Computational research on microstructure evolution and mechanical properties of martensitic stainless steel during welding and heat treatment processes
This project aims at the prediction of microstructure evolution and thermomechanical properties in and around a laser weld seam of a martensite steel plate.
Starting point for this joint activity was a slab like geometry of a martensitic steel grade and the parameters of the laser welding process.
Essentially following activities have been performed:
- Simulation of the laser welding process including the specification of a bi-modal heat source with the scope of calculating the time dependent temperature profile for subsequent microstructure calculations. The resulting thermal profiles showed good agreement with the melt pool profile as determined by metallurgical sections.
- Experimental characterization of the microstructure of the base material by LOM/EDX/EBSD to generate a suitable initial, digital microstructure for subsequent microstructure simulations.
- Microstructure simulations in 2D to model melting, re-solidification, phase transformations and grain growth in the different zones of the weld, i. e. in the base material, the HAZ, CGHAZ and melt-pool.
- Development of criteria functions to model the martensite to austenite transformation in a phase-field model (MICRESS software).
- Development of an FEM model for the austenite to martensite transformation involving both temperature and stress.
- Compilation of the thermo-elastic properties of the individual phases (ferrite, austenite, martensite, MC-carbides) based on literature research.
- Calculation of effective thermo-elastic properties of the different RVE’s (Base Material, HAZ, CGHAZ and melt-pool) from 2D microstructures and above literature data using a homogenization tool (HOMAT software).
- Generation of synthetic 3D RVEs based on and corresponding to the 2D results.
- Determination of flow curves of the different 3D RVE’s (Base Material, HAZ, CGHAZ and melt-pool) based on simulated microstructures using an FEM model (ABAQUS software).
- Mapping the different RVE properties to the entire component.
- Performing virtual tensile tests on the entire component.
- Performing tensile tests on the real component.
- Comparing experimental and virtual tensile tests of the component.
The major lessons learned are:
- For this ICME set-up the hand shake data transfer was sufficient for a comprehensive modelling of the microstructure response on laser welding.
- Need of a bi-modal heat source for the laser taking into account volumetric heating in the keyhole and surface heating in the laser diameter to be able to correctly reproduce the experimental pool profile.
- Formation of coarse grains in the CGHAZ is most probably due to growth of grains in the semi-solid region and not due to normal grain growth in the solid state.
- The composition of the different phases - besides their volume fractions, morphology and distribution in the microstructure - has a significant impact on the mechanical properties.
- In order to predict accurately the effective thermo-elastic and elasto-plastic behavior, a random 3D orientation has to be specified in multi-phase field simulations, even if those are performed in 2D.
- 3D initial microstructures including grain orientations would be beneficial for further improvements of model predictions.
|Microstructure Simulation Group,
|Research & Development Group,
|1-1 Omika-cho 7-chome,
Hitachi-shi, Ibaraki 319-1292,
Steel Institute IEHK,
RWTH Aachen University
- G. Laschet, B. Böttger, K. Komerla, Y. Kanegae, M. Ogata, M. Park, U. Prahl and G.J. Schmitz: Prediction of microstructure evolution and thermomechanical properties of a martensite steel plate during laser welding, IWCMM25 Conference, October 1st to 2nd 2015, Bochum, Germany
- G.J. Schmitz, M. Apel and B. Böttger: On the role of solidification modeling in ICME settings, Materials Science and Engineering 117 (2016) 012041, DOI: 10.1088/1757-899X/117/1/012041 | RWTH Publications