A cost-function driven adaptive welding framework for multi-pass robotic welding. (July 2021)

• Record Type: Journal Article
• Authors: Loukas, Charalampos; Williams, Veronica; Jones, Richard; Vasilev, Momchil; MacLeod, Charles N.; Dobie, Gordon; Sibson, Jim; Pierce, Stephen G.; Gachagan, Anthony
• Is Part Of: Journal of manufacturing processes. Volume 67(2021)
• Page Start: 545
• Page End: 561
• Publication Date: 2021-07
• Subjects: Robotic arc welding -- Multi-pass -- Weld sequence planning -- V-groove -- MAG; Production management -- Data processing -- Periodicals; Manufacturing processes -- Periodicals; Procestechnologie; Productietechniek; Production -- Gestion -- Informatique -- Périodiques; Fabrication -- Périodiques; Manufacturing processes; Production management -- Data processing; Periodicals; 670.5
• DOI: 10.1016/j.jmapro.2021.05.004
• Languages: English
• ISSNs: 1526-6125

Abstract

Highlights: An adaptive cost driven welding framework for automated multi-pass welding is introduced. Welding parameters can be optimized by minimizing a user selectable weighted cost function. The generation and adaption of welding parameters per layer for varying V-groove joints is presented. Robotically experimental verification on welded joints of single sided V geometries is shown. Ultrasonic inspection is utilized to validate the quality of the weldments. Abstract: Manual teaching of robot paths and welding parameters for multi-pass robotic welding is a cumbersome and time-consuming task, which decreases the flexibility, adaptability, and potential of such systems. This paper introduces and presents a new automated weld parameter and pass deposition sequencing framework, which builds on the current state of the art developments and enables automatic planning of multi-pass welding for single-sided V-groove geometries. By integrating a novel cost-function concept that permutates and identifies the welding parameters for each layer through a user-driven weighting, the framework delivers the minimum number of passes, filler material and welding arc time based on application requirements. A mathematical model relating the cross-section area of beads with the pose of the torch and weaving width was built upon to allow full-process automated welding parameter generation and adaption for different geometric characteristics of the groove. The concept methodology and framework were then developed and verified experimentally, through robotically deployed Metal Active Gas (MAG) welding. For a given representative joint, the arc welding time and amount of filler wire were found to be 32.9 % and 26.18 % lower respectively, than the worst-case available welding parameter combination, delivering a corresponding decrease in direct automated welding manufacturing costs. Lastly, an ultrasonic inspection was undertaken to verify the consistent quality of the weldments validating the framework outcome and enabling welding pass automation through robotic systems.