High speed sealing application

[sean01]

Hi all,

The company I work for has been approached to re-engineer a test bed for a helicoptor turbine fuid coupling. The test bed needs to be capable of rotating each side of the coupling at two different speeds (>25000 r.pm.) independently.

The reason for the re-engineering is that previously the bearings the component were mounted in were oil lubricated, due to the high operating speeds - this meant the shaft they were supporting needed to be sealed. With operation of the rig at these speeds the seals inevitably let some oil past - a situation that the operating company now finds unacceptable.

Is there any way of sealing a shaft at these speeds - besides traditional brush seals ? The current design features brush seals - which are now viewed by the operators as unacceptable - their design requirement is that the seals should have a life of >500 hours continuous use !!!

Any help would be gratefully received !!!

sean

 

[JJPellin]

A mechanical seal could be used. Either a convention contacting face seal or a gas seal would work. Sundyne gearboxes have output shafts at the sort of speeds you are talking about. They use a specially designed contacting face seal to seal the oil into the gearbox. We are looking into designing a nitrogen purged gas seal for some of these where the gearbox seal is the limitation to run length. Depending on the size, a gas seal might be very expesive. And it would require a support system for the nitrogen. A standard mechanical seal would be less expensive and probably require no external support system. Is the seal submerged or just splashed with oil? How large is the shaft? Do you have enough axial space for a typical mechanical seal?

 

[EdDanzer]

The difficulty you face in this project is the reality of physics. In order to seal liquid or air tight, two surfaces must touch. The trick is to balance the contact pressure so that the P/V of the two contacting materials is not exceeded. P/V is the contact pressure relative to the surface velocity of the two surfaces. There is very little accurate P/V data available for dry operation, which is what happens at the sealing point. Other problems are dissipating the heat developed at the contact point, and controlling heat created size change. Different materials have different rates of expansion, as a system goes from cold to hot and back to cold the contact pressure can change dramatically, causing failure.

 

[JJPellin]

I may be making this more complicated that it needs to be. We have many compressors that have oil lubricated bearings. Some of them turn at speeds comparable to what you are dealing with. The overwhelmingly most common seal for these bearing housings is a simple aluminum labyrinth seal. In some extreme cases when we had oil leaks that were chronic, we would go further. A rotating defector on the inside of the laby helps to shield it from oil jetting out of the end of the bearing. In only a couple of instances, we drilled a port and injected an air purge into the center of the laby to create a pressure dam that keeps the oil in. This has worked very well in machines at 5 to 10 thousand rpm with some large shafts. Depending on your shaft size, the relative velocities could be even lower than our examples. So often in our business, the simplest answer is the best. When we have tried to design more complicated seals for bearing housings, we have had more trouble and ended up going back to the old tried and true aluminum laby seal.

 

[sean01]

All,

Thanks for your suggestions - there is a good chance that we maybe nominated on this project soon - so no doubt will be in touch shortly !!

Sean

 

[bingopin]

I have seen good results using N2 purged labyrinths on ~3" shafts at 30k rpm. The application was an Atlas Copco centrifugal compressor. The setup included multiple labys to create a buffer zone where the purge was injected. Obviously, the idea is for some of the purge to leak in and some to leak out. The purge pressure was controlled by a DP regulator using a reference signal from the compressor case to keep the purge pressure at a few psig above the compressor pressure.

 

[shspucad]

helix seal