Double vibratory stress relief treatment of welded pipes as an alternative to heat stress relief treatment: Effects on hardness, Charpy impact energy and CTOD

Abstract

This work analyzes the effects of intermittent resonant vibration treatment in pipes welded to connectors applied in the field of oil and gas wells. The vibratory stress relief (VSR) process was applied intermittently under two conditions: the first one, called Simple Vibration Treatment (TVS), which consisted of a resonant vibration treatment with the vibrational system at the mean length of the product, a second step, with the product being rotated 90°. The results obtained with this process led us to study a new resonant vibration process with the purpose of intensifying the previous one, taking care that no defects would be introduced into the product. This condition, called Double Vibration Treatment (TVD), consisted of a treatment formed of three steps: the first two steps were performed in the same way and with the same procedure used in the Single Vibration Process (TVS). The third step consisted of carrying out a second resonant vibratory treatment with the vibrator positioned on the longitudinal weld of the pipe, close to the circumferential weld of the pipe with the connector. These VSR treatments were compared with the stress relief heat treatment PWHT (Post Weld Heat Treatment) which is normally used after welding the connectors to the pipe. The resonant vibration treatments (VSR), did not introducing any damage to the product, presented reduced treatment time, and tended to positively affect the evaluated mechanical properties, exceeding or equaling the same tendency presented by the heat treatment. The resonant Doble Vibration Treatment (TVD) showed the best results, affecting the evaluated properties more positively than the TVS, which practically promoted the same trend of results produced by PWHT. The studies were characterized by Vickers hardness tests, Charpy V impact fracture tests and CTOD fracture toughness tests, and the results were analyzed using ANOVA statistical modeling.