Impact on Duplex Stainless Steel Pipes with and without Precipitated Sigma phase
MetadataVis full innførsel
A manufacturing defect in pipe fittings supplied to the oil and gas industry has resulted in installed components with unknown impact capacity. If the altered capacity is not sufficient, production stops and expensive replacements of parts are necessary. Erroneous heat treatment led to precipitation of a brittle sigma phase in duplex stainless steel fittings. The overall goal for this thesis is to increase the knowledge of how the impact capacity of fittings with this production error is affected. The main objective in the performed study has been to create a finite element model and simulate impact loading on pipe bends both with and without precipitated sigma phase. Special attention has been given to the initiation and propagation of fracture and how this can be modeled. The model can in further studies be used to find the impact capacity of bends with various levels of sigma content. The objective was reached by material testing, full scale component tests and non-linear finite element analyses of both the material tests and the component tests. All analyses were carried out using the finite element code LS-DYNA. Material tests are performed previous to, and during this study, on both specimens affected and unaffected by sigma phase. The material, with three different amounts of precipitated sigma phase, was modeled by use of the modified Johnson-Cook constitutive relation and the Cockcroft-Latham fracture criterion. Full scale component tests on pipe bends were carried out under both dynamic and quasi-static load conditions. The results were compared, and the effect of sigma phase in the component was addressed. During simulations of the full scale tests, attention was given to recreate the global response of the component at various impact velocities and to predict the initiation of fracture. The effect of high levels of precipitated sigma phase was found to reduce the impact capacity of the tested components. No visual fracture was found on any of the components containing low amounts of sigma, i.e. 0-5 %, while the components with higher content fractured. The same tendency was found in the quasi-static tests. Components containing 15 % sigma phase were found to have a reduced impact capacity of more than 80 % compared to components unaltered by sigma phase. The finite element model was able to predict the global response of the components sufficiently accurate. The Cockcroft-Latham fracture criterion, when calibrated according to the tensile tests, was not able to predict the initiation of fracture in the finite element analyses of the component tests. However, by reducing the critical Cockcroft-Latham parameter by a constant factor, the initiation of fracture was accurately predicted both for quasi-static simulations and dynamic simulations with various impact velocities. The simulations were not able to recreate the rapid fracture propagation seen in the full scale tests.