Metal additive manufacturing techniques opened the door for rapid prototyping and on-demand manufacturing while improving supply chain resiliency through decentralizing Manufacturing. However, uncertainties in operating conditions and machine reliability can limit these advantages by increasing lead times and degrading the mechanical properties of fabricated components. Specifically, process interruptions influence the melting-solidification cycles during layer evolution, leading to microstructural changes and variations in Mechanical characteristics of 3D-printed metal parts. These interruptions can compromise structural integrity, resulting in the formation of defects that weaken the mechanical strength of the final component. Accurately assessing these flaws is critical to ensuring part reliability. There are a variety of techniques which can remove or reduce the amount or size of flaws. Post- processing methods, such as Hot Isostatic Pressing (HIP) and other heat-treatment techniques, enhance the mechanical properties and structural integrity of AM parts. The choice of post-processing approach depends on the intended application and the material properties of the printed component. Since variations typically occur at the layer level, it is essential to analyze mechanical properties in a manner that effectively identifies and quantifies these deviations, guiding appropriate post-processing decisions. Hence, this study investigates the effects of process interruption on the mechanical properties of the metal 3D printed parts for stainless steel parts with and without post-processing. To evaluate these effects, various stainless steel 316L parts are fabricated using powder-bed fusion-based selective laser melting(SLM), manufactured under various process parameter conditions The printed samples undergo sectioning, mounting, and polishing to achieve a mirror-like surface finish, allowing for a detailed investigation of their micro-mechanical characteristics. Nano-indentation testing is employed to measure key mechanical properties, including elasticity and hardness, in a matrix grid format across the interruption-affected regions.
To evaluate the effectiveness of post-processing, a subset of the samples undergoes heat treatment in a vacuum furnace, and their micro-mechanical properties are reassessed. A statistical analysis is conducted to examine variations within the heat-treated samples and compare them to their as-built counterparts. Additionally, machine learning techniques are integrated into the analysis to predict mechanical property variations based on processing conditions, enabling data-driven insights for optimizing AM parameters and post-processing strategies. Results demonstrate that process interruptions impact mechanical properties; however, post-processing treatments significantly mitigate these variations. Furthermore, predictive modeling using machine learning provides a powerful tool for anticipating mechanical performance, ultimately improving the reliability and consistency of metal AM components.
