bearing cone fit influence on axial play 1

[321GO]

Hi Guys,

This is a taper roller bearing with a preset spacer for a truck wheel end.

Does any of you have experience with how much the bearing cone fit influences the axial play in such setup? I have the clearance of the bearing from the bearing supplier, but this is a 'non-fitted', i.e. 'on-desk' dimension.

I suppose the fit of the cone has a fair amound of influence on the axial play after assembly, but to what degree??

The cone has a R7 fit, which is -0,050/-0,090 in mm, so I quess a fair amound of 'oemph' is needed on assembly.


Again, any input is more than welcome!

 

[Tmoose]

Timken used to provide a worksheet and maybe a program for calculating spacer dimensions to achieve a particular preload. Inputs were actual shaft and housing dimensions, bearing ID/OD dimensions, actual "length" of assembled bearing cup-n-cones, etc.

 

[rmw]

I am not sure what you are asking. What I see here is .002" to .004" end play if I understand what you have presented.

You can fit the bearing to those dimensions and then mount a dial indicator on the axle housing and indicate the hub rim. You have to be meticulous when you take your reading in that you have to push and roll the hub one direction to "walk" the cone up the incline and then pull it back toward the indicator and roll it again to get a good reading. You have to pick a clean and machined surface on the hub and watch the arc of your swing as you 'roll' the hub back and forth.

It takes a little practice before you can 'roll' all the variables out, but I have done this to .0005" measurements. I used to deal with a European Mfgr'd axle that had this set up and their recommendation was about what you stated. I found that we got better life (this hub had a planetary in it subject to lots of stresses when the bearings were too loose) out of the assembly when the bearing was set to about a half a thou.

In that it is measured and set and held by the spacer and shims, you can set to that close of a clearance with confidence. Other methods don't let you have that luxury.

rmw

 

[321GO]

@rmw,

thanks for your reply, but what i meant was something different. First of, the R7 Fit was for the cup in the wheel hub, i.e. the press-fit of the cones.

These are 2 taper roler bearings for which the axial play is determined by a spacer between the two cones(and the shoulder dimension in the hub itself), so basicaly it is set during fabrication and cannot be 'set' in any sort of way during assembly.

With the ´shoulder dim.´ i mean the axial distance between the two cones.

Now, i´m in the process of determining this axial distance between the 2 cone for correct axial bearing clearance after assembly(0-0.1mm).

I have this 'shoulder dim' from our bearing supplier, but I'm not sure how much this dim. is effected by the actual assembly due to the press fit of the cones.

I suppose the press fit of the cones has a destinct effect on the axial play.

p.s. i know this is a fage question, but i highly appreciate any input regardless

 

[tbuelna]

321GO,

As you noted, an r7 fit on your cones would give substantial interference. The effect of such an interference on your cone race would probably be a change in diameter. The magnitude of this change would be a function of the relative stiffness between your cone race and hub shaft. Usually, the race is much less stiff than the shaft, especially with a large diameter race and solid shaft cross section. So the installed race will likely end up being slightly larger in diameter.

While race dimensional changes due to fits must be carefully evaluated with bearing types that are sensitive to radial clearance (ie. radial ball or cylindrical roller), tapered rollers can compensate for this effect at assembly by adjusting axial setting.

If you know the amount of interference and have dimensions of the race and shaft cross sections, you should be able to calculate the change in diameter of the installed race. Once you have that, you will need the cone angle of the race surface itself, which is based on the contact angle and taper angle of the rollers. Based on the race surface cone angle, you can calculate an axial shift value for a given change in diameter.

Hope that was helpful.
Terry

 

[321GO]

Thanks Terry,

yes, I will try to do the hand calc's. and look for some correlation with Cosmos.

 

[321GO]

I did the following, please check the attachments.

I assumed all deflection to be in the cup(housing not effected).

The trigonometry calculation approaches the Timken formula well, but 0.1mm axial displacemt seems a lot to me though.

And since this is a 'back to back' taper wheelbearing setup, the total axial displacement would even be 0.2mm

p.s. the bearing cone are preset with a bushing(for correct axial clearance), so it cannot be set afterwards


What do you guys think, are these axial displacement value's for real?

 

[321GO]

here's the trigonometry calculation.

 

[321GO]

Posting my findings for future reference.


I did a FEA on the cone with a 0,058mm initial interference.

The cone contraction is roughly 60% of this initial interference.

In this particular instance it resulted in 0,017mm radial contraction on the race incline with resulted in roughly 0,06mm axial(horizontal) displacement, i.e. 0,06mm decrease in axial clearance after fitting.

This seems to correlate to the Timken formula for thin walled/hollow-shaft hubs(not the formula posted by me before, that one is for heavy cross sectioned hubs).

 

[tbuelna]

321GO,

It's good to see that you are evaluating your bearing fits using different methods, and then checking them against each other. Most people don't fully appreciate the effects of race fit on bearing performance.

In the FEM displacement result you posted, you can see that there is a small change in the cup race surface taper angle (due to the difference in hoop stiffness along the race length). While the installed race surface deflection in your example is small, if it was greater it would result in edge loading of the rollers. Edge loading of rollers is especially problematic with rollers that have a high L/D aspect ratio, like most tapered roller bearings. These types of rollers usually must have a fair amount of crown to compensate for this.

Thanks for posting your work.
Terry