Abstract

Traditionally fusion welding of manufactured components ‎and quality control inspection of such welds are distinctly ‎separate processes in the supply chain, which ultimately limit ‎productivity, throughput and increase re-work. The concept of ‎combining both of these practices directly at the point of ‎manufacture offers the potential to produce superior, globally-‎efficient fabrications. This paper presents the results of a study ‎investigating such a strategy, where a multi-pass weld is ‎autonomously deposited and, in parallel, an autonomous ‎inspection is deployed for real-time Non-Destructive ‎Evaluation (NDE). A real-time sector scan is implemented after ‎each of 21 weld passes and in three inspection positions, i.e., ‎the first position is 50 mm from the weld start, the second ‎‎being in the center of the weld length and the last is 50 ‎mm ‎distance to the weld endpoint. ‎Only in the second position, an ‏intentionally embedded defect, a tungsten rod, is introduced ‎into the multi-pass weld to allow subsequent in-process ‎calibration and verification. Based on the phased array ‎inspection results, the tungsten ‎rod is successfully detected in ‎the real-time NDE of the deposited ‎position. Furthermore, the ‎reflection ‎due to the partially ‎filled groove is captured during ‎the ‎inspection of the filling ‎passes. ‎
Original languageEnglish
Number of pages3
Publication statusPublished - 18 Jul 2019
Event46th Annual Review of Progress in Quantitative Nondestructive Evaluation
- Portland, Portland, United States
Duration: 14 Jul 201918 Jul 2019
Conference number: 46
https://event.asme.org/QNDE-2019
http://event.asme.org/QNDE-2019

Conference

Conference46th Annual Review of Progress in Quantitative Nondestructive Evaluation
Abbreviated titleQNDE 2019
Country/TerritoryUnited States
CityPortland
Period14/07/1918/07/19
Internet address

Keywords

  • weld inspection‎
  • in-process NDE
  • real-time Inspection‎
  • ultrasonic phased array‎

Fingerprint

Dive into the research topics of 'In-process inspection of multi-pass robotic welding'. Together they form a unique fingerprint.

Cite this