You did not provide enough information. But, you gave a few vague clues. I will attempt to guess the configuration. You describe a coupling between the compressors. I assume you have a single driver with two compressor cases. A centrifugal compressor will normally have the thrust bearing at the non-drive end. The compressor case will be pinned at the drive end and have guide keys that allow it to grow away from the drive coupling. The case grows away from the coupling. The shaft grows toward the coupling. The two somewhat cancel each other out. This works great for a single driver driving a single compressor.

If you add a second compressor case driven off the back end of the first one, you have a challenge. The thermal growth of the first compressor is all directed at that second coupling. Depending on the length of the first compressor and the temperature it runs, this can easily be several millimeters. To accommodate this movement you either have to pre-stretch the coupling, or use a gear or spline coupling that allows for that much movement.

Provide the configuration of your machine train with the locations of the thrust bearings and the case keys and pins along with the process temperatures if you want a more detailed explanation.

Johnny Pellin

Thanks Johnny. Our machine consists of a single driver (Steam turbine) which is used to drive three compressor barrels and there is a gear/spline coupling between each two barrels. The thrust bearings are located at the non-drive end of each barrel.
So as I understand from your post, the rotor of each barrel expands depending on the temperature along with the casing. This is why the couplings have large axial floats. But does this thermal expansion not affect the relatively smaller thrust bearing clearance of each barrel?

Thanks again!!

Thermal expansion of metal is in proportion to the length of the part. The distance between the active and inactive thrust bearings might be 25 mm. The growth in thickness of the thrust disk is relatively insignificant. The case growth I am referring to is between where the case is pinned on the drive end and the second coupling on the non-drive end. This could be two meters or more. Two meters of steel has the potential to grow much more than 25 mm. That second coupling has to accommodate the growth of that entire casing. We have machines where the casing grows by more than 12 mm.

Johnny Pellin

Got it!! thanks again Johnny