Balancing Productivity and Product Quality in Welding

Conference poster, 2016

The use of Design of Experiment (DOE) can have vast impact on the measurement need. A study conducted at Volvo CE by Hammersberg and Olsson was aiming at finding settings for parameters influencing weld toe radii, throat size and penetration. The performed DOE concluded a process window where all three requirements were achieved. The parameters identified were related to the welding geometry rather than the commonly varied welding parameters. Positioning of parts and weld torch as well asgap size had larger impact than productivity parameters such as amperage and weld speed. This is surprising since most investigations do not uncouple parameters for weld geometry from those controlling productivity. A possibility to control the weld quality by adjusting indirect parameters independently from the productivity emerges. From an inspection perspective that is very influential. In the studied case that means monitoring angles and gap size proactively before welding instead of inspecting weld toe radii, throat size and penetration after welding in a reactive manner. There are difficulties connected to the measurement of penetration and weld toe radius. It is in comparison relatively easy to control angles and gap sizes. This observation means new challenges for the organization on the meta-level. It seems necessary to have incremental improvement of competence and joint understanding and to have a critical mass of people having the same view. Not until 7 years after the first study of lack of measurement precision the organization was ready, as an organisation, to apply the optimized design of experiment, even though single individuals were trained earlier. When educating the team as a team the best results were achieved, probably because a common language and new culture emerged. This indicates that the challenge of improved welding performance lay in the elevated organisational plane, rather than on individual competence and technology level. This research addresses the corner stones for welding process development that turns the perspective around from average to variation, regarding balancing process performance and product quality assurance.