Thrust Stand Design

[Slagathor]

I am hoping for some experienced feedback on a thrust stand design for a vertical pump application

Axial Load = 22,000# steady state down (there is a change of 0-5 second upthrust, up to 3,000# during start up)
RPM = 510
Radial Loads at each shaft end = minimal. Mainly due to misalignment and drive shaft vibration. No steady state radial loads
Duty cycle = infrequent. 8-12 hours every few months.
Lube = Grease
Cooling - If required will be via an integral fan that pulls air in, routes across bottom of upper bearing housing, and expels out of the frame
The large shaft is a function of the high torque being transmitted - 8,000+ ft-*lb

Design requires spherical roller (SKF 29326E 130mm) to get the required life
Deep groove bearings (SKF 6026 130mm) to take radial load top and bottom.
If shaft grows due to heat, the lower bearing can move down in the housing

This might be overkill, but here is the preliminary design:

Here is the upper bearing detail. The inner races are locked together, with a spacer, and an SKF nut, locking them to the shaft. The housings are machined so there is a .000-.002" gap above the deep groove to the cap. If there is upthrust, the upper bearing will push against the top cap. While not shown on these details, there will be preload die springs in pockets under the stationary race of the thrust bearing. If it upthrusts, the springs will keep the required load on the thrust bearing (about 4000#). The radial fit on the outer race of the thrust bearing will be loose enough to allow the pre-load springs to work. The radial fit on the upper bearing housing to radial bearing OD will be tight enough to give good radial support, but it can move if upthrust occurs.

Here is the lower bearing detail. The fit to the housing will again allow the bearing to float axially, so it carries only radial load.

This is sort of a cross between the bearing configuration on a vertical motor, and a large bearing frame end suction pump.

Am I on the right track?

 

[JJPellin]

I don't see any need for the radial bearing in the upper location. The roller bearing can take the radial load. The radial bearing is in the housing cover which will not be perfectly aligned with the rest of the housing. I expect that radial bearing to have a very short life.

We housings that are similar to this in some ways. Our coker drill stem gearboxes have a normal downward thrust load of about 150,000 lbs. They use a single spherical roller bearing on one end. I have included a drawing of that housing.


Johnny Pellin

 

[Slagathor]

Thanks for your perspective JJ. It sounds like you may deal with some much tougher applications that this!

The reason for having that radial bearing at the top is there is a potential for a temporary upthrust, of about 10-20% of the normal steady state downthrust. The fit on that bearing to the housing will be H6...so pretty loose. A few thousandths gap above that radial bearing will allow the whole shaft to move up a till that bearing takes axial load for a few seconds. Then it transition to downthrust. This arrangement is used on large end suction sewage pumps, and I have never seen that upper guide bearing have issues.

I could use the lower guide bearing to handle that upthrust, but that will require careful control of the dimensions of the whole assembly so that when upthrust occurs, the shaft does not move too much before the lower bearing can take move against the cover and carry axial load. That may be a better, and lower cost way to go as we can eliminate that upper bearing.

Thanks.

 

[tbuelna]

Slagathor,

The arrangement shown in your OP appears to be over-constrained. The arrangement shown in JJPellin's post is better, but I would suggest a couple changes to it.

- First, I would change the lower bearing from a two row spherical roller bearing to a single row spherical roller bearing. A single row spherical roller bearing axially preloaded by a spring, will maintain better load distribution within the roller complement under the conditions described than a two row spherical roller bearing will.
- Second, I would re-locate the axial preload spring(s) to the lower bearing outer race. Machine the lower bearing housing bore for a sliding fit, and machine the upper bearing housing bore for an interference fit. Eliminating sliding/rotation of the highly loaded upper bearing outer race in its housing bore will minimize the potential for fretting damage.
- Third, I would add a seal or shield just below the upper bearing (assuming there isn't one). This will keep the grease where it's needed.

One last thing you should consider with your housing design is minimizing tolerance stack-up. For example, compare your housing design to the one shown in JJPellin's post. In the latter example, all the bearing bores are located in a single piece of the housing, which makes it easier to maintain precise alignment.

Hope that helps.

 

[geesamand]

I had good success putting the thrust bearing alone at one end to handle radial and all thrust load. On the other end of the stand I used a single taper roller bearing. We set the end play as if it were a pair of separated taper roller bearing. Elegant and effective.