Simulation of Heat Flow During the Welding of Thin Plates
- PDF / 656,302 Bytes
- 6 Pages / 594 x 774 pts Page_size
- 102 Downloads / 297 Views
THE fusion welding of thin plates, besides being involved in practical welding application, is often used as a tool to study both weld-pool solidification and solid-state phase transformations during the fusion welding of engineering alloys. 1-5As a result, heat flow in thin-plate welding has been a subject of active study during the past two decades. Quantitative heat flow information in thin-plate welding can be very useful from both the metallurgical and the welding design viewpoints. The analytical solution to the steady state, twodimensional heat flow problem of thin-plate welding was first derived by Rosenthal 6 and was later reviewed by several other investigators. T M The major assump* Only references concerning thin-plate welding are cited here. Literature related to 3-dimensional heat flow will be given in a separate paper dealing with thick-plate welding.
tions made were: l) a point heat source, 2) no heat of fusion and no melting, 3) no surface heat loss due to convection or radiation, and 4) constant thermal properties. Due to some of the rather unrealistic assumptions mentioned above, heat flow and solidification in the weld pool can not be predicted, and poor agreement exists between calculated and experimental thermal histories in the heat affected area immediately adjacent to the weld pool. As a consequence, many investigators have tried to modify Rosenthal's analytical solution. For example, Grosh 1~and Swift-Hook N considered the temperature dependence of the thermal properties. Jhaveri ~2took into account the surface heat loss. Ghent ~3included the heat of fusion. Trivedi 14considered the finite size of the heat source. However, due to the complex nature of the problem, the modified SINDO KOU is Assistant Professor, Department of Metallurgical Engineering and Materials Science, Carnegie-Mellon University, Pittsburgh, PA 15213. Manuscript submitted February 13, 1980. METALLURGICAL TRANSACTIONS A
analytical solutions so far have had rather limited success. Recently, with the help of digital computers, numerical methods have been applied to the study of heat flow in fusion welding, and significant improvements have been made. For example, based on the measured shape of the weld pool, Pavelic 15calculated the temperature distribution in a thin plate of steel using the finite difference method. Friedman, ~6,~7on the other hand, used the finite element method to calculate the temperature and stress distribution in a thin plate being welded. The measurement of the weld-pool shape was not required in his calculations. However, heat conduction in the welding direction was neglected. More recently, Sharir la employed the finite difference method to calculate the unsteady heat flow during the fusion welding of thin tantalum sheets. The heat conduction in the welding direction was taken into account, which is an improvement over Friedman's study. However, unlike Friedman, who used a grid mesh of variable spacing, Sharir employed a grid mesh of constant spacing, resulting in a rather coarse mesh near the h
Data Loading...
