keyway cutting with edm

I am looking for a way to cut either a round hole into a bearing and differential carrier as a method of repairing it. The problem that I am having is finding a way to cut into the bearing as it is hardened steel. My idea is that I would install a bearing onto the diff. carrier and then take an edm and burn a hole into the bearing and carrier at the same time. Once the hole has been burned into the carrier and bearing I would then take a roll pin and drive it into the carrier to hold the bearing in place. My main concern in doing this is that I do not alter the temper of the bearing and I do believe that an edm would do this. Anyone with any ideas please let me know.

EDM would not ruin the temper of the bearing, at least not beyond a depth of a few ten thousandths of an inch (.0001")

Perhaps the better question is; why would you need to key a bearing? Should be a light press fit. Did the bearing lock up causing the shaft to spin in the bearing wearing the shaft? Might be better off welding the shaft to rebuild it and have it reground.

The biggest problem with pinning a bearing is you will weaken the race, and could cause it to deform. If the race deforms, it will creat a hot spot while running and quickly fail.

What we do where I work is rebuild differentials for cars and trucks. The main thing that we see that happens on the carriers that the ring gear bolts to is that the bearing on the driver's side spins. Some of these carriers cost anywhere from 100 to 300.00 each depending on what they fit and once they spin they are headed to the scrap pile. We have knurled them on the lathe and this holds for a while, but it is a temporary fix, so, what I am hoping to do is come up with a way to permanently fix them.

Todd,

I assume you're referring to the inner race of a tapered roller bearing?

Does a new inner race on a new carrier have interference? And if so, how much?

PM

If the forces on the bearing are great enough to cause it to spin a good interferance fit asuming that it is good then do you think that a small roll pin is going to last long.

As for using and edm. it has aready been said, it will work fine, I bored a hole though a large ball bearing to make a swivel with edm, the edge of the hole was still hard, if it is a tapered roller you could try 2 or 3 grooves in the face up the thick end of the bearing say 1/8" with 2 or 2 small dowel pins put into the shaft near the seating face shoulder to engage those grooves. you will need a dividing head or the like to get this precise all holes are blind. There should be enough meat on the thick end of a tapered roller to keep the effect of the grooves well away from bearing race. I have seen this done on one machine.

There is something odd going on bearings shouldn't spin.

The carrier has probably about a .002 press fit on it or something like that, we use a tool to drive them on, so I actually have never really checked the tolerances on the fit. It is a tapered roller bearing that goes on, and it has one on each side (one on the left and one on the right.) For whatever reason, the one on the left is the one that spins about 80 percent of the time. The carrier is cast iron, I am absolutely positive it's not ductile as it is fairly soft metal and is easily broken with a sledge hammer (after hitting it several times). If anyone has any more questions please let me know. I really wouldn't see where using edm would be a major problem. I do know that the inner bearing race and the rollers are hardened. Another thing that I would like to know is if anyone makes a square disintrode to go in the edm machines as I don't know if square keystock or a round roll pin would be the way to go.

All the diff carriers I have come across have been grey cast iron. EDM in itself is no problem it is ensuring that the material removed from the bearing doesn't weaken the bearing, there is not a lot of wall thickness in a bearing, section one and you will see combine that with 200K psi or so rolling over the races, it is like making a tunnel under a road carrying heavy trucks, do it wrong and it will fail. As for keys you will have to keep the corners rounded, stress raisers in hard materials is asking for trouble. If you want to erode square holes buy some square copper rod and use that.

.002" sounds OK, heavy force fit, I would do some random sampling of carriers just to check that that interferance is there and pay attention when installing in case there any loose fitting carriers. You don't use chinese bearings do you? they may have improved but they used to be such trash that anything was possible.

I know what you mean about the thickness of the bearing, it isn't a whole lot there that you can play with. I have been racking my brains trying to come up with a way to repair these carriers and so far this is the only thing that I can come up with. My dad has told me to find a way to do it and he says that he will buy the equipment to do it with. The problem is trying to cut a keyway into the bearing itself, the only way that I can think of possibly doing this is by edm as it aint happening with a keyseater, carbide cutter or even ceramic. I have thought of another way of doing it maybe, make a repair sleeve to fit the carrier and then put the bearing on. I do not think this would work, though as the fit that would be required for the sleeve to work would not be possible (the sleeve probably would not be over .040 or .050 max. thickness and it would probably distort when it was being pressed. If you have any more ideas let me know, this is a tough problem to figure out how to fix. The bottom line is this, the customer wants to save money and if we can do this and provide a real fix for this problem and charge about 50 bucks or so and warranty it, everyone would be happy, all about the mighty $$. Oh and by the way, the bearings we use are koyo or timken, we had several instances using chinese bearings and I am not going to take another chance to see if the quality has improved.

I don't think cutting a keyway in a bearing is a good idea. Eventually the bearing race will fracture thru the key.

With the proper interference fit a bearing race will not spin.

Instead of driving on the bearing it should be heated with a 'bearing heater' no hotter than 250-275f and slid onto the shaft or carrier. Then on cooling it will grip the shaft.

Can you take a picture of the bearing and carrier and post it?

I will post a picture of the bearing and carrier and post it, it will be on Monday as I played hookie today

My gut feeling is that if circumstances are causing an inner race with .002" interference to come "loose" somehow now... it will do the same thing when there is a key or a rollpin in it.

If the inner race gets loose and there is a locating pin in there, that pin will be quickly "devoured". (IMO)

Something doesn't make sense... that bearing with the proper interference on that carrier should not be coming loose.

PM

These carriers do have a lot of stress on them as they are in the housings that drive an entire automobile. Believe it or not the ones that spin the bearings the most are chevrolets and toyotas. I have been thinking about how to repair them, but have yet to come up with an idea. To be honest, I do not actually know how much of a press fit that they have. I haven't ever measured the id of the bearing and the hub on the carrier. The .002 was a guess; the reason I have never measured the tolerance is that when you drive a bearing on you can tell if the carrier is good or not by how hard you have to drive it on, and also you can look at the radius where the bearing seats on and tell if it has spun. If you have an idea on how to repair these things please let me know. I know it probably sounds crazy, but it's one of those things. Also, I have been trying to figure a way to knurl the caps in some of these housings and had no success on this either. The problem with trying to do this is that you have to go through an axle shaft tube that is about 28 to 32 inches long. I know that this has been done as I have seen some housings that had been rebuilt by Jasper Rebuilders and the caps had been knurled.

Get the makers specs for the carriers which are giving problems, check the bearing seating surface against those specs. Check for any service bulletins regarding these makes. Fixing the symptom of a problem is only managing a problem rather than mitigating it. Unless you start measuring there is too much assumption, I have been there and sometimes what you measure is not you expect.

A few numbers I will not go into the the mathematics, simple but long. The shear strength of the bearing shaft interface of a 50mm taper roller (30210) with .001"/" interferance and a 5/16" wall grey iron shaft
is roughly 2700 Lb a round dowel pin made of high tensile steel (100Ksi UTS) would need to be 1/4" in diameter to have the same strength as the interference fit and I was being conservative with the interference fit parameters, I bet it takes more than 2700Lb to press on a bearing.

If you still want to pin the bearing then the method I described in the first post is a good option as it has rounded grooves well way from the bearing race and by having several of them it has good shear strength.

As for knurling the caps you will have to make or buy or have made an expanding knurling head. I havn't seen one but I can guess the constuction, 3 rollers in a round carrier with a long removable drive shaft which is supported at it's far end by a bearing which locates in the axle bearing bore.

Here is a question; why are they failing? Is there a higher load on the failing side? Lack of lubrication?

I doubt any quick fix will prevent it from happening in the future. If the bearing race is spinning on the carrier, it is because the bearing failed. No amount of pinning, gluing, keying, welding, or otherwise retaining the race will solve the real problem. If it fails with a pinned race, you're just going to transfer the problem to another area and cause even more damage. Worst case is a siezed bearing that can't spin on the carrier can lock an axle, causing the customer to loose control of the vehicle and crash.

I'm going to suggest that you post this over in the General Forum. I agree with the guys here that modifying the bearing and carrier could be dangerous. And will also mention that (IMO) changing the mfgr's specifications could open you up for liability issues if something were to happen.
There are a lot of automotive types there that might give you some more ideas.

I would simpily sleeve it, and hold the bearing on with green locktite, the light green kind you can see thru. I've put races into Harley cases that were almost loose, and locktited them in, and never heard any more about it. I'd think these races get quite a beating. They're Timken bearing races, by the way. It ouldn't matter if the sleeve was only .050 this is fine. Press and locktite the sleeve on, then machine it to the final size, and press and locktite your bearing on, and you're good to go. I've sleeved the bores on a lot of Roadranger truck transmissions with no problems.

"Dutch pinning" is fine for some things but, this isnt one of them. If the bearing isnt snug on the carrier and you pin it,every time the pin is at the 6 oclock position as the carrier rotates the full load goes on the pin,and through the race in a very narrow area. This cycling and flexing will crack the race quick. Even if there was some press fit left,half a pin hole is a weak spot in a hard,brittle ring that is in tension already. Many have tried all kinds of ways around this type of problem for over a hundred years or more. Without seeing it,sleeving the o.d. of the carrier would be worth looking at,but may not be economical. The race needs to be snug and truly round to survive. You cant go around this fact.