Robust Design of Friction Stir Welding Process for Joining Polymer and Aluminum with Adhesive Bonding Using Taguchi Method, Data Envelopment Analysis and Neural Networks
Artykuł w czasopiśmie
MNiSW
0
brak dyscyplin
| Status: | |
| Autorzy: | Kasgari Ahmadpour S., Aliha Mohammad Reza, Pietras Daniel, Choupani Naghdali, Sadowski Tomasz |
| Dyscypliny: | |
| Aby zobaczyć szczegóły należy się zalogować. | |
| Rok wydania: | 2026 |
| Wersja dokumentu: | Elektroniczna |
| Język: | angielski |
| Numer czasopisma: | 3 |
| Wolumen/Tom: | 62 |
| Strony: | 606 - 624 |
| Impact Factor: | 1,6 |
| Web of Science® Times Cited: | 0 |
| Scopus® Cytowania: | 0 |
| Bazy: | Web of Science | Scopus |
| Efekt badań statutowych | NIE |
| Materiał konferencyjny: | NIE |
| Publikacja OA: | NIE |
| Abstrakty: | angielski |
| This research addresses the experimental results conducted joining of dissimilar aluminum and polymer materials. Friction stir welding (FSW) joints were manufactured using aluminum alloy (Al 6061-T6) and polycarbonate (PC) sheets, and hybrid joints (a combination of FSW and adhesive) were fabricated using the same Al-PC materials, and a two-component epoxy paste (MasterBrace® 1438) was used as the adhesive. The experiment involved placing these sheets together, conducting FSW without additional material, and repeating the process with an epoxy layer in between. Parameters were set using Taguchi methodology and partial factorial design, varying spindle speed, welding speed, and welding type. With the Taguchi L27 Orthogonal matrix, twenty-seven tensile strength experiments were conducted on the joints with different input variables. Optimal manufacturing conditions include: a spindle speed of 1400 rev/min, welding speed of 50 mm/min, and repair welding on the epoxy adhesive sample, which were identified through (signal to noise) S/N analysis and confirmed by a data envelopment analysis (DEA) network model. In morphological analysis, longer void defects are created during conventional FSW, whereas application of the second FSW process to repair (improve) welding with increased shoulder diameter significantly reduced the size of these void defects. Additionally, repair welding decreases joint deformation and increases the ultimate tensile strength. Predictive analysis using ANN (artificial neural network), ANFIS (adaptive neuro-fuzzy inference system), SVM (support vector machine), and KNN (K-nearest neighbors) methods favored ANFIS, exhibiting the best improvement in comparison to direct prediction by the DEA network that was determined by mean square error (MSE). |
