Measuring a radial loaded shaft

[CarlesFEA]

Dear experts,

In order to measure radial load on a shaft, 8 strain gauges on two full Wheatstone bridge circuits have been used.

Each bridge circuit refers to a measuring plane (namely I and P).
Radial force as well as voltage signal from two planes is measured for 5 different shaft angular positions (alpha=45, 90, 135, 180 and 225deg).
When load is aligned with plane I (alpha=180deg), we get a certain gain relating measured volts on plane I and measured force (factor_I= 0.0157V/kN). No signal is measured on plane P, as it corresponds to the shaft's neutral axis.
When load is aligned with plane P (alpha=90deg), we get a certain gain relating measured volts on plane P and measured force (factor_P= -0.0178V/kN). No signal is measured on plane I, as it corresponds to the shaft's neutral axis.
When load is 45deg from both planes, we get a gain of +/-0.0097 (depending on plane and shaft position).
Is there a way to calculate the gain at any shaft angular position based on factor_I and factor_P?
Thank you very much.

Regards,

 

[rb1957]

you'd've thought that the gain should be the same in both directions, yes?

so then what's different ? I suspect you're using single channel s/g. I suspect the there is a very slight misalignment between the gauges, causing different strains being recorded. In your graphs I'm not seeing a 90 degree check ... where the I gauges are moved 90 deg and so reading what the P gauges were reading, yes??

another day in paradise, or is paradise one day closer ?

 

[CarlesFEA]

Hi rb1957,

Thanks for jumping in this thread.

In fact I was expecting to have different gains for the two planes (factor_I and factor_P) but then also have two different gains when planes were at 45deg from the radial load (sqrt(2)/2*factor_I and sqrt(2)/2*factor_P)...
Concerning your "90 degree check": do you mean when alpha=-45deg?
I do not have measures for that but I would expect the same shape as for alpha=135 but with reversed gains (factor_p=0.0095 and factor_i=-0.0096)

Gauge misalignment is certainly possible but the ratio between factor_P for alpha=90deg and alpha=45deg shouldn't be still something like sqrt(2)/2?