A theoretical and experimental analysis of variances in weld bead morphologies

Abstract

Predictable and reproducible weld bead morphologies and dimensions are a major concern in welding. In bead-on-plate welding, the heat flow is controlled by the heat source parameters (power, speed, and radius) and the physical properties and dimensions of the workpiece, especially its thickness. Complex models that account for weld pool circulation have been developed to quantify welds. However, to some extent, fluctuations in weld dimensions can be explained with conduction models of moving Gaussian heat sources. In early investigations with point and line heat sources, relationships between process parameters and plate thickness were derived to differentiate between two- and three-dimensional heat flow. To date, the heat source radius (R) has not been taken into account. The dimensionless ratio (D*) of the plate thickness (D) to the heat source radius (R) is actually a variable to consider. With the introduction of additional dimensionless parameters (*)—speed (v*), power(q*)— relationships among governing variables, heat flow dimension, and weld bead dimensional fluctuations can be derived. Weld bead fluctuations are found to depend on dimensionless variables (v*, q*, D* ) and occur when the heat flow dimension is intermediate between two- and three-dimensional. Occasionally, experimental data exhibit trends that differ from predictions. This article presents a dimensionless version of a heat flow model and discusses the discrepancies between experimental and theoretical results.