Prediction of laser-spot-weld shape by numerical analysis and neural network
- PDF / 418,782 Bytes
- 9 Pages / 612 x 792 pts (letter) Page_size
- 115 Downloads / 174 Views
I. INTRODUCTION
LASER beam welding is a field of growing importance in industry. It is particularly useful in cases where localized heating is desired. There are many outstanding advantages in using laser welding as the bonding method of choice over other bonding technologies. Compared to common adhesive bonding or soldering used in joining processes, laser welding offers a number of attractive features such as high weld strength–to–weld size ratio, reliability, and minimal heataffected zone. These provide the benefits of low heat distortion, noncontact process, repeatability, ability to automate, and high throughput. A wide range of research activities has been undertaken, including laser beam delivery systems and mechanical behavior of laser-welded sheet steels. However, research on the dimension of laser beam welds for a given metal thickness has not been extensively studied. The weld pool dimension and weld quality of spot welds produced by using a pulsed Nd:YAG laser welding machine depend on various process parameters such as spatial intensity distribution of the incident laser beam, peak power of pulse, pulse energy, pulse time, and temporal shape of beam power during the pulse. When developing the welding procedure for a specific application, each of these parameters must be characterized and fully specified. This is usually done by using empirical techniques, since most analytical and numerical models for laser weld pool development ignore various aspects of the laser-material interaction,[1] thus making it difficult to accurately predict the weld pool shape.
W.-S. CHANG, Senior Member of Research Staff, is with the Precision Machining Group, Korea Institute of Machinery & Materials, Taejon 305343, Korea. S.-J. NA, Professor, is with the Department of Mechanical Engineering, Korea Advanced Institute of Science and Technology, Taejon 305-701, Korea. Manuscript submitted September 29, 1999. METALLURGICAL AND MATERIALS TRANSACTIONS B
Previous Model of Laser Welding Many models were proposed on laser welding to produce reliable predictions of temperature distributions that are of great importance for in-depth analysis and eventual improvement of the process. The quasi-stationary temperature field in the workpiece was calculated analytically by introducing a moving line source of heat. In 1977, Cline and Anthony[2] integrated the point source over the workpiece surface to yield a Gaussian power distribution, eliminated singularities in the temperature calculations, and showed that the spot size has a strong influence on the maximum temperature occurring in the workpiece. The model assumed 100 pct absorption and calculated an exponential decrease of temperature in the vertical direction. Both the conduction and keyhole welding condition were considered, whereas weld pool convection was not included in the model. A simpler approach is to compensate for the weld pool convective heat transfer with an artificially high thermal conductivity of material in the weld pool. Typically, a thermal conductivity of 10 time
Data Loading...
