Strain gauge circuit for pressure vessel

[bschroet]

I want to measure the pressure inside of a small thin-walled cylindrical pressure vessel with strain gauges. It's a battery can, and during safety tests it heats up, so I want to lay out my Wheatstone bridge circuit such that thermal effects can be canceled. The only way I can think of doing this is to assume Sx=2Sy (hoop stress = 2 x longitudinal stress), and that gets me a relation between the strain in the X and Y directions as a function of E and Poisson's.

Does anyone know if this is standard practice when measuring the stress of a pressure vessel, or would it be advised just to use 1 strain gauge and a theromocouple reading to back out the thermally induced strain?

 

[TheTick]

A long time ago, I was shown that the way to lay out strain gages was to have one mounted to measure strain, and another loosely mounted to the same surface such that it responds to temperature and not to strain. They should be placed such that temperature effects cancel (common node, opposite sides of bridge).

 

[BobM3]

It's been awhile but I think if you can mount the 4 gages of the bridge close to each other (so they see the same temperature) then the temperature effects nearly cancel out.

 

[amorrison]

Try

http://zone.ni.com/devzone/cda/tut/p/id/3432

and/or Google strain gage temperature compensation

 

[rb1957]

sort of like temperature compensated rossettes ?

 

[Philrock]

Contact Vishay Measurements Group technical support.