Compression helical spring

[rafaellicursi]

Hi everyone!

I work with pressure vessels inspection and it includes pressure safety valves (PSV) calibration.
We have a new standard that recommend us to perform two tests on the springs that are part of those PSV. They are:
- Check if the spring axis is 90º +- 2º of the end plane;
- roll the spring over a plane surface and check if every coil is in contact with the surface for a complete tour, i.e. it must be a perfect cylinder.

The service is considered static. The spring is preloaded between 10% to 80% of the deflection range.

Those restrictions are promoting a rejection rate about 40% of the tested springs.

The question is: Are those tests too strict to this aplication?

Best Regards!!

Rafael Licursi

 

[MikeHalloran]

Only the PSV manufacturer could know with any certainty just how crooked a spring their design can tolerate, so I'm hoping that your new inspection criteria came from them.

Be sure to inspect the new replacement springs to the same standard, and to be religious about rejecting the 'bad' ones. That may cause some readjustment of the standard.



Mike Halloran
Pembroke Pines, FL, USA

 

[israelkk]

Standard commercial tolerance is 90º +/- 3º asking for tighter tolerances require sorting and raise the cost. The requirements you subject the spring indicate a bad design of the valve. A good design of the valve should use means inside the valve that doesn't care for such requirements. For example transferring the spring force via a small steel ball which assures always an exact direction of the spring force with "almost" no matter how the perpendicularity deviate from 90º +/- 3º. The second requirement indicate that there is a concern that the spring will buckle sideways and rub on the housing walls thereby introducing friction. The good design should use a slightly tapered spring which avoids the buckle risk or use a shorter spring compared to the spring length.

 

[israelkk]

Sorry, it should be "shorter spring compared to the spring diameter".

 

[Ripz]

These checks seem more necessary for spring manufacturing, than Relief Valve certification.

I've never heard of a test like this done on PRV calibration and API-576 makes no reference to it.