Owing to the rapid advancement of composite materials technology, composites are not only replacing conventional materials in many applications but also creating new areas unique to themselves. In recent years, there has been increasing interests in textile reinforced composites due to the attractive capabilities for a variety of applications. Coupled with the ability of tailored design for composite materials, the unique properties of textile reinforced metal matrix composites (MMCs) include increased through-thickness moduli and strength, fracture toughness, and danage tolerances. Other distinct benefits of textile composites are their near-net-shape capability and the automated manufacturing. Due to the advantages of high temperature performance, superior dimensional stability, and cost-effective manufacturing, textile reinforced MMCs have high potential in the application of space structures or electro packaging parts
Considerable efforts have been made by many researchers in manufacturing of MMCs and in evaluating their various properties. However, reports on the integrated study from the processing of textile reinforced MMC to analytic characterization of the composites is very limited. In this dissertation, emphasis is placed on the development of optimized process conditions and prediction of thermomechanical properties of two-dimensional (2D) plain woven and three-dimensional (3D) orthogonal textile reinforced aluminum matrix composites. PAN-type carbon fibers and pure aluminum were utilized as reinforcements and matrices, respectively. In the fabrication of 3D orthogonal textile preforms, the multi-layer weaving machine has been constructed, and the fiber architecture was designed by considering the sequence of harness lifting. The PAN type carbon bundles of size 12K has been utilized for the stuffer and fill yarns, and 3K bundle of the same material has been used for warp yarns, which penetrate into the preform thickness. The number of stuffer yarn layers is 6 and that of fill yarn layers is 7. The thickness and the width of the preform were 7mm and 50mm, respectively. From the geometric model, the percent ratio of volume content of fill yarns, stuffer yarns, and warp yarns was calculated as 66:32:2.
In order to obtain the optimized processing conditions of MMCs, infiltration mechanism of melt aluninum has been identified from the analysis of the permeability and deformation of textile preforms and viscosity and surface tension of melt aluminum. In the flow model, both the one-dimensional flow and radial flow have been considered. From the flow analysis the major factor in the pressure loss was found due to the surface tension, which invokes bundle shrinkage and decrease the permeability. The theoretical minimum pressure for melt infiltration into one bundle was as low as 1MPa. In the process of pressure infiltration casting, the most important process parameter is the timing of pressure application. Several tests and the solidification analysis utilizing ProCAST program revealed that the optimum timing of pressure application corresponded to the time period of solidification state on the mold wall and liquid state of preform surface. In order to examine the pressure loss, the flow stress of the solid elements in the mold has been calculated. The required stress was 52.5 MPa, which is much higher than the theoretical pressure of 1 MPa. It was concluded that the most applied pressure has been lost in plastic deformation of the solidified layer in the mold.
From the oxidization test, carbon fibers were oxidized at the temperature above 400℃. Thus, in manufacturing of carbon fiber reinforced MMC by the pressure infiltration casting, the pre-heating temperature for the preform should be kept below 400℃. Above this temperature an inert gas or vacuum environment is required to prevent the carbon fiber damage.
The composite samples of 2D plain woven and 3D orthogonal textile reinforced aluminum matrix composites were fabricated by vacuum assisted pressure infiltration method using a permanent metallic mold. Plain-woven carbon fabrics of an approximate unit cell size, 10mm × 10mm, were used as reinforcement preforms. Ten layers of woven fabric were stacked. The preform diameter of 2D and 3D textile MMCs was 75mm, and the matrix material was a pure aluminum having a nominal composition of 99.8 wt%. The pure aluminum was melted at 800℃ in an electric furnace under vacuum atmosphere, $10^{-3}$ torr. After the preform was positioned, the molten aluminum was cast into a heated metallic mold at a temperature of 500℃. The hydraulic pressure of 60MPa was applied to obtain melt infiltration during casting and to maintain the preform geometry. Microstructure of the samples were observed by using an scanning electron microscope (SEM) to examine the matrix infiltration and to measure the geometric parameters of the fiber architecture. The fibers were fully infiltrated with aluminum without any visible defects. The fiber volume fraction of the sample was determined by calculating the weight of the fabrics per unit volume of the composite sample divided by the fiber density. The fiber volume percents of 2D and 3D textile MMCs were 42.2% and 51.3%, respectively.
In the characterization of MMCs, analytic models based on the unit cell have been developed to predict geometric characteristics, three-dimensional engineering constants and coefficient of thermal expansion (CTE) of 2D plain woven and 3D orthogonal textile composites. Based upon the assumption that the yarn cross-section of 2D plain woven composites is a lenticular shape, the fiber crimp angle and the fiber volume fraction were obtained. For the case of 3D orthogonal textile composites, cross-sections of each yarn were assumed rectangular. These assumptions of yarn cross-section were clearly observed from the microscopic examination of the samples. The thermoelastic model utilizes the coordinate transformation and the averaging of tiffness and compliance constants based on the volume of each reinforcement. In the prediction of thermal conductivity, the thermal-electrical analogy method has been applied. Apart from the previous geometry model, the fiber undulation has been simplified as a straight inclined shape. In the modeling, one quarter of the unit cell size was considered for the simple analysis. Each yarn was assumed as a block of homogenized properties, and the thermal resistance was calculated based on the electrical resistance averaging. In order to verify the model prediction, samples of 75mm diameter have been fabricated and five specimens have been obtained for the determination of Young`s modulus, coefficient of thermal expansion, and the thermal conductivity. Mechanical test and RUS (Resonant Ultrasound Spectroscopy) method were used for the measurement of elastic constants. The CTE measurement was conducted using straingages in a heating oven. The thermal conductivity in the thickness direction of the sample was measured using hot disk method. Relatively good agreements between the model predictions and the test results have been observed. The main reason for some discrepancies in the experimental results may be due to the minor warpage of preforms during the sample fabrication. Another possibility may be the existence of aluminum carbide $(Al_4C_3)$ which is formed at the surface of the fibers. It can be concluded that the proposed models can predict the geometric characteristics, three-dimensional engineering constants, CTE, and thermal conductivity of 2D and 3D textile reinforced Al matrix composites. By examining the effects of various parameters such as constituent materials, fabric patterns, and microstructures through a parametric study, optimal parameters for composite’s performance can be properly selected. From models of MMC processing and analytic characterization suggested in this study, composite material designer can make a judicious selection of fiber and matrix system, preform geometries and microstructures, and processing conditions for meeting specific composite performance requirements.
