Repeated Disaster Bushing Failure

[rota1]

Hello everyone
We have a between bearings Union centrifugal pump in a high temperature (700deg F) high viscosity service.After shop overhaul the disaster bushing failed during startup.The bushing was fused against the seal sleeve.This disaster bush is made of bronze and has a diametral clearance of 0.035 thou.You would think that with heat, the disaster bushing would expand and rub the seal flange, but here we are seeing the bushing fusing onto the sleeve.Can hydraulic instability cause the shaft to deflect 0.017 thou causing the shaft to rub against the bush?
How critical is the ring clearances for a pump in this type of service in relation to hydraulic stability?
Thanks in advance

 

[MikeHalloran]

I'd bet that _somebody_ in the overhaul shop knows exactly what they did that they weren't supposed to do, or what they didn't do that they should have.



Mike Halloran
Pembroke Pines, FL, USA

 

[rota1]

Thanks Mike ! for the quick response, but could you elaborate some more!
Concentricity and perpendicularity checks was carrried out during the overhaul and was found to be 0.003 thou.

 

[MikeHalloran]

I claim no special expertise, and no familiarity with your particular pump.

The word "Repeated" suggests that the pump was overhauled because the "disaster bushing" (whatever that is) died ... and it died again on startup.

Did anyone figure out what really failed in the first place?

I.e., from here it smells like your shop replaced all the parts they had spares for, and called it an overhaul ... but maybe it needed a part you didn't have. The maintenance guy who expresses the worst attitude about having to r&r the pump again may know or suspect which part that is.





Mike Halloran
Pembroke Pines, FL, USA

 

[rzrbk]

Just some starting points.

Is most of the diametrical clearance above the shaft? I would expect the shaft to lift off the bearing sleeve on start-up, but don't know by how much.

Were the bearing brackets removed and replaced during the overhaul, possibly changing the elevation between bearing and seal register fit?

 

[Artisi]

It seems fairly obvious that there is insufficiant clearance betwen the 2 components, this is irrespective of what the manual or the experts have to say. Whether the problem is thermal expansion or shaft deflection it makes no difference to the outcome ie,failure. Open the clearances- it might only need a couple of thou. to overcome the problem but only you are in the position to make this judgement as you have the failed components etc assertain what is reasonable.

 

[rota1]

Again thanks to everyone for commenting.
Mike Halloran :The component that I am referring to (Disaster bushing) is also referred to as the throttle bushing. No we didn't figure out what went wrong in the first place, these two failures took place within seconds in one instance, and about a couple of minutes in the other.

rzrbk :This pump has ball bearings,the brackets were adjusted and dowelled into position during concentricity checks during assembly.

Artisi :API states that the throttle bush clearance should be 0.025 thou, we are presently at 0.035thou, which is a generous clearance.I don't think we would want to increase this clearance anymore as we would be defeating its purpose, if we haven't already.
What about the point of the thermal expansion of bronze?, it is confusing to us that the throttle bushing would fuse on the sleeve and not on the seal flange.

 

[rzrbk]

Many throttle bushings are pinned to the seal gland to prevent rotation, meaning it would fuse against the shaft before breaking loose and then rubbing against the gland. Are there signs of rubbing between the bushing and the seal gland in one spot or all the way around?

Making this bushing out of carbon would allow some contact during running, but would not really solve your problem.

Is this seal the same type as was running in the pump before the overhaul? (has satisfactory history before the recent problems)

 

[ApC2Kp]

rota1,

The 700F sevice would make me wonder about piping expansion and extra loads on the pump after heating up system. Check that the piping supports were re-installed correctly. Casing distortion from excessive piping loads usually first cause bearing failures, but in your situation it is an immediate failure. Some field techs will put dial indicators on casing or piping connections to show any casing movements that could cause loss of clearances inside of pump. All the concentricity checks on the work bench would mean nothing if the piping adds several thousand pounds of loads on the casing. If possible, check the shaft clearance by hand rotation for any rubbing, before starting pump but after filling piping to heat up the system.

Which brings up another point when you ask about ring clearance and hydraulic stability. Are you sure that fluid has filled the seal cavity? If the bushing is dry, then any rubbing could quickly result in enough heat to fuse parts.

 

[JJPellin]

I have seen this problem when there was a machining error with the seal gland or seal chamber. You mentioned that the face and bore run-out were good. Did you verify that the fit between the pilot on the seal gland and the seal chamber bore are within specifications? This can be difficult to measure in the field because the shaft is in the way. We made up a special bracket so we can use an inside micrometer around the shaft to measure the seal chamber bore. If this fit was excessively loose, the seal would drop to the bottom of the sloppy fit and could rub hard on the top of the shaft. I have seen this occur for two reasons. In one case, the seal chamber had been bored out oversized during a previous repair and this modification was not documented. In the other case, the seal manufacturer had made a error when converted old paper drawings to CADD and had changed a dimension. In this second case, the error was on a machining drawing for the gland but did not show up on the seal assembly drawing. Verify the pilot fit and make sure it is within the specification called out by the seal manufacture (perhaps 0.0015" to 0.002" loose).

One other possibility occurs to me. If the shaft was bent during the first failure or first repair, there could be excessive shaft run-out. This would not show up on the seal chamber concentricity check since the dial indicator would still spin a true circle on the bearing centers. Did you check shaft run-out in the seal area?


Johnny Pellin