作者: Necat Altinkok , Rasit Koker
DOI: 10.1016/J.MATDES.2004.02.014
关键词:
摘要: Abstract In this study, Al 2 O 3 /SiC particulate reinforced (aluminium matrix composites) AMCs, which was produced by using stir casting process, bending strength and hardening behaviour were obtained a back-propagation neural network that uses gradient descent learning algorithm. First, to prepare the training test (checking) set of network, some results experimentally recorded in file on computer. experiments, dual ceramic powder mix prepared chemical route then inserted A332 alloy melt condition process. While SiC particles supplied commercially, chemically from aluminium sulphate. prepared, heated up crucible furnace. Then cake [1] . This milled adjust particle size before casting. with different range added into liquid alloy. The effect reinforcing hardness resistance AMCs investigated. Mechanical tests showed 10 vol% composites decrease increasing size. networks module, (μm) used as input strength, also porous properties outputs. Then, trained (also known set). At end data check system accuracy. As result found successful prediction for any given AMCs.