The following warnings occurred:
Warning [2] Undefined variable $search_thread - Line: 60 - File: showthread.php(1617) : eval()'d code PHP 8.1.27 (Linux)
File Line Function
/inc/class_error.php 153 errorHandler->error
/showthread.php(1617) : eval()'d code 60 errorHandler->error_callback
/showthread.php 1617 eval




Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Crown Wheel Run-Out
#1
In the throes of installing a new crown wheel and pinion.
I have measured axial run-out on the rear face of the crown wheel with all bolts nipped up = around 0.004".
Arguably acceptable but having already had the mounting face gently skimmed I'm frustrated that it hasn't dropped to nil!

I have tried rotating the two halves of the diff carrier relative to each other but it doesn't materially alter the reading, nor the angular location of the 'high' spot on the carrier (gear side). n.b. It is simple 1st order run-out we are talking about i.e. 1 high spot and 1 low spot per revolution, which are diametrically opposed.

I also tried rotating the crown wheel through 180 degrees relative to the carrier, but the high spot stays in the same position on the carrier, suggesting the gear itself is true.

I'm forced to accept then that the skim has not perfectly trued-up the gear side of the carrier.

I can of course accept this and continue the build, but I'd much rather get it spot on while the opportunity is there.

Can anyone suggest (preferably backed up by experience!) how to get rid of that last 4 thou?

Shimming is of course one option but I'm not entirely sure it's a good idea...
Reply
#2
4thou is a bit much. Are you sure there isn’t a burr or something somewhere in the joint? I’d start by giving the mounting face a good rubbing with scotchbrite or similar and try again.
Alan Fairless
Reply
#3
Hi Chris

No solutions based on experience I'm afraid but.....

Have you made sure the diff bearings are hard against the carrier? 

I've done an axle rebuild recently and found a lot more than 6 thou run out. It wasn't just simple 1st order run out and worked back to pitting on the back face of the crown wheel.

Hope you sort it!

Cheers

Howard
Reply
#4
Thanks to both. Yes, the bearings (albeit old ones) are hard up against the shoulder. I've gone over both crown wheel and mating face with a magnifying glass and found no burr or witness mark, and gently rubbed over the carrier with a fine file regardless. Also given both surfaces a good wipe down before assembly so both are clean.

How do people go about truing them? Presumably hold the cylindrical surface the bearing inner race sits on in a chuck or collet and skim the mounting face(s)? Or is there another way?

I'm currently clocking 0.11mm = 0.0043" (and as Howard points out, also trying to ignore minor surface deviations on the back of the crownwheel).
Reply
#5
Ah, now then. It’s not unknown for diff carriers to be bent. Have you checked that each half is running true to its bearing diameter before assembly?
Alan Fairless
Reply
#6
When I last worked on a diff the pattern was consistent and I had not heard of the tendency for carrier to be bent so did not check. From tales of engine lockups etc I can see how it occurs...
Sorting this sort of problem can require machining in toolroom territory and somewhat beyond the realm of many “mechanics”. Desired run out is better than that of many lathe chuck jaws.
 
How do rebuilders check runout prior complete assembly? Just sitting the side bearings on V blocks or equivalent hardly seems adequate. Turning without bearings on V blocks would not adequately involve the bearing shoulders. I suppose the ideal would be to hold side bearings stationary between chucks or equivalent in a medium lathe  and rotate the carrier. For most probably simpler to assemble.
 
If bent, machining of halves independently using the bearing mount set to run absolutely true seems promising but cannot touch diameters so unlikely to totally avoid problems when bolted together.
 
Much depends on how the carrier was mounted for skimming. Whoever did the work, what were their instructions? Simply packing in chuck to remove the “error” probably the simplest.
 
A lot can be achieved with a scraper, engineers blue, fine emery paper, straight edge, a piece of cadged plate glass etc....
 
Re the roller bearing question a few days ago I looked up the 1928 Austin reprint diag on p204  of Wyatt 3rd Ed. Checking again I note that .003 is given as runout limit and scraping is suggested. This reprint must be available on the many websites.
Reply
#7
I have skimmed a couple of  A7 diffs that had runout on the Crownwheel face. To do this I turned up a couple of top hat shaped plugs to be a close fit in the halfshaft bores. these were centre drilled at the flange end of the top hat.  With the plugs fitted in the halfshaft bores, the diff was mounted between centres in a lathe. The diff being driven at the headstock by a lathe dog mounted on the bearing boss.

As long as the halfshaft bores are concentric to the bearing bosses, after machining, this will deliver a true running crownwheel mounting face.  Chucking one end in a three jaw or similar self centreing chuck is likely to contribute some runout even if a fixed steady is used at the tailstock end.
Reply
#8
That's interesting. I am reassembling a 1928 axle and have clocked the run out at 0.008"   . The measurement made whilst one carrier bearing outer held in the lathe 3 jaw chuck , and pressure brought to bear on the other bearing outer race  via some spacers and the tail stock. mag mounted DTI at on saddle and  onto back face of the crown wheel.

It is a bit awkward allowing for the pitting in areas but, my run out appeared to be coincidental with one of the three, short studs.

I was measuring the run out as when bluetoothing, I felt a resistance in one area of crown wheel rotation.

I have now  dressed round the stud hole with a dremmel and mini wire brush and lightly filed the same area- or that third of the carrier face. That is the face that is on the half of the carrier that encases the planet gears . First attempt reduced run out to 0.003".   I may try boxing up and seeing if I can feel a tighter point when checking the mesh with engineers blue........ again.

I wondered if anyone else has a better method of correcting run out


 regards

Charlie
Reply
#9
Stuart,
I was recently given a very used diff carrier, and tried a similar way as you to ensure accurate results.
I turned a mandrel in the lathe to a good fit in the bush unfortunately the bush must not have been concentric, it proved too far out to continue.
I then realised that if I turned a bore to fit the a/c bearing seat in a piece of bar it would guarantee to hold it accurately. Having another carrier with 0.020” runout I used that and faced the crown wheel seating area, also the adjoining faces of both halves of the carrier. I machined as little as possible of each face.
After assembling the parts I then replaced the carrier in the mandrel and supported the other bearing so that I could measure the run out.
The reading was taken at 90degree intervals on the circumstance, three readings were 0.000” one was 0.004”.
This measurement was taken quickly as it was late in the day so will check again tomorrow.
Thank you again for prompting me to try this method.

           

Roger
Location:- Haverfordwest, Pembrokeshire.
Reply
#10
Have read recent input with interest, thanks guys. It is a little out of my reach though, not having access to a lathe. Trying to get someone else to do this kind of work while paying him an hourly rate quickly exhausts one's patience and budget both. I have for the time being therefore fallen back on the time-honoured expedient of looking for a better one.
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)