Modelling of Stable Tearing in Aircraft Structures

This report summarises the cooperative research program on stable tearing between DSTO and CEAT. The main objective was to study the conditions under which aircraft materials fracture by stable tearing and to develop a predictive capability for the process under operational conditions. The experiments on both CCT and CT specimens were to assist in validation of numerical modelling. Tear bands were successfully reproduced on CCT specimens with different specimen thickness by experiments at CEAT. The results were used to assess empirical models Schijve's and Forsyth's models, and R-curve methods as well. Stable tearing feature was successfully simulated by a commercial finite element package ZENCRACK. Due to lack of local failure criteria, ZENCRACK cannot be used to predict whether stable tearing would occur or arrest under cyclic loading. But it appears to be useful for modeling such phenomena for indicative purpose only. A new 3D numerical model was proposed using a cohesive zone approach. This model can predict features similar to stable tearing and agrees well with the published experimental data. However, more research work needs to be done. RELEASE LIMITATION Approved for public release Published by DSTO Platforms Sciences Laboratory 506 Lorimer St Fishermans Bend, Victoria 3207 Australia Telephone: (03) 9626 7000 Fax: (03) 9626 7999 © Commonwealth of Australia 2005 AR-013-350 March 2005 APPROVED FOR PUBLIC RELEASE Modelling of Stable Tearing in Aircraft Structures Executive Summary Fatigue crack growth in metallic structures remains one of the main threats to aircraft structural integrity, and is the subject of extensive research worldwide in an attempt to assess existing fatigue cracks, and predict their future growth under operational conditions. Most of this research has focussed on crack growth, which occurs at a rate of less than one micrometre for each application of loading. However, the prediction of such growth rates is often complicated by the presence of tearing fracture (“crack jumping” or “stable tearing”), in which cracks can extend by millimetres under the influence of a single high load. These tearing fracture bands are commonly seen on aircraft fracture surfaces. This tearing fracture is caused by parts of the crack front becoming unstable, and advancing rapidly, thus changing the crack front shape. This change of shape can provide increased resistance to crack extension and in many cases can cause the crack to revert to the more usual slow fatigue growth. Therefore in order to improve aircraft structural integrity and safety it is important to develop an understanding of the stable tearing process, and to have access to useful engineering models which allow analysis and, if possible, prediction of the event. The process of stable tearing is a complicated phenomenon, which occurs during fatigue loading, and at present no fatigue crack growth models address the occurrence of tearing, despite the fact that it may make up half of the fracture surface. DSTO and CEAT have been collaborating on a possible predictive capability for tearing, involving testing specimens to produce tearing, and assessing various models. This report presents the summary of the modelling and experimental work on stable tearing. Experimental work at DSTO and CEAT has proven that the simple tearing models formulated by Schijve and Forsyth are good predictors of the stress intensity factor (SIF), which causes tearing. But both models are only valid for post fracture analysis as they rely on measurements taken from the fracture surface. R-curve analysis is an alternative, but relies on performing time-consuming R-curve tests using material of the appropriate thickness. The use of simple crack-front stress intensity analysis offers another possibility, and a simple approach based on elastic FEA is assessed against the experimental results. DSTO-TR-1657