Please use this identifier to cite or link to this item:
http://bura.brunel.ac.uk/handle/2438/31672
Title: | Intelligent development of high strength and ductile heat treatment-free Al-Si-Mg alloys for integrated die casting through the machine learning of experimental big data |
Authors: | Dong, X Liu, Q Han, W Yang, H Shan, Z Ji, S |
Keywords: | aluminum alloys;die casting;natural ageing;mechanical properties;machine learning;artificial intelligence |
Issue Date: | 15-Mar-2025 |
Publisher: | Elsevier |
Citation: | Dong, X. et al. (2025) 'Intelligent development of high strength and ductile heat treatment-free Al-Si-Mg alloys for integrated die casting through the machine learning of experimental big data', Journal of Alloys and Compounds, 1021, 179769, pp. 1 - 13. doi: 10.1016/j.jallcom.2025.179769. |
Abstract: | A group of twelve Al-xSi-yMg (x = 7–10, y = 0.3–0.6, in wt%) alloys were prepared by high pressure die casting, and the microstructure and tensile properties of the alloys were evaluated in the as-cast state and after natural ageing for 14 and 30 days, respectively. Based on the experimental big data, several machine learning (ML) models were applied for learning the relationship between the composition, natural ageing time and tensile properties of the die-cast alloys, and the performance of the ML models was evaluated by four parameters including mean absolute error (MAE), mean squared error (MSE), root mean squared error (RMSE) and R-square (R2). Among the applied ML models, the adaptive boosting (AdaBoost) was found as the most accurate and intelligent prediction of the yield strength (YS), ultimate tensile strength (UTS) and elongation (EL) of the die-cast alloys because of the lowest MAE of 1.06, 0.72 and 0.38 and the highest R2 of 0.96, 0.98 and 0.84. A high strength and ductile heat treatment-free Al8Si0.45Mg die-cast alloy was intelligently predicted by the AdaBoost model. The experimental validation showed that the alloy delivered the YS, UTS and EL of 150.6 ± 2.8 MPa, 288.2 ± 3.1 MPa and 11.74 ± 0.94 % in the as-cast state, and 160 ± 2.4 MPa, 294.5 ± 2.8 MPa and 10.87 ± 0.91 % after natural ageing for 30 days. The errors between the prediction and the experimental results were < 0.5 % for the strength and < 3.9 % for the ductility. This work provides a pathway for the intelligent development of high strength and ductile Al-Si-Mg heat treatment-free die-cast alloys for integrated die casting. |
Description: | Data availability:
The data that has been used is confidential. Supplementary material is available online at: https://www.sciencedirect.com/science/article/pii/S0925838825013271?via=ihub#sec0125 . |
URI: | https://bura.brunel.ac.uk/handle/2438/31672 |
DOI: | https://doi.org/10.1016/j.jallcom.2025.179769 |
ISSN: | 0925-8388 |
Other Identifiers: | ORCiD: Xixi Dong https://orcid.org/0000-0002-3128-1760 ORCiD: Shouxun Ji https://orcid.org/0000-0002-8103-8638 Article number: 179769 |
Appears in Collections: | Brunel Centre for Advanced Solidification Technology (BCAST) |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
FullText.pdf | Embargoed until 15 March 2026. Copyright © 2025 Elsevier B.V. All rights reserved. This manuscript version is made available under the CC-BY-NC-ND 4.0 license https://creativecommons.org/licenses/by-nc-nd/4.0/ (see: https://www.elsevier.com/about/policies/sharing). | 5.81 MB | Adobe PDF | View/Open |
This item is licensed under a Creative Commons License