This was originally discussed here:
In that thread John Regan showed pictures of an axle which had been ruined by the cage of a solid roller bearing rubbing the axle. Below is a copy of one of the pictures he posted. The end plates on the cage have been wallowed out by the pins on the end of the rollers which allowed the plate to cut a groove in the axle.
I have been scratching my head for a couple months, trying to think of a way this could happen. In another thread there was mention of the importance of not having tapered bearings, shafts, or sleeves. I am wondering if this could have caused the failure that John showed.
If there is a taper in any of the parts, the rollers will not be running true. For instance, if the rollers are tapered, the larger end will be trying to catch up with the smaller end or the smaller end will lag behind the larger end. This will cause the rollers to move out of alignment and wallow out the holes in the end plates, causing in turn the plates to rub the axle.
John posted that the rollers were .500 diameter. If the rollers are tight between the shaft and sleeve the taper effect could be exaggerated. It may not take much of a taper to cause this problem.
What do you think? Does this sound reasonable?
As I commented last time, I wonder if the housings were straight. Last year I made some fixtures for checking rear axle housing (both standard and Ruckstell) and found a significant number were substantially bent. Not too hard to straighten with just a oxy-acetylene torch. Just a thought
Another point is the original twisted spring steel Hyatt Bearings were of a reverse alternating pattern. This design took up expansion, as well as re dispensing the grease and heavy gear lube as needed. Take a good look at any quality needle bearing with a load application and you will see it is never in a retaining configuration with a pinned cage.Full float yes, retention float no. Those copies were a nice attempt in the replacement of the original, but the damage they cause is always the same. Don't reinvent the wheel, quality will with stand the test of time.
All the Best,
Hank in Tin-A-See
The axles were ruined because of the groove cut into them by these bearings. I rebuilt the rear end with NOS original Hyatt bearings and new hardened sleeves and new axles. The rear end went together very tight but was smooth turning by hand via ujoint being turned with nothing tight or binding. The car went back on the ground and has been on several long tours and is very quiet. The engineering on these solid rollers is so substantially different than the original setup as to need separate analysis as to whether they will even work. Replacing 8 roller Hyatt with 6 to me is a substantial difference. The axle clearance hole at the ends is substantially smaller than the hyatt version and it would take very little wear of the cage end holes and the roller ends before the cage would be riding on the axle and the sawing operation begins. I don't have much interest in pursuing them as a viable approach to replacing the hyatt bearing design but others may want to do that. I simply posted what I found. The solid rollers show no sign of wear other than at the ends and the axles are also not worn at the place where the bearings ride at both inner and outer locations. If not for the groove cut into them they would be perfectly usable measuring 1.062 all across the bearing surface. The keyways are straight and true at the ends. They were very good axles and are not bent.
Here's a solid roller bearing (left) with seven rollers, not six. I have no intention of ever putting it into a car. I've been told by one Model T guy who should know that these are not as bad as their reputation, but I see no reason to risk it when plenty of usable originals are available.
When I study the pictures, it seems like the axle was only contacting the bearing rollers at each end of the rollers. Which would be logical if the housing was bent.
This problem is why I had my fixtures made for checking rear axle housings for straightness.
If good Hyatt bearings were readily available, then this condition would only be of academic interest, but they are becoming difficult to find and this is only likely to get worse.
To the extent that I only use Hyatt roller bearings on the inboard end and use ball bearing hubs on the outboard end to preserve the supply as best I can.
I am no great fan of the solid rollers, but with straight housings they seem to be okay. So my second choice for sure