Change search
ReferencesLink to record
Permanent link

Direct link
Robustness analysis of welding simulations by using Design of Experiments
2008 (English)Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

In this master’s thesis robustness in welding has been investigated by means of simulations. A simple welding case has been simulated with the FE program MSC.Marc. If stability can be discovered during a simulation, then the welding process can be redesigned. However, instability in the simulation must not be due to instability in the real welding process. It can be due to pure numerical and modelling causes also and therefore care must be exercised when evaluating robustness. The study also includes a review part about the meaning of robustness and different approaches for evaluating this. In this work, a parametric study was used to investigate the robustness for an existing model. The material used in the simulations is called Greek Ascoloy, a martensitic stainless steel and the geometry consists of two plates which are welded together without additional filler material. The aim was to investigate which of the chosen parameters has the largest influence on the model, i.e. how sensitive or robust the model is when it is subject to disturbances. The chosen parameters to investigate were Poisson’s ratio, Young’s modulus, thermal expansion, initial yield stress, minimum yield stress, welding power, and the heat source geometrical parameters. To evaluate the results and determine whether or not the model is robust for a given set of parameters, the responses of the residual stresses in the sheet plane as well as gap and out of plane deformations were studied. The study was performed by using a factor screening method and each investigated parameter has been investigated at three levels. The methodology has been to change one parameter at a time while keeping the other parameters constant. The results have shown that the simulation model is quite sensitive to changes in welding power which resulted in increased deformations when the power was set to a high level. This will of course also occur in a real welding operation since the welding power has a big influence on the deformations of the welded component. The mechanical activation and thermal expansion, both resulting in large differences of stresses and deformations between the different levels, have also shown to have a negative influence on the robustness of the model. The heat source geometrical parameters have not influenced the model to large extent, although reducing the cr parameter showed an increased gap deformation. The model has also shown to be robust regarding changes in Young’s modulus, Poisson’s ratio and yield stress since very small changes in stresses and deformations occurred when changing these parameters. One possible continuation of this work could be to study the influence of interaction between the different parameters on the robustness of the model. Also, an in-depth investigation to explain why the model is more sensitive to changes in certain parameters would also be of interest.

Place, publisher, year, edition, pages
Keyword [en]
Technology, welding, FEM, Design of Experiments, robustness, parametric, study
Keyword [sv]
URN: urn:nbn:se:ltu:diva-52551ISRN: LTU-EX--08/100--SELocal ID: 9adef9fd-04c0-404e-ad52-afa8183e6e2eOAI: diva2:1025921
Subject / course
Student thesis, at least 30 credits
Educational program
Engineering Physics, master's level
Validerat; 20101217 (root)Available from: 2016-10-04 Created: 2016-10-04Bibliographically approved

Open Access in DiVA

fulltext(815 kB)0 downloads
File information
File name FULLTEXT01.pdfFile size 815 kBChecksum SHA-512
Type fulltextMimetype application/pdf

Search outside of DiVA

GoogleGoogle Scholar
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

ReferencesLink to record
Permanent link

Direct link