Shunt Calibration of a Strain Gauge

El Moriana

I've recently been researching how to measure strain using strain gauges. I came upon the subject of shunt calibration and hit a mental roadblock.

From what I understand, shunt calibration simulates a strain across a strain gauge using a comparatively large resistor placed in parallel with said strain gauge. This virtual strain can be worked out rather simply and the readout obtained from the shunt load (i.e. the value output from the measuring system) can be correlated to the strain.

My problem is that, from what I can see, shunt calibration is done with one resistor. How does one account for the deviance of the strain gauge through a range of possible loads if one only has one calibration point?

Studiot

You didn't say where you had been reading, but I expect the example was diagramatic.
There is nothing stopping you having many (switched) parallel calibration resistors and in fact that is what commercial systems do.

El Moriana

Sorry about prior vagueness.

The system I have my hands on a.t.m is one provided by a university (i.e I cannot make alterations to it).

System readout is digital numeric and is uncalibrated (values attained are in the hundreds whilst I know for a fact that stress is around 10MPa).

As far as I can see, the setup is one in which the shunt resistor is soldered into place with a switch to cut it off from the system when it is not being calibrated, I also see no other connections available for parallel resistors.

Is there a way to do calibration of a range with only one shunt resistor? ... though I suspect the accuracy of such a method would be doubtable.