For your preladed set up a dial indicator on the end of the shaft, and add enough shims till you get some freeplay. Then add the amount of free play to the amount of preload you want, and remove that amount of shims. If I remember right from my time at John Deere most of the time it was around .oo8-.0010. It's better to have less than too much.
It depends on what you mean by "calculate". If you mean calculate what the clearance will be, given the component dimensions (including spacers etc), that's one thing. If you mean calculate what the clearance or preload should be, that's another.
If you have a high speed application, in many cases it is impossible to run a simple taper roller with a fixed preload - the bearing will burn up. In general, to actually calculate the minimum acceptable clearance or how much preload you can tolerate from first principles is extremely difficult - it requires a coupled transient thermal- mechanical FEA analysis. So you are usually left with rules of thumb and manufacturer's recommendations, tempered by your own and others experience. Poplawski Associates have recently developed a computer program which runs in conjunction with Ansys, to enable detailed calculations to be carried out, but it's a complicated exercise.
That's fine theothercarl, but what exactly i needed to know was the theoritically how can we decide about how much amount of pre-load to be set up i.e like yo replied pre-load to be 0.008 - 0.010. How we arrive at this pre-load?
Actually, unless it's a very large bearing or the speed is very low, I would be surprised if you could use as much as .008" - .010" preload, but it all depends on the stiffness of the supporting structure. In some cases, the structure is flexible enough that "thermal runaway" will not occur with high preloads. In other cases, even if you start out with clearance, you can run into trouble. The material of the housing (steel or aluminum for example) also makes a difference because of the thermal properties.
There are several approaches that come to mind for finding the ideal clearance :
1. Ask the bearing manufacturer (probably Timken). Sometimes they will run a detailed computer analysis for you - it depends how important you are to them as a customer as to how much detail and time they will spend, and also what your relationship is with the local representative, and how motivated he happens to be. If they just give you a general recommendation, it is unlikely to be optimum, and they will probably suggest you try it and revise if necessary using approach 2 below in any case.
2. Trial and error. Start off with a small clearance, do heat checks, and try reducing and increasing the clearance, performing more heat checks etc. If you are using grease lube, you have to be careful to let the grease "channel" before taking the temperature readings seriously.
3. Use a program called "Shaberth", if you can find it. This is an old mainframe era program for thermally optimizing bearing preload and is very hard to use by modern standards, although there is a DOS version.
4. Use Poplawski's new Cobra-AHS program (if it's ready) in conjunction with ANSYS. He will also consult in tricky cases. See
I agree fully with English Muffin. You might want
to start out with .0002 preload and run the bearing
and keep increasing the preload until you start
experiencing too much heat generation. A lot of the
theoretical calculation considers speed as well
as the loads. Too much preload ends up in overloading
the bearing and an early failure. Whoever designed
the bearing application must have had some spec to
work with. More is not always better.
