Trends in Welding Home      Event Overview      ASM Homepage
Back to "POSTER SESSION" Search
    Back to Main Search

Tuesday, June 3, 2008

Efficacy of various modelling techniques for predicting Weld Bead Geometry

R. Wattal, Delhi College of Engineering, Delhi, India; S. Pandey, Indian Institute of Technology Delhi, New Delhi, India

Mechanical strength of welds is influenced by the composition of the metal and to a greater extent by the weld bead geometry and shape relationships (BG&SR) and in turn the weld bead geometry is influenced by the direct and indirect welding parameters. Also in the recent years, there has been a significant growth in the use of automated and robotic welding systems. These systems can be used effectively when mathematical models that correlate welding process parameters to the weld bead geometry and shape relationships are available.

 The study of weld bead geometry and shape relationships is important as these dimensions and ratios decide to a very large extent, the load bearing capacity of weldments. This includes the study of penetration, bead width, bead height, ratio of bead width to penetration (w/p) also known as weld penetration shape factor (WPSF), ratio of bead width to bead height (w/h) also known as weld reinforcement form factor (WRFF), and % dilution (%D).

 Factorial and dimensional analysis techniques were successfully used for developing mathematical models for predicting the weld bead geometry & shape relationship responses. However the main issue was to decide as to which one is more flexible as applied to modelling problems in welding. Some state of indecisiveness prevailed when it came to deciding a technique for developing models since both of these are used by the researchers quite often. This comparison helped in overcoming the state of confusion. Regression analysis offered the ability to summarize the collection of sampled data by fitting it to a model that accurately described the data.


Summary: Factorial and dimensional analysis techniques were successfully used for developing mathematical models for predicting weld bead geometry & shape relationship responses. However the main issue was to decide as to which one is more flexible as applied to modelling problems in welding.