Wheatstone bridge - strain gauge layout

Hi,

I want to use a set of strain gauges to measure the tension in a metal link, although I've used strain gauges successfully to measure bending, I've not used them for pure tension (i.e. stretching only) and am unsure of the layout.

Consider this layout (classic wheatstone bridge)...

When I was measuring bending, I had one gauge in tension (R1), one in compression (R4), and the other two not flexing at all, they simply provided temperature compensation.

To measure pure tension, looking at the circuit above, if all 4 resistors (strain gauges) are under equal tension the value V0 will remain unchanged (I think) similarly if R1 and R4 are under equal tension, and R2 and R3 are just for temperature compensation, then V0 will also remain unchanged. So I think what I need to lay the strain gauges out so that R1 and R3 are in tension, and R2 and R4 are left unstressed (temperature compensation only) is this correct? If not, what is the correct layout

Thanks

If R4 and R3 are normal resistors and R1 and R2 are strain gauges, then it is still compensated for temperature changes of the strain gauges.
Put one strain gauge in the length of the metal link, and the other one 90 degrees rotated.
It’s not perfect, but at least you can get some data.

With bending and torsion, one strain gauge compresses and the other one stretches. That is the best situation. If you can not add a load cell, and want to measure something in an existing device, then those two in 90 degrees can do the job. I don’t know an other way.

I understood that to get accurate temperature compensation I should use 4 identical strain gauges, even if two are mounted cross-wise and don't actually get stretched or compressed. A pair of resistors will react differently to temperature changes than a pair of strain gauges.

Put one strain gauge in the length of the metal link, and the other one 90 degrees rotated.
It's not perfect, but at least you can get some data.

This is what is described as a 'quarter bridge' isn't it?

If I mount R1 and R3 in tension, and R2 and R4 are mounted cross-wise to give temperature compensation only, then won't I have a half-bridge, which should give me twice the accuracy?

thanks

Yes, I think so. Twice the accuracy

If R3 and R4 are identical resistors at the same temperature, it should also work. They would be to divide the voltage in half.

Its normal to arrange the sensors at 90 degrees to measure an axial strain field, they need to be tightly
thermally coupled to avoid temperature differences swamping the measurement, so right next to
each other is good.

This guide seems handy: http://www.omega.co.uk/faq/pressure/pdf/positioning.pdf