Machine learning models based wear performance prediction of AZ31/TiC composites

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F61989100%3A27230%2F25%3A10260311" target="_blank" >RIV/61989100:27230/25:10260311 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.webofscience.com/wos/woscc/full-record/WOS:001671760600003" target="_blank" >https://www.webofscience.com/wos/woscc/full-record/WOS:001671760600003</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1038/s41598-025-33417-5" target="_blank" >10.1038/s41598-025-33417-5</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Machine learning models based wear performance prediction of AZ31/TiC composites

Popis výsledku v původním jazyce

This study presents the fabrication of AZ31 magnesium matrix composites reinforced with 5, 10 and 15 vol% TiC particles using the Friction Stir Processing (FSP) technique and evaluates their wear behavior under varying loads (10-50 N) and sliding speeds (75-225 mm/s). The incorporation of TiC significantly enhanced the microstructural and mechanical properties of the composites. In particular, the AZ31/15 vol% TiC composite exhibited a refined grain structure with an average grain size of similar to 8 mu m, compared to similar to 60 mu m in the unreinforced AZ31 alloy. The same composite also demonstrated a substantial increase in hardness from 62 HV (base alloy) to 116 HV, highlighting the effectiveness of TiC reinforcement in improving strength. A key innovation of this work is the application of five machine learning (ML) algorithms, trained on experimental data using input features such as load, sliding speed and reinforcement content, to model and predict wear performance. After rigorous hyperparameter optimization, the Gradient boost algorithm achieved the highest predictive accuracy (R-2 = 0.9987), with errors falling within the range of experimental uncertainty. The study further includes residual analysis and computational efficiency assessment, supporting interpretable and robust AI-driven modeling. This integrated experimental-ML approach establishes a new benchmark for predictive modeling and data-driven material design in magnesium-based metal matrix composites.

Název v anglickém jazyce

Machine learning models based wear performance prediction of AZ31/TiC composites

Popis výsledku anglicky

This study presents the fabrication of AZ31 magnesium matrix composites reinforced with 5, 10 and 15 vol% TiC particles using the Friction Stir Processing (FSP) technique and evaluates their wear behavior under varying loads (10-50 N) and sliding speeds (75-225 mm/s). The incorporation of TiC significantly enhanced the microstructural and mechanical properties of the composites. In particular, the AZ31/15 vol% TiC composite exhibited a refined grain structure with an average grain size of similar to 8 mu m, compared to similar to 60 mu m in the unreinforced AZ31 alloy. The same composite also demonstrated a substantial increase in hardness from 62 HV (base alloy) to 116 HV, highlighting the effectiveness of TiC reinforcement in improving strength. A key innovation of this work is the application of five machine learning (ML) algorithms, trained on experimental data using input features such as load, sliding speed and reinforcement content, to model and predict wear performance. After rigorous hyperparameter optimization, the Gradient boost algorithm achieved the highest predictive accuracy (R-2 = 0.9987), with errors falling within the range of experimental uncertainty. The study further includes residual analysis and computational efficiency assessment, supporting interpretable and robust AI-driven modeling. This integrated experimental-ML approach establishes a new benchmark for predictive modeling and data-driven material design in magnesium-based metal matrix composites.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

20301 - Mechanical engineering

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2025

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

Scientific Reports

ISSN

2045-2322

e-ISSN

—

Svazek periodika

16

Číslo periodika v rámci svazku

1

Stát vydavatele periodika

GB - Spojené království Velké Británie a Severního Irska

Počet stran výsledku

22

Strana od-do

nestránkováno

Kód UT WoS článku

001671760600003

EID výsledku v databázi Scopus

2-s2.0-105028744541