Errors, Corrections and Limitations

The obvious aim of experimental stress analysis is to determine the significant stresses in a test object as accurately as necessary to assure product reliability under expected service conditions. As demonstrated in the preceding sections of this publication, the process of obtaining the principal stresses involves three basic, and sequential, steps:

1. Measurement of surface strains with a strain gage rosette.
2. Transformation of measured strains to principal strains.
3. Conversion of principal strains to principal stresses.

Each step in this procedure has its own characteristic error sources and limits of applicability; and the stress analyst must carefully consider these to avoid potentially serious errors in the resulting principal stresses.

Of first importance is that the measured strains be as free as possible of error. Strain measurements with rosettes are subject, of course, to the same errors (thermal output, transverse sensitivity, leadwire resistance effects, etc.) as those with single-element strain gages. Thus, the same controlling and/or corrective measures are required to obtain accurate data. For instance, signal attenuation due to leadwire resistance should be eliminated by shunt calibrations , or by numerically correcting the strain data for the calculated attenuation, based on the known resistances of the leadwires and strain gages.

( continued ... )

Page 14 of 19