Experimentation, Validation, and Uncertainty Analysis for Engineers, 4th Edition

APPENDIX EASYMMETRIC SYSTEMATIC UNCERTAINTIES

In some experimental situations, there are significant systematic errors that displace the experimental result in a specific direction away from the true result. These asymmetric systematic errors may vary as the experimental result varies, but they are fixed for a given value of the result. Moffat 2 calls this type of error “variable but deterministic.”

Transducer installation systematic errors discussed in Chapter 6 are usually of this asymmetric type. They tend to cause the transducer output to be consistently either higher or lower than the quantity that is being measured. If the effects of these asymmetric systematic errors are significant, then they must be estimated and included in the determination of the systematic uncertainty of the result. Note that in some cases the effect of these asymmetric errors may not be large enough to justify the effort of dealing with them rigorously. In these cases, an effective symmetric uncertainty limit is used. The methodology for handling asymmetric systematic uncertainties is given below for TSM propagation in Section E-1 and for MCM propagation in Section E-2 and is then illustrated in the examples that follow.

In many cases an analytical expression can be formulated for these asymmetric systematic errors, and this expression can be included in the data reduction equation [as in the case of conduction and radiation losses in Section 6-1.2, Eq. (6.5)]. This technique will usually allow ...