Prediction of elastic modulus, yield strength, and tensile strength in biocompatible titanium alloys
J. Min. Metall. Sect. B-Metall., 60 (2) (2024) 273-282. DOI:10.2298/JMMB240221019M
Full text (pdf)
Export manuscript information:
RIS Format (EndNote, Reference Manager), BibTeX
Available online 26 September 2024
(Received 21 February 2024; Accepted 09 July 2024)
doiSerbia
Biocompatible titanium alloys possess a balanced set of improved mechanical properties and good biocompatibility, making them crucial materials in biomedical engineering. There is an increasing demand for these new alloys with superior properties. Furthermore, there is a need to understand the relationship between parameters and properties, and machine learning is being applied to make the whole process cheaper and more efficient. The aim of this study is to develop accurate machine learning models for predicting mechanical properties: modulus of elasticity, tensile strength, and yield strength, specifically using the Extra Trees Regressor model. Compared to the previous results, an improvement of the elastic modulus prediction model was observed after the inclusion of data on heat treatment parameters and Poisson’s ratio, as seen in the reduced MAE from 7.402 to 7.160 GPa. Models were built to predict the values of tensile strength and yield strength, where iron and tin were shown as most important features respectively, while the correlation coefficients for the test set were 0.893 and 0.868.
Keywords: Modulus of elasticity; Tensile strength; Yield strength; Biocompatibility; Machine learning
Correspondence Address:
G. Marković,
Institute for Technology of Nuclear and Other Mineral Raw Materials, Belgrade, Serbia;
email: g.markovic@itnms.ac.rs
This work is licensed under a
Creative Commons Attribution-
ShareAlike 4.0 International License