a heat gun to heat up the cover prior to installing. Or just stick it in the Texas sun for a while. 150-200 deg is all I want. When using epoxy the cover is only in place long enough to tighten the bolts,take the bolts out,remove the cover and crownwheel assembly and tap the bearing loose with a rubber hammer. Never considered the harm to the bearing because up to this point the bearing is going to be replaced anyway.
I may be reading this wrong, but are you saying you are checking the shim measurement with the OLD bearing rather than the new one?
 

·
Premium Member
Joined
·
2,086 Posts
Discussion Starter #13
Voyager said:
I may be reading this wrong, but are you saying you are checking the shim measurement with the OLD bearing rather than the new one?
Yes, that way I will know what shim to order. The bearings do not vary by much in thickness. The 2 I checked this weekend were different by .0005 at the most. I will install the new bearing and verify shim thickness before installing the shim and closing it up.

These 2 drives were overshimmed by .30mm each. This goes along with what Curtis has been finding on his rebuilds.

dan
 

·
Registered
Joined
·
637 Posts
Don't get fooled into measuring and comparing just the bearing race thicknesses. You also have to measure the net result of bearing ball-to-race clearances on the thrust face alignments. This would require some sort of a fixture, and can vary substantially from bearing-to-bearing. It could easily vary more than say .004 inch - especially between manufacturers.

I really wish there was a way to measure the axial (thrust) stiffness of an assembled FD as this would be a better way to make sure the crown bearing balls are rolling on the correct lines on the races. It would eliminate the effect of varying shaft and housing diameters. It also would allow a FD to be evaluated without disassembly.

My guess is that the axial stiffness of the assembled FD axle with respect to the housing (K in lbs/inch), would increase sharply with applied load, but that with say 500 lbs of thrust will be on the order of 10^6 lbs/inch. The opposing tapered roller bearing would be in parallel to that but I suspect its axial stiffness is an order of magnitude less, so it wouldn't dominate the measurement.

Those are my WAG numbers.....
 

·
Wrencher Extraordinaire
2005 K1200LT
Joined
·
13,938 Posts
DanMartin said:
The bearings do not vary by much in thickness. The 2 I checked this weekend were different by .0005 at the most. dan
The one I am doing now has a solid 0.003" difference in race thichness from old to new.
 

·
Premium Member
Joined
·
2,024 Posts
John, would you say that given Dan's process, you could make the measurement of bearing races and then add or subtract the delta depending on if the new bearing is thinner or thicker to determine the shim thickness?
 

·
Registered
Joined
·
2,746 Posts
I have to agree completely with Mr Petersen......
Now, what did he say???? :think: :think::think: :eek:

Sure is good to have such a group of smart fellas subscribing to this forum!

Loren

niel_petersen said:
Don't get fooled into measuring and comparing just the bearing race thicknesses. You also have to measure the net result of bearing ball-to-race clearances on the thrust face alignments. This would require some sort of a fixture, and can vary substantially from bearing-to-bearing. It could easily vary more than say .004 inch - especially between manufacturers.

I really wish there was a way to measure the axial (thrust) stiffness of an assembled FD as this would be a better way to make sure the crown bearing balls are rolling on the correct lines on the races. It would eliminate the effect of varying shaft and housing diameters. It also would allow a FD to be evaluated without disassembly.

My guess is that the axial stiffness of the assembled FD axle with respect to the housing (K in lbs/inch), would increase sharply with applied load, but that with say 500 lbs of thrust will be on the order of 10^6 lbs/inch. The opposing tapered roller bearing would be in parallel to that but I suspect its axial stiffness is an order of magnitude less, so it wouldn't dominate the measurement.

Those are my WAG numbers.....
 

·
Premium Member
Joined
·
7,546 Posts
DanMartin said:
Yes, that way I will know what shim to order. The bearings do not vary by much in thickness. The 2 I checked this weekend were different by .0005 at the most. I will install the new bearing and verify shim thickness before installing the shim and closing it up.

These 2 drives were overshimmed by .30mm each. This goes along with what Curtis has been finding on his rebuilds.

dan
This is a poor practice in general. Bearings CAN vary much more than what you measured, especially if there is a change in manufacturer from one to the next. It is also a challenge to measure a bearing consistently across both races and the rollers/balls and ensure all slack is removed. However, I agree that generally bearings are made to pretty precise tolerances.

However, since you are checking the new bearing you should catch any changes before you button things up.

As an aside, given that 0.30 mm is about 12 thousands, half a thousandth is still a 4% error, but well within the 2 thou or so that BMW allows. However, should you get a difference between bearings of 3 thousandths or so, the error is now 25% and substantially exceeds the tolerance BMW allows.
 

·
Premium Member
Joined
·
2,086 Posts
Discussion Starter #19
Voyager said:
This is a poor practice in general. Bearings CAN vary much more than what you measured, especially if there is a change in manufacturer from one to the next. It is also a challenge to measure a bearing consistently across both races and the rollers/balls and ensure all slack is removed. However, I agree that generally bearings are made to pretty precise tolerances.

However, since you are checking the new bearing you should catch any changes before you button things up.

As an aside, given that 0.30 mm is about 12 thousands, half a thousandth is still a 4% error, but well within the 2 thou or so that BMW allows. However, should you get a difference between bearings of 3 thousandths or so, the error is now 25 and substantially exceeds the tolerance BMW allows.
Agreed. I had to order new bearings and went ahead and ordered a new shim. I can reuse one shim hopefully.
 

·
Premium Member
Joined
·
2,086 Posts
Discussion Starter #20
DanMartin said:
First: Dman dial indicator .020" measured at least 12 times
epoxy also showed .020" measured 8-10 samples
This converts to .51mm
add .05 or .10mm preload
Shim should be .55mm or .60mm, existing shim was .90mm
Did some more work on this drive this week.
With a new bearing installed,the dial indicator shows a .021 movement. Epoxy shows a .031 gap. Also tried using solder for the first time today. Picked a solder that
would need to compress to no less than half thickness. Found a .047 solder
in my stash. It compressed to a .031 thickness. Same as the epoxy.
Picture attached of the solder setup. I used six 1/2" long pieces set in grease to hold.
Did 4 pieces of epoxy at the same time.

Shim needed:
.021=.53mm + 5-10 for preload =.60 shim
.031=.79mm + 5-10 for preload = .85 or .90 shim

I am thinking that maybe the solder is too hard to compress. Am going to try
the epoxy again by itself w/o the solder.
 

Attachments

1 - 20 of 35 Posts
Top