An algorithm to prioritize welding quality deterioration factors

Abstract

Purpose -The purpose of this paper is to present and implement an algorithm to prioritize welding quality deterioration factors for improving welding personnel performance. A case study is carried out in a piping components fabrication process which supplies these components to oil and gas production and processing facilities on the Norwegian continental shelf (NCS). Design/methodology/approach -The quality deterioration factors' prioritization is carried out using statistical methods in conjunction with the data recorded in the welding inspection database (WIDB) of the case study company. Data cleaning and rearrangements were performed to reflect final objective. Based on the welding procedure specifications (WPSs) and quality imperfection groups classified in NS-EN ISO 6520-1, the analysis is performed to prioritize the welding quality deterioration factors. Findings -Based on the WPSs and quality imperfection groups classified in NS-EN ISO 6520-1, it is possible to prioritize the welding quality deterioration factors. These factors are possible to use for improving the performance of welding personnel to assure the quality of welds in steel fabrications. Practical implications -The factors prioritized are possible to use for improving the performance of welding personnel to assure the quality and reliability of welds in a steel fabrication. Social implications -Assuring quality as proposed in the manuscript, the catastrophic failures that are potential in production and process plants can be mitigated. This enhances health, safety and environmental performance of welds in steel fabrications. Originality/value -The value of this paper is to illustrate an innovative approach to a real life quality problem; it demonstrates how the application of qualitative and quantitative quality instruments in accordance with technical specification can help in increasing and maintaining product compliance and in optimizing the management of resources.

Weld defect detection model based on machine vision

“…Welding is an exceptionally pivotal process in diverse domains including mechanical manufacturing, aerospace, hydroelectric power, marine drilling, shipbuilding, and electronics [1] . The process of welding is fundamentally intricate [2] , involving numerous variables [3] that exhibit significant interdependence, thereby rendering the entire welding process highly nonlinear. Variations in one or several variables during the welding process can result in deviations from the desired geometric dimensions of the post-welded seam or surface defects like undercutting, porosity, cracks, indentations, uneven weld widths, and lack of fusion.…”