A ductile fracture methodology is presented here that takes the load-versus-displacement data from a precracked laboratory test specimen, such as a compact specimen, and through a series of analysis steps that will allow direct prediction of the load-versus-displacement behavior of a structural component containing a crack-like defect. This is based on a ductile fracture methodology originally presented by Ernst and Landes that was used to evaluate maximum load in a structure under complex loading. The load versus displacement from the test is analyzed by using the methodology in reverse so that it can be divided into two outputs, calibration functions, and fracture toughness. The calibration functions used here are based upon the load separation principle; the fracture toughness is given in terms of the J-R curve. The calibration functions and fracture toughness are then transferred from those relevant for the test specimen to those for the structure to be analyzed. Having done this, the methodology is used to predict the load-versus-displacement behavior of the structural component.
A critical step in the methodology is the transfer of the calibration functions and fracture toughness from those for the specimen to those for the structure. The fracture toughness transfer must be made knowing the effects of three categories on toughness; size, geometry, and thickness constraint. Presently, the data in the literature show conflicting trends for the effect of size and constraint. Therefore, a transfer based on known principles is not possible. The approach used here is to try to get a conservative value of the fracture toughness. Fortunately, for many structural components, the fracture toughness is not the controlling input, and the calibration functions have more influence on such things as maximum load prediction.
A procedure for transferring the test specimen calibration function to the structural calibration functions has been developed. Two important things must be addressed for this transfer. A functional form for the calibration function is needed for the structure. This form can be based on the J-calibration equation and is known for common geometrical shapes. For more complex geometries, this form must be determined.
In this paper, several example structural components are taken to illustrate the methodology. A compact specimen is used for the laboratory specimen geometry and six structural component models are analyzed to illustrate how well the method works and to evaluate the importance of the various steps in the method. From these examples, suggestions are made for further work that could improve the overall method.