Conventional and spark plasma sintering of alumina parts produced by micro powder injection molding

• Main Author: Khong, Elmer Jun Ming.
• Other Authors: Khor Khiam Aik
• Format: Final Year Project
• Language: English
• Published: 2010
• Subjects: DRNTU::Engineering::Mechanical engineering; DRNTU::Engineering::Materials::Plasma treatment
• Online Access: http://hdl.handle.net/10356/39824
• Institution: Nanyang Technological University

Micro powder injection molding (μPIM) is a technique frequently used in the fabrication of ceramic parts with ceramic powders as the starting material. After the appropriate powder has been selected, the powder is mixed with binder components to form the feedstock. The presence of a multi-component binder among powder particles enables moldability to its desired shape in the subsequent injection molding process. Debinding is then performed on the green part to remove the binder components while retaining its desired shape. Debound samples are delicate due to the absence of binder components to hold the powder particles together; hence, sintering is an essential step thereafter to enable the powder particles to neck together at their interfaces. The fabrication of alumina samples with structures was analyzed in this study. Although focus was centered on the sintering stage in the μPIM process, green and debound samples were observed under the Scanning Electron Microscope to analyze the quality of the starting samples and for surface abnormalities which might be present. Conventional and spark plasma sintering were performed on pre-sintered samples at various sintering temperatures followed by characterization through measurements in relative density, microhardness, surface roughness and structure, channel and sample overall dimensional shrinkage. In this study, the feedstock was pre-mixed and the green parts were pre-molded. Complete binder removal was ensured during debinding to minimize the formation of defects after sintering. Debound samples were observed to have rounded corners and edges implying that incomplete filling of the mold cavity had occurred during injection molding. Dense, regular horizontal protrusions were also observed on the channel surface adjacent to the edge of the structures opposite the gate of the green part. Powder-binder separation was found to have occurred during the injection molding stage thus causing the formation of such defects. Conventionally sintered samples were observed to densify with increasing sintering temperature. Densification rates decreased when near theoretical density was achieved above temperatures of 1350°C where a relative density of 99.74% was obtained. Microhardness was found to increase correspondingly. Overall dimensional shrinkage was larger at higher temperatures due to increased densification while that of structures were found to be less significant. Roughness of both channel and structure surfaces were found to increase with increasing temperatures while roughness of channel surfaces was found to be distinctly lower than that of structure surfaces. Spark plasma sintering, advantageous for its lower sintering temperature and shorter dwell time however, produced samples of darker shades as sintering temperature was increased. Tests showed that contamination increased with increasing temperatures due to carbon diffusion from the graphite die and punches used. Observations made in the analyses of spark plasma sintered samples were similar to that in the conventional sintering study except that large standard deviations in microhardness and dimensional shrinkage were observed on samples sintered at lower sintering temperatures, indicating inconsistent results. Near theoretical density was achieved at temperatures above 1300°C with a corresponding relative density of 99.38%. The effect of number of sintering cycles and dwell time was found to be insignificant on spark plasma sintered samples.