Design and evaluation of support structures in selective laser melting: A practical engineering approach to improved performance

Abstract

Additive Manufacturing (AM), particularly Powder Bed Fusion-Laser Beam for Metals (PBF-LB/M) or Selective Laser Melting (SLM), has revolutionised various industries, including aerospace, automotive, and medical, by enabling the production of complex, thin, lightweight, and customized metal parts. Despite its advantages, key challenges remain in optimizing and designing support structures, which are critical for ensuring part quality, reducing defects caused by thermal stresses, and minimizing post-processing efforts and overall costs. This thesis systematically investigates block-type support structures, evaluating their performance through a Multi-Response Optimization (MRO) approach, experimental studies, and numerical simulations, while comparing them with alternative support geometries, including line, contour, and cone-type supports. It aims to address key challenges in SLM, such as support generation, support removal effort, material consumption, and relevant defects. The research is organized around three key objectives. First, it introduces a framework using multi-response optimization approach to evaluate block-type support structures, aiming to propose optimized geometries that minimize support volume, prevent warping deformation, and enhance support removal efficiency. For layers aligned parallel to the build plate (e.g., 0° overhangs), the findings revealed that block-type supports with a tooth height of 2.7 mm, a tooth top length of 0.2 mm, and a hatch distance of 0.7 mm provide an optimal balance between mechanical stability, minimal warping deformation, material consumption, and ease of removal. Conversely, for layers inclined relative to the build plate (e.g., 25°- 45° overhangs), the optimal balance was observed with a tooth height of 4 mm, a tooth top length of 0.05 mm, and a hatch distance of 2.5 mm. Second, thermo-mechanical simulations examine the thermal behaviour of alternative support geometries, including block, line, contour, and cone-type supports, with the goal of mitigating defects caused by thermal stresses and identifying configurations that offer optimal thermal performance. The outcome, consisting of plots and tables, provide valuable guidelines for achieving effective printing and ensuring the production of defect-free parts. Additionally, the numerically optimized results of this study were validated through experimental testing. It was found that block-type support structures, despite their larger volume and the challenges associated with their removal, demonstrate slightly improved thermal behaviour compared to the other support types analysed. Finally, the study presents an innovative design framework for optimising the SLM workflow, introducing a web-based platform for automated support generation, optimization, and thermal performance assessment. This platform serves as a valuable research tool for managing and visualizing experimental data, allowing researchers and professionals to improve AM production and produce defect-free, high-quality prints while reducing printing time and costs. By integrating and visualizing experimental data and simulation results, the platform allows users to import 3D models, adjust orientation, generate and visualize optimized support structures, and export ready-to-print designs. Moreover, this tool is designed to be accessible to non-expert users, simplifying complex support design decisions while effectively reducing trial-and-error approaches and streamlining the SLM process. Validation on a small L-shaped mounting bracket demonstrates that the platform effectively generates and visualises block-type support structures for imported parts while facilitating successful SLM printing. The printed outcome exhibited excellent dimensional accuracy, minimal warping deformation, and easy support removal (achieved within 2-4 minutes of manual effort) along with satisfactory surface roughness. By leveraging experimental, optimization, and computational tools, such as SLM 3D printing machines, Design-Expert, and COMSOL Multiphysics, this thesis presents a structured methodology for optimizing metal support structures, reducing defects, and enhancing the efficiency of the SLM process. These contributions not only address current challenges but also create opportunities for future advancements in the field, such as improved quality control during pre-printing preparations, ultimately establishing SLM as a viable production method.