Cracked Rotor End Ring 2

[rockman7892]

I wanted to share some pictures of the rotor we recently has removed from a 1250hp 4.16kV motor. This motor was sent away to a motor shop for scheduled inspection and it was found that it had a cracked rotor end ring. This is a new one to me so I found these pics quite interesing.

Was curious to see if anyone else ever seen this before and what the causes were.

I'm guessing that we should have seen some electrical losses associated with this crack.

 

[rockman7892]

Heres another one

 

[edison123]

Looks like a crack in the copper casting. Did this reveal while in service like swinging stator current or did you find out only on the routine inspection ?

 

[electricpete]

I will offer a guess. As we know the most common source of stress for rotors is differential expansion during the rapid heating during motor starting.

Through repeated thermal cycling, the position of one or more bars can change relative to the core... a process called thermal ratcheting. To imagine it, remember that copper and iron have different coefficients of thermal expansion (copper moresor than iron) on top of which the copper becomes hotter than the iron. So the copper within the core expans more than the associated core. We can imagine that perhaps at one single location (center of the core) there is not relative motion between copper and core but on both sides the copper grows outward relative to core. Likewise during contraction one location has not relative motion and the rest have relative motion. If the location of no relative motion is different during expansion and contraction then at the end of the process we have movement (or at least stress). The swaging pattern of the bars has a profound effect on this phenomenon.

The crack shown here is directly above a bar. It looks to me as if the bar directly under the crack has migrated more toward this end over repeated cycle. Now when the rotor heats, there is a relative push upwards by this bar and downwards by it's two neighbors.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

If you do a runout check it might shed some light on my "guess".

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

I mean of course runout which indicating in the axial direciton on the face (vs rim) of the endring. Would probably require removing that fan to get access.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[rockman7892]

This crack was found during a normal inspection. There really wasn't any swinging stator current that was noticed and when I asked production if there was any unusual current increase they said it appeared to be drawing a higher current lately but its hard to say exactly because this motor is on a fan controlled by dampers and obviously damper position plays a role in the current changing. Would there be such a noticiable current swing with a crack like this.

Electricpete it sound like what you are alluding to is the fact that this damage came about as a result of excess heating of the rotor possibly due to excess starting or starting to frequently.

 

[electricpete]

As well as I can remember, the fan usually has a surface that lays flat against that endring to assist heat transfer from end ring to fan, different than what is shown in your photo. Is the photo shown the as-found configuration? Does this look normal to you guys?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

Electricpete it sound like what you are alluding to is the fact that this damage came about as a result of excess heating of the rotor possibly due to excess starting or starting to frequently

Yes.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[rockman7892]

electricpete

yes this was the as-found configuration.

 

[gsjhand]

Few years ago I came across same kind of problem on 1800 KW, 11KV motor. It happened during running. Suddenly there was huge variation in current and motor became noisier. We got that repaired but manufacturer of motor suggested that whole end ring should be replaced.

 

[lansford]

Frequent starting is a major cause of induction rotor end ring failure.
JIM

 

[ankervik]

Were any of the rotor bars found to be loose, cracked or showing signs of local heating/arcing? Is the rotor a double cage type? it is difficult to ascertain from the photos.

 

[Marke]

Has this rotor been repaired before?

Some repair shops have a practice of shortening the bar overhang when doing rotor repairs and this can cause this tye of problem.
Look for transients during start, i.e. open transition star delta (wye delta) starter etc.

High inertia loads require special rotors with a high thermal inertia. It is possible to have rotor damage from too high a load inertia.

Best regards,
Mark

Mark Empson
L M Photonics Ltd

 

[oftenlost]

This looks like a GE custom 8000 frame rotor. you say it is on a fan. how long does it take to get up to speed? Is there a chance that this machine is an extended start machine which could have been energized repeatedly? If it is a GE, the endring is not cast but milled and tig or mig welded before final machining.

 

[rockman7892]

I'll try to answer some of the questions above but I'm off site and dont have all the information in front of me. I can get more later.


I belive the motor is a single cage type. I dont belive any of the rotor bars were damaged or loose.

From what I am told this motor was re-wound some time ago in order to increase the hp of the motor. I'm not exactly sure how it was re-wound and am only told that it was re-wound at some point.

The few times I have seen this motor start, I did not notice any starting transients. How would this crack cause such transients?

I'm not sure what the inertia of the fan is. I can pull the fan data when I get back and look for this information. If my memory serves correct I do believe it had somewhat of a high inertia.

This motor is a GE 8000 custom fram motor.

The motor does have an extended starting time. From the few times I've seen it start I noticed that the acceleration time is around 32 seconds. When looking in the multiline 269 relay I noticed that a custom curve was programmed for this relay to account for this long starting time. I want to find motor datasheet to see where the rotor damage curve falls in relation to this curve.

There was an incident recently where we had a hard time getting this motor started and there were several repeated attemps to start this motor.

 

[electricpete]

There was an incident recently where we had a hard time getting this motor started and there were several repeated attemps to start this motor.

It sounds like a safe bet that it was related. Either the degradation had begun and was degrading starting torque, or else the event of excess start attempts/duration contributed to the degradation.

30 second start is very long for most of the DOL equipment I work with (severe start).

Is this a DOL start or soft start or vfd?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[rockman7892]

This motor is a DOL start. It is a critical process fan so once it starts it usually run continusously and is never shutdown. It is only usually only shutdown bi-annally for shutdowns or when the motor trips on overload which rarely happens. So this motor is not started very often but like I mentioned we had some troulbe getting it started last time it was down.

How would this crack degrade the starting torque as you mentioned and lead to starting difficulties?

 

[ankervik]

This article might answer some of your questions.

 

[electricpete]

Yes, broken bars and rings will make the motor start slower. The relevant quote from ankervik's link would be:

Second, reduction in full current flow throughout the cage will reduce motor output torque as well. The motor probably won't stall under normal load, but may be unable to accelerate, or take much longer than usual to start (which can hasten further deterioration by adding to rotor heating during each start).

Here's my two cents on the subject.

As a first approximation, we consider it to resemble an increase in rotor resistance. If you look in a textbook at varying rotor resistance for wound rotor motors, you'll see the effect of changing rotor resistance on torque speed characteristic: it slides the maximum torque toward lower speeds and decreases torque performance at all speeds above the max-torque speed and increases torque at all speeds below max-torque speed.

But broken bar is a little more than an increase in rotor resistance. It creates an unbalance. We can model the unbalance as the sum of forward rotating field and backward rotating field. The backward rotating field draws current and causes voltage drop with creating useful work and in fact can create negative torque (assuming the rotor unbalance creates a stator current unbalance so we have backward rotating field on both) which decreases the net positive torque.

In the olden days it was one of the common things to look for when looking for rotor bar problems: increased starting time and oscillation on the current meter. Now we have more sophisticated predictive maintenance tools and we don't rely on increased starting time as a primary indication of broken bar or ring.

You mentioned 30 second DOL start and infrequent starting. Does the thermal damage curve really allow you to go that long?

Here's my frame of reference. We have 30 motors > 2500hp, 70 motors 300 – 1250hp, and around 1000 motors 5 – 250 hp. None of them takes 30 seconds for DOL start so by my standards it is a long start. Of course large motors are custom-designed, so in theory the rotor is matched to the task. We only have two families of motors that take more than 5 seconds to start. We have one 250 hp that starts in around 17 – 22 seconds. It is a standard motor, operated very close to manufacturer's limit. We have one 9000 hp that starts around 22 seconds (has to accelerate massive flywheel). That particular rotor has a very thick copper end ring (lots of thermal inertia as well as reduced electrical resistance) reinforced by a steel end ring for strength.

I believe for large motors this long type of start will require a very unusual rotor such as the one I mentioned. Your copper end ring doesn't look large relative to the size of the bars. There is no reinforcement.

My gut feel is that your motor is certainly underdesigned for the application as others suggested. The application includes: inertia to be accelerated, load torque during start, voltage at motor terminals during start, and frequency of starting.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.