This isn’t quite automotive, but was the closest I could find for a John Deere lawn tractor.
The front tires spin on the axles/spindles on bushings that are pressed into the inside and outside of the tire rim. The rims are held in place by an e-ring at the outer edge of the spindle. I just discovered that the e-ring/washer/dust cover was missing from the right front wheel after the last mowing activity. Strangely enough, the tire never came off!? I think that’s because the outer end of the axle is slightly lower than the inner end, thus keeping the tire on the axle.
As I was looking at the wheel with the missing e-ring, I discovered that it had lots of play. It felt (and looked) like 100-150 mils. After some googling I found that the bushings can be replaced by sealed bearings, which I intend to do. My concern is whether the spindle is too worn and needs to be replaced. Here are the measurements that I took on the spindle in question.
Center: 0.749" (this is unworn, as no bushing is in this area)
Inner: 0.745" (roughly scored)
Outer: 0.743" (lightly scored)
The bearing to replace the bushing has an ID of 0.751".
Since I’m a sparky, I don’t have a good sense for spindle-bearing fit. From the above center spindle measurement of 0.749" and the bearing ID of 0.751" yields a clearance of 2 mils. Is that a nominal clearance for a bearing-spindle fit? What’s a good range for this kind of fit?
The calculated spindle-bearing gaps are 9 mils (outer) and 7 mils (inner). This seems excessive so I’ll probably just replace the spindle unless the price is excessive.
There is a copy of the machinery handbook in the machine shop. There is an entire chapter in that book that has tables listing criteria for fits. I would start with that.
Update: I was writing my reply when wandrson posted his answer. I agree, check out what the recommended clearances are.
Curious: How long have the sleeve bushing lasted? Was there any “wobble” noticeable? If the bearing is replaced, would the expected life exceed the rest of the machine, primarily the engine, blade bearings, etc.? The new sleeve bushing should tighten things up a bit since they won’t have any wear increasing the ID. I would also assume that the roller type bearing would have a smaller ID so it fits the spindle tighter if not snugly. What is the wear on the other side like?
Normally, the races on bearings should fit fairly snug so they don’t rotate on wheel or shaft, the bearings races should be engaged so the rolling elements roll over the races. Typically the outer races are pressed in and the inner are held tight by the spindle nut pressing on the outside of the inner race. I’m assuming that the replacement bearing you’re looking at installing is just a spherical/ball element (I can’t imagine the hubs being designed to take tapered rollers like found on automotive applications)
If the tractor is very valuable and expected to last a long time, replace the spindles and install rolling bearings. Without addressing the spindle you’ll have a gap either way. Sort of a cost/benefit or perfectionist decision. But upgraded bearings on a worn shaft isn’t addressing the clearance fit issue.
A roller bearing should greatly reduce further spindle wear. I’d probably go with roller, they aren’t expensive. I know what I’d do to lock it in place given to rotational speeds … but will decline to state. (It doesn’t involve duct tape).
My practical opinion is that 2 thou is a bit snug, and 10 thou is a bit loose for a high speed spindle application. But for a lawn tractor, you are probably OK for quite a while with new bearings or bushings.
This is very low rotation speeds, it’s a tire on riding lawn mower. If he were mowing at 5~6 mph (jogging speed) that would be approx. 7.3~8.8 ft/sec., a tire 10" in diameter would be turning 850-1,050 rpm. If there was no noticeable wobble/shimmy before then either solution should work without need to replace spindle, IMHO. If spindle is reasonable, replace it.
wanderson - thanks for the machinery handbook reference. I did try a google search before posting my question but my choice of search words didn’t yield any results that were meaningful. What I was looking for was a rule of thumb (similar to kbraby’s reply) for diameter differences in a ‘bearing on a shaft’ arrangement, realizing that general recommendations are usually sub-optimal for any specific application
I realize that’s like asking “What’s the recommended tolerance of a resistor in an op amp circuit?”. I would say that, in general, 5% would be OK in a non-critical (low wheel speed equivalent) while 1% (high wheel speed equivalent) should be OK. Proper disclaimers apply, such as depends upon your application, check available references, apply worst case analysis,…
I suspect too many people have been burned with obsolete versions of books about software, networking, and programming languages, and forget that not all fields are still developing as rapidly. I can think of a few IT books that are more stable than the majority but probably the one that should be most stable is “Numerical Recipies”. And even it has gone though 3 revisions aimed at target audiences with goals that evolved over the years.
And much of the IT knowledge has been rewritten many times since on line publishing became common. But other more mature fields don’t redevelop their reference material anything near as often as IT, and have not made the same transition.
To recap, pretty much anything O’Reilly press offers I want as current as possible. Other fields, I’m quite happy with a 25 year old copy of “Machineries Handbook” or a 25 to 50 year old copy of the CRC Handbook.
In fact, about 20 years ago, I went hunting in used book stores to find one of the older editions of the CRC handbook that were still in the small format printed on the really fine paper.
I would argue that the Knuth “Art of Computer Programming” books are the most stable and of the longest term use (I think they will still be valid in a hundred years).
I know I have owned my copies of volumes 1-3 for 4+ decades.
That is the fundamental problem with electronic books. You have to know not only what to look for but how to specify the search. A manual next to a machine is the best and most intuitive user interface.
Yep, it is and always has been a pricey set of books, just well worth it for anyone in the field.
That said, I really don’t care for the Numerical Methods book. I haven’t looked at it in years, but as I recall a number of the ‘recipes’ it provided has some subtle but important flaws that could bite the unwary.
