Calculating Tension Life of a Spring for Valve Application

[araanandv2]

hi,

Our counterparts have designed a spring for a valve application.
the entire component depends on the tension of the spring
We need to calculate the tension life of a spring(i.e the duration for which the tension in a spring remains a constant)
Please provide me a solution

Thanks

 

[Gokul43201]

I don't know if there's specific studies on springs, but there are more likely to be creep tests for different materials under different fixed stresses. I would imagine the decay of the spring restoring force to be strongly correlated to the creep-related stress relaxation of the parent material.

 

[Astronuc]

Normally a spring would be used well within its 'elastic range', i.e. such that the peak local stress was well below is yield strength. Creep would only be an issue if it was in a heated environment with the operating temperature above about 0.35 of melting temp.

Fatigue may be the limiting case, and that depends on the fracture toughness and fatigue resistance of the material, as well as the largest allowable imperfection in the material.

See if you can find a Langer-O'Donnell curve (fatigue cycles to failure for a given stress level). Sometimes is called O'Donnell and Langer model (or curve). However, those are names in the nuclear industry.

Alternatively, look for S-N curve, or search google for "S-N","fatigue", e.g.
http://www.ndt-ed.org/EducationResources/CommunityCollege/Materials/Mechanical/S-NFatigue.htm

 

[Gokul43201]

Oops. I misuderstood the OP (particularly the bit in parentheses) to mean the spring was under constant tension during its entire life (i.e., no cycling loads).

 

[Astronuc]

Oops. I misuderstood the OP (particularly the bit in parentheses) to mean the spring was under constant tension during its entire life (i.e., no cycling loads).

Well it's not really clear. The actual operation needs to be discussed in more detail. Depending on the valve, it might be under constant tension with very infrequent change. If it is a control valve, then there is a good chance that there is some type of variation in load - even if it's high frequency, low amplitude. I believe Richard Hertzberg (Lehigh Univ) has looked at this phenomenon.

https://www.amazon.com/dp/0471012149/?tag=pfamazon01-20

I met him when during a presentation to our local ASM chapter. Nice guy.

http://www.lehigh.edu/~inmatsci/faculty/hertzberg/Hertzberg.htm