The primary objective of this work is to provide the theoretical computation of load-point compliance for the arc-bend/arc-support specimen and some experimental verification of these predictions. This specimen is a segment of an annular disk loaded in three-point bending with the roller supports positioned at the inside surface.

Load-line compliance was determined principally by utilizing Irwin's equation that relates the compliance rate of change with crack length to the strain energy release rate. The strain energy release rate was determined from the stress intensity solution available in the literature for this sample. Corroboration was also provided by utilizing boundary collocation, boundary integral element (BIE), and finite element (FEM) methods. Representative cases were experimentally tested to confirm theoretical and numerical results. During the tests, it was also found that duplication of idealized support and fixture conditions was extremely important with regard to accurate measurement of load-point compliance.

Two statically equivalent models were used to calculate the load-line compliance using the numerical methods. The first we call Model 1, where the load point is fixed and the roller reactions are applied at the appropriate location. The second we call Model 2, where the load is applied at the load point and the roller reaction points are fixed in the appropriate manner. Although both models are statically equivalent, significantly different load-line compliances are obtained depending upon the model chosen.

Experimental results were also obtained. Specimens were tested with machined slots to simulate a crack and other specimens tested had fatigue cracks. The measured compliances were found to agree with Model 2 results in some cases, with Model 1 in others, and with neither in still others.