TY - GEN
T1 - Wear performance analysis of Aluminum matrix composites using Artificial neural network
AU - Idrisi, Amir Hussain
AU - Hamid Ismail Mourad, Abdel
N1 - Publisher Copyright:
© 2019 IEEE.
PY - 2019/5/14
Y1 - 2019/5/14
N2 - This paper represents the study wear characteristics of AA5083/SiC micro and nanocomposite using the artificial neural network. The aluminum matrix composites with different wt % of SiC micro (5% and 10%) and nanoparticles (1% and 2%) were fabricated using stir casting route. The gears were fabricated using developed MMC and tested under wear application at room temperature. The rotation of the gears during the wear test was kept constant to 1450 rpm. The wear performance of gears was investigated with four levels of experiment time (30, 60, 90 and 120mins) and three levels of applied load (10N, 20N, and 30N). The response for the test was wear wt. (%). The wear (%) obtained for each MMC was used to train the neural network for evaluating the performance and prediction capability of the model for AMCs reinforced with micro- and nanoparticles using MATLAB's neural network toolbox. The best validation performance of the network was obtained at 12th (5.8×10-8) and 28th (9.1×10-8) epoch for micro and nano particles composite respectively. The MSE for the predicted regression plots was found to be 2.0×10-4 and 1.2×10-8 for micro and nano-particles composite data. Furthermore, the experimental and predicted results compared to check the accuracy of the network.
AB - This paper represents the study wear characteristics of AA5083/SiC micro and nanocomposite using the artificial neural network. The aluminum matrix composites with different wt % of SiC micro (5% and 10%) and nanoparticles (1% and 2%) were fabricated using stir casting route. The gears were fabricated using developed MMC and tested under wear application at room temperature. The rotation of the gears during the wear test was kept constant to 1450 rpm. The wear performance of gears was investigated with four levels of experiment time (30, 60, 90 and 120mins) and three levels of applied load (10N, 20N, and 30N). The response for the test was wear wt. (%). The wear (%) obtained for each MMC was used to train the neural network for evaluating the performance and prediction capability of the model for AMCs reinforced with micro- and nanoparticles using MATLAB's neural network toolbox. The best validation performance of the network was obtained at 12th (5.8×10-8) and 28th (9.1×10-8) epoch for micro and nano particles composite respectively. The MSE for the predicted regression plots was found to be 2.0×10-4 and 1.2×10-8 for micro and nano-particles composite data. Furthermore, the experimental and predicted results compared to check the accuracy of the network.
KW - Levenberg-Marquardt algorithm
KW - Metal matrix composite
KW - SiC nano and micro reinforcement
KW - feed forward back propagation neural network
KW - gears
KW - wear test
UR - http://www.scopus.com/inward/record.url?scp=85067005452&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85067005452&partnerID=8YFLogxK
U2 - 10.1109/ICASET.2019.8714330
DO - 10.1109/ICASET.2019.8714330
M3 - Conference contribution
AN - SCOPUS:85067005452
T3 - 2019 Advances in Science and Engineering Technology International Conferences, ASET 2019
BT - 2019 Advances in Science and Engineering Technology International Conferences, ASET 2019
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2019 Advances in Science and Engineering Technology International Conferences, ASET 2019
Y2 - 26 March 2019 through 10 April 2019
ER -