Thermally Sensitive Salient Pole Rotor 2

[fflinders]

We have had a 12MW alternator rotor recently rewound. 4 pole, Salient type, brushless excitation system. After installation it was found to have vibration which whist cold was of relativly low level but increases as the machine is loaded. The vibration level appears to increase as the temperature increases. However the effect is irrevesible in terms of it does not imediately decrease as the machine is unloaded. But once taken off line and allowed to cool completely and then restarted it will again run with low vibration while the load is low. Therefore the rotor would appear to have thermal sensitivity. I know this can be a problem with Cylindrical pole rotors due to differential expansion and bowing of the rotor causing unbalance. However can this happen with a salient pole rotor?
I have not been able to find any reference to thermally sensitive salient pole rotors.

 

[ScottyUK]

This has the sound of a thermal bend, which is usually a result of a shorted turn on the field causing abnormal heating of that portion of the rotor. The area of the rotor affected by the shorted turn expands more than the unfaulted parts, and the expansion causes the rotor to develop a slight bow. On a cylindrical rotor this type of fault certainly shows up as increased vibration. I guess the principle will apply to salient pole machines too, although they are typically of a shorter fatter design than a cylindrical machine and so I would assume the effect is less pronounced.

Can you determine whether the effect is MW dependant or MVA dependant? If the latter it is almost certainly a field problem similar to that described above.

----------------------------------

I don't suffer from insanity. I enjoy it...

 

[fflinders]

Thanks ScottyUK,

I will try to organise a test to check for relationship between excitation and vibration at constant MW and also constant MW and variable excitation. I agree that you would think that salient pole would be less effected, unless there was multiple shorted turns which is not likley.

Another effect that seems to occur is that the vibration level tends to ratchet up. That is it increases and then decreases but is on average rising as load is increased. I am thinking maybe the coils are getting restricted from thermal expansion (Axial)and moving in a stick slip manner. Maybe all coils are not expanding the same thus causing an unbalance. What do you think?

Thanks for the advice!

 

[ScottyUK]

Hi fflinders,

My experience of salient pole machines is limited, so I'm trying to extrapolate my knowledge of cylindrical turbo machines to (relatively) low speed salient types. Perhaps that is a little risky, but I want to be up front about it.

Here's another thought which might not apply in your case: when the vibration levels step up, do the steps roughly coincide with OLTC operation on the GSU transformer?

If you don't have one already, get a current transducer fitted to the field of the machine and take it back to a spare analogue channel of your control system if you have one. Temporarily sacrifice another instrument if you need to - just don't pick anything too critical to 'borrow'! I think the trend might prove very useful.

What vibration analysis equipment do you have? If it is Bently Nevada or CSI - the two I'm familiar with - or something similar, see if the vibration equipment supplier can offer a vibration signature analysis. Bently have helped us out a few times by identifying likely causes. Often when we check their suggested trouble spots we find the problem. Most of the major turbine & generator manufacturers have departments which specialise in vibration analysis too. We are supported by Machine Dynamics from Parsons at Heaton on Tyneside. These guys do all the post-outage machine balancing, and have enough experience to look at a vibration trend and take a fair guess at the origin.

I'm not 100% behind the theory of axial stiction of the coils in the slots, mainly because the coils are presumably multi-turn and fairly short. The forces generated by thermal expansion are huge, and constraining them in a stick-slip manner seems improbable to my turbo-alternator mind. But maybe that's my ignorance - you know much more about your machine than I. Does the physical arrangement of the coils lend itself to this failure mechanism?

----------------------------------

I don't suffer from insanity. I enjoy it...

 

[powerjunx]

flinders,
I had closely similar experience with yours. From our customer observation; As the load increase, vibration proportionally increases - it was found during condition monitoring activities.
Since the vibration level is on the critical values, we resort to dismantle the rotor assembly and conduct a dynamic balancing check. Upon doing it, we found uneven eccentricity of poles. We opted to correct the non-uniformity through machining a certain part of pole. Then we had final balancing. Fortunately, the abnormalities due to vibration was corrected.
Evidently this time, alternator is still in service without reported abnormality as before.

 

[fflinders]

ScottyUK,

Thats a good idea. I am going to log the field current under MVAR regulation and constant power after we increase the load to full power. If there are temperature sensitive turn shorts then we should see the excitation trying to compensate for this as they occur. At the same time I will log the vibration level. This is probably the only way to effectivly test this since these shorts may not show up when the machine is offline and stationary.

If no shorts are found, the next test will be the constant excitation and constant power tests to establish if the vibration is a function of field heating or shaft torque.

If it is due to field temperature then it would seem there must be some issue with thermal growth of the coils causing a balance problem. I am wondering if we have got one coil growing in one direction and an opposite one growing in the other (axial growth). The resulting unbalance might cause the vibration which is seen. The length of the coils is 1850mm, so the thermal growth for 100C would be approx 3mm. This would give approx 1kg of unbalanced mass on opposite sides at either end. (Causing a moment which might bend the shaft into an S)

Do you think my reasoning is sound?

This unit is connected directly to an 11kV system. There are no tap changing tranformers on site. The site is fed at 11kV from the grid.

The vibration equipment is CSI. As you say it would be worth while consulting with them to obtain the possible cause of the data which has been measured.

I will keep you posted on my findings.

 

[fflinders]

fbcybil,

Thanks for your comments. Yes at some point if this problem cannot be resolved the rotor will need to be pulled again, and as you say, it should go for dynamic balancing. From your comments on ecentricity it will be wise to check the run-out of the shaft surfaces and the poles shoes to check for this plus any permanent bows in the shaft.

Thanks