Optimum location of a three strain gauge rosette for measuring mixed mode stress intensity factors

This paper analyzes the errors inherent to the determination of mixed mode stress intensity factors from data obtained by using a three strain gauge rosette. The analysis shows that the errors are mainly due the third characteristic value (3/2) and its corresponding coefficients. It is also shown that the errors do not depend on the orientation angle of the rosette, the angle between the strain gauges and the material properties. The error mainly depends on its location (radius, angle), being linear in the radius. For pure mode I, an angle of 90° will completely eliminate the error due to the angle, while for pure mode II, a 0° angle will minimize it. The normalized variation of the errors with the angle at any radius is shown for different ratios of the corresponding coefficients of the third characteristic value. The analytical results are applied to a numerical example of an edge crack subjected to mixed mode loading. From the numerical example, it is recommended to use two strain gauge rosettes at the same angle, and linearly extrapolate their results, if errors less than 15% for a mixed mode field are desired.