PDA

View Full Version : press fit advice from the press fit guru's



A.K. Boomer
01-31-2012, 07:34 PM
Im trying to save the cases on a ZF-6 transmission from a ford F-450

the front main (which is a tapered bearing) has spun it's outer race and has gained clearance leading to excessive gear noise.

there are no oversize outer races that I know of so im going to be making up a sleeve for the new bearing - It's an aluminum case - the bearings OD is 4.125"

Im going with a 6061 T-6 sleeve to be left thick on the ID of the sleeve and pressed in first - im then going to bore the ID of the sleeve to proper bearing OD dimensions after its pressed and settled --- so there's two press fits to consider - one is the aluminum to aluminum sleeve which will probably be around 4.225" - the next is the outer race of the taper roller into the aluminum sleeve @ approximately 4.125"

everything is going together with locktite sleeve retainer


something I don't understand about this trans --- these are the first tapered rollers Iv seen in an aluminum trans case and it did surprise me because the case is quite a span --------- which means that when it heats up the shaft loses its tolerance due to it of course being made out of a material that expands at about 1/3rd the rate...

anyways - advice is needed and im all ears, thanks.

Shuswap Pat
01-31-2012, 07:58 PM
When in doubt consult the eperts - SKF / Timken or NTN' I have sleeved lots of tapered bearings - if you do it right, you can finish the sleeve, and press/locktite it in and you are done. Check the SKF link, and it will guide you though the fit tables - they are basically the gospel!

http://www.skf.com/portal/skf/home/products?maincatalogue=1&lang=en&newlink=1_0_74

Carld
01-31-2012, 08:18 PM
How much is the gap and have you thought about using gap filling locktite?

wierdscience
01-31-2012, 08:49 PM
That's a regular job on electric motor endbells.I just roll up rings out of flatbar and weld the ends together.I have a 8" round piece of 1/2" plate welded to a 3" OD chunk of cold rolled round that I chuck in the lathe for a fixture.

I tack the ring to the plate and turn the ID and OD to the required size in one setup.On that size ,002"under on the bore and .003"-.004" over on the OD assuming the Aluminum housing is beefy enough to handle that much.Part the ring off the finished length and deburr.

Even with the shrink fit on Aluminum I will still drill and pin the sleeve into the castings in a few spots.

A.K. Boomer
01-31-2012, 09:54 PM
Thanks W.S. , wow - .003" to .004" on the OD on over a 4" part? really?

Damn that sounds steep bro, im seeing an aluminum sleeve buckling at that kind of press and if it does go would it not just gall it's way in?

Carld yes thought about the sleeve retainer but this has got to be a one time fix - the outer race has gained at least a couple/three thou and rotates by hand freely - there's so much labor in pulling one of these beasts - my bro wants it done with no worries...

thanks for the info Shuswap, can't find outer race specs into aluminum yet but will check in the morning with a clear head, (or at least one hopped up on coffee):p

Rich Carlstedt
01-31-2012, 10:19 PM
004 " is not bad.

The expansion on a 4" diameter piece of aluminum is .0044 for a 100 degree (F)
That means, if you freeze the bushing in the freezer ( zero F) and heat the transmission body to 200 degrees, you will have a .0088 differencial in size, so you should have .004" clearance for the fit. You have to move fast, but it will work

Rich

becksmachine
01-31-2012, 10:24 PM
Is there an overriding reason that you want to use aluminum for the sleeve? I would use steel. I agree that using that much fit with an aluminum sleeve would be an invitation to review your non politically correct vocabulary. ;)

The fits weird gave are standard "heavy press" fits, eg .001" tight per inch of diameter. Might be a bit much for an aluminum case. Or maybe not, as it is a sleeve and it will not have the rigidity of a solid shaft.

Might use loctite whatever the fits you decide, it seems to make a nice anti-seize/gall material.

Dave

wierdscience
01-31-2012, 10:30 PM
Thanks W.S. , wow - .003" to .004" on the OD on over a 4" part? really?

Damn that sounds steep bro, im seeing an aluminum sleeve buckling at that kind of press and if it does go would it not just gall it's way in?



Oops,forgot to mention steel sleeves.

No press fit either,shrink fit is the way.Aluminum in that size expands considerably.I did some parts in 6061 this summer,2-3/8" ID 6" OD rings shrink fitted over a 2.379" male 6061 boss.That was .004" interference,yet heated to 275f it was a drop on fit.

Not knowing how heavy your housing is use your own judgement,don't want to split anything.Getting bearing races to stay put in Aluminum bores can be a hair pulling PITA.

Here is a run down on the whole process of sleeving an endbell-

http://www.shaverkudell.ca/sleeve.html

Willy
01-31-2012, 11:19 PM
something I don't understand about this trans --- these are the first tapered rollers Iv seen in an aluminum trans case and it did surprise me because the case is quite a span --------- which means that when it heats up the shaft loses its tolerance due to it of course being made out of a material that expands at about 1/3rd the rate...


I have read a rebuild manual for this transmission and I can't remember all the details but there is a specified mainshaft preload that has to be checked if any major component has been changed, also Ford is quite fussy in this regard stating only their speced sealant should be used on the case in order to preserve proper end play/preload. Of course you'll have to check this very carefully as after this repair these dimensions will have changed. Unfortunately I don't have this manual.

GM also used this transmission on older Corvettes and later in their 8L and Duramax equipped pickups. In order to make them more quiet they used a different helix angle on the gears. GM also chose to not use the oil pump/oil cooler, instead opting to allow the transmission operate at a higher temp and using a better synthetic lube to handle the heat.

Not sure what kind of service this trans will see but under long pulls I have regularly seen 13 and 18 speed cast iron manual transmissions reach 235-250F with synthetic lubes. Knowing how much aluminum expands the .001 per inch interference fit does not seem at all out of line.

I hear ya on the PIA portion of a Re & Re, do as much research as you can until you feel comfortable about not having to do it again.

A.K. Boomer
02-01-2012, 08:55 AM
Good stuff u guys thank you very much,

My reasoning for going with the aluminum sleeve is simply due to not having the proper boring tool for my boring head for steel - the diameter is bigger than my boring heads travel but I already made up a nice extender bit that should cut the aluminum no prob.
Could make up a steel sleeve and do ID and OD and then press but it's going to be so thin (to keep the case strength strong) that I know it would have a taper before part off...

the aluminum to aluminum does present a problem when heating the case up for the first press - its' going to be instant transfer to a small part - not good

So here's what im thinking guys - and please feel free to critique the hell out of this cuz if im off base on this now's the time for me to re-calibrate,

Im really not seeing an advantage for heat with the first press - and I don't want any trouble with galling cuz the last thing I need is a bunch of highs and lows with empty space in there - so - I press the aluminum into the aluminum with the sleeve retainer locktite (good to know that becksmachine also thinks its got anti-gall properties as that's been my experience too)
with only a couple thou interference...

I then machine the ID to have a .004" press/shrink fit for the outer race --- then when the steel race heats up whilst pressing and positioning (u know how it goes) im not losing all my expansion values - and its going in with the locktite sleeve retainer - it's also making the press fit of the aluminum parts very stout as the .050" sleeve just transfers the press.
the race is nice and ground hardened steel and more compatible for anti-gall and is designed for this so should be good.

Willy I read that same thing about GM --- these are noisy gearboxes and they went with a steeper helical angle to keep the noise down --- I don't like the idea of just running them at higher temps due to the expansion values of the aluminum -- GM also states that their fluid is "lifetime" or the vehicle -- to me that's crazy - it's lifetime alright cuz if you hash a trans then that's the lifetime of the vehicle - lol

I like fords approach better - keep it cool as you can - and no reason why you cant run special synthetics in them also....

Thanks for all the advice...

Carld
02-01-2012, 10:20 AM
I'm with you on everything except the .004" fit of the bearing race. I would go with .001-.002" interference fit and the locktite loose fit retainer. The cooler the tranny runs the better it is.

Willy
02-01-2012, 12:06 PM
I like fords approach better - keep it cool as you can - and no reason why you cant run special synthetics in them also....


That's why I mentioned the GM 'lifetime' fluid. I would not suggest removing the pump and cooler.
I would though, as you surmised, use the best of both and use the Ford pump/cooler combo in conjunction with GM's better synthetic fluid.

Busy day for me today AK but I'll see if I can find the end play/preload setting procedure as this will be key to a long quiet life for the mainshaft.

cijuanni
02-01-2012, 12:28 PM
Usually if the trans has a tapered roller bearing,,,,,, it is a tapered on one end of the shaft and a ball or roller bearing on the other. The force of the helical gears keep the thrust directed to the tapered bearing end.

Is that how your trans is?

Willy
02-01-2012, 12:30 PM
AK, before I go let me leave this with you.
This is a link from ZF on the S6-650 six speed service manual as used on GM vehicles. I haven't had time to read it all but it seems very comprehensive and does include the all important mainshaft preload set up procedure.

I know you're on dial-up but grab a coffee, it's only a little over 2 mb so it will be there before too long.

ZF service manual (http://www.zf.com/na/content/media/united_states/corporate_replacement_parts_1/service_portfolio/lcv/GMServiceManual.pdf)

There may be a Ford one but I don't have time to look any longer, and as far as I know the only difference between the two is Ford's oil pump/cooler and GM's gear helix angle change.

Rich Carlstedt
02-01-2012, 08:06 PM
.004 is tight to do in a press fit.
.002 is more conducive, especially if it is steel
Reconsider a shrink fit.
For the same interferrence amount, a shrink fit is twice as strong (no spin) as a press fit

Rich

Arcane
02-01-2012, 09:50 PM
Never mind then...

A.K. Boomer
02-01-2012, 11:28 PM
Dang Willy I hope u didn't hunt too much --- I have the ford manual - got the adjustment shim kit coming tomorrow (all $160.00 bucks worth from ford:rolleyes: )
but thanks bro...
Cijuanni the main has two opposed tapered rollers - one at the front main and one in the middle main -- then there is a ball bearing at the tail end.

Arcane this main spun --- it has .005" clearance --- I really do like the loctite and permatex products but this can't come down again - it has to be right or my bro and I would worry about it too much - plus with that much clearance there's no knowing if the bearing will settle out cocked or hug the gear side or opposite - just don't like the idea of it esp. with all the heating and cooling cycles and loads. can't do it...
This trucks going to be hauling a snorkel lift around and stuff...

posting about the boring procedure now in "look what swallowed my mill" :p

A.K. Boomer
02-04-2012, 11:24 AM
OK --- as of yesterday the press-fits a done deal and im posting on here the results for future reference for others (including myself after I forget all this stuff in a week or two) to maybe use as kinda a hillbilly shoot from the hip guide for the future...

Keep in mind that my machining techniques are far from traditional - they are more like what to do with one machine when stranded on a deserted Island with all kinds of free time :p

The bore job turned out perfect - for a homebrew setup with just about everything it really turned out nice.

this is the final job at approximately .125" oversize

http://i146.photobucket.com/albums/r249/AK_Boomer/DSC02532.jpg

Next up was making the sleeve --- this took me quite sometime as I don't have all the tool bits needed and only have insert cutters - so to get a sleeve built with .0625" wall thickness out of aluminum and not have a taper is a bit of a hat trick without real sharp finishing cutters - so what do u do ? you become a "sleeve wisperer"

with the OD already finished to good size I then bored close to tolerance with the ID and got taper at the beginning of the cut due to all the flex - but the OD was nice/smooth and polished so I lubed up my fingers with light oil and pushed the material from that side whilst cutting for the final - had to do this a few times to get to know the proper pressures --- then it was a done deal - it was a very nice uniform sleeve...

http://i146.photobucket.com/albums/r249/AK_Boomer/DSC02536.jpg


Now for the biggie -- the tolerances and what I decided to do.

I decided to go for a .002" shrink/press for the aluminum sleeve into the aluminum tranny case ---- I wanted enough interference to squeeze all the extra permatex sleeve retainer out and yet enough predictability to be able to "pre-build" the sleeves ID so I would not have to put the case back on the mill for re-boring cuz it was such a hassle ------- this is the tolerance that I really pondered and in fact I slept on it -------- I settled on an interference fit of .003" between the bearing race and the sleeves ID,
I decided a little lighter here than some of the recommendations due to the "compound effect" of both the presses combined.

The sleeve install went like a cake walk --- due to them both being aluminum I took advantage of it and froze the sleeve and heated the trans to about 200f,,, applied the permatex - literally dropped the sleeve in - gave it a twist and then it grid locked and I watched all the extra permatex ooze out of the fit with gusto,,, very very nice...

This also spoke to me that the ID of the sleeve expanded and that of course was going to give me less than .003" interference for the bearing race when the trans was heated of course --------

so there - I leave this for u guys to guess how the rest went --- was it to easy ? was it too tight ? was it perfect ?

I do know the answer will surprise some...


This last pic is a doozie --- it's the only true mounting surface I had on the trans for indicating it in - but it was spot on from the top of the case to the bottom --- the only problem - One bolt was holding the entire procedure together - But I about deformed my mills table with that one bolt - and I also took only .005" per side cuts to reach my destination

http://i146.photobucket.com/albums/r249/AK_Boomer/DSC02534.jpg

A.K. Boomer
02-04-2012, 08:31 PM
Im surprised no one want's to take a stab at this?

You guys had opinions and I thank u for them but I would think you would want to know how close your opinion was --- by the way - no matter what your particular "outcome" is I appreciated your input...

so don't be shy - lets get this crystal ball rolling....

with what I did - was it too tight? too loose? or just right?

and I did have an infrared heat temp gun so I know I was in the ball park with temps, plus I did the ole spit test now and then too ;)

alright ----------- who's going to be the first to step up to the plate?

becksmachine
02-04-2012, 10:36 PM
Im surprised no one want's to take a stab at this?

You guys had opinions and I thank u for them but I would think you would want to know how close your opinion was --- by the way - no matter what your particular "outcome" is I appreciated your input...

so don't be shy - lets get this crystal ball rolling....

with what I did - was it too tight? too loose? or just right?

and I did have an infrared heat temp gun so I know I was in the ball park with temps, plus I did the ole spit test now and then too ;)

alright ----------- who's going to be the first to step up to the plate?

Sure, I have had 2 beers already. :D

My first impression of your description combined with my natural (un-natural??) pessimistic tendencies say that the fit was too tight as the sleeve was compressed by the .002" fit into the case, thus giving a total fit of .005"-.006" tight on the bearing race.

Dave

lakeside53
02-04-2012, 11:03 PM
I've had a couple too :)

To avoid an "under/oversize result", I would have taken a slightly different approach - made a very tight shrink fit plug (bigger than your 0.0625 wall), then bored it in place.

Measure what you have; maybe you got lucky.

wierdscience
02-04-2012, 11:10 PM
Well you said you went lighter,but didn't say how much lighter or how many Beers you had while doing the job:D

So I would say if you chilled the race and warmed the case it was a drop in.

A.K. Boomer
02-05-2012, 10:00 AM
Sure, I have had 2 beers already. :D

My first impression of your description combined with my natural (un-natural??) pessimistic tendencies say that the fit was too tight as the sleeve was compressed by the .002" fit into the case, thus giving a total fit of .005"-.006" tight on the bearing race.

Dave



Dave ---- U have been the most consistent and the closest in your predictions --- So im labeling u "press fit guru" :)

I read everyone's reply's and kinda went "middle path" with a little of my own technique thrown in,
by compound figuring I was right at about .005" but actually thought that maybe after the sleeve was installed that it would put up a little resistance and maybe keep the final press slightly lighter than .002" under some mild case heating,
this was somewhat verified after it dropped right in and then tensioned up, (but some of that illusion was due to having the aluminum sleeve in the freezer) I thought my race press was going to go perfect as I guessed I had far under .002" to press against (and more like .001" or less:eek: ) - and just by the way the sleeve slid in with all that permatex and then watching all the excess pump out of the interference bore as the sleeve heated I actually thought for a second that I might be screwed for any kind of serious resistance - but - even though it was steel I underestimated the expansion rate of the race - and instead of having it in the freezer it was sitting @ about 40f in my bro's massive unheated shop - The trans was just heated to about 200f with a torpedo heater so I grabbed the race and final adjustment shims and went for it - I don't think it would have mattered much if I had it in the freezer but it might have dropped in with dry ice... ?

The verdict is if I had to do it over again I would have went .001" less on the race to sleeve fit, I knew I had some le-way with taking the trans up to higher temps and that's exactly what I did --- @ about 250f we finally got it all the way in with a home brew all-thread press through the center bore...

I also had a limiting factor not just in case material but with this little plastic catch/drip tray that had to be installed with the sleeve - so I did not have a whole lot more of "comfortable" wiggle room above 250f temperature wise...

this was a real learning experience for me as iv never pressed anything so large - thanks for all the advice --- it's in and it's there to stay forever (it's permatexed too:eek: ) - which is not ideal if those main bearing shims ever have to be changed again... but with the way my bro is with fluids this tranny should outlast the truck (hopefully)

becksmachine
02-06-2012, 12:21 AM
Dave ---- U have been the most consistent and the closest in your predictions --- So im labeling u "press fit guru" :)

Thank you, thank you, thank you!

I am not convinced that I am worthy, but all gratuitous six packs will be accepted. :p

One thing to keep in mind, a shrink fit will usually loose less/none of it's interference initially, as opposed to pressing two parts together and the resulting sliding of the opposing surfaces which knocks the high spots off each.

Of course this is dependent on a number of factors including amount of interference, surface finish, lubrication etc.

Dave

A.K. Boomer
02-06-2012, 09:12 AM
It's that last .001" that gets u in a press fit --- to me press fits should be judged kinda like earthquake ratings - one notch more interference creates 10 fold more press needed.

So - U, Rich C. and Carld were all leaning to lighter and were correct - but from what you have heard would .002" and .002" been about perfect?
that's what im thinking...


A quick note to anybody doing something similar - some of us tend to think bearings are a total standard for press fit situations and are held down to very close tolerance on the ID's --- Iv always believed that to some degree, but instead of building the sleeve ahead of time (and using the old race as a guide) and maybe leaving a little extra (like .001") in the bore for clean up I waited for the new bearing to show --- there it was - a Timken with the same identical part number and made in the same country (U.S.A.) but instead of the bearings outer race being 4.127" it was 4.124":eek:
Had I decided to just "glue" it in That would have put me well out of the loctight/prematex range for sure (as the old bearing had .005" of clearance) so im glad I did what I did and so is my bro,

Wonder how many ZF-6 "rebuilt" trannies are out there rotating their front mains...:rolleyes:

Thanks again everybody for the input - yes it could have went better but it was an overall success...:)

Edit; One last note of a technicality - In having the case bored out I thought I had the ideal situation for setting up main-shaft pre-load, instead of having to press the bearings race in with "guess" shims and then pull it back out with ford tool puller # 4,367 (which by the way would have most likely proved to been impossible cuz the press was so tight) - I inserted a piece of .065" cardboard into the bore and used it to keep the new race centered with the "guess" shims behind - assembled, took my end play measurements - pulled the unit down and put the appropriate shims behind and reassembled and verified the proper pre-load by tightening the inner case allen's down... Then I pulled the unit down again and removed the cardboard centering devise and installed the sleeve and the appropriate shim and bearing race.
All was good ------ everything made perfect sense, So why was this method somewhat of a mistake ? had someone asked me before I did it it would have made me think - but I did not even give it a second thought while doing it...

Doozer
02-06-2012, 09:31 AM
One thing I know about, and others with more experience please chime in, is press fit motorcycle engine cases.
Some aluminum bike engines are press fit to the ball bearing races that support the crank and other shafts.
You need a big bolt-on puller to seperate the cases.
This is because the aluminum cases expand more than the steel bearing races when they get hot,
so the bearing races are a very tight press fit in the aluminum case halves.
I guess this is also sorta what you had to consider with this transmission.

--Doozer