Split Bearing Cap Housing with Rolling Element Bearing

[cybanical]

I don't see this in practice much so I wanted to know: if anyone as experience with using a split bearing cap to make a housing for a rolling element bearing. In particular I'm looking to use a relatively narrow section (6903) deep groove ball bearing with a rotating shaft for oscillatory motion. The housing has some additional loads that would prefer it to be a single piece, so I essentially have a yoke supporting the two bearings (B below). One thought is to assemble the bearings onto the shaft, then drop this into the split bearing housing. The cap can then be tightened down to take up the internal clearance from the outer race.

I usually only have come across such a design in engines, and those are using journal bearings. I could guess that non-circular deformation could certainly cause problems for roller elements. The design lets me save some axial width. The alternative (A below) is that I effectively build shoulder screws to come in from other side to become the shaft. I would seat one side against the inner race, and preload the opposing one.

any thoughts, recommendations?

 

[tbuelna]

Should not be a problem if you match machine the housing carefully. Rough machine the housing and cap bores/shoulders with extra stock. Finish machine the housing and cap mating faces and fastener holes. Bolt the caps to the housing with the assembly torque that will be used. Then finish machine the bores and shoulder faces to the dimensions required. In order to obtain the best load distribution of your bearing elements under operating conditions, a FEM would be helpful to optimize the cap structure and machined bore profile.

With the mounting arrangement shown in the sketches, be careful not to apply excessive axial preload with the shaft screws.

 

[cybanical]

Yeah okay that seems pretty reasonable. Thanks for the comment.

 

[cybanical]

Btw, would you recommend another way to provide the preload?

 

[tbuelna]

Your sketch appears to show a pair of deep groove radial ball bearings, which may or may not need some axial preload, depending on the operating conditions. With your oscillatory inner race motion and radial loading, some axial preload would improve the load distribution among the ball complement. And if the shaft oscillatory motion produces high accelerations at the ball/race contacts, some axial preload might be needed to minimize skidding of the balls.

In any case, I would recommend modifying your mounting arrangement so that the inner races are clamped axially between shoulders on the shaft and bolts. This will minimize fretting between the shaft and bearing bore surfaces.