Experimental and artificial neural network evaluation of aluminum matrix composites
Other Title(s)
دراسة عملية و باستخدام الشبكات العصبية الصناعية لتقييم المواد المركبة ذات أساس المنيوم
Dissertant
Thesis advisor
Comitee Members
al-Muzaffar, Mujtaba A.
Jayyad, Jumah Salman
Ulaywi, Jawad Kazim
Nassar, Amin A.
University
University of Basrah
Faculty
Engineering College
Department
Department of Mechanical Engineering
University Country
Iraq
Degree
Ph.D.
Degree Date
2013
English Abstract
The present study used the powder metallurgy technique to prepare a composites material using aluminum powder as the matrix with additions of : The binders/lubricants [stearic acid, zinc stearate, and wax] at 1.5wt% for each.
Reinforcement materials [Cu coated ZrO2, dual effect of Y2O3 and graphite] at different weight fraction.
After blending and mixing, the materials were compacted at different compaction pressures, then sintered at different sintering temperatures using a vacuum tube furnace under argon gas for one hour.
Various physical and mechanical properties measured such as density (green, sintering), porosity, roughness factor, hardness, tensile strength, number of cycle to fatigue failure, microstructure examinations by using transmitted polarized microscope, digital microscope and scanning electron microscope (SEM), and identification phase by using X-Ray diffraction.
The present investigation of primary concern with prediction of the characters of aluminum matrix composites with additions of different materials using artificial neural networks.
Two types of artificial neural networks were design and constructed : 1.
Prediction of porosity, hardness, tensile strength, under different compact pressure and sintering temperatures.
2.
Prediction of the number of cycles to failure under different bending stress.
The results of X-Ray diffraction shows that there is new phase exist after sintering for all sintering temperature and weight percentage, also that the hardness, tensile strength and number of cycle to fatigue are increased with increasing of the amount of ZrO2- Cu coated particles up to 6%wt.
The number of cycle to fatigue failure reduce to zero when the 15%graphite added to Al with addition of 10%wt Y2O3 at 640oC, 320MPa.
The artificial neural networks using training a logarithm of multi layer preceptron back propagation showed successful in prediction of aluminum matrix composites characteristics under different process conditions.
Main Subjects
No. of Pages
159
Table of Contents
Table of contents.
Abstract.
Abstract in Arabic.
Chapter One : Introduction.
Chapter Two : Literature review.
Chapter Three : Experimental work.
Chapter Four : Design and construction of ANNs models.
Chapter Five : Results and discussion.
Chapter Six : Conclusions and scope for further work.
References.
American Psychological Association (APA)
al-Hasan, Nuha Hadi Jasim. (2013). Experimental and artificial neural network evaluation of aluminum matrix composites. (Doctoral dissertations Theses and Dissertations Master). University of Basrah, Iraq
https://search.emarefa.net/detail/BIM-744764
Modern Language Association (MLA)
al-Hasan, Nuha Hadi Jasim. Experimental and artificial neural network evaluation of aluminum matrix composites. (Doctoral dissertations Theses and Dissertations Master). University of Basrah. (2013).
https://search.emarefa.net/detail/BIM-744764
American Medical Association (AMA)
al-Hasan, Nuha Hadi Jasim. (2013). Experimental and artificial neural network evaluation of aluminum matrix composites. (Doctoral dissertations Theses and Dissertations Master). University of Basrah, Iraq
https://search.emarefa.net/detail/BIM-744764
Language
English
Data Type
Arab Theses
Record ID
BIM-744764
