DEVELOPMENT OF RESIDUAL STRESS DURING PART BUILD IN METAL ADDITIVE MANUFACTURING

The Additive Manufacturing (AM) process is the scale-able, flexible and prospective way of fabricating the parts. It forms the product of desired design by depositing layer upon layer of material to print the object in 3D. It has a vast field of applications from forming prototypes to the manufacturing of sophisticated parts for space and aeronautical industry. It has even found its way into the domain of biological research and the development of implants and artificial organs. Depending upon the form of the raw material and the mechanism of printing it layer upon layer, there are different techniques of AM in metal parts production. One of them is Selective Laser Melting (SLM). This process involves the raw material in the form of metal powder. To manufacture the product, this powder first undergoes melting through a moving laser. Afterwards, it solidifies and joins with the already solidified structure in the layer below. The movement of the laser is carried out in the shape of a 2D cross-section design which has to be consolidated at the corresponding height. This process involves the repetitive heating and cooling of the material which causes sharp thermal gradients in the object. Because of such gradients, the material during manufacturing consistently undergoes thermal loading. Such thermal loading, therefore, induces the residual stress and permanent distortion in the manufactured part. These residual stress and thermally induced distortions affect the quality of the part and cause the mismatch in dimensions between the final product and the required design. To reduce the waste of raw material and energy, therefore it is important to predict such problems beforehand. This research work presents the modelling of a numerical simulation platform which simulates the part-scale AM SLM part manufacturing process and its cooling down in a virtual environment. The objective of establishing this platform was to evaluate the residual stress and thermal distortion. It included the modelling of thermal & structural analysis and their coupling to establish the multi-physics simulation tool. The transient thermal analysis with the elastoplastic non-linearity of the material model was implemented to capture the permanent deformation behaviour of a material under thermal loading. The modelling was done on the Finite Elements Method (FEM) based open source numerical analysis tools to incorporate the flexibility of numerical modelling in the project. The modelling strategy of solidified material deposition was incorporated via the elements activation technique. To synchronize the activation of the elements in the multi-physics FEM solver with the laser movement, the interfacing with AM G-code based data was performed. With this modelling strategy, the simulation experiments were conducted to analyse the evolution of thermal gradients, residual stress and deformation in part manufacturing. The study also highlights the challenges of the applied elements activation technique and its limitations. It was also studied how the prediction of simulation results vary with the different material deposition methods. Moreover, the resulting numerical analysis of the established simulation platform was also compared with the experimental and validated simulation data to ensure its reliability. In this comparative study, the current numerical strategy replicated the trends of stress and deformation from physical experimental data and represented the expected material behaviour in the manufactured part. Additionally, during this study, the skill gained for results handling and their validation was also applied in the other field of numerical modelling e.g. in the numerical analysis conducted for a blast furnace with Computational Fluid Dynamics - Discrete Element Method (CFD-DEM) coupled multi-physics platform of eXtended Discrete Element Method (XDEM). The current working simulation platform, via its AM G-code machine data interface with numerical solver, can facilitate the manufacturing engineers to predict earlier the possible thermally caused residual stress and deformation in their AM SLM produced product via simulation. On the other hand with the identified challenges in the virtual depiction of material deposition, the simulation developers may also be able to expect such limitations and make relevant decisions in the choice of material deposition technique in their AM SLM process modelling. Moreover, with the potential of this simulation tool being the basic building block, it may also provide the opportunity to build upon it the multi-scale numerical techniques and add them to the multidisciplinary research work of Artificial Intelligence based digital twins.