Abstract
The acronym IBESS stands for "Integrale Bruchmechanische Ermittlung der Schwingfestigkeit von Schweisverbindungen" which, translated from German, means "integral fracture mechanics determination of the fatigue strength of welds". the method introduced in this study is the outcome of a German Research cluster in which eight partners were involved. A list of them is found at the end this study. The IBESS method is characterized by a number of partially novel aspects and elements of fracture mechanics applied to the evaluation of fatigue stength of welds. The most important ones are: (a) Determination of fatigue crack propagation for mechanically/physically short and long cracks. (b) Determination of an elastic-plastic crack driving force for the treatment of mechanically short cracks. To that purpose an analytical expression for the cyclic J-integral was developed and validated against finite element results. (c) The gradual build-up of the crack closure phenomenon is determined by using cyclic R-curves which describe the crack size dependency of the fatigue crack propagation threshold in the physically short crack growth regime. (d) A physically meaningful initial crack size is defined for total life consideration. It is based on a two-criteria approach. Based on a cyclic R-curve analysis, the crack size at crack arrest is determined as a lower bound. If, however, a pre-existing crack-like defect is larger than this, its dimensions define the initial crack size. (e) Multiple crack propagation at the weld toe is considered. (f) In conjunction with this, the variation of the weld toe geometry is considered in a stochastic model. (g) As a result, both the fatigue limit (defined for 107 loading cycles) and the finite life (high cycle) fatigue S-N curve are obtained statistically. (h) At various analysis steps, parametric equations have been developed which allow for analytical calculations instead of complete stochastic analyses based on finite elements which are unrealistic even at present. (i) The method has been validated with a large number of S-N curves including two materials, three weldment types with two geometries, each referring to differnt manufacturing technologies and the as-welded and stressrelieved state. (j) Althrough not finally solved, an extended discussion is provided on the issue of welding residual stresses including their redistribution under cyclic loading. (k) A number of simplifications is proposed at lower analyses levels which, however, partly lack complete validation by now.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have