School of Engineering

Projects

Predicting and managing weld induced distortion of thin walled steel structures

One meter square test samples

Many metal structures are assembled from thin plate with welded supports for stiffness to resist local loadings. However, welded joints require a large heat input which may cause significant distortion to the finished product. In the shipbuilding industry, distortion is generally associated with thin plates, defined in this case as 8mm and less.

In this project, a neural network approach, in conjunction with experimental measurements and a finite element method will be used to study the relationship between distortion of welded ferritic steel plates and the design parameters. The two key aspects of the problem, which will be investigated in this work, are the interaction of process and production parameters in causing distortion and the influence of pre-existing residual stresses in the plate.

By modelling the distortion process using material, design and welding parameters, the parameters can be optimised to minimise the resulting distortion. The approach is a powerful one as it provides a means of taking into account the variability and uncertainty in the values of the input parameters and on the basis of the uncertainty in the input data, the neural network will determine the uncertainty in the resulting distortion and will display the surface topology.