A Reliability Engineering Framework for Evaluating Creep Limited Welded Structures in Power and Process Plants

摘要

The steady state creep response as well as the total creep life of a material is related to the operational component temperature through, respectively, the exponential and inverse exponential relationships. Creep is a viscoelastic property exhibited by all materials under sustained loading at an elevated temperature. The primary variables affecting the creep process are the values of the principal stresses as well as the operational component temperature. The kinetics of the creep process is controlled by the topological dynamics of the various defects structures existing in the material. The overall creep deformation process is approximately determined by a linear sum of the deformation rates associated with all the defect structures. Voiding at grain boundaries and inter-phase interfaces starts early in the creep process. Eventually, the nucleated voids grow link up, and form micro-cracks. The growth of the micro-cracks results in the emergence of extensive macro cracking in the material leading to eventual failure. Fabrication of high temperature power & process plant components in volves extensive use of welding as the fabrication process of choice. Naturally, issues related to the creep life of weldments have to be seriously addressed for safe & continual operation of the welded plant components. Unfortunately, a typical weldment in an engineering structure is a zone of complex micro structural gradation comprising of a number of distinct sub-zones with distinct meso-scale and micro-scale morphology of the phases and (even) chemistry and its creep life prediction presents considerable challenges. The present paper presents a stochastic algorithm, which can be used for developing experimental creep-cavitation intensity versus residual life correlations for welded structures. Apart from estimates of the residual life in a mean field sense, the model can be used for predicting the reliability of the plant component in a rigorous probabilistic setting.