Asymmetrical Loading of Helical Spring

[enirwin]

I have two discs clamped together by a helical spring. Pressure is applied to a semi-circular area between the discs. As a result, the separating force is not applied uniformly around the circumference of the spring. Is there a way to calculate the pressure at which the discs will begin to separate?

Thanks.

 

[israelkk]

When you say Pressure you mean Gas of liquid pressure?

 

[enirwin]

I'm not sure it matters, but you can assume that it's liquid pressure sealed between the plates until they begin to separate. Separation results in leakage.

What I'm really after is how to predict load and deflection for a helical spring that is asymmetrically loaded.

Thanks.

 

[CoryPad]

You need another object in addition to the spring to help clamp the pieces, yes? Are the discs and spring contained within a housing? Are the disc contact surfaces planar?

 

[israelkk]

My question regarding the pressure was exactly from the reasons CoryPad mentioned. If the two discs are guided similar to pistons in a cylinder then the side force which create moment on the discs in addition to force will create friction between the discs and the housing. The shape (length and diameter) of the discs dictates the reaction forces between the discs and the housing and thereby dictates the friction forces. Without knowing the detail design there is no way to say what will be the pressure to separate the discs.

 

[enirwin]

Let me try this another way. Please see the attached sketch showing a helical spring between two plates. One plate is rigidly fixed and the other plate is hinged. Due to the constraints on the motion of the plates, one side of the spring will deflect more than the other. Is there a was to calculate the spring load?

Thanks.

 

[israelkk]

I would sugest using FEA analysis of such use of the spring. However, be carefull that the spring may buckle.

 

[btrueblood]

There is a weak treatment of the theoretical moment couple required for keeping the ends of a coil spring planar during deflection, in Wahl's book Mechanical Springs. I would not buy the book just for his derivation, however, since he points out that the theory and real-world test results can vary by "25 to 30%", and my own experience suggests that he is overstating his results for higher deflections and higher spring L/D's. It depends so much on hard-to-model details such as contact between the end coils, that (as israelkk suggests) an FEA is probably the best way to predict it. I'd say it's probably cheaper (than FEA) to make up a sample run of springs and test them - you'll get real-world variations included for free that way, rather than having to model them.

 

[israelkk]

Here is an article I have found

http://www.springerlink.com/content/v3m7v7k8k0620k85/fulltext.pdf

 

[enirwin]

Thank you all for your help. I can perform a test, and have already done so, but I was trying to determine if there was a proven way to address this problem analytically for future use. It appears that FEA is probably the best analytical approach. Thanks again.