Loading…

Quantifying the effects of cooling rates on Fe-rich intermetallics in recycled Al-Si-Cu alloys by machine learning

In this study, three casting processes named gravity casting (GC), vacuum casting (VC), and centrifugal casting (CC) were introduced to prepare the recycled Al-Si-Cu alloy in order to cover a wide cooling rates range of 15.77~75.25℃/s. Machine learning methods have been used to establish a regressio...

Full description

Saved in:
Bibliographic Details
Published in:Journal of alloys and compounds 2025-02, Vol.1014, p.178718, Article 178718
Main Authors: Li, Quan, Wang, Junsheng, Xue, Chengpeng, Miao, Yisheng, Hou, Qinghuai, Meng, Yanan, Yang, Xinghai, Li, Xingxing
Format: Article
Language:English
Subjects:
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this study, three casting processes named gravity casting (GC), vacuum casting (VC), and centrifugal casting (CC) were introduced to prepare the recycled Al-Si-Cu alloy in order to cover a wide cooling rates range of 15.77~75.25℃/s. Machine learning methods have been used to establish a regression model for predicting the Fe-rich intermetallic size of the recycled Al alloys under different cooling rates and verified by experiments. Further, the effects of cooling rate on the size, morphology, and percentage of Fe-rich intermetallics has been investigated, and the correlation between microstructure with mechanical properties were analyzed. With increasing cooling rate from 15.77℃/s to 75.25 ℃/s, the morphology of β-Fe gradually changed from coarse platelet to thinner and shorter shape, and finally to granular α-Fe particles. The cooling rate to fully suppressed the formation of β-Fe has been found to be 75.25℃/s, at which the percentage of α-Fe increased to 2.01%, which was increased by 25.65%, leading to the YS, UTS, and EL of the alloy increased by 12.56%, 11.01%, and 8.61%, respectively. After T6 treatment, the percentage of α-Fe in CC process reached 2.15%, which was increased by 12.05%, resulting in the improvement in elongation by 57.36%. By X-ray computed tomography, the statistical number density of Fe-rich intermetallics and microporosities were 4141.25mm-3 and 241.48mm-3 in VC process, while those were increased by 21.29% and 56.12% in CC process, respectively. Moreover, the quantified volume fraction of Fe-rich intermetallics and microporosities with equivalent diameter greater than 30μm were 0.20% and 0.59% in VC alloy, while those were decreased by 95.10% and 9.91% in CC alloy, respectively, indicating that high cooling rate promotes the nucleation of Fe-rich intermetallics and microporosity, thus refining its size. This study provides a guide for regulating the morphology of Fe-rich intermetallics in recycled Al alloys by controlling cooling rate. [Display omitted] •A machine learning model was established to predict the Fe-rich intermetallic size in recycled Al alloys.•Influences of various cooling rates (15.77~75.25℃/s) was successfully predicted by the model and verified by experiments.•3D distribution of Fe-rich intermetallics in different cooling rates was quantified by using X-ray CT.•The morphology evolution of Fe-rich intermetallics was revealed and the threshold cooling rate for suppressing detrimental β-Fe was found.
ISSN:0925-8388
DOI:10.1016/j.jallcom.2025.178718