Volume 23, Issue 9 (September 2023)
Modares Mechanical Engineering 2023, 23(9): 521-530 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Modanloo V, Mashayekhi A, Akhoundi B. Prediction of formability of the metallic bipolar plates for PEM fuel cell in stamping process using meta-heuristics algorithms. Modares Mechanical Engineering 2023; 23 (9) :521-530
URL: http://mme.modares.ac.ir/article-15-70622-en.html
URL: http://mme.modares.ac.ir/article-15-70622-en.html
1- Mechanical Engineering Department, Sirjan University of Technology, Sirjan, Iran , v.modanloo@sirjantech.ac.ir
2- Mechanical Engineering Department, Sirjan University of Technology, Sirjan, Iran
2- Mechanical Engineering Department, Sirjan University of Technology, Sirjan, Iran
Abstract: (1560 Views)
In addition to the need for lightweight properties, the metallic bipolar plates in the PEM fuel cells should work in a humid and acidic environment. Due to its low density and excellent corrosion resistance, titanium is a proper candidate for manufacturing bipolar plates. In this paper, the manufacturing of bipolar plates made of commercially pure titanium with an initial thickness of 0.1 mm was investigated using the stamping process. A four-channel die with a parallel flow field was used in the experiments. To estimate the formability of microchannels of the bipolar plates, the response surface method, genetic algorithm, and adaptive neural fuzzy inference system were employed. Die clearance, stamping speed, and friction coefficient between the sheet and die were considered input variables, whereas the die filling rate was as output. The designed experiments using the response surface method were used to train the meta-heuristic techniques. The results showed that the regression model obtained from the response surface method predicts the die filling rate with acceptable accuracy. Furthermore, the coefficients of the equation obtained from the regression have been improved using the genetic algorithm and the error rate has been reduced by about 53%. Finally, an adaptive neural fuzzy inference system was used to predict the die filling. The results showed that the proposed system is very feasible and approximates the maximum filling rate with high accuracy.
Article Type: Original Research |
Subject:
Fuel Cell
Received: 2023/07/19 | Accepted: 2023/08/25 | Published: 2023/09/1
Received: 2023/07/19 | Accepted: 2023/08/25 | Published: 2023/09/1
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |