identify the cause of the failures

[77851802]

The purpose is to identify the cause of the failures and make corrective repairs. The pumps have been in service since the late 1970's and have exhibited high vibrationa at vane pass frequency but operated with normal maintenance without failure for approx 10 years at a time. Since 2003, five of the pumps modified with new diffusers to reduce high vibrations at vane pass frequency. All had balancing device and thrust bearing failure. the modified diffusers have 9 vanes instead of 8. This is a boiler feed 5 stage pump. I did the CFD analysis of considering the cases of 8 & 9 vanes but as per that analysis hard to say anything.

 

[Artisi]

It seems that the change from 8 to 9 vane diffuser has resulted in some mismatch beween impeller outlet and diffuser inlet increasing the axial thrust and needs to be the focus of investigation. Any feed-back from the manufacturer.

 

[77851802]

No feedback from the manufacturer. Actually original OEM was differnt and then later the design ( 8 to 9 diffuser vane) revised by different manufacturer. It is a 5 satge pump & 1st stage is a double suction impeller. In the 5th stage the A gap (with 8 vane ) was .070 inch. After modification they reduced the A gap to .030 inch. Is this be a couse of failure ? This is the information i have.

 

[MikeHalloran]

It appears from here that manufacturer B has a less perfect understanding of what the pumps require than does manufacturer A.

Is there some reason you can't get them rebuilt by manufacturer A? ... and restored to their original design, or one that's evolved in the correct direction?



Mike Halloran
Pembroke Pines, FL, USA

 

[Artisi]

what was the effect on the blade pass frequency after changing to 9 vanes.
Anyway- it seems that even though you have high frequency vibration with the OEM design, the life of the pump was better and you should consider reverting to the original and look into the frequency problem again.

 

[dickon17]

Just a wild thought here, we have never actually tried this. What about cutting new key ways through the impeller hubs of the last 4 stages. This would allow you to progressively rotate/offset slightly these impellers on the shaft so that the blade passing does not occur all at the same time for all of the impellers. The blade passing frequency would increase but the amplitude should decrease. This would be a simple machining operation plus a strip and rebuild. Just a thought.

 

[77851802]

There is no problem to fix it as OEM A but I am interested in the cause of failure. This is the analysis i am doing right now. We have taken the project to find out the possible cause of balance disk failure. There are only three changes as I inderstood.
1. The second OEM reduced the A gap to .03 inch in the last stage.
2. Change in the diffuer design (8 to 9 vanes).

As per my thinking diffuser is not playing any role here. May be the 7 impeller vanes & 9 diffuser vanes combition is not right. There are some charts but i am not having right now. A Gap can play an influnce on this problem as it is reduced considerable (from .079 to .039) but I need some proof to proov this. The most of the OEM's are using the values like .040 & sometimes even less value. I have calculated the resultant thrust also. The total resultant thrust is 11.21 E04 lbs. By considering this thrust, i have calculated the counter bore diameter of balance drum. As per my calculation that diameter is 9.65. As per the drawing the diameter is 9.40. This may also be the possible cause of failure.

3. There was no taper on balance drum & balance sleeve. There is a full sucface contact.