Numerical procedures for enhancing the evaluation of elastic-plastic fracture toughness test results

J-R curves characterize the fracture toughness property for materials which fail by a ductile modeMost current J-R curve testing uses the single specimen technique. Crack size is measured indirectly by elastic compliance. Elastic unloading cycles, during the test, provide local stiffness, which are related to crack size through calibration functionsAlthough this technique has been standardized for many years, it still remains a somewhat difficult test to conduct, one that requires accurate equipment and experienced technicians. Compliance, potential drop and other methods have being used, however they may lead to large uncertainties or do not apply for some materials.

Load separation and normalization permit J-R curves to be estimated using only load versus displacement data and calibration points. Initial and final crack sizes can be measured directly from the specimen. They are used as initial calibration points for normalization methods. Two approaches to normalization were developed:

• direct normalization - the normalization function is estimated and crack size distribution is adjusted to this function;
• indirect normalization - crack size distribution is estimated and normalized load is adjusted to this function.

Two boundaries are set by the approaches. Actual crack extension will occur inside those boundaries. It will be shown in this work that the difference between the predictions from the two approachs is smaller than the scatter usually found in crack size measurement by elastic compliance. The approaches were also successfully applied where other methods failed. Polymer materials and dynamic loaded tests are good examples of those cases.