Large motor endplay requirement 1

[electricpete]

I am looking at a particular large vertical open (WP2) motor with spherical roller thrust bearing on top and deep groove bearing on bottom. Distance between top and bottom bearing is approx 50"

The upper bearing takes downthrust.
The lower bearing takes upthrust.

OEM specification for endplay is not available.
I believe EASA specifies 0.005" to 0.010" endplay for vertical motors.
Do you think that 0.005" to 0.010" endplay would be ok for this motor? (assuming driven equipment has no requirement).

The endplay is of course set and checked at room temperature.
If the rotor expands more than the stator, the endplay will increase.
If the stator expands more than the rotor, the endplay will decrease

Assuming both stator and rotor have temperature coefficient of steel.
Assuming endplay set at 0.005" at room temperature
How much more would stator have to increase than rotor to lose endplay (and start loading the lower bearing):
deltaL = L * deltaT * alpha
deltaT = deltaL/(L*alpha) = 0.005"(50"*0.000007/F) = 14F
That doesn't seem like much.

It seems like typical assumption is that the rotor willl heat more than the stator, but I'm not sure if that's true under all transient conditions. What do you think?

=====================================
(2B)+(2B)' ?

 

[Motorwinder]

We were rebuilding 1000HP and 1200HP GE verticals with 29434 spherical top bearings. There was much discussion in the shop, and with the customers as to what to set the bearing end play. We called GE, and was told to set them at .012-.015" Never had any problems with that setting. We tried one at .008" once, and it was too tight.

 

[electricpete]

We tried one at .008" once, and it was too tight.

Can you explain what you saw to lead you to believe it was too tight?
Lower bearing heated up? During what machine configuration? (uncoupled, coupled, loaded?)
Just curious.

=====================================
(2B)+(2B)' ?

 

[Motorwinder]

Just tight to the feel. Hard to turn by hand. Uncoupled, during assembly. We test them unloaded.

BTW, the lower bearing isn't suppose to take up-thrust. It's basically just a standard deep groove/guide bearing.(6220 C3) The setting of the top bearing is to control up-thrust by raising the rotor till the lower bearing preloads against the bearing cap, then lowering it to the predetermine setting. I suppose the tightness comes when the lower bearing starts to preload.

 

[electricpete]

Just tight to the feel. Hard to turn by hand. Uncoupled, during assembly. We test them unloaded.

Thanks. I'm wondering how low endplay (but still greater than zero) could cause difficulty turning, resolved by increasing endplay (assuming it's not contact of rotating/stationary parts)....

I think with the smaller endplay, the upper bearing springs must have lifted the rotor and loaded the lower bearing excessively. When you increased the endplay, you allowed the springs to expand a little which decreased that load. Would you agree?

BTW, the lower bearing isn't suppose to take up-thrust.

I agree. When I said the lower bearing takes upthrust, I meant the lower bearing takes any upthrust that is present. We don't expect continuous upthrust from the pump.


=====================================
(2B)+(2B)' ?

 

[Motorwinder]

Would you agree?

Makes since.

I've been pondering this question for a long time. GE always said .003-.005 for their angular contact bearings.(ball bearings). US motors set theirs .008-.010. They say that's for spherical bearings too. Every motor mechanic has his own opinions from 0-.020. That's why I called GE, and talked to their engineers. Who said .012-.015 for their spherical. I also got this in writing to show the customer.

It might have been okay to leave it tight at .008, but I felt better with it spinning free at .012-.015. Never had one come back at that setting.

I'm not an engineer, but have been in this trade for 40 years.

 

[dArsonval]

A, "tight to the feel. Hard to turn by hand..." situation can mean other factors not associated with
bearing end-play at all.

1) Is the shaft bent?
2) Are the Carrier/ Bearing Fit(s) and their relation to the shaft "Fit" correct?
3) Is there paint, rust, or some other debris obstructing the rotating movement of the shaft?
4) Are the register fits on the stator parallel and concentric with the end brackets?

What is the Total Indicated Run-Out of all the mating components?

If something is "binding", something is either not fully "seated", or out of tolerance.

If the apparatus is in pristine condition, the shaft should turn freely no matter what end-play is set.

Always enjoying the forum,

John

 

[electricpete]

Thanks... good comments. I agree there can be other possible causes of tight / difficult to turn symptom.

I'll say the purpose of this thread is to help me understand the considerations that go into selection of endplay (let's say by the OEM for argument's sake.... there are situations where it is very important for users to understand the basis for endplay selection and effects of changes). I am interested most in one particular motor, which happens to be the one described above... didn't mention it's a Reliance.

The discussion of my initial post viewed endplay in a simplistic way without thinking about how it can interact with the springs. I think I have learned a little since then by asking and thinking. I came up with attached powerpoint to analyse this further.

Slides 2 and slide 3 show features of our motor already described.

Slide 1 is my attempt to qualitatively analyse the relationship between endplay and load carried by the upper and lower bearings under various conditions of downthrust.

Continuing with slide 1...
Top left stick figure shows stator in black, rotor in red. Rotor moves up and down, and its position is described by h. h varies depending on total applied force Fapplied shown at bottom. The reaction forces on rotor from upper bearing and lower bearing are also shown with arrows (upper bearing force in green and lower bearing force in red). These same colors green/red will be used for plotting these forces in the plots below.

The graph on lower left hand corner shows both bearing forces (upper bearing force in green, lower bearing force in red) as function of rotor position h. (h will take on whatever value is required to make the three forces sum to zero).

The graph on the lower right eliminates h as independent variable, and makes Fapp the independent variable instead. We now see how the forces (upper in green, lower in red) change with applied force. Rotor weight W is sketched in, and represents the value of applied force which would exist when uncoupled. Position of W shown would be consistent with a rotor where the springs are strong enough to lift the rotor to the top limit of endplay h=EP (otherwise W should be shown to the right of the change in slope of the curves occuring at Fapp = K*[Lsf-Lsmin- EP]).

The purple lines are intended to show which where the two bearing load curves would shift if we decrease endplay for a given motor. We can see that decreasing endplay increases the upper and lower bearing loads for all conditions except where the applied load bottoms out the upper bearing (h=0 <=> Fapplied >= K*(Lsf-Lsmin- EP)). Additionally , the load required to bottom out those springs shifts higher with lower endplay.

So, we can see for uncoupled conditions (under assumption upper bearing is not bottomed out, i.e. h>0, i.e. W < K*(Lsf-Lsmin- EP)), increasing the endplay would tend to lower the upper bearing load... which could be a bad thing since the spherical roller bearing depends on minimum axial load in order to maintain its radial stiffness. If endplay were increased too much, it would tend to orbit more during uncoupled run. If endplay were decreased too much, we are also loading up the lower bearing more during uncoupled run which can be a problem. And possibly with very low endplay setting we may even load the lower bearing during normal operation if the external pump downthrust load is light.

The one other concern I see with too high endplay is that it can give the rotor more room to get a running start during momentary upthrust which could make that even more traumatic for the lower bearing.

What do you guys think... the analysis is ok? Any other considerations


=====================================
(2B)+(2B)' ?

 

[electricpete]

I've been reviewing endplay and just stumbled accross this quote in a G.E. motor manual (GEI-M1017 page 10) which suggests that low endplay can cause difficulty in rotating by hand as suggested by Motorwinder:

http://www.geindustrial.com/publibr...NR=Installation and Instruction|GEI-M1017|PDF

When slight preloading is experienced, there is no endplay and the rotor will not turn as freely by hand.

=====================================
(2B)+(2B)' ?

 

[rhatcher]

Pete,
If you post the motor nameplate information, including serial number, I may be able to get the information you need.