Motor / coupler / pump alignment vs. current draw

[ScottI2R]

Hi everyone,
We recently had a coupler fail on a prelube pump for a 5 stage turbine air compressor. That style of coupler has been replaced with a lovejoy coupler.
This pump is driven by a 1750 r/m 3ph. 5 HP 480V motor. No load, the motor draws 2.5 A/ phase. Loaded as it is now, it draws 4.2A/4.4A/4.5A for each phase respectively. Nameplate states 6.2 Amps and the overload is set for 7 amps. While the alignment of the motor is not PERFECT (off horizontally maybe a mm) vibration is minimal. Actually, it seems very smooth. I don't feel current draw is too much either. My question is: Does anybody see a problem with this?

Oh, while I do not know the GPM of the pump, the main oil lines are 2 inch and the pressure is 28-30psi.

Thanks in advance,
Scott

I really am a good egg, I'm just a little scrambled!

 

[MikeHalloran]

The motor seems happy enough...

Mike Halloran
Pembroke Pines, FL, USA

 

[oldfieldguy]

It's always good to ask the question, "Where's the power going?" in situations where current goes up.

In the case, for example, of a one amp increase after changing a coupling, as yourself "Does that coupling look like it's soaking up few hundred watts?" because that would be the amount of power going SOMEWHERE.

In your case, though, the current draw seems okay. On the misalignment, I would expect to see the coupling suffer some excessive wear.

old field guy

 

[edison123]

A slight difference in the motor voltages could explain current differences. A misalignment would not create a current unbalance.

Muthu

http://www.edison.co.in

 

[ScottI2R]

HI everyone and thank you,
Let me re-phrase my readings of the current. The low values were taken when the old coupler was broken so essentially the motor had no load. I figure the increase in current is just the load of the oil pump. And, I also figure it would probably take a severe misalignment to show up in excess current draw I.E. binding. Anyhow, The Supe is happy, (he is electrical as well) but he is going to have one of our mechanical guys give it a once over, and we can put this puppy to rest. And oldfieldguy, the lovely bride said the same thing:"That motor could give a $hh^, the spider isnt going to like it though." ....gotta love a woman with smarts.

Thanks again,
Scott

I really am a good egg, I'm just a little scrambled!

 

[MikeHalloran]

Sometimes couplings fail just from old age, but more often there is a contributing factor, like fluid contamination that destroys elastomers, or some odd event that locks up the load.

Did you investigate why the old coupling died?



Mike Halloran
Pembroke Pines, FL, USA

 

[electricpete]

I agree with all the conclusions you reached. I don't really see enough to suggest a problem.

1mm = 0.040" is on the high side for misalignment as often people apply much tigher tolerances, but it can be hard to apply traditional alignment techniques on very small machines (can you even get your dial indicators mounted and reading properly), hence there is more tendnency to use less accurate methods such as straight-edge/eyeball.

The spider/coupling are stressed by misalignment. The bearings also can be stressed.

vibration is minimal. Actually, it seems very smooth.

In general, the absence of vibration does not prove the absence of large reaction forces from misalignment because those forces can have very large "dc" (constant with time) component which does not directly cause vibration. Imagine you have a perfectly symmetrical machine that is stationary and coupled... now move the motor 1mm or more offset toward the north and you have a dc force tending to push the motor back to the south. Now rotate the shaft by some amount (90 degrees) and repeat your measurement... what changes? Not much... the force is still pushing the motor back to the south. So you can see there can be a very large dc/constant-direction force even as the machine rotates... that constant direction force does not show up on vibration.

At this point I should mention there are a number of secondary effects which do enter into the picture allowing misalignment to cause vibration. One would be varying stiffness of the key and coupling as I rotate the shaft. Let's say I have a keyway, the shaft is stiffer in one direction than 90 degrees rotated, so the dc force caused by offset misalignemnt can result in 1x rotating speed and 2x rotating speed movement/response/vibration. Additionally, considering angular misalignment, the restoring stiffness of the coupling can vary 3x per revolution with a 3-jaw Lovejoy coupling depending on wether the jaw is in the plane of the anuglar misalignment or not. It is very common to see 3x, 6x, 9x etc multiples of running speed in the vibration spectrum for Lovejoy 3-jaw couplings and I generally interpret these as a symptom of misalignment or coupling distress. But again it is a secondary effect and touch to judge the magnitude of the large dc forces that may be present based on smaller time-varying responses.

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

 

[electricpete]

Attached is a spreadsheet that I developed to get a rough estimate of static forces associated with misalignment.

The inputs are the green cells.
MOB Brg Position Deviation 0.039370079 inch
MIB Brg Position Deviation 0.039370079 inch
PIB Brg Position Dev 0 inch
POB Brg Position Dev 0 inch
Motor Outboard Brg Stiffness 1.00E+07 lbf/inch
Motor Inboard Brg Stiffness 1.00E+07 lbf/inch
Pump Inboard Brg Stiffness 1.00E+07 lbf/inch
Pump Outboard Bearing Stiffness 1.00E+07 lbf/inch
Motor - distance betw bearings 6 inch
Motor - shaft extension length 3 inch
Pump shaft extension length 3 inch
Pump - distance betw bearings 6 inch
Coupling Spacer Length 2 inch
Cplg flex elmt ang stiffness 1.00E+07 ft-lbf/rad
Coupling Spacer Diameter (steel) 1 inch
Motor shaft diameter 1 inch
Motor Shaft ext diameter 1 inch
Pump shaft diameter 1 inch
Pump shaft extension diameter 1 inch
Motor Weight 0 lbf
Pump Weight 0 lbf

So the problem I tried to analyse is 1mm = 0.039" offset of motor compared to the pump. The information regarding distance between bearings, shaft extension length, coupling space length, shaft diameters I have estimated, but should be very easy for you to get reasonably accurate estimates for these.

The remaining items focusing on support stiffness and coupling stiffness are very difficult to know without some effort, however I tend to think we can develope a "worst case" estimate by putting in very high values for these stiffness.. In otherwords we assume rigid mount and rigid coupling and look at the bearing reaction forces which are directly associated with shaft characteristics. It is definitely on the high side because as you know the whole reason we use flexible couplings is to reduce these reaction forces.

Under these assumptions and dimensions the conservatively high estimate for bearing reactons for the static problem is 3919 Newtons. You could compare this to bearing ratings and estimate reduction in life using the beairng life equation something like Life = 1E6 revolutions * (C/Crated)^3 where C is actual rated load and Crated is rated radial load.

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

 

[electricpete]

Correction in bold:
You could compare this to bearing ratings and estimate reduction in life using the beairng life equation something like Life = 1E6 revolutions * (C/Crated)^3 where C is actual rated load and Crated is rated radial load.
should've been:
You could compare this to bearing ratings and estimate reduction in life using the beairng life equation something like L10Life = 1E6 revolutions * (Crated/C)^3 where C is actual rated load and Crated is rated radial load and L10 life is time by which 10% of bearings would be expected to fail

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

 

[ScottI2R]

Wow Pete! Now that IS analysis! Here are a couple of photos. The new lovejoy and the old disassembled coupler. Plenty of room for indicators here Pete. These are really big compressors.
Mike, judging from the large amount of rust inside the old coupler, I would say age and lack of lubrication / maintenance caused failure.

I really am a good egg, I'm just a little scrambled!

 

[ScottI2R]

OOPS! Here is the old coupler.

I really am a good egg, I'm just a little scrambled!

 

[jraef]

Looks to me as though the old one was riding on the ends of the splines and they eventually just gave way. Nothing lasts forever...

I didn't see where you posted an increase in current from one to the other. I don't see any issues here, but as was said, a slight misalignment is a precursor to early failure, or so my ME buddies tell me. Electrically though, no issue is evident from this data.

Other than this:

Nameplate states 6.2 Amps and the overload is set for 7 amps.

Having the OL setting 12% high will likely cause significant motor damage in the case of an overload. Most (if not all now) Overload Relays ALREADY have the pickup point pre-determined in the relay itself, so you are REQUIRED to set the relay for what the motor nameplate FLC is stated as.

"If I had eight hours to chop down a tree, I'd spend six sharpening my axe." -- Abraham Lincoln
For the best use of Eng-Tips, please click here -> faq731-376

 

[ScottI2R]

Alrighty,
OL has been set accordingly and my mechanical guy got things aligned better now.
Jraef, I did not have any current measurements prior to the failure of the original coupler. Current draw now is still as posted in my OP.

Thanks everyone

Scott

I really am a good egg, I'm just a little scrambled!