I'd like to measure the force that a human leg impacts on a knee brace during normal gait. I have a problem in finding a suitable sensor for the job…
The leg (and hence the brace) are not flat at their contact surface and it seems FSRs change their response dramatically when bent… and load cells seem quite bulky and also have the same restriction.
I looked at more exotic solutions, e.g. Creating my own sensor out of Velostat (http://www.robotshop.com/uk/pressure-sensitive-conductive-sheet-velostat-linqstat.html) but my concern is it won't be accurate or reliable.
Can strain sensors give me low creep and good repeatability?
Best Answer
Use a strain gage.
Super-glue a full wheatstone bridge onto the structure. Amplify it by 100 to 1000x and your good to go.
Calibration is easy with a precision shunt resistor (to get strain). Then use some maths to estimate load.
It's a bit pricey for quality gages, and you will still need to acquire the voltages. But once it's in place, it will last a long time.
If you design it as a in-series module, then you could directly calibrate it with dead-weights or a press, and it's even easier (I think that it's still a good idea to have a shunt resistor so you can detect if the gages are drifting or damaged).
Vishay has some good app notes on the subject.