loctite grades

what is the most appropriate grade of loctite for fixing my layshaft and mainshaft bearings bearings into the gearbox case? My mainshaft bearing can be easily pushed in and out by hand, the layshaft bearing is a little tighter but could still be pressed in by hand without heating the case. Mick Hemmings recommends the 638 but he may be assuming the bearings are looser than they are also I believe this particular grade is a little thick with a very short cure time and I envisage problems applying it successfully and I think it may be extremely difficult/impossible to remove once set. I would appreciate input from people who have used loctite for this purpose, thanks.

Your shell bores are trashed and I'm still waiting after 15+ year reading tales on a number of lists of loctite of any grade in engine or tranny stil locking on next inspection so forgetabout doing that way. Googe or search here with key words AMC gearbox bore repair JBW JB weld or bearing race staking or set screw to see alternatives that can save the shell. Might be cheaper to get Atlas shell at 1/2 the cost or just run like it is till if won't no more. When ya marry a once prior married gal ya can expect some baggage to get past.

Loctite 641 should be fine. It's for fitting bearings that are a little loose on the shaft or in the case. It fills gaps up to 8 thou but I doubt if you have that much gap. You just need to be sure that the bearing's not so tight in the case that it squeezes all the loctite out.

638 is higher strength faster curing. Either would do the job.

I have seen many bearings held on all kind of shafts an casings with all kinds of loctite an they worked fine. It is important to be carefully when preparing the surface that all oil is removed as this is normally the biggest cause of failure in my experience.
There are other loctite products such as cleaners an activators which may help but the best advice on such things would be to contact loctite or a supplier recommend by them.


If the bearing sticks when fitting then a blow torch will kill the loctite an should no harm the bearing hardening so after good clean you can try again. If it works and am sure many people will say it does then you have saved the casing. If it fails an like Hobot am sure many will say it does not work then you tried but just putting it back an running it as it is will end badly sooner or later. You could pin the bearings and also loctite to be sure but loctite on its own will hold if the damage to seating area is not to bad an if the rest of box is working as it should an not over stressing the joint so check shaft are straight etc.

Loctite 680. The ultimate apart from epoxy type resins..
good for 2 to 8 thou gaps. this is classed by the Loctite literature as a Permanent Assembly.
it requires nearly 200 centigrade to soften it for release. very good shear and compressive strength. can go off very quickly if high humidity. being classed as a permanent assembly, it will require heat to remove.
I have recently rebushed a bronze Acme nut .. the threads were worn but the shape of the nut was cast. so bored out old threads, made a bush, (plain OD, threaded ID) and loctited using 680 with 2 thou clearence. as a test, I put 1.5 tons pressure in a press on it after curing. got worried about causing damage so did not put any more force. 1.5 tons did not even look like moving the sleeve. it is good stuff.
If I ever need to, this is what I would try on a sleeve gear bearing.

Has anyone ever tried specifically Loctite 680 in the gearbox and had it fail. I have heard all the failure stories about any grade, but still wonder if anyone has used 680 here and had it fail. my bet would be that 680 would work. Given that the preparation is faultless. Bradley

Hope its that easy but might try heating outer surround of the loctited bush and put air impact on it at low pressure and work up to expected engine pulse torque and see what happens. My meager experience with red permanent stud grade is if much vibration occurs when its cold binding gap opens up thermally by disimilar metals then it kind of cracks up to a fluid powder state. At higher heat it turns gooey for removal. Try it and see please as its a no brainer if holds up short of racing straining.

hobot said:

Hope its that easy but might try heating outer surround of the loctited bush and put air impact on it at low pressure and work up to expected engine pulse torque and see what happens. My meager experience with red permanent stud grade is if much vibration occurs when its cold binding gap opens up thermally by disimilar metals then it kind of cracks up to a fluid powder state. At higher heat it turns gooey for removal. Try it and see please as its a no brainer if holds up short of racing straining.

Click to expand...

Gidday Steve, just letting you know mate, this is a Loctite green product.

I have also used this industrially lots. when a thread absolutly must not come undone, 680 is the go. if there is a problem with this stuff, it is breaking the joint and actually getting things apart. it really is a permanent assembly.

Big hydraulic cylinders running hot oil will not break it. Also big Allis Chamers vibrating screens. 600kg out of balance, 1500RPM, running on bearings mounted with Loctite 680, no problem. Thats why I reckon it will work. there are big forces on this stuff, bigger than a Norton box.
As I said, I would not hesitate to use this in a Norton box if I ever needed to.
I am still awaiting reports here of 680 actually being used in this application and failing.

Opinions to the contrary and corrections welcomed. Bradley

I talked to a loctite tech representative and he suggested 609 since the bearings are still a light press fit when cold and it is important to have good surface contact, the 638 is too thick. With the 680 I would be worried about the problem of dismantling if required since it does seem permanent. Of course if the 609 is not strong enough I guess I run the risk of the whole box dismantling in a major way at 100 mph! Decisions, decisions.

As I said I have yet to read of any loctite product still locking in *Norton* engines or AMC shell on next inspection and seems same with comnoz and why he uses JBW. I also said try it and see so we-me can do the same, if found it holds up in our elastic components. I may be lazest mechanic here always looking for short cuts so if was me I'd have machinist put set screw and bush in the thin cleavage area as a bugger to dig into tranny again to find out, yet on the other hand AMC boxes not used in race anger seem to function a long time with creeping or spinning races. If glue'd texture the race and bore to help what ever glue used to hold on better as shell expands more than the race and try to limit my time and power loads in lower gears unless mechanically fixed. If green loctite works long term then there are piles and piles of AMC around the world's vendors I talked to a decade ago that could be saved if not already sold for scrap. In my Peel case I bought an Atlas shell for $250+shipping vs C'do version with the notch that needs the thick washer spacer and costs 2x's as much for that hassle feature.