An Investigation on Effect of Process Parameters on Weld Bead Profile of Modified 9Cr-1Mo Steel by Pulsed GMAW

Abstract

Environmental regulations requiring reduced C02 emissions coupled with demand for greater efficiency are demanding greater use of advanced creep resisting steels such as modified 9Cr-lMo (P91) for high temperature components in the power generation industry. Traditionally, modified 9Cr-lMo steels are welded using gas tungsten arc welding, shielded metal arc welding and submerged arc welding processes. In order to reduce costs and downtime, particularly for site repairs, recently there is much interest in the use of high productivity welding processes; particularly gas shielded welding processes such as Gas Metal Arc Welding (GMAW). However, the situation with GMAW, particularly with active gas mixtures, is more complex because of the variable recovery of key elements such as Mn, Si, and Nb/Cb. It has been reported that the solid wire Gas Metal Arc Welding (GMAW) process has not found widespread use in the industry mainly due to concerns over lack-of-fusion, sensitivity to welder error and demands for more sophisticated power sources. However, in recent past, as a process of potentially high productivity, interest in GMAW for welding P91 materials is increasing for Industrial applications.

In the present investigation, bead-in-groove trials are performed on a 12mm thick modified 9Cr-lMo material by pulsed current gas metal arc welding (GMAW-P) process. Two primary parameters like current and weld travel speed are considered in the present experiment and their effect on weld bead geometry, weld metal toughness and weld metal microstructure are discussed. The samples are also subjected to chemical analysis, oxygen content measurement, and inclusion level&hardness survey after PWHT. Response surface methodology (RSM) design approach is used to develop a mathematical relation between the input variables (current and speed) and responses. The developed model is then compared with the experimental results; it is found that the deviation falls within the limit of a 95% confidence level. The direct and interactive effects of the process parameters are also discussed in the present paper.