load sharing on multiple pinions 1

[klarskov]

thread406-196324

The above mentioned thread describes a problem with load sharing on multiple pinions. The distribution of loads on each of six pinions can be unfavorable when the system is locked.

I'm now faced with a similar problem, anyone got any suggestions or references to literature that describes this problem in details, since at the moment I'm not 100% sure why the load is some times different on the pinions when the system starts rotating.

Thanks for any comments,
Klarskov

 

[Windward]

Is your application in a multi-branch parallel shaft gearbox, or like the one described in the reference?

 

[klarskov]

This is a machine placed on top of a large gear and using 6 separate motor/gear-box/pinion configurations to rotate the machine.

From measurements of torque on each pinion, it can be seen that sometimes the load is distributed very different. Usually you see a peak in torque to start the rotation in all pinions, but sometimes 1 or 2 ot the pinions are missing this peak, could this me related to different meshing of the gears?

 

[tbuelna]

klarskov,

The situation in the linked thread described good load sharing when the motors/pinions drove the ring. It was only when the ring tried to backdrive the motor/pinions that there was apparently a load sharing issue. I would assume that the motors incorporated some sort of brake that prevented them being back driven.

If the pinions have a high pressure angle and/or a very small number of teeth the mesh contact ratio may be very small (ie. very close to 1.0). Combine that with large backlash, high run out and loose index tolerancing and it seems quite possible that the pinions would not all be equally loaded when braked and back driven by the ring gear.

When the motors/pinions are driving the ring they should all load fairly equally unless there is some ring gear geometry condition that causes excess friction. However, such a problem in the ring would produce a cyclic effect as the ring rotated, and would be fairly easy to identify.

AGMA 6123-B06 sec. 9 gives a good explanation of load sharing in epicyclic gear sets, which is somewhat relevant to your situation.

Hope that helps.
Terry

 

[Windward]

It appears that the loads equalize after the system is running because the six motors eventually converge to equal power output.

If you can start the motors slower, maybe the unbalance between the gearmotors at startup would be reduced to an acceptable level.

 

[klarskov]

Thanks guys,

I will be looking into these suggested explanations. I thing there might be something about the mesh contact ratio, since it is low in this application.

Klarskov