Abstract

Materials and parameters for conventional arc welding with the welding torch in vertical (PA) position are well known and investigated. However, apart from the ideal PA-position, not much is known about a deposition in so called ‘forced’ positions. This is becoming increasingly important in the case of particularly large components, as these can efficiently neither be clamped nor moved on a rotating and tilting table. More flexibility is achieved when the welding gun is moved as a tool on curvilinear paths to build up complex parts without necessary part movement. This link to the component surface means to leave the ideal vertical position. To make that possible, optimal parameters must be found for each angular position to enable high-quality and reliable build-up welding. In this work, a fuzzy logic-based system is designed based on set of experiments in single and double-layer weld beads structure using mild steel wire to predict the bead widths and heights for three different positions (Horizontal, Rising, and Falling torch movement). A comparison between the fuzzy and experimental values is studied. The averages of width mean error between experimental and fuzzy logic model values for the three positions were (1.3%) and (0.85%) for the single and double-layer weld beads, respectively. Averages of width mean error for the bead heights were (0.97%) and (0.73%) for the two structures. The current proposed study demonstrates a good agreement between the predicted fuzzy values and the experimental outcomes for the bead width and height.

Details

Title
Out-Of-Position Bead Geometry Prediction in Wire Arc Additive Manufacturing (WAAM) Using Fuzzy Logic-Based System
Author
Mustafa Mohammed Abdulrazaq 1   VIAFID ORCID Logo  ; AL-Khafaji, Mohanned M H 1   VIAFID ORCID Logo  ; Salahaddin, Abdulkader Kadauw 2   VIAFID ORCID Logo  ; Krinke, Stefan 3 ; Zeidler, Henning 3   VIAFID ORCID Logo 

 University of Technology, Iraq 
 University, Erbil TU Bergakademie Freiberg 
 TU Bergakademie Freiberg 
Pages
46-53
Publication year
2025
Publication date
2025
Publisher
De Gruyter Poland
ISSN
22990461
e-ISSN
24505781
Source type
Scholarly Journal
Language of publication
English
ProQuest document ID
3188371382
Copyright
© 2025. This work is published under http://creativecommons.org/licenses/by/4.0 (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.