Date of Award


Publication Type

Doctoral Thesis

Degree Name



Mechanical, Automotive, and Materials Engineering


Engineering, Mechanical.


Watt, D. F.,




Laser welding has firmly entrenched itself as an attractive manufacturing process technology in the industry. Despite its role of a viable manufacturing process, its overall efficiency is not significantly higher than other joining processes, in particular resistance welding. It is apparent that the high traverse speed of laser welding would increase the manufacturing process efficiency and productivity. One aspect of the research is to optimize process parameters for the physical performance of automotive body-in-white sheet steels joined by high speed welding. For this purpose, exploratory phases of pre-design experiments, bead-on-plate welding and optimization of welding in butt joint configurations are adopted. Parametric design schemes are undertaken in this research. The parametric design methodology is adopted to optimize the product and process design prior to recommendation for production implementation. This methodology takes into account the quality in the experimentation phase since so-called real-time quality control and/or SPC (statistical process control) can never fully compensate for all the physical process variables. Taguchi methodology, though used for other joining processes, is applicable to laser welding but has never been studied. This research discusses the laser welding optimization using Taguchi methodology. The Taguchi methodology employed a set of orthogonal arrays. The orthogonal array used in this experiment is L$\sb8$ (2$\sp7$) orthogonal array. The L$\sb8$ orthogonal array employs only 8 prototype trials while the equivalent conventional full factorial experiment would be 128. Though the designs sacrifice the effects of complex interactions of process variables, but the number of tests has been reduced 16-fold. The two levels of process variables such as speed, shield gas, flow rate, nozzle diameter are used along with three beam parameters each also at two-level. Two material conditions are also studied. The welding experiments were conducted on sheet steels in the butt joint configurations. A 10 kW fast axial flow continuous-wave CO$\sb2$ laser was used for all the trials. The sheets were butt welded transverse to the joint and parallel to the sheet rolling direction. The bead profile of the laser welds were investigated through visual and optical examinations. The most significant process parameters that provided the highest quality of welds were established. The quality characteristics were determined through destructive testing. It is observed that the high speed laser welds are associated with defects such as bead protrusion commonly known as humping, hole formation, undercutting, etc. Though these phenomena have been observed previously, in high speed laser, electron beam and arc welding, there is a scarcity of reported physical explanation of the cause of these high speed weld nonlinearities. A mathematical model is developed to correlate the instabilities of high speed weld melt pool. The weld molten pool behavior is explained in terms incompressible Newtonian fluid flow. The continuity and momentum equations are formulated for surface perturbation or instability of the melt. The surface perturbation is defined as the quasi-periodic protruded bead structure measured after solidification. The surface perturbation is not measurable within the weld speed that does not create weld defects and is distinctive of striations. The onset of instability is determined employing dynamic similitude that correlates the weld penetration and weld speed. A comparison of experimental welding speed showed a good approximation to the weld surface perturbation. This thesis detailed the mathematical formulation of propagation of the perturbation on the melt at high weld speed and experimentally determines the stability/instability criteria.Dept. of Mechanical, Automotive, and Materials Engineering. Paper copy at Leddy Library: Theses & Major Papers - Basement, West Bldg. / Call Number: Thesis1994 .U32. Source: Dissertation Abstracts International, Volume: 56-11, Section: B, page: 6356. Adviser: D. F. Watt. Thesis (Ph.D.)--University of Windsor (Canada), 1995.