Doctoral Dissertations

Zhen Zhou

Date of Award

5-1992

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Major

Engineering Science

Major Professor

John D. Landes

Committee Members

J. A. M. Boulet, R. S. Benson, R. D. Krieg, A. Mathews

Abstract

An advanced ductile fracture mechanics methodology is presented here. This methodology takes a load versus displacement result from a laboratory specimen test to predict the load versus displacement for a structural component containing a crack-like defect. To use this methodology, three steps need to be followed. First, a test record from a laboratory experiment needs to be reduced into two pieces of information, the calibration function and the fracture toughness. The calibration function relates load, displacement and crack size. The fracture toughness relates crack driving force and crack extension in terms of a J-R curve. The second step is to transfer this information to a structural component to be analyzed. After the calibration function and the fracture toughness for the structural component are determined, the last step is to use a previously developed methodology to predict the load versus displacement behavior of the structural component. Traditionally the J-R curve has been developed according to the ASTM standard. Crack growth is usually estimated by instrumentation. This does not always work well, especially for non-side grooved specimens and polymeric materials. To ensure accurate data reduction, the normalization method for J-R curve calculation is further developed. Its accuracy in estimating crack extension is greatly improved by introducing a new deformation function, the LMN function as the calibration function. The new version of the normalization method has many advantages over the elastic unloading compliance method. It eliminates the need for the crack extension monitoring equipment which the unloading compliance requires. Therefore it is easier to apply. It gives accurate J-R curve predictions in cases where the unloading compliance fails to predict crack extension accurately. Therefore it is more reliable. It works well for both metallic and polymeric materials, whereas the unloading compliance method is only suitable to metals. Therefore the normalization method is more versatile. The normalization method has proven to be a reliable method for J-R curve development. Two specially designed procedures are developed for the determination of the three constants in the LMN function. The first procedure is based on the observation of fracture toughness behavior of many different materials. The second one incorporates the tensile properties of the material through a limit load solution. Both procedures work equally well. The values of the three constants L, M and N in the LMN function are found to be closely related to material tensile properties for compact specimens. Based on this discovery, the J-R curve could be obtained by direct integration of the LMN function with the constants coming from tensile properties. The effect of material property like yield stress and strain hardening on the ηp1 factor is investigated. It is found that hp1 factor is independent of material properties for bending type specimens such as the deeply cracked compact specimen (CT), deeply cracked single edge notched tension specimen (SENT) and single edge notched bend specimen (SENB). Whereas for pure tension type specimens like the center cracked tension (CCT), double edge notched tension (DENT) specimens, T]pi factor is dependent upon both material properties and gage lengths. These results suggest that different plastic deformation patterns exist in these pure tension specimens for different materials. The calibration function for a structural component to be analyzed can be obtained by following two different procedures developed in this thesis. One procedure can be used to transfer the calibration functions for the test specimen to those for the structural component. The other procedure is based on numerical simulation incorporating the load separation method. The transformation procedure is more suitable for the structural components with known limit load solution and geometry function. The second procedure is more general; it can be used for any complex structural component whose behavior can be represented by load versus displacement. Different geometries and materials are chosen to test this methodology and the proposed intermediate procedures. The selected materials include both metals and plastics. The geometries cover commonly used test specimens and real structural components. The predictions are compared with the experimental results for all the cases. In general, the predictions are shown to be accurate and successful. The methodology presented here has many advantages over other currently used fracture analysis methods. It gives complete description of the structural behavior. The inputs can be obtained by simple finite element analysis. It can be applied to wide range of materials and structural components. Its simplicity, its versatility and reliability will certainly be attractive to many engineering designers.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS