Use of computational and physical simulation on arc welding heat affected zone microstructure evolution studies

Use of computational and physical simulation on arc welding heat affected zone microstructure evolution studies

  • Date: Sep 11, 2018
  • Time: 11:00 AM c.t. - 12:00 PM (Local Time Germany)
  • Speaker: Dr. Viktor Oliveira
  • Federal University of Rio de Janeiro (UFRJ)
  • Location: Max-Planck-Institut für Eisenforschung GmbH, Seminar Room 1
  • Room: Seminar Room 1
  • Host: Prof. Dierk Raabe
The heat affected zone (HAZ) is most commonly the critical part of welding joint and the comprehension of the thermal cycle it suffers during welding and its effects on the final microstructure is fundamental to predict and reduce the properties degradation on that zone. The traditional approach to study the HAZ involves several welding tests varying the principal parameters (voltage, current and welding speed) with subsequent mechanical testing. These welding trials could be very time, material and cost demanding; could not replicate the plant/field true welding conditions (need for small scale/plant no available for research tests) and still may not provide a profound insight on the mechanisms in play as the thermal history would not be evaluated. In this context, it is very interesting to use simulation techniques that have evolve significantly in the last two decades to optimize the research effort. In one side, we have the material computational simulation development, with the use of finite element methods and double ellipsoid heat source model to describe the process (thermometallurgic – mechanical coupling) and methods like CALPHAD, Phase Field and Cellular Automata to describe the microstructure evolution in details. One the other side, there are equipment (Gleeble) capable of applying very rapid and controlled thermo-mechanical cycles (acquired in the computational simulation) to a sample, so to produce physical simulated specimen that represents the HAZ region of interest, enabling more detailed characterization and some mechanical testing in isolated microstructures. This permits some validation of the computational simulation too. Seizing these techniques potential, LNTSold have been developing a series of studies in welding simulation to characterize the HAZ of different steels for oil and gas industry applications. For the X100M API 5L steel pipe, it was simulated on FEA software (Sysweld) the welding process of the pipe (SAW) and the field pipeline assembly (GMAW). The main concern for this steel is the toughness reduction it may be subject to in the HAZ, with possible formation of local brittle zones due to the evolution of very sensible constituents as the martensita-austenite (MA) constituent. From the bibliography reference, the two HAZ critical regions are the coarse grain region and the intercritically re-heated coarse grain region, so it was studied the thermal cycle of these regions with heat input variation in the FEA software. The thermal cycle was then reproduced in Gleeble samples to produce specimens for microscopy (focus on the MA constituent morphology and quantity analysis) and for Charpy impact test, to assess the toughness losses. The results indicate that the MA morphology depends very much on the peak temperature and that its quantity does not seem to control directly the impact resistance. For an AISI 4130 steel connector, it was performed a study with FEA software (Sysweld) and CALPHAD software (JMatPro) of the coarse grain HAZ region of the last welding passes, focusing in the hardness prediction and considering the post-weld heat treatment. A simulated CCT diagram and an experimental one were developed to include phase and hardness prediction in the FEA modelling. Then some heat treatment conditions (temperature x time) were evaluated using CALPHAD, trying to optimize the production time. All welding and the best heat treatment conditions were physically reproduced in Gleeble. The simulated CCT showed initially a good correlation with the experimental one, but the FEA hardness prediction was more precise using the experimental CCT. It was possible to achieve the hardness requirements and even increase the impact resistance with a faster heat treatment with close relation to simulation results. Finally, the welding of a 9% Ni steel pipe with Ni 625 alloy filler metal was also simulated in the FEA software and the different HAZ regions reproduced in Gleeble with dilatometry analysis to study the reversion and retention of austenite, which plays an important role in this steel tenacity. The goal it is also to isolate the microstructure and study its hydrogen embrittlement susceptibility.
Go to Editor View