Fusion-based manufacturing techniques, such as laser welding and additive manufacturing, are revolutionising how we produce complex components across industries. Yet, one critical challenge persists: how do we reliably control and predict melt-pool behaviour to ensure high-quality outcomes? This question was at the heart of my PhD research, completed at Delft University of Technology in the Netherlands. The full dissertation is available (open access) here: https://doi.org/10.4233/uuid:06ff0b5e-d5da-4149-a90f-62064c29f238.
Why Melt-Pool Oscillations Matter
The stability of the molten metal melt pool during fusion-based processes is key to producing defect-free, structurally sound parts. However, traditional approaches—largely based on trial-and-error—are costly, time-consuming, and often lack generalisability, especially with new materials or configurations. My work aims to change that through a simulation-based approach, offering a robust and physics-informed alternative for process development and optimisation.
What’s New in This Research?
At the core of the dissertation is a generic computational model built around the enthalpy-porosity method. The model captures key physics of complex heat and fluid flow, melting, and solidification, and the dynamics of heat source modelling in arc and laser processes, and can be adapted across a range of welding and additive manufacturing processes by modifying boundary conditions—eliminating the need for process-specific formulations.
Some highlights include:
-
Systematic study of the mushy-zone constant (permeability coefficient) and its effect on solidification predictions.
-
Critical assessment of “enhancement factors” (often used to tweak thermal conductivity/viscosity) and their unintended impact on simulation fidelity.
-
Incorporation of time-varying energy distributions and surface deformation effects, improving realism in modelling melt-pool dynamics.
-
Wavelet analysis (CWT) used to capture non-stationary features of melt-pool oscillation signals, surpassing conventional FFT techniques.
Applications in Action
Beyond theoretical development, I’ve extended this modelling framework to several industrially relevant case studies:
-
Absorptivity Modeling: A more realistic absorption model accounts for surface temperature, laser angle, and material properties—offering a path toward data-driven and calibrated-free simulations, aligning with Industry 4.0 goals.
-
Laser Beam Shaping: Explored how tailored laser intensity profiles affect melt-pool flow and microstructure in stainless steel 316L. Beam shaping significantly influences melt-track geometry and grain morphology—critical for microstructure control.
-
Laser Butt Welding of Thin Sheets: Simulations of symmetric/asymmetric welding setups reveal the dominant role of advection in energy transport and highlight how weld defects can arise from thermal and flow instabilities.
-
Powder-Induced Attenuation in Directed Energy Deposition (DED): A coupled DEM–FVM model captures how powder streams alter laser power distribution—addressing a long-standing gap in DED simulations.
Why It Matters for the Mechanics Community
This work bridges mechanics, thermofluid modelling, and process engineering, offering tools and insights that can reduce development cycles and enhance predictive capability in advanced manufacturing. For mechanicians, it opens new pathways to model non-linear, multi-physics problems in a modular and adaptable framework. For industry, it contributes to digitisation and process automation, ultimately helping bring robust, high-performance parts to production faster.
Learn More
You can explore the full dissertation and detailed case studies here:
https://doi.org/10.4233/uuid:06ff0b5e-d5da-4149-a90f-62064c29f238
I look forward to feedback and discussion from the iMechanica community. Feel free to reach out or comment below if you'd like to discuss applications, model extensions, or collaboration opportunities!