2 Speed 2 Wind Motor Won't Cold Start After 9 Months in Full Operation 2

[GrimesFrank]

Could use some 3rd Party Expertise please.
New motor replacement for 30yr application (Circ Pump LP/HP); 60 Hz, 575V, 13.3/40 HP, 1200/3600 RPM, 17.5/37.4 FLC, Wye Motor
- In 30yr nary an electrical issue; only bearing failures.
- New motors run for ~9 months then 1200 RPM won't cold start; gets to ~100 RPM before O/L kills. Loud electrical noise (screech not hum)
- Can run with bump from 3600 RPM. No loud noise just normal starting harmonic.
- Cold Start has balanced Voltage/Current as measured by Baker/Meggar Explorer Dynamic @ motor (490 V RMS ph-ph and 45A RMS ph)
- When bump started; Starting voltage dips to 305 V RMS ph-ph but and SS(s) @ 340 V RMS ph-ph.
- When bump started; Starting current maxs at 34A RMS ph and SS(s) @ 2.4A RMS ph

I have a bit more data from the Explorer but I don't want to influence opinion. My rudimentary reading has formed an opinion but would like to hear from all.
What are your suspicions, can we expect 3600 RPM to fail too, what physically would you inspect and finally are there specific tests you would perform to confirm suspicion?

Thanks.

Today is gone. Today was fun.
Tomorrow is another one.
Every day, from here to there,
funny things are everywhere. ~'Dr.' Theodor Geisel

 

[GrimesFrank]

Re-used meaning that frames, etal are being sourced by non-N.A. workshops are stripping old motors for parts and recycling for new purchases.

Posting an operational video of motor coupling to hear harmonic and see speed.
I hope it is able to be played.

Today is gone. Today was fun.
Tomorrow is another one.
Every day, from here to there,
funny things are everywhere. ~'Dr.' Theodor Geisel

 

[edison123]

And Toshiba put their name to it? If yes, it's clearly a malpractice.

Muthu

http://www.edison.co.in

 

[electricpete]

So the current went almost immediately to 80Apk = 56Arms (~ 3 x FLA) and stayed there for 6.4*0.167sec ~ 1 seconds and then it tripped on overload.

A few things don't seem quite right.

1. I'd assume the 56A must be your locked rotor current but 3 x FLA seems a little low to be locked rotor current. Voltage is above nameplate. Is there a KVA code assigned to the low speed winding or other indication of expected locked rotor / starting current in slow speed?

1. 56A vs 7.5ish A = 750%. We don't run at nameplate FLC....very much underloaded when in the LP condition

The running current is irrelevant. The thumbrule is that locked rotor current is 5-7 times nameplate full load amps.

Ordinarily if the motor is not drawing locked rotor current I might expect a power supply problem. But that doesn't fit anything else. You measured the voltage. There was no current unbalance (which rules out voltage unbalnce).

And then there's that noise in the video.... I think it's around 480hz. If you're in the US that is 8 times line frequency.... a very peculiar frequency to show up. I certainly wouldn't expect that for voltage unbalance or low voltage.

You mentioned crawling and edison mentioned slot combination, I wonder if that type of noise is expected for that. The fact that it started rapidly and jumped right to that speed where it remained... that really seems to be exactly what would be expected from crawling.

Of course crawling is built into the design (slot combination). There's nothing to suggest it would get worse over time after installation. Degradation over time of course brings to mind rotor bar/circuit degradation, but for rotor degradation I'd expect a slow start from the beginning, not a rapid start and then lock in at that speed.

I wonder if there is something about the sequence of the events that is tricking us into thinking that it's degradation when it's really just random or dependent on starting conditions. Or is it the case that once this problem occurs on a given motor, it keeps on occuring on that motor during every start?

 

[edison123]

Since the 'new' motor runs fine for a few days/months and then starts to crawl and it was built with crappy re-used parts, I think the 'designer' put the original rotor bars too close to the crawling zone and a few rotor bars failed over time thereby putting it in the crawl zone.

Once we have the stator & rotor slots and rotor photos, the picture could become clearer.

Muthu

http://www.edison.co.in

 

[iop95]

May be an issue related to power supply and/or cable and/or contact connections + variable load start torque request.
At low start, voltage dip to 500V and motor torque may remain below needed start torque requested by load.

 

[edison123]

Today, our repair shop received the following crappy chinese 16 KW, 6 pole high slip (13.5% slip, probably NEMA D motor) Al die cast rotor with six bars broken clear of the end ring. High slip means higher rotor losses and higher rotor temperatures. Melted Al sprayed over the stator winding and killed it too.

Customer complaint was the motor was running slow and taking high current and then tripped on OL. That rotor is toast.

Muthu

http://www.edison.co.in

 

[LionelHutz]

I fully expect Toshiba will take a standard 1800rpm or 3600rpm motor from their robot factory and then strip the winding to re-wind it into this dual speed motor. That's what other manufacturers do. I've been in the local shop that modifies Marathon motors. The TECO motor that I mentioned before certainly appeared to be done that way. It was wound to a lower rpm than their standard catalog offerings.

 

[electricpete]

The "bad" rotor/stator slot combinations (which combinations cause crawling, among other things) depend on the number of poles.

One implication of that is that a 2-speed motor has more bad slot combinations than a one speed motor (so it might be a bit more challenging for designer to select a good combination and meet his other constraints)

Another implication is that it's a big problem if they don't carefully consider the number of rotor/stator slots when rewinding an existing motor for a new speed or for multiple speeds. I gather that was Lionel's point. It's a good point, I'm just spelling it out in detail in case it wasn't obvious to op.

 

[edison123]

Lionel

Per OP, the motor worked fine first in both speeds. So bad slot combination, which I am sure Toshiba knows, was not an issue. A bad rotor with broken bars could result in that bad slot combination resulting deteriorating performance as times goes on.

pete

The number of bad slot combinations for cogging and crawling are the same for all the poles since the rules/formulas are the same for all poles.

Muthu

http://www.edison.co.in

 

[electricpete]

bad slot combination, which I am sure Toshiba knows, was not an issue.

Every manufacturer makes mistakes from time to time.
I acknowledge there was more to your statement, which is discussed below...

Per OP, the motor worked fine first in both speeds. .... A bad rotor with broken bars could result in that bad slot combination resulting deteriorating performance as times goes on.

It was acknowledged by myself and others above that deterioration is a feature of rotor bar circuit degradation and not slot combination. Nevertheless the total collection of symptoms presented do not seem match any of the scenario's exactly so one of the reported symptoms may be incorrect (which is why my post 11 May 23 22:29 questioned the op about the details suggesting it is a degrading phenomenon).

The following factors steer my personal opinion away from rotor bar circuit degradation and towards rotor / stator slot combination causing crawling
[ul]
[li]1. It appears from the video that speed initially rapidly accelerates to a few hundred rpm and then appears to stay steady at that speed... with rock-solid audio pitch. That seems to be what we expect for crawling from rotor/stator slot combination, but imo that is not what we expect for rotor bar circuit degradation[/li]
[li]2. The motor cold starts fine in fast speed but does not cold start in slow speed. That's exactly what we expect for crawling (since the particular rotor/stator slot combinations that cause crawling DO depend on speed or poles, more below), but this is not expected for rotor bar degradation (fast and slow speed share the same rotor, and starting in fast speed is generally more challenging). [/li]
[li]3. The motor starts fine in slow speed during coastdown but not in slow speed from cold start. That's exactly what we expect for crawling if the coastdown does not drop below the crawlng speed, but it's not particularly what we expect for rotor bar circuit degradation.[/li]
[li]4. No abnormal symptoms reported for steady state running with a load. The lack of symptoms while running loaded is exactly what we expect for rotor/stator slot combination resulting in crawling, but not for rotor bar circuit degradation. (op seems very familiar with condition monitoring and I certainly assume he would have mentioned any running anomalies that were observed, but I guess at this point it's appropriate to double check with op: are there is any pulsing in noise, vibration or current indication while running under load? Are there any elevated pole pass sidebands around running speed in vibration and around line frequency in current?) [/li]
[/ul]

To be clear that's just my opinion, but I don't think anyone can guarantee a diagnosis from what we have been presented since none of the scenarios seems to explain all the facts (so everyone is entitled to their own opinion on this particular question and it's difficult to proclaim any one opinion as right or wrong)

pete

The number of bad slot combinations for cogging and crawling are the same for all the poles since the rules/formulas are the same for all poles.

I believe you are mistaken.
See figure below from EPRI 5036 Volume 17 (a very useful reference imo and also available for free).
Each row corresponds to a given number of poles P given in the left hand column.
The right hand column "C" represents "cusps" which are identified elsewhere as cusps in the torque speed curve that may dip below zero (I interpret that as crawling)
The entries in that column C are not all the same, they change depending on number of poles. Relevant to the current discussion, the table entries in column C for 2 and 6 poles are different.

for op - The figure also tells us in advance what you might be looking for (when you find out stator/rotor slots) to cause cusps/crawling for the 6-pole winding. Difference in rotor/stator slot combination of 5 in either direction, or else rotor bars exceeds stator slots by 30 (rotor bars exceed stator slots by 12 would also be problematic, but it would be problematic for both speeds). I would recommend you make that count on at least one of these failed-start motors.. either ask your repair shop to do it, or else ask your own maintenance staff to do it

 

[electricpete]

A bad rotor with broken bars could result in that bad slot combination resulting deteriorating performance as times goes on.

I don't rule out there are more than one thing going on.

SCENARIO 1 - Rotor Degradation first, harmonic crawling 2nd. I think that's what edison was suggesting. Let's say we have a 6 pole motor where the table shows us a bad combination is R-S=12 and it is built with R-S=13. If one bar manages to open circuit, that would not give the same results as the table combination of R-S=12. The table combinations are based on spatial harmonics which depend on angular spacing between bars which would be unchanged (total number of bars is used as a proxy for spacing between bars). But certainly for one or more open bars, the rotor spatial harmonic profile would change in very complicated ways so I can't rule out that it might somehow create the condition for crawling although I have not heard that once reported in the literature (so call me skeptical).

Prelude to scenarios 2 and 3: The next two scenarios require us to pre-suppose that the crawling based on factory combination is intermittent based on starting conditions. That doesn't sound unreasonable to me. It might not sound intuitive to you that a motor can successfully start passing through a cusp where motor torque dips to/below zero, BUT the torque speed curve represents a quasi-static model of the motor which neglects certain dynamic behavior. When we consider the full dynamic model of induction motor it is well known an induction motor can actually overshoot sync speed during an unloaded DOL start as I discussed here thread237-248895 That is a behavior which defies the intuition you might have based soley on the quasi-static torque speed curve, and yet it accurately describes motor behavior including the dynamic electromagnetic effects (currents and torques are a function of not just speed, but also recent history). I think it's logical that similar behavior can propel the motor to successful start past a zero/negative cusp of the torque speed curve IF conditions are favorable (some combination of low load torque, low load inertia, high voltage). Alternatively maybe the steady-state torque anomaly doesn't go below zero but goes very low, so that successful start depends on load torque below that value at that speed. Either way during the successful starts it might slowed passing through the crawling speed but never be noticed until it actually gets hung up there. So it's possible this factory-initiated condition only exhibits intermittently. That leads to two more scenarios:

SCENARIO 2 - Slot-combination-based crawling (no rotor degradation). The failed starts only occur intermittently as discussed above. Once it occurs you call the motor failed and move on to swap the motor. That's why I'm particularly interested to know how many more unsuccessful starts you have after the first unsuccessful start and trying to nail down whether there is repeatable change in behavior after a certain point in time.

SCENARIO 3 - Slot-combination-based crawling first, rotor degradation 2nd. Harmonic crawling occurs intermittently as discussed above. When it occurs the first time, it resembles a locked rotor condition which damages the rotor and makes subsequent successful starts less likely. If there is truly a change in behavior over time this could be the mechanism. Although the fact that the motor tripped after only one second in the posted waveform makes it seem less likely that rotor could be damaged by a single failed start unless it is a really bad rotor.

so in summary I think they're all somewhat unlikely. Scenario 1 and 3 are unlikely because they require rotor degradation but we see no sign of rotor degradation during running. Further we see no signs of rotor degradation based on successful starting after coast and successful start in fast speed. Scenario 2 is unlikely to the extent the phenomenon described by op is changing over time (rather than just intermittent depending on start conditions).... I haven't nailed that down and 2 is still my personal favorite unless/until op clarifies the extent to which it is repeatable after first "failure"

So with everything somewhat unlikely, what is the action. I see op already plans to inspect the motor with expert assistance and edison already suggested count slots and inspect rotor. So I'm in agreement with all that, and I have no different actions to suggest. Unless maybeyou can somehow measure speed during failed start to check if it is close to 1200rpm/7.... like maybe high frame-rate video of start for slow-motion playback at a known rate. I see you reported speed at 150 to 200rpm... I'm curious how did you determine that.

 

[waross]

The definition of insanity is doing the same thing over and over again and expecting a different result.

I hope that you have found a source for replacement motors other than Toshiba.
A number of very smart and experienced engineers are not agreeing on a definitive source of the issue.
In plants that I have seen, production was king.
Management was more concerned with results (eg: reliable production with no down time) than with details.
Have you considered stepping away from the problem by replacing failed motors with a different brand?
Or
A seat of the pants solution:
I remember a two speed machine in a mill that used two motors belted together.
A large motor for high speed and a smaller motor for slow speed.
Select a 1200 RPM motor built on the same frame as 3600 RPM motors.
Belt that 1:1 to the existing large motor.
Use a reversing contactor to select one or the other motors.
Your present motor is 13.5/40 HP.
You may not need the full 13.5 HP on slow speed.
The solution won't be pretty but it may work and will appeal to the bean counters who just look at the numbers and not at the actual machinery.
Then you will have all the time in the world to investigate the Toshiba problem.



--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[edison123]

pete

I have literally posted the example of Scenario 1 of rotor bars degradation affecting the motor speed and performance, which happens often, regardless of you having not read any literature on it or not.

And no, it's not a challenge at all for a winding designer to pick the correct slot combination for 2 speeds or even 3 speeds. I often convert speeds (the above rotor being the most recent example of conversion from 6 pole to 4 pole) and I faced zero issues with any of the conversions.

Muthu

http://www.edison.co.in

 

[electricpete]

pete

I have literally posted the example of Scenario 1 of rotor bars degradation affecting the motor speed and performance, which happens often, regardless of you having not read any literature on it or not.

Good grief. Yes, I'm well aware rotor bar degradation affects motor speed and performance as I'm pretty sure everyone reading this thread is. You're focusing on scenario 1, so let's look at what it is that you seem to have skipped over:

SCENARIO 1 - Rotor Degradation first, harmonic crawling 2nd

Your posted rotor has nothing to do with harmonic crawling. Harmonic crawling is the scenario where the torque speed curve has a cusp associated with spatial harmonics. It is not the same as simple rotor degradation which just lowers the whole torque speed curve. I have enumerated things that don't fit simple rotor degradation and one of them is the video which shows evidence of harmonic crawling rather than rotor degradation (rapid initial speed increase followed by sudden speed plateau at approx 1/7 sync speed... the most common crawling speed). Another is that it appears to only affect low speed (expected for harmonic crawling, not for rotor degradation). Scenario 1 would be where opening of a rotor bar shifts the spatial harmonics in such a way as to create spatial-harmonic-induced crawling (that is the type of crawling that would create a torque torque cusp and only affect one speed). I have not heard of that and I'll assume you haven't either.

And no, it's not a challenge at all for a winding designer to pick the correct slot combination for 2 speeds or even 3 speeds. I often convert speeds (the above rotor being the most recent example of conversion from 6 pole to 4 pole) and I faced zero issues with any of the conversions.

Are you sticking with your previous claim that problematic rotor/stator slot combinations do not depend on speed (15 May 23 14:06), or have you revised your position on that?

 

[electricpete]

I was wrong to suggest the crawling speed would necessarily be exactly 1/7th of sync speed in the case of 7th spatial harmonic crawling. It would be exactly 1/7th of sync speed in the case of synchronous cusps, but it would be somewhere above 1/7th of sync speed in the case of asynchronous cusps associated with 7th harmonic

https://www.eeeguide.com/crawling-and-cogging-in-induction-motor/

I had mused about the possibility of dynamic electromagnetic effects allowing the motor to accelerate past the torque anomaly as one possible explanation for behavior that I postulated was intermittent/nonreproducible (I postulated it was intermittent to come up with an alternative to degrading). If that particular scenario is at all plausible, it seems it would be more plausible for the synchronous cusp, where the torque anomaly only exists in a very narrow band of speed and so a very narrow window of time during acceleration.

I hope op will feed back his findings because this is an interesting one. If it does turn out to be a slot combination crawling problem, that's something few of us ever get to see.

 

[GrimesFrank]

Thank-you all. I took a much needed vacation to refresh my mind from this.
Before I reply/go through your valued responses; an update:
[ul]
[li]As was stated by me earlier the motor was replaced with a fresh identical. I have come back to find out it has "intermittently" failed to cold start. Was told it failed to achieve min flow/start (no one present to view in field) then re-attempted and started on 2nd pass. Operations are now gun-shy and have temp procedure in place to perform pump duty swaps in HS ONLY and then revert to LS until solved. No cold LS starts...no issues. Production pressure. [/li]<br>
[li]Motor expert has not been to site yet NOR motor pulled apart even to inspect stator/rotor. Trying to decontam motor I am told to ship to rewind shop for inspection. Not likely successful once it goes in containment it can't leave station. Frustrating.[/li]<br>
[li]Motor project is rescoping. Looking to source either a VFD to make this HS winding do dual duty OR (my choice) match new non-Toshiba motor/VFD set to cover duty. 2 speed motors are so passé. [/li]
[/ul]
Now onto your posts;
[ul]
[li]I have been in the edison123 camp from the beginning on this; see my original posts. The "designer" tried to balance the LS/HS winding slot combinations; leaning heavily to the edge of LS acceptability and if the rotor end tacking is poor the cold starting torque degraded it to pop (not as bad as edison123's photos though ) Great slot combo table electricpete to use when we get to finally look at this thing. All the info posted here has been quite helpful in evaluating this.[/li]<br>
[li]If we have installed a motor that works from first the behaviour is...it works, then occasionally misses a start but goes on the re-go, then mis-start frequency ticks up slightly, then no joy at all. Previously as stated this was over a 6-9 month period but for this install we have been in for a month and it is exhibiting the behaviour already. Also remember we have 2 other instances of this motor (same duty; different Bus) that have never given us a problem for 2 years.[/li]<br>
[li]For electricpete: Shaft speed was measured with a hand held laser tachometer so accuracy is what it is.[/li]<br>
[li]For waross: Like those ideas; I tell our interns we're farm tractor engineers...make what works and then justify it. In this case we have some time and for some reason this time a budget to fix it. Likely Capital Finance folks are feeling a pinch.[/li]<br>[/ul]
I will keep you all posted as info comes in; we only cool irradiated fuel with these pumps, nothing too important.

 

[waross]

Also remember we have 2 other instances of this motor (same duty; different Bus) that have never given us a problem for 2 years.

How is your bus voltage?
Low voltage may be exacerbating a bad design.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[GrimesFrank]

How is your bus voltage?
Low voltage may be exacerbating a bad design.

We have had failures from different pumps on different Busses. Each time we measure the Bus Voltage it is as near to 600V as I think you can be.
We had (1) other path we were chasing but it reached a dead end as the voltage at the motors, when working, was well enough to start this thing. From MCC contactors to motor goes from #4 AWG (500 ft) to #12 AWG (1000 ft on mobile flexible cable carrier) to #4 AWG (65 ft). We originally thought that the #12 was causing too high a volt drop on a new high efficiency motor to start but we measure and we get what you see above in the charts, we even swapped them for spare #6 AWG on a failed motor and still couldn't start it.

 

[edison123]

Yup, their 15 minutes were up years ago. Especially, with the advent of sophisticated VFD's, which more than anything else, provide smooth acceleration with no inrush current spikes and related mechanical shocks to the system, I doubt OEM's will make 2 speed/3 speed motors any more. I used to rewind many 2 speed motors from sugar industry (notoriously conservative when it comes to adapting to new technology) till nineties and almost all of them have been replaced with drives. Sure VFD has its own drawbacks, but then which system is perfect?

Hope you get the chance to tear down this joke of a motor and look at its entrails.

Muthu

http://www.edison.co.in

 

[edison123]

Below is the original good rotor of an unused spare motor received today.

I think the rotor bars started failing due to high slip (high rotor loss) and frequent starts (high inrush currents).

Muthu

http://www.edison.co.in