ResearchBib Share Your Research, Maximize Your Social Impacts
Sign for Notice Everyday Sign up >> Login

Investigation of Friction Coefficient of β Titanium Alloy Aged At Different Heating Rates Using A Python-Based Support Vector Regression Model

Journal: Süleyman Demirel University Faculty of Arts and Science Journal of Science (Vol.17, No. 1)

Publication Date:

Authors : ;

Page : 239-246

Keywords : Heating rate; Wear resistance; Metastable β titanium alloy; Machine learning; Support vector regression model; Python;

Source : Downloadexternal Find it from : Google Scholarexternal


In this study, the effect of heating rate on the microstructural and wear properties of Ti-15V-3Al-3Sn-3Cr metastable β titanium alloy was investigated. Four heating rates, namely 0.4 °C/min, 4 °C/min, 25 °C/min, and 50 °C/min, were used during heating to the aging temperature. After heat treatment, the microstructural properties of the alloy were investigated by Scanning Electron Microscope (SEM) analysis. The mechanical properties of the alloy were determined by applying microhardness and wear tests. The finest α phases (393±43nm) were observed in the sample's microstructure, whose heating rate was 0.4 °C/min.. With the increase in the heating rate, the α phases grew, and thus the highest and the lowest microhardness were observed respectively at 0.4°C/min and at 50°C/min heating rates. Wear tests were carried out at loads of 3N and 10N and with two different sliding distances, 150 meters, and 400 meters. Mass loss after the wear test increased in all sample groups with increasing load and sliding distance. The friction coefficients were determined by applying statistical analysis to the data obtained from the wear tests. The data was divided into two sets, such as 40% test and 60% training. Model performance was evaluated by considering the mean square error, root means square error, and regression score value. The model was able to predict the friction coefficients of the samples heated to the aging temperature at different heating rates with an accuracy of above 76%.

Last modified: 2022-12-08 18:06:46