Hardness prediction for the repair welding of 2.25Cr-1Mo pressure vessels

Reactor vessels used for the hydrotreating of heavy oils and tar sand bitumen are frequently made of 2.25Cr-1Mo steel in thicknesses of 150 to 300 nm. Defects developed during installation or service are often repaired by welding. For practical reasons, postweld heat treatment of the repair welds is undesirable. This has led to continued effort to develop weld repair techniques that do not involve postweld heat treatment. Recently a six-layer automatic gas tungsten arc welding (GTAW) technique has been proposed for the repair welding of nuclear reactor vessels made of SA508 Class 2 Steel. In this technique, the second and third passes refine the microstructure of the first pass, and the last three passes temper the first pass. Alberry has developed a set of empirical rules predicting the hardness after each pass in multipass welds made in SA508 Class 2 Steels. This algorithm has been used to predict the number of layers required to achieve desired hardness. A transformation and tempering algorithm for 2.25Cr-1Mo, similar to that of the above steel, is presented. The tempering algorithm of Alberry suffers from several minor problems and can be improved. A mathematically correct method for the calculation of the tempering occurring in an anisothermal cycle is demonstrated. In addition, the rules used to relate the softening that occurs during temperature are heuristic. Separate rules are proposed for the kinetics of softening depending on the peak temperature. A re-examination of those rules reveals that they can be recast in the form of a single rule for the material examined. Reassessing the basic data presented by Alberry leads to a single softening rule with better theoretical justification