We read threads on here about bearing outers being loose in the crankcase housing. We read about Loctite being used to fix the bearing outer in the crankcase with varying degrees of success. We then read (I think from Pete R) that at running temperature the outer bearing turns in the crankcase anyway. Certainly the fit of the bearing in the crankcase housing, which is great at room temperature must be much less so at running temperature. So with tens of thousands of these engines having run millions and millions of miles do we need a better retention method for the bearing outers in the crankcases or should we just carry on as we are? Should, for example, a mechanical retention device be used as well as the interference fit. Would be interested in your views and especially in your experiences ref this matter.
As long as you have the original interference fit on the outer and the correct running clearance inside the bearing there is nothing to worry about.
Adding interference would require a C4 bearing fit, not sure if the alloy could handle the higher interference and lose the greater interference fit over time and you would be back where you started, a pin half in the alloy case and half in a ground bearing outer would work but doubt its needed except for racing maybe.
Assuming the elements of the bearing are in good shape and it doesn't have too tight an internal clearance for the application and has suitable lubrication, the rolling elements of the bearing do not transmit any significant turning force to the bearing outer. If the clearance fit of the outer in the housing is roughly that as specified when cold, it will still prevent outer "creep" at running temperature (I guess roughly in the 80-95 deg region)as the bearing will be at a higher temperature than the case around it, due to the work it is doing displacing oil etc. Of course, when the housing is worn it may allow the outer to creep, in which case loctite may be the best solution. I can't see an obvious way of putting a peg in there, unless along the lines of a woodruff key, then how do you stop it from slipping out? Dave
This discussion is as old as as the first time someone restored a motorcycle with an aluminum crankcase. I remember reading some pages penned in the late 1930's by P.E Irving about techniques used to repair crankcases crankcases made in the 1920's. Several race bikes, like the Matchless G45 500 race bike had their bearings retained by three counter sunk screws sitting in flats ground into the outer face of the bearing. And the G45 had a center bearing that eliminated most of the crankshaft flex.
British engineering designs developed to meet the needs for performance, and later fuel economy, provided for light, almost fragile crankcase castings. While the US had all of the iron ore in Minnesota and crude in Texas engineering followed a different path. As time went on the lightness followed in the design of the flywheels, and crankshaft flex is one of the factors that create this problem. I think the worst design is the Vincent where with 4 widely spaced bearing they routinely come loose and spin in the crankcase.
I fear our grandchildren will be having the same talk if they are fixing T140's 20 years from now. JMO John
Last edited by John Healy; 02/23/132:41 pm. Reason: PE Irving 1930's
the rolling elements of the bearing do not transmit any significant turning force to the bearing outer.
If the clearance fit of the outer in the housing is roughly that as specified when cold, it will still prevent outer "creep" at running temperature (I guess roughly in the 80-95 deg region)as the bearing will be at a higher temperature than the case around it, due to the work it is doing displacing oil etc.
The rolling elements in a wheel bearing do not transmit any significant torque to a tyre,but the tyre still rolls on the road.The main bearing rolls around in a housing that has become bigger than the outer bearing race.It won't move much and it won't always be rolling in the same direction.It won't be skidding (unless it siezed);it will be rolling.
A 75 C. increase in temperature will loosen the housing 0.002"-0.0025".Most Triumph mains don't have that much interference at room temperature.Even if the outer race were 20 C hotter,it only changes by 0.0006". The bearing people will tell you it's OK if the outer race is not a tight fit in the housing.They prefer it that way.If it moves (and they do),the load is not always on the same part of the race. The inner race should be tight or clamped tightly to the shaft.A loose inner race will create problems.
About 25years ago I pegged a drive side roller outer race in a competition motor. The method was simple and ensured the race wasn't going to turn. I drilled a small hole in the c/case bearing tunnel and dropped in a small steel ball, making sure it protruded just a tiny amount. Then I ground a small corresponding slot in the race outer. To fit the bearing I simply aligned the slot with the ball. It didn't give any trouble though I never used the method again.
So, if as Pete R says, the outer bearing race moves around in the crankcase bore, what does this do for crankshaft location? We talk in terms of tightening the crankshaft onto a main bearing inner race as though that rigidly fixes the crankshaft in the crankcase via virtually no clearance between the inner and outer races of the bearing and assuming the outer race is firmly fixed into the crankcase bore. So in reality we have the outer bearing race moving round and round and laterally in the crankcase housing (moving in and out)so the crankshaft is moving from side to side subject to whatever end float is built into the bottom end, taking with it the conrods and the pistons. All moving in and out and round and round. No wonder we get vibration in these engines! Seems to me to be a good idea to firmly fix one at least of the main bearing outers firmly into the crankcase in a way that prevents all movement between the outer bearing race and the crankcase bore. John mentions that this was done on the G45. I have certainly also seen this on some Manx Nortons. What ideas are there for a good method to easily incorporate this on, say, a unit Triumph twin engine? As John says---this may be a perennial question but this is the right time of the year to give it some thought---if only as a way of fighting "cabin fever"!
You're asking a lot from Loctite in this situation.6 square inches of loctite will have good shear strength,if the surfaces remain in contact without differential thermal expansion.
What is the tensile strength of Loctite? Would you Loctite a broken bolt together and expect it to hold? You're expecting this sort of tensile strength if it can stop the crankcase expanding away from the bearing.Is the Loctite strong enough to stretch the bearing to a bigger diameter,purely by adhesion?
Apart from expansion,you also have some serious forces which could stretch the housing about 0.001",if not more.
Loctite doesn't work.
The "Superblend" thread on the Norton Bulletin Board proved this. Superblend
Whatever end-float you have at room temperature,you can expect 0.005" more when the temperature increases by 75 C.(main bearings 6" apart).I would only expect the bearings to be restrained by the crankcase shoulders,not by any shrink-fit or Loctite.
0.010" crank end-float at running temperature should be fine.The conrods can only move 0.0075" either side of cylinder centres (the rod has 0.002-0.005" side-clearance on the crankpin).
I don't think any of this is a big problem if the rods are central on a cold engine,allowing for end-float.If you found 0.020" end-float,it could be worth fixing to keep everything central.
Anecdotal evidence of one person applying Loctite on one bearing and reporting it didn't work doesn't really prove a point to me. Also what does a broken bolt have to do with an interference fit bearing in a housing, I fail to see the comparison there. One thing I can say first hand is when I removed my drive side bearing in this last engine I rebuilt much heat was required to finally release it and green Loctite was in evidence all around the bearing O/D and the housing with no sign of it having ever spun, so as far as I'm concerned my jury is still out on this one on Triumph engines, 500's in my case.
Ray--you are right---it must be either POR fumes or old age! No--thinking about it--probably a combination of the two! I must be real sad but I keep thinking about it. Worked on Brit bikes now for over 50 years and always until the last couple of years under the impression that the interference fit of the bearing outer in the crankcase provided a firm no movement fixing upon which the crankshaft orientation etc could be based. Now I guess struggling to get my head around the fact that everything---crankshaft, bearings, crankcases are all in a state of perpetual motion relative to each other. Having thought it through, though, still think it must be a good idea to firmly fix one of the main bearing outers to the crankcase. They didnt do it on G45s and Manxes for nothing I suspect. So how do we go about it on a Triumph twin? Excalibur has given us one idea (although that wouldnt stop in/out movement). Set screws are another. I think I remember the Manx design involves a "keeper plate" screwed to the crankcases. Any other ideas?
Also what does a broken bolt have to do with an interference fit bearing in a housing, I fail to see the comparison there
Controlling or preventing the separation of parts that radially expand at different rates requires some serious tensile strength.The Loctite needs to either pull the housing radially to a smaller size,or pull the bearing to the larger size of the housing.
This is not a small amount of tension.It wouldn't be much different to Loctiting a broken bolt back together and expecting it to hold tension.It wouldn't be much different to Loctiting a bolt head to the roof and expecting it to hold up a 1000 lb weight.
30 years ago it was thought that Loctite could be used to fix loose bearings in Vincents. It never worked, as the stresses appear to be beyond the strength limits of the Loctite.
I tried it on a B40 driveside bearing 25 odd years ago and it did not work, my guess is that the differential expansion between the steel bearing in an alloy case destroys the loctite grip over a number of heat cycles and the bearing spins again. The bearing was definitely solid in the cases when fitted but dropped out when the cases were split 1000 miles later.
Davy: The fact that you were successful using Loctite green, and you didn't experience the same problems as has been talked about in the other post, bearing locking up, is also anecdotal. While our friend, who's being locked up, could have got a bearing at the lower limint of pre-fit internal clearance, you on the other hand could have got a bearing that was at the high (loose) limit of pre-fit clearance. What I am saying is I'm my experience repairing Vincent crankcases with loose main bearings, and most are, that people who tried to repair them with Loctite products were not successful. I have done more than 25 cases over the years, as well as my mentors who have done a lot more, and we find that repairing the crankcase to renew the original interference fit by sleeving has proved, with time, to be a satisfactory repair. For the inner bearings I use a steel sleeve often seen in Matchless engines retained by three screws. I just sleeved a brand new Vincent crankcase, bought from a noted British supplier of such things, that found its way into the American box where the outside, timing side bearing had .050" clearance and the two inner bearings had less than .001". Now you would assume that the timing cover would be near parallel with the crankcase parting halves, but it wasn't, but that's another story.
That Loctite "Do it right user's guide" provides few answers, and indirectly confirms what I have experienced with Loctite products in the past 40 years. British engineering practice, when it comes to these old motorcycle engines, is contrary to they way most engineering captures the outer bearing race. Although not fully discussed in the text, most applications for Loctite bearing retaining products are steel, or iron housings that clamp the outer race. Any rocking action, or heavy load is better controlled than with a bearing just pressed into a crankcase. This is engineering is better suited to control costs (those 6 screws and two plates cost money) and issues related to weight and fuel economy, both of which were big factors when these bikes were being made. It was not better engineering by any stretch of the imagination, but served a real purpose at the time.
It is because the bearings interference fit, both on the shaft and in the crankcase, is an integral part of the engineering, it must be renewed as a part of a satisfactory repair. Loctite, whether it works (and in my experience it doesn't) or doesn't has nothing to do with repairing a bearing that is loose in one of these engines. If you are not addressing the bearing's fit running clearance you are not doing your job. And to do this you must address the bearings interference fit both in the housing and on the shaft, as well as the bearings pre-fit running clearance.
If you look at bearing books of the day you will only find varying pre-fit internal bearing clearances in bearings being made by British bearing industry. Originally they were identified as "one spot", "two spot", etc identifying the amount of pre-fit clearance, and if you walked into a US bearing supplier in the 1950's - 1960's and asked for a "one spot" bearing they would look at you as if you were nuts! It was a long time before engineers World wide adopted this system which became the CN, C2, C3, etc. system we are using today. Using Loctite, without considering the bearings fitted running clearance is not a repair, but a bodge.
As far as pinning a bearing to prevent rotation the Japanese adopted this practice on most of there horizontally split crankcases.
Now on a Triumph there are 4 points that I always use Loctite (low strength): The three flywheel bolts, rotor nut, main shaft clutch nut, and main shaft kick starter side nut. Just how I see it John
Either works for you or it doesn't and if your B40 bearing dropped out next time you split it I suggest there could be a simple engineering problem there. I know I'm not the first person to have considerable trouble removing drive side bearings in 500's after considerable distances and stresses and believe me as I raced them in the 60's I did a few.
Nortons and Vincents have absolutely nothing to do with the bearings in a Triumph, the cases are different and so are the stresses, I have no experience there so I won't discuss them here.
There is no way in the world I'm going to believe correctly applied Loctite expands strongly enough to crush a hardened steel bearing outer race and believe me I've done a lot of research into this very subject when a prototype product my business makes failed due cracking a thin walled plug with an M3 stainless screw down the center. We initially blamed the sealer but that wasn't the fault but it did cause me to start the fact finding about Loctite and it's specialised application. Yes it does expand, that's how it works and seals gaps in threads to lock them but with nowhere enough force to deform the metal. The only way your going to crush a bearing is if you overheat the case, freeze the bearing and apply an oversize amount of Loctite in the gap which forms a thick ring which then sets solid quickly due to the heat involved then crushes the bearing when the case contracts.
To successfully Loctite a bearing in a case it has to be done properly. First, if the case and bearing are at elevated running temperature or above when the bearing is fitted and Loctited and held at that temperature whilst the Loctite cures (not very long at that temperature) any expansion of all components involved (case, bearing, Loctite) has already occured and everything is normalised at running temperature and is taken out of the future equation, differential expansion rates accepted as they are fixed beyond our control. Second the surfaces need to be properly prepared texture wise not just wiped with some handy solvent, Third the correct priming procedure needs adopting, Fourth the correct grade of FRESH Loctite needs to be used (loctite in my experience doesn't store too well especially large bottles after they have been opened in humid climates. I had a retainer failure on a Ducati using old stock once which was successfully rectified with fresh Loctite. Fifth, I never freeze bearings, just heat the cases until the correct cavity size is reached to allow the bearing to be gently tapped into place due to expansion And Sixth, the whole assemble needs to be left in the turned off oven until it has naturally cooled to room temperature over a few hours.
Back in the 60's we had one particular race engine that repeatedly spun timing side bushes (and then starved the big ends of oil) After repeated attempts to prevent this by various means with the standard peg and copper plating the bush carrier o/d we finally used Loctite with complete success. I can only speak for my own personal anecdotal experience with these engines and Loctite but so far it has been a happy relationship.
Now that's how I do things and it works for me but if others chose to do it other ways and it doesn't work for them I'm OK with that. If it didn't work for someone else, that's their problem not mine and perhaps they can peg locate their main bearings or use some other method but the bearing will still be rattling around, just prevented from spinning by the peg once the case expands. That might not even matter in the grand scheme of things as these bikes have mostly stood the test of time just as they are.
Serious question here, how much do you think a T100 drive side main bearing expands on it's outer race outside diameter when heated to 240 degrees C? It's a lot more than most people think as if you cut the race and straightened it out you would have a piece of metal 8.5 inches long.
I just heated and measured one and it expanded 0.004" on the o/d
So the housing doesn't simply expand away from the bearing with temperature and leave the bearing for dead. With an interference fit initially there should still be some grip between the two components when they both rise to running temperature.
No-one is denying that the bearing expands. It actually should have expanded the diameter a little over 0.006",for a 200 C. rise in temperature,if expansion rates haven't changed lately.The circumference would increase by 0.019",or so.
The aluminium housing diameter would expand by twice that amount (0.012"),with the same temperature rise.
Davy I am not unfamiliar with the vagaries of little Triumph 500 having built my first one for the Daytona amateur race in 1959. The last we built to race was in the 1990's. Except for the flywheel bolts (obviously not needed with the Nourish crankshaft), engine sprocket nut, and both main shaft nuts, we never felt any need to use any of the Loctite products elsewhere on the engine. We have an engine that was robust when designed as 350cc and 15 hp. Push it to 500cc and find try to find 50hp and it is a different story.
What I have seen is a lot of people who have come to believe that this product works miracles. What I have seen is a lot of people who have caused more problems for themselves than were cured.
" That might not even matter in the grand scheme of things as these bikes have mostly stood the test of time just as they are." It is just that, time, and what time and stress does to these castings. We are starting to see with the 1970's bikes what we have seen with bikes made in the 1940's and 1950's.
People read these posts, and without the knowledge or the tools to take accurate measurements, out of hand apply Loctite.
The bearing's .0015" to .003" interference in the crankcase, and the inner race press fit on the crankshaft, and the inevitable decrease in internal running clearance is everything you need to know about how the bearing engineering in these engines works. Failing to understand this, and randomly applying Loctite think it is a cure is purely a bodge. When you are done you want a fit bearing internal clearance of .001". More and you risk bearing noise. Less and you risk ball, or roller, skating and bearing failure.
And if you think the little 500 engine's crankshaft doesn't flex (and even the Nourish) putting a rocking force on the bearings, well... IMHO it is better too fix it rather than patch t...