Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Register Log in
  • Congratulations MintJulep on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

1500 HP Internal Cooling Fan Failure

Status
Not open for further replies.
dArsonval

dArsonval

Electrical
Mar 21, 2010
375
Photos attached show what’s left of an Internal Cooling Fan on a 1500 HP gas compressor motor recently placed in service.
The relatively new Baldor/Reliance brand motor may have less than 500 hours on it.
Oil appeared very clean, and the sleeve bearings were not wiped out.
Internal fins connecting the “keyed” fan shaft hub to the outer fan blade(s) were completely ripped away
causing the fan to bang around inside the winding portion of the motor, hence a winding failure as well.
At this writing… I reluctantly admit I left the shop without looking at whether the motor is a two, or four
pole apparatus. (I’m thinking it’s a 4 pole unit but it may be a 2)
What kind of mechanical stress would cause such a failure? The light weight aluminum fan would seem to be
merely going for a ride affixed to the shaft. Yet it failed in the manor shown in the photos.
The end user appears to be pointing at the manufacturer as to the cause of fan failure.
Any additional ideas, opinions, questions would be helpful.

John

 
Replies continue below

Recommended for you

Sort by date Sort by votes
I see damage to the ends of the blades, looks to me as though something hit the fan first and then the attachments had no real strength. Was there a guard shroud around / over the fan? I have seen where excess acceleration causes the fan to warp slightly and hit the shroud.

"Will work for (the memory of) salami"
 
Upvote 0 Downvote
My initial thought on the "ends of the blades" damage is that it took place when the fan destructed hitting the winding.
There is a "non-metallic" shroud guiding airflow at each end of the motor. The fan-hub location on the shaft in relation
to the shroud sits back further into the motor winding area /space... yet, I'll look at that tomorrow again for clarification.

Exploring the "excess acceleration"... I'm not sure at what point during the motor's operation that would occur.
If that is an issue to consider... one would ask does it occur at "start up" of the motor? And would that be a motor control issue if so?
And thanks for the your reply jraef.

Always enjoying the forum,

John
 
Upvote 0 Downvote
It’s an oddball construction. You’ve got a thick hub on the shaft and a thin cylinder with fan blades. It looks like someone placed square keystock in the gap between and welded or brazed along the accessible corners.

I’m no welder, but the large, irregular goops of metal doesn’t suggest good workmanship to me. Also, I wonder if the outer ring was aluminum and keystock is steel, and they're trying to weld diferent metals. That usually doesn't work too well.

The welding itself if done poorly can create stresses in the thin piece, making it want to warp and pull away (putting a stress on those welds).


=====================================
(2B)+(2B)' ?
 
Upvote 0 Downvote
Use a magnet to check which parts are steel.

=====================================
(2B)+(2B)' ?
 
Upvote 0 Downvote
Just brainstorming other less likely possibilities.

Heating when the hub was initially installed onto the shaft? (especially if the outer thin ring was heated).

Driven by vfd, in theory there could be torque pulsations and minute speed pulsations that excite mounted resonance of that piece. But that seems like a real longshot to me because the mass of that fan is so small compared to everything else.

Both are longshots. Otherwise shouldn't be any problems unless really bad construction or else jeff's theory that the fan blades hit first. Get a welder in your shop to give you his opinion on those welds.

=====================================
(2B)+(2B)' ?
 
Upvote 0 Downvote
electricpete; VERY Useful considerations. Thank You for your replies.

John
 
Upvote 0 Downvote
Speaking purely from impression rather than experience, I can't imagine enough shear force to break those keys unless the fan blades contacted something. Is it possible the damage to the blades occurred before the breakage? What are the clearances from the blade tips to the enclosure? If the bearings are good then it's unlikely due to shaft movement. Is a foreign object, or enclosure misalignment possible instead?
 
Upvote 0 Downvote
You may have torque pulsations coming back from the compressor as each vane or impeller blade passes the discharge that are exciting a resonant frequency in the fan.
This is a long shot if the motor is line driven but more of a possibility if the motor is on a VFD.

Bill
--------------------
"Why not the best?"
Jimmy Carter
 
Upvote 0 Downvote
Hi

The fracture faces to me at least look very brittle i.e. a fast fracture occurred with no warning, further it appears that the fan blades failed just outside the edge of the fillet weld in the several places, I would therefore be thinking that it was some weld defect or possibly fatigue at play.
In the pictures posted by the OP the second to the last pic shows some marks of the fracture, if we could have another picture with the fracture face zoomed in close and viewed from above we might get a better idea.

desertfox
 
Upvote 0 Downvote
Thanks to Everyone who have posted a reply to this thread. Additional photos will be posted relating to the Fan Failure later in the day on 8-19-2013.
The hub portion of the fan has been removed from the Opposite Drive End of the rotor (where failure occurred) to allow further inspection and better photos to be taken.
A photo of the "in tact" Drive End Fan will also be posted to show its construction style and comparison to the failed piece, mounting position related to the rotor, etc.
It turns out, the motor is likely a 6 pole unit... ( not a 2 or 4, as alluded in original post) which I'll also confirm in Monday's Post.

Always enjoying this Forum,

John
 
Upvote 0 Downvote
Additional photos attached show the Opposite Drive End rotor fan failure. The fan hub is shown removed
from its keyed/set screw fit and resting on the rotor shaft along with the outer fan portion.
Drive End photos of the intact rotor illustrate the fan(s) welded aluminum construction and its mounting.

I don’t have any further info related to the motor’s power supply other than its name plate rating of 4000 volts
and rated speed of 1190 RPM.

Thanks in Advance for any, and all of your ideas related to this failure.

John

 
Upvote 0 Downvote
Now I get the picture that all 6 spokes broke and each of them broke at both ends. And that construction is a lot more fragile looking than I was thinking with those long spokes.... and the orientation of the spokes isn't favorable for strength to resist torsional bending (would be rotated 90 degrees for that type of strength).

I'm not sure what to make of it now. Before we were thinking maybe the blades hit or bad weld or (longshot) torsional resonance. Now I'm inclined to add possibly a bad design to the longshot list.

One thought that comes to mind to increase strength would be to put some gusset/stiffners in the perpendicular direction to those spokes, BUT that would have an adverse effect on axial airflow into the rotor for rotor/stator cooling which would have to be looked at very carefully. Certainly would want to get OEM involved (I assume you're already trying to get their input).

Another possible inspection would be NDE inspection of the welds on the non-broken fan side. If they're starting to crack then at least you can probably rule out the fan blades hitting something.

Maybe deserfox and others more mechanically inclined than me will have better ideas about this failure.

=====================================
(2B)+(2B)' ?
 
Upvote 0 Downvote
ISTR that straight radial blades are the #1 "Don't do" in the Boy Scout Handbook for Casting and Foundry Practices.

The same caution applies to weldments, for nearly the same reasons.

Those welds are UGLY. I figure it's a low production part, but can't anybody just take a snag grinder to the rust before striking an arc?

The OEM should be on the hook for fluorescently crappy workmanship.
Yeah, good luck with that.



Mike Halloran
Pembroke Pines, FL, USA
 
Upvote 0 Downvote
Hi

Thanks for the additional pic's!

My thoughts are this, looking at the last two pictures the vane to the left of vertical and adjacent to the hole in the hub appears to have some kind of undercut on its front edge face i.e. the narrow thickness of the vane, if this is so then it's very possible that cracks can grow along the weld and manifest itself in the type of failure you have at present.
It is hard to be absolutely certain because when I zoom up the picture it's goes blurred.

The site above shows a diagram of a weld undercut which will give you some idea of what I mean.


The second link gives a list and explains different weld defects and if you scroll down it mentions undercutting in relation to reducing joint strength.

I think you need to get some proper material analysis done on the failed parts and after that contact the vendor if you haven't already done so.
 
Upvote 0 Downvote
My first thought was that the construction was bad. Straight radial blades and poor welding/fitting - good way to get a failure. I noticed the under cut or back cut on the edges of the blades where they were welded too and it didn't impress me much.
 
Upvote 0 Downvote
Hi dArsonval;
Can you tell us something about the load? Direct drive or belted? piston, vane or fan?
I notice that despite the crappy welds, the breaks look like they are beside the welds. I don't see weld failures. We have mentioned resonance and torsional vibration a couple of times. I suspect that load induced effects caused early failure of a bad design.
I have seen pump casings with a hole the size of the palm of my hand blown out of a good casting by inadvertent operation at a "forbidden frequency". If you are inadvertently operating at or near a forbidden frequency even perfect welds may fail.
Forbidden frequency. A speed or frequency where periodic pulses from one part of a mechanical system excite a resonant vibration or oscillation in another part of the same system.

Bill
--------------------
"Why not the best?"
Jimmy Carter
 
Upvote 0 Downvote
Hi dArsonval

Is your 6P-1500HP gas compressor motor driving a reciprocating (IE: Areil) type compressor?
Does the compressor have a flywheel for the torsional pulsations or is motor just coupled to the compressor?
How is motor started - DOL or VFD? What standard was used for motor design (Nema only or Nema/API-541?
Is there is vibration equipment mounted on the motor (accelerometers or proxy probes?)

I agree the fan design /welding is weak, but there missing pieces of the puzzle for a NDE internal cooling fan to fail.
I could see the DE version fail as it is much closer to the load and potential vibration / resonance issues.

TIA
Mac
 
Upvote 0 Downvote
Better Late, then Never?

In an effort to not leave those who followed this thread “hanging”, the failed motor has since been repaired.
The stator was rewound. New cast aluminum fans were installed in place of the welded versions,
new sleeve bearings acquired from the manufacturer… were replaced, and so on.

The repaired motor comprises 1 of 2 new units driving new “two cylinder” reciprocating compressors. (When this thread was originally posted, the failed motor had approximately 400 hours of service on it, and not the 500 mentioned earlier. [The failed motor was built new in 2012].)

Motor back in service, all is well.

Well… Not so fast.

Both motors on these compressors exhibit “random” axial hunting.
The repaired one [more so] than the other unit. (Hence, the Repair Center’s continued involvement.)

In the early diagnosing stage of the motor failure, most everyone was looking at the fan, its poor design,
construction, etc. (The fan design and its construction were established in 1983, with revisions in the ‘90s
according to the manufacturer’s mechanical drawing of the fan.)

To date, no clear answer has been declared as to what caused the internal fan to fail other than pointing
at the fan itself.

One of the End User’s comments related to the original motor failure that raised my attention was that the motor took a long time to settle down with its axial hunting when the apparatus was originally being installed the first time. Well, here we are again… the motor is exhibiting similar characteristics.

If we rule out all the text-book reasons why a motor may hunt for its electrical center, and grant that the device(s) has/have been installed level, aligned, and being fed with etiquette power… (4123 Volts “measured”) What’s left?
As a Motor Person wanting to solve a problem, I don’t want to point at anything other than the solution.
What’s being overlooked here? Any ideas?

John
 
Upvote 0 Downvote
Does the subject motor have some kind of thrust bearing within it?

Should it?




Mike Halloran
Pembroke Pines, FL, USA
 
Upvote 0 Downvote
Status
Not open for further replies.

Similar threads

unclebob
  • Question
500 HP, 440V motor
2
Replies
22
Views
5K
waross
waross
MitchusMaximus
  • Locked
  • Question
General VFD Enclosure Cooling Question
Replies
19
Views
541
Heaviside1925
Heaviside1925
tjmurf
  • Locked
  • Question
Powerflex 755 (15 hp) IGBT SMOKED!
Replies
3
Views
170
electricuwe
electricuwe
petronila
  • Locked
  • Question
How many hp or kW 1
Replies
3
Views
141
edison123
edison123
edison123
  • Locked
  • Question
VFD for a 100 Amps, 1500 Hz application 2
Replies
5
Views
249
bacon4life
bacon4life

Part and Inventory Search

Sponsor

Back
Top Bottom