Description:

This problem our site is experiencing relates to a layshaft mounted between an electric motor and a right angled gearbox, which is thus connected to a conveyor drive pulley. The layshaft is supported by a bearing at each end and has mounted a brake disc in between. Couplings are overhung at each end, providing the connection between the motor and the gearbox. The arrangement is thus; motor; coupling; bearing; disc; bearing; coupling then gearbox.

The braking torque is provided by a Svendborg double acting 300 series failsafe hydraulic calliper. This calliper is non floating - meaning the air gap between each piston/pad needs to be measured and synchronised periodically. The brake disc is 650mm dia. and 35mm thick and balanced to 3g.

The bearings are a SKF 22220C double spherical roller type, one end being fixed and the other free to float in its plummer block housing. Taper locks are used to fix the inner race to the layshaft. Each outer race has a H7 transition fit in it's housing.

The couplings are a Centaflex CF-E-350-2-285 type. The lower mass halves of each coupling are mounted to the layshaft, and thus the higher mass halves are mounted to the motor and the gearbox.

The layshaft itself is 100mm diameter, with the couplings and disc being attached by the use of taperlocks.

Problem:

Around 20 of these layshafts have a vibration problem. Vibration analysis has been carried out and one common amplified signal exists (up to 15mm/s), this occurs at 4x the running speed (~25Hz), which is around 100Hz. The result is that the outer race of the bearings become loose and thus tend to spin in their housings. The fault frequencies of the bearings have all be looked at and none correspond to this 4x signal. Significant work has been carried out including, loctighting the outer race and even pinning the outer race and the result is worst; the amplification just goes up. The pedestals supporting the bearings have even been stiffened, changing their natural frequency away from the 25Hz and 100Hz, no advantage has been achieved in doing so. The alignment of the layshaft between the motor and the gearbox has also been verified and seems to be ok in horizontal, vertical and angularity.

Ideas at this stage:

The only things I have been able to come up witha are;
1. The non-floating brake calliper may be forcing the layshaft in an axial direction, causing the outer races to be shifted axially in their housings. This is feasible at the floated bearing but not the fixed bearing, could be a contributer though.
2. Simply the alignment is not sufficient. (Angularity is questionable).
3. The balancing of the brake disc mounted with the layshaft may not be sufficient.
4. Dont think the couplings have an affect, Nuprex types have been substituted with no change.

If anyone has any ideas on this please let me know.

Thanks,

Chris Muller

Is the frequency:
- exactly 4 * running speed - might point toward misalignment if coupling has 4 bolts per coupling or something similar.
- exactly 2*line frequency? (seems unlikely at this location , but I have to ask)
- Other non-syncronous?

May be it just runs close to structural/shaft resonance excited by 4x ?

Chris,

My guess would be a problem with the caliper being fixed. I would suggest checking axial phase between the bearings. I would also check for resonance in some component at 4 x running speed.

I also wonder why a brake is being used rather than a backstop. There are limitations on the number of loaded stops for a high speed backstop and speed limitations for a lowspeed and maybe you are beyond those. What are the speeds and torques involved?

Are the "taperlocks" actually Taperlocks (no hubs, using axially mounted set screws) or QD bushings (hubs with hex head bolts through the flange) and are they properly torqued? I assume no neverseize has been used or you would be breaking bushings.

Good Luck,

You probably know some amount of movement of outer ring in housing is normal. But if it is an amount that requires sleeving the housing, I would say it is a symptom of a heavy rotating load of some kind (rather than static load like machine weight).

You'll notice in the charts the housing fits get tighter (higher letters J K M P) for indeterminate or rotating load or shock load... you can't do that and still keep your floating bearing capability so you need to try to eliminate the cause of the rotating load, whatever it is.

On vertical motors where we have had spinning of lower rolling element bearing in the housing, our repair shop tells us it is caused by misalignment, assuming fit was correct (and if they did it last, of course it was correct ;-) ). I assume the misalignment could create the rotating load that causes the spinning.

The axial phase checks mentioned may be helpful in checking for misalignment as well.

By the way, how high are the 1x peaks?

I should mention I'm not at all familiar with the setup you mention... just talking in general terms.

Just an idle question. Have you done any bump tests to determine what the natural frequencies of the unit are? And a bump test of the shaft itself to see if it may have a resonance near 100 Hz?

Jon
Spintelligent Labs

I assume a bump test was done based on the original poster's discussion of changing natural frequency away from 25hz and 100hz but maybe he can clarify.

Along the same lines, it would still be interesting to know if the vibration is directional.

Have a look at the drive pulley vibration readings, sometimes you will see problems in the motor/coupling areas but these are not the source of the problem. Testing the drive roller bearings can identify balance/misalignment cracked shafts etc. Usually the levels are not as high because thay are mounted more securely and the vibration gets pushed back through the gearbox to the high speed area due to the size and stability. What ratio is the gearbox running at?