Search

Residual macrostresses in a multipass circumferentially butt-welded P91 ferritic steel pipe have been determined numerically and experimentally. The welded joint in a pipe with an outer diameter of 290 mm and a wall thickness of 55 mm is typical of power generation plant components. An axisymmetric thermomechanical finite element model has been used to predict the resulting residual hoop and axial stresses in the welded pipe. The effects of the austenite to martensite phase transformation have been incorporated into the simulation. Residual stresses have been measured using the X-ray diffraction technique along the outer surface of the pipe and using the deep-hole drilling technique through the wall thickness at the center of the weld. Good correlation has been demonstrated between the residual hoop and the axial stresses obtained numerically and experimentally. The paper demonstrates the importance of using a mixed experimental and numerical approach to determine accurately the residual macrostress distribution in welded components. [DOI: 10.1115/1.4000347] Keywords: P91 steel pipe, welding, finite element simulation, X-ray diffraction, deep-hole drilling, martensite transformation, residual stress

Past studies from creep tests on uniaxial specimens and Bridgman notch specimens, for a P91 weld metal, showed that anisotropic behavior (more specifically transverse isotropy) occurs in the weld metal both in terms of creep (steady-state) strain rate behavior and rupture times (viz., damage evolution). This paper describes the development of a finite element (FE) continuum damage mechanics methodology to deal with anisotropic creep and anisotropic damage for weld metal. The method employs a second order damage tensor following the work of Murakami and Ohno (1980, 'A Continuum Theory of Creep and Creep Damage,' Creep in Structures, A. R. S. Ponter and D. R. Hayhurst, eds., Springer-Verlag, Berlin, pp. 422-444) along with a novel rupture stress approach to define the evolution of this tensor, taking advantage of the transverse isotropic nature of the weld metal to achieve a reduction in the number of material constants required from test data (and hence tests) to define the damage evolution. Hill's anisotropy potential theory is employed to model the secondary creep. The theoretical model is implemented in a material behavior subroutine within the general-purpose nonlinear FE code ABAQUS (ABAQUS User's Manual Version 6.6, 6006, Hibbitt, Karlsson and Sorenson, Inc., Providence, RI). The validation of the implementation against established isotropic continuum damage mechanics solutions for the isotropic case is described. A procedure for calibrating the multiaxial damage constants from notched bar test data is described for multiaxial implementations. Also described is a study on the effect of uniaxial specimen orientation on anisotropic damage evolution. [DOI: 10.1115/1.3007429]