Quantifying Variability in Laser Welding of Stainless-Steel Alloy 304

Open Access
- Author:
- Finch, Nicholas
- Graduate Program:
- Materials Science and Engineering
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- October 15, 2024
- Committee Members:
- Tarasankar Debroy, Thesis Advisor/Co-Advisor
Darren C Pagan, Committee Member
John Mauro, Program Head/Chair
Todd Palmer, Thesis Advisor/Co-Advisor
Jingjing Li, Committee Member - Keywords:
- Numerical modeling
heat transfer and fluid flow
welding
laser
stainless steel
process control
uncertainty
variability
thermophysical properties
material properties - Abstract:
- Control of weld characteristics has conventionally employed a trial-and-error approach to establish acceptable ranges in variability for process parameters and composition limits. This approach has limited predictive capability for how changes in either composition or process parameters will quantitatively impact weld characteristics. Composition limits set by industrial standards groups encompass broad composition ranges that result in broad material property ranges. Using limits, medians, or other statistical improvisations of these set limits is not necessarily representative of the material that has been welded in scientific literature or in industrial applications. A survey of reported alloy compositions used for crack-free welding of austenitic stainless-steel alloys 304 and 304L clarified what expected values of the composition of these alloys are. The compositions found in the survey were used with commercially available thermophysical modeling software to develop property ranges and models of composition-property relationships. The property ranges were input into established numerical models of heat transfer and fluid flow developed for laser welding to establish ranges of weld characteristics as a function of the properties and composition ranges found in literature. A one-at-a-time sensitivity analysis of property-composition dependence was used to establish the relative importance of each alloying element on material properties pertinent to laser welding numerical models. Then another one-at-a-time sensitivity analysis was used for both the material properties and the processing parameters independently with respect to the established welding model results to determine a hierarchy of contributions to variability in weld characteristics. The material properties found to be of most importance to variability in welding are the enthalpy of the solid and liquid, thermal conductivity of the liquid, heat capacity of the solid, solidus temperature, and liquidus temperature. The composition changes that strongly influence these parameters are changes in chromium, nickel, manganese, and silicon. The process parameters with the strongest influence on weld characteristics are defocus distance, laser beam power, and laser travel speed. Quantifying these relationships between composition, material properties, process parameters, and the resulting weld properties will allow for the development of more robust laser welding process quality control requirements.