Predictability – an enabler of weld production development

Abstract: Predictability enables proactive control of the production instead of reactive inspection. The research presented in this thesis provides insights and tools to enable predictability in weld production. The research shows that sufficient use of process information facilitates predictability and can support the reduction of unnecessary and costly safety margins in production. This will, in the short term, affect the productivity of the company. Even more influential is the effect predictability can have on product and process development in the long term, enabling welding companies to reduce weight of their fatigue loaded structures. The result can be lower fuel consumption and increased payloads, which will increase the companies’ competitiveness. Knowledge of the process is the key to evolve towards predictable weld production. To understand variation and how the evaluation process should be designed is therefore important. A model describing evaluation process development is presented. It emanates from the internal customer’s need for information about the product or the process. Several aspects are decided on before the evaluation method is chosen, such as what information is needed and suitable ways to present that information. Understanding the variation occurring in all parts of the process is necessary to achieve predictability. Tools long known in the quality society have proven to be useful. MSA, measurement system analysis, has shown to be a useful tool to understand the variation stemming from the evaluation process. Control charts seem to be effective to visualize variation and facilitate improved decision making by creating a common understanding. To have a critical mass of people sharing the same nomenclature and definitions is also necessary. The studies show that soft issues are often as important as technical solutions. Implemented results from the research indicate a significant financial effect. The research could have a bearing in a wider context, which has been demonstrated by interview studies including companies in the welding industry.