DC motor risers break 2

[Skogsgurra]

Two 500 V 1500 kW DC motors for a reversing rolling mill have had catastrophic failures in that the risers have broken and (as it is assumed) caused a collector round-flash. In both cases it happened after around three months of operation and there is no sign of arcing or other electrical damage to the breaks. They look like fatigue breaks.

There are some speculations as to the reason for the breaks. I do not disclose what the thoughts are, because we would like some unbiased and fresh thoughts. What are the more common reasons for a case like this?

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[dArsonval]

Among speculations and opinions, one could point toward flexing risers due to inadequate blocking installed during previous repairs.
Breaks in risers have been documented to occur where they (connective-ly) mount, and not necessarily where they flex.

Attached photo of stripped armature showing commutator risers for forum reference.
Are the breaks near the commutator, or up near the winding connection?

John

 

[waross]

Conductor issues?
Is there any chance of conductor movement.
Heat expansion? Do the conductors follow a zig zag or curved path to the risers to allow for heat expansion.
Do the conductors go straight to the risers so that heat expansion will cause pressure on the risers?

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[MikeHalloran]

The risers form an array not unlike the face of a centrifugal pump impeller.
They have to induce some radial airflow as the rotor rotates.
... which is periodically frustrated by the brush ends passing nearby.
A siren-like noise, and fatigue to the risers, would result.
Do you have fatigue life data for the riser alloy?
The fatigue would go faster if the excitation is close to a natural frequency for the riser elements as beams.



Mike Halloran
Pembroke Pines, FL, USA

 

[lukin1977]

could you post some pictures of the damage?

 

[Skogsgurra]

Thanks to all! I have not been on site, yet. Will take some pics Monday.

Mike's thinking is new input. There is some doubt as to annealing after cold-working the risers.

Bill, there are curved sections to take care of thermal expansion.

Breaks are close to middle of risers. There are supporting "ropes" with cured impregnation keeping the risers from heavy vibration. A few of those ropes have broken.

The motors came directly from the manufacturer and are brand new.

As I said, I will take pictures Monday. Thank you for the comments.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[waross]

Maybe in the weeds here, but part of trouble shooting is eliminating possible causes.
With two almost new motors failing, can we consider that the risers may have been damaged during the manufacturing process? Is it possible that the risers were cracked and waiting to fail even before the motors were put into service?
" A few of those ropes have broken."
Let's consider if this is cause or effect.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Skogsgurra]

Yes, that is good input. I have some indications that this manufacturer had problems also in other installations. So, it may be that the workmanship isn't the best. Shall keep an eye open for that.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[Skogsgurra]

Mike, I think that you are right!

I found a picture that I magnified so that the Brush, brush holder and risers can be seen. The structure is definitely "Siren like".

See:

Furthermore, the breaks are in the curved region running close to brush holders:

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[lukin1977]

Has the motor been running above max nominal speed?

 

[waross]

A question Gunnar; re the second photo showing the breaks. Is that the risers or the conductors leaving the risers and entering the armature core?
I wonder, are the drive ends of the windings restrained so that most of the thermal expansion must be absorbed by those curves?
There is differential movement between the windings and the core during warm-up. The heat is generated in the conductors and then conducted to the core.
I once had an exciter that developed shorted windings due to differential movement. The rotor tested good when it was cold. When the rotor was heated in an oven there was no differential movement and the short would not develop.
The shop was reluctant to rewind a rotor that appeared to be good.
We ran the machine until the short developed and then immediately did a brush null test. That is, we excited the field with AC and measured the voltage across the brushes. The varying voltage as the machine was rotated proved a winding fault.
The point is that heating and thermal expansion as the motor warms up is not a simple process. There is probably differential movement and depending on the construction of the rotor the curves may not be enough to absorb the cycling movement.
You mentioned that the manufacturer may have problems with other motors.
I would expect that the time to failure would correlate with the number of starts more than with the number of hours run if this is a heat expansion issue.
Extreme load-unload cycles may skew the results somewhat.
If you are able to contact the person concerned at the manufacturer they may be able to confirm a correlation, even a loose correlation, with the number of starts. Yes, I know.- It may not be possible to talk to anyone with first hand knowledge.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Skogsgurra]

lukin: It is being run above base speed. But not above maximum allowed speed.
Bill: It is a reversing mill. So there are continuous starts/stops and starts/stops in reverse direction. The speed is ramped up/down and the thyristor rectifier is current limited, so there are no extra high currents during operation. And, yes, the movement is restrained and expansion is absorbed by the curved parts.

I tend to follow Mikes tip. There are sound recordings where a whining sound that comes and goes. Much like a siren sound. And there is a steady 300 Hz (50 Hz grid, hence 6x50 Hz ripple) sound in the background.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[dArsonval]

Rapid acceleration and deceleration contributes to tall freestanding riser flexing.
It appears the surge rope laced around the risers shifted from its possible initial midpoint location
(perhaps by centrifugal force) leaving the middle portion of the risers inadequately braced.
The blocking shown closer to the commutator in the first photograph would have contributed to increased
rigidity had it been located more centered along the riser length where flexing is prone to occur.

John

 

[waross]

Hi Gunnar. I'm not concerned with reversing unless it contributes to heat expansion cycling. I am thinking starts from cold rather than fast stop/starts and reverses. If I was on site I would try to get a ball park estimate of the amount of heat expansion those curves must absorb during a heating/cooling cycle.
Possibly earlier designs had a larger curve and so did not experience the stresses that the current motors endure.
But, I am not discounting Mikes tip either. A ball park estimate of expansion may rule out my suggestion.
And, looking at design and construction issues:

The fatigue would go faster if the excitation is close to a natural frequency for the riser elements as beams.

A change in the shape of the risers from the original design would probably change the natural frequency also.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[edison123]

Looks like stress cracks to me. Might be that the riser material is suspect. Some commercial grade copper enforced by bean counters?

Or they were cold bent instead of annealing and then bending?

Muthu

http://www.edison.co.in

 

[edison123]

The cracks are closer to the commutator end. May be at around 2 to 3 inches from the root. Hence flex could not be the problem.

Muthu

http://www.edison.co.in

 

[Skogsgurra]

I ran the sound file through an FFT and got the spectrum below. The 300 and 900 Hz are fixed, of course, but the other peaks move with speed. Those peaks represent the whining sound. Not present in motors built with more distance between brush holders and risers.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[Skogsgurra]

Question for Mike:

We have checked the motor in a winder's shop and ran it through the whole speed range. The resonances are there, but we have a problem seeing exactly what is happening. There are several reasons for that - the stroboscope we had available is rather weak, and it wasn't synchronized with with the rotor. There is also very little space between risers and brush holders. We are trying to find a better strobe, synch it with the rotor and find a good mirror so that we can see better what is happening behind the brush holder and stabilizing ring.

My question is: Do you have any references, pictures or recordings that can help us explain the phenomenon to the motor manufacturer? They are sceptic and not willing to accept our explanation without any hard evidence.

We need to finalize the report by Tuesday and it would be extremely helpful to have your input Friday or Monday - if at all possible.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[MikeHalloran]

Send the motor guys to a fire station and have them examine a mechanical siren.

It also occurs that if the VFD is not completely stopping the motor before electrically reversing it, there could be large currents induced by said 'plugging' of the motor. I did that with a 1/15 HP ac gearmotor, and blew a 15A breaker right out of the breaker box. I did not photograph the remains of the breaker, which broke into little pieces on hitting the floor.

Maybe you can find a high speed video camera, which may capture flexure even without being synchronized.

For a small motor, you could rig a dove prism to rotate at ~half the motor speed, and look at the assembly along the motor axis with a telescope. The illumination would not have to be strobed, but your motor may be too big and too enclosed for any such rig to be practical.

Maybe you could rig up some soft plastic, maybe fibers like PlastiGage, adjacent the riser bars, and examine any witness marks left from flexure of the riser bars.




Mike Halloran
Pembroke Pines, FL, USA

 

[Skogsgurra]

Thanks, Mike.

I was more looking for a documented case. Also, this is a DC machine so I think that mentioning a VFD was a slip of mind. And the high torque scenario (bridge failing to commute during braking) is not applicable. Operation is smooth and an occasional bridge short circuit would not produce fatigue breaks. There was no tripping and no blown fuses until the failure occurred.

The PlastiGage or similar technique is something to ponder. Will think more along that line.

We have been turning this problem inside out and tried to find a plausible explanation. There are torsional and flex resonances, but they are damped with a Q somewhere around 5 or less (estimate from impulse test) and they also do not produce forces that act in the plane where the risers are.

Your suggestion is the only plausible one, so far. But we need to either show the effect in a convincing way or find a case where it has been documented.

Anyhow, we now have a Shimpo DT 31 strobe that is rather powerful (xenon, not LED) and can easily be synchronized to the rotor and we will run a test on Monday. The synchronization will allow us to get a steady picture while changing speed slowly across the whole working range. A high speed camera is available at the nearby ABB HV breaker lab, but we haven't taken that step - yet. That would involve bean counters and would complicate matters more than we wish.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.