triple offset butterfly valve leakage

[bao2ye]

Hi, Gentlemen

There is a 12 inch class 300 triple offset valve under cycle tests. This is a solid seal ring valve, different from commonly used laminate seal.

The valve performs very well with water and gas test when is assembled.

However, after 50 cycles, it starts to leak.

Then we assemble the valve, and do the same test for four times. After 50 cycles, the valve starts to leakage.

Anyone has an idea, why this happens?

 

[gerhardl]

Could be (as usual) one or several reasons alone or together. As this seems to be a repetitive occurrence (without any traceable damage to seat or sealing, or change of seat or sealing???), it would seem natural to ask how the valve is moved, and what limits the stroke. Force or mechanical limitation? Differences in accuracy in force, temperature, limitation of stroke or necessary compensation for differences of same occurring in as cycle is repeated?

Test limits, conditions and allowable and observed leakage? It is almost impossible to have a complete 'non leaking' valve, bot possible to have a leakage within limits.

 

[TOVmania]

This is usually associated with the permanent deformation of sealing elements. Time to go back to calculation.

 

[bao2ye]

Thanks Tovmania and gerhardl,

We checked all parts and did not find any wear or damage on the sealing couple. Also no yield on the parts.

Then we reassemble the valve using the same parts, the valve works again.

After that, we drive the valve for 200 cycles, it failed...

 

[al1979]

Bao2ye,

Are you opening the valve under differential pressure? Which gasket is used below seat / seal spiral wound or flat graphite sheet?
Gasket relaxation and loosening of retainer screws are known culprits.

Cheers !!!

 

[gerhardl]

Bao2ye,

See my previous post. Really all checked?

If everything is OK again when reassembling all parts 'as is' after leakage, something is different after 200 cycles and when reassembled. Bolts or tightening parts or sealing/sealing support giving tenth's of mm's or less leeway? Try to locate the leakage precisely when leaking, and control all details of this part 'under microscope' if possible.