[b]Spring Torsional Stress[/b]

[howkers]

I was asked to check the torsional stress in the wire of a spring and plugged the numbers into the standard equation from the text books:
shear stress = (8 x load x mean diameter x w-factor)/(pi x wire-diameter cubed)

I was surprised to get a result which exceeded the shear ultimate stress for the solid length and the estimated shear yield at the working length. I've done the calc in both metric and imperial units.

These springs have been in service for a number of years and we have had no experience of either breakage or permanent set. Can anyone explain this?

 

[israelkk]

They probably were designed for a preset (scargging, setremove) process which. However, without information regarding exact dimensions, forces, deflections and material it is difficult to be sure.

Are those springs for static or cyclic use?

If they are for static use, for how long they can "see" the compression load?

http://israelkk.googlepages.com/home

 

[howkers]

Valve is a check valve. Doesn't move very often.
Spring parameters are:
Free length 101.6mm
Working length (closed valve) 53.82mm
Working length (open) 30mm
Solid length 26mm
Wire diameter 2.84mm
Spring rate 2.02 N/mm

Material is a Monel alloy with UTS 1000 MPa and tensile .2% proof stress 759 MPa.
We were expecting the stresses to be below .57 of proof and .35 of UTS.

 

[israelkk]

What are the forces at oth working points?
How many active coils (or total coils) the spring has?
What is the outside diamter?

 

[howkers]

Sorry forgot the outside diameter - 37.08mm.
Not sure about the number of coils - the only drawing definition is ends closed and ground square.

Ran parameters through

http://www.efunda.com/DesignStandards/springs/calc_comp_Fstress.cfm

which gave:
load at closed position 96.5N and stress 411 MPa
load at open position 145 N and stress 616 MPa

This is about what I got by hand calc at work.

The stresses are above our estimated shear yield stress ( .577 of tensile yield). How do you assess acceptable stresses in springs? Can these continue working after an initial plastic set - is this to do with residual stresses after the first compression?

 

[israelkk]

The stresses you received are OK. However, such a spring is illegal. The stresses are too high even for a preset spring. With the preset process there is a theoretical (33% more than yield stress and practical (~20%) limit to the favourable residual stresses that can be induced into the wire. The calculated stresses in this spring are far more higher than can be achieved with a preset process.

Good design procedure is to refer minimum wire tensile properties. However, the actual tensile stress can be even 20% higher this may explain why these springs have been in service for a number of years and have had no experience of either breakage or permanent set.

Maybe, the spring manufacturer/designer made a special order for a wire with maximum properties (quite expensive).

Anyway this spring had to be designed for a prest process (and especially ordered/sorted from the maker of the spring wire to achieve the maximum wire properties).

There is one more option and it may just a good luck (that the wire properties for this spring batch were at maximum) for an improper design.

If the springs are working OK why you were asked to check them? If it is for reverse engineering then you have a problem as I explained.

http://israelkk.googlepages.com/home

 

[howkers]

The valve is supplied to us by a reputable valve manufacturer. We are back-fitting a safety justification.

Thanks for the assistance.

 

[israelkk]

Can you explain the term "back-fitting"?
Are you sure the spring material is Monel? Maybe it is a stronger material such as Inconel-x?