Obtaining residual stress along weld beads by means of analytical and experimental methods

摘要

The nodal values of the residual stresses in terms of their relationship with the displacements in the FEM model were corroborated with those achieved by experimental means using the technique of blind hole stress relief. The percentage of error in the results across all cases was minimal for both the longitudinal and transversal stresses (The maximum error produced was 5.2 percent). This demonstrates the relevance of the techniques employed, both analytical and experimental. The validity of the results and the reliability of the innovative FEM computational methods used are shown by comparing the values of the residual stresses obtained using these experimental and statistical methods (high correlation and concordance). It can be concluded that an a priori value can be given to the technique before the practical repair, adjusting parameters, such as the entry heat and weld speed, allowing further definition of the process to the most suitable condition, reducing the external rigidity of the joint wherever possible, such as in the optimal welding sequence, among other factors, and avoiding a technique of trial and error. FEM weld modelling may represent a transient nonlinear process, and achieve accurate results for the thermal fields and residual stresses associated with the system of constraint or rigidity of joints, in an unusual fashion for the repair of bulky, costly and in many cases unique objects.