Verifying Measurement Methods
I am looking for a bit of advice regarding some analysis I am required to undertake.
The organisation I work for measures large items (600m + lenghts), incrementally using a 30m measuring device, and I have been asked to analyse the amount of variation we could expect to see over the large length, based on the individual measurements. We know the measurement error on individual measurements, but over the complete length stack-up, how does this equate. I was thinking of conducting a cap study, to understand the amouont of variation inherent in each measurement, and apply this to the overall lengths required. I don't think an MSA would give me the required results as we are not measuring the unit as one.
Any Ideas.
Propagation of Errors
This is a "propagation of errors problem". You are saying that the total length L consists of the sum of individual measurements a,b,c, etc:
L = a + b + c
The standard deviations of the individual measurements add in quadrature ... equivalently, the variance in L will be the sum of the variances of the individual measurements.
These types of calculations are commonly required by scientists to assess the effect of measurement errors within complex formulae. You would therefore find good coverage of the subject in statistics books directed at scientists and engineers.