The influence of prior thermal and mechanical loading on fracture

• Main Author: Sisan, Ali Mirzaee
• Published: University of Bristol 2005
• Subjects: 620.176
• Online Access: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.424411

A research programme was carried out to explore the influence of prior thermal and mechanical loading on brittle and ductile fracture behaviour. The programme concentrated on the behaviour of type 316H austenitic stainless steel and A533B and A508 low alloy ferritic steels. Finite element analyses were employed to explore different methods of introducing well-defined residual stress fields in laboratory specimens. Two different approaches were followed to introduce tensile residual stress fields. Mechanical loading introduced a residual stress field as a result of local plastic deformation in specimens. Thermo-mechanical loading introduced a residual stress field into specimens as a consequence of a severe thermal gradient. A novelty of this study was to examine the effect of the sequence of cracking and load history (prior mechanical and thermo-mechanical loading). A comprehensive range of experiments are presented. The fracture behaviour of cracks in the presence of an initial tensile or compressive residual stress at room and low temperature was investigated. Different methods of residual stress measurement including incremental centre hole drilling, neutron diffraction and deep hole drilling were used to validate results from finite element analyses. Another objective of the project was to study global and local approaches to the prediction of brittle failure when residual stresses are present. The challenge was to predict not only the effect of residual stress on brittle fracture but also the scatter in fracture toughness data. A local approach based on the Beremin model and a global analysis based on a modified l-Integral were implemented in this study. The transferability of the Weibull parameters in Beremin model between cracked speCImens with different constraint, test temperatures and also from unstrained specimens to specimens with residual stresses is explored. The conventional assessment code, R6, together with advanced numerical methods was also used to assess the experimental data. Finally the conservatism of R6 assessment code in failure prediction as result of interaction of primary and secondary load is discussed for a set of experimental test data.