Development of non linear thermal expansion model for casted aluminium silicon carbide

• Main Author: Roni Sahroni, Taufik
• Format: Thesis
• Language: English
• Published: 2012
• Online Access: http://psasir.upm.edu.my/id/eprint/48469/1/FK%202012%20115R.pdf; http://psasir.upm.edu.my/id/eprint/48469/

The complication of casting processes are known to be a significant influence on the overall of production cost. The successful of casting processes need knowledge in preparing of molds and patterns, melting and pouring, thermal and molten metal flow, solidification, and casting quality. Thermal expansion in the casting process is one of the most important parameters that influence the casting quality. Metal matrix composites (MMC’s) are engineered materials comprised of an alloy matrix and composite reinforcement that is embedded and transformed to improve the material property. Among several metal matrices composites, aluminium silicon carbide is selected because of its advantages such as elevated temperature, weight reduction and fatigue life improvement. However, the actual cast specimens or products are needed to be fabricated to analyze the mechanical properties of the cast metal matrix such as impact strength and yield strength. The purpose of this practice is to ensure casted aluminium silicon carbide to meet the expected strength and fatigue life. This lead to higher cost consumption and lead time at the design stage of casting product. Furthermore, the thermal behaviors in casting processes follow the non linear condition. The future trend of the current issue is how to predict the thermal behavior of metal matrix composite material in non linear condition by casting process without any experiments. The main objective of this research is to develop the non linear thermal expansion model for casted aluminium silicon carbide. In this research work, the non linear thermal expansion model for casted aluminium silicon carbide material is developed by the squeeze casting method. Three casting processes are applied in this research project, namely sand casting, low pressure die casting, and investment casting. Design case studies in casting processes are developed to present the thermal expansion behavior in low pressure die casting, investment casting, and sand casting. Experimental work of production tooling for aluminium silicon carbide is performed by sand casting process. The casting quality on surface roughness and dimensional accuracy are performed by using the appropriate equipments. It is observed that the result of testing on surface roughness and dimensional accuracy are complying with the standard in the sand casting process. The non linear thermal expansion model is developed for metal matrix composite material. The Coefficient of Thermal Expansion (CTEs) of aluminium silicon carbide fiber reinforced material is significantly influenced by the thermal stresses and interfaces between matrix and fibers. It is found that the thermal expansion behavior of the casted aluminium silicon carbide fiber reinforced composite relies on the thermal expansion of the fibers, and influenced by the onset of interfacial strength and residual stress state. The validation is conducted among the model, Rule of Mixture (ROM), and experimental result of LM6 alloys silicon reinforces, and showed a good agreement. The obtained Pearson’s, Kendall’s, and Spearman’s correlations value are 0.740, 0.949, and 0.975, respectively. It is concluded that coefficient of thermal expansion (CTE) has positive correlation with surface roughness (Ra). There is a statistically significant negative relationship between CTE and dimensional accuracy. The obtained Pearson’s, Kendall’s, and Spearman’s correlations value are -0.670, -0.949, and -0.975, respectively. It is concluded that when the amount of CTE increases, the dimensional accuracy is improved. In order to determine the performance of the model, the analysis of variance (ANOVA) is presented by using SPSS. It is concluded that there is no statistically different in accuracy between the experiment and the model. In addition, the test of inter-item correlation matrix shows the correlation at the high level of accuracy 99.9% (confidence level of 95%) between the experiment and the model.