Fraunhofer Institute for Mechanics of Materials IWM
Fraunhofer Institute for Mechanics of Materials IWM

Weld screening: Digitalization of weld seam surfaces for automated, practical detection of weak points and service life assessment

Completed research project

Welding is the most widely used joining process and is applied across all industries. Particularly in regulated areas (construction, shipbuilding, crane construction, plant engineering, etc.), the quality of weld seams must be assessed and classified in accordance with DIN EN ISO 5817. There are many SMEs working in this field that must meet these requirements but do not currently have the capability to perform automated evaluation of welded joints. The Weld Screening project created a database on the geometry of welded joints and developed automated software solutions for quality and service life assessment.

Project description

Welded joints have irregular surface geometries. The seam transition geometry has a direct influence on fatigue strength. Notch radius and seam rise angle are key parameters that can be used to evaluate fatigue strength and support effect. Current standards such as DIN EN ISO 5817 require detailed knowledge of the local and global seam geometry in order to evaluate the quality classes of welded joints. In practice, however, such evaluations are usually only carried out on a random basis using gauges or angle measures.

As part of the Weld Screening project, a database on the local geometry of welded joints was established to serve as a basis for evaluating welded joints using AI approaches. In addition, concepts for automated quality assessment according to available quality standards were tested, and tools for location-specific fatigue assessment were tested on a digital twin of the weld seam. 

Transfer of project results to the following Fraunhofer IWM R&D services for companies:

  • Development of a standardized procedure for digitalizing welded joints using optical methods (stripe light projection, laser triangulation)
  • Automated evaluation and assessment of geometric parameters of welded joints according to available quality standards
  • Generation of input variables for probabilistic models for fatigue strength prediction
  • Determination of local stresses for the service life analysis of welded joints and their local and targeted post-treatment to increase service life