Chris,
My apologies for an oblique reference to the problem posed by Deep. I tried to take it off line because I did not want to take up too much white space with what in essence is a Taguchi approach to solving this problem.
I believe that an analogy for Deep’s four screws corresponds to a car with four steering wheels – one for each wheel. But in order to reduce the problem to that of an ‘ideal function,’ it is necessary to modify the way measurements are taken. It might also be further complicated by the camber of the road so that the four steering wheels are not independent.
According to Taguchi, differential measurements have low energy (negentropy) and a first step would be to refer all measurements to a ‘datum’ on a measurement jig. My approach would then be to attempt to construct a linear relationship between the displacement of the bar from the datum and the distance around the bar. If successful, then each screw would correspond to a signal factor, and the response would correspond to an ideal function (ideal component response.)
My interest in this problem arose because I think this example would provide a simple classroom exercise to demonstrate a dynamic DOE.
Another approach to this problem – an heuristic approach – might be to conduct a jig with a in-built distortion which can be relaxed by successive grinding. Of course this approach would require that there are no other sources of variation in the process as pointed out by Mike.
Regards,
Andy
.
.