Annie Wang

(Advisor: Dr. Robert Speyer)

will defend her doctoral thesis entitled,

Sintering Methodologies For Silicon Carbide Ceramics

On

Thursday, November 30th at 11:00 a.m.

Love 295

Committee

• Prof. Robert Speyer – School of Materials Science and Engineering (Advisor)
• Prof. Rosario Gerhart– School of Materials Science and Engineering
• Prof. Naresh Thadhani– School of Materials Science and Engineering 
• Prof. Arun Gokhale– School of Materials Science and Engineering 
• Prof. Richard Neu– School of Mechanical  Engineering 

Abstract

Silicon carbide (SiC) ceramics are known for their high hardness, light weight, high strength, high oxidation resistance, high thermal shock resistance, low elevated temperature creep, and chemical inertness. Sintering of powder compacts has been via both eutectic liquid-phase and solid-state processes; both were investigated in this study.   

Solid-state sintering, following the method of Prochazka, requires both carbon and boron (or B4C) sintering aids.  In this work the use of C additives alone was shown to be necessary but insufficient for sintering.  The mechanical properties of SiC with varying B4C and C were studied with results of 98.31 to 99.66% relative density,  22.76 to 27.66 GPa for Vickers hardness and 3.0 to 4.18 MPa⋅m1/2 for Vickers indentation fracture toughness. The work showed that the merits of increasing B4C addition stopped at the solid solubility limit of B4C in SiC, demonstrated to be at ~0.26 wt%. 

To investigate the liquid-phase sintering methodologies for silicon carbide, 10 wt% of AlN and Y2O3 were added with a molar ratio of 3:2. The effect of different powder beds for the specimens to be immersed in, and different sintering atmospheres were studied. Four types of powder beds were investigated: pure SiC, 1:1 (wt%) SiC and AlN, the same composition used to make the samples, and pure AlN. It was found that the pure AlN powder bed yielded the highest relative density and finest grain size. This indicated that without the powder bed, the relatively high vapor pressure of AlN (or its vapor decomposition products) in the compact favored either evaporation/condensation particle coarsening or grain growth over sintering; the overpressure provided by the AlN powder bed surroundings thus improved sintering conditions.  

Four different atmospheres were then studied with the use of a 1:1 SiC and AlN powder bed. The results showed that different sintering dwell temperatures were required for optimum relative density using these different atmospheres.  Flowing He requires the lowest sintering dwell temperature (around 1700°C), followed by Ar, static vacuum, and then N2 requiring the highest temperature (~1950°C). These higher dwell temperatures were required from the more difficult diffusivity of larger molecular/atomic sized trapped gases out of sintered bodies of closed porosity.  Significant grain growth was observed for temperatures higher than their optimum temperatures, with associated decreasing sintered relative density. The highest relative density (96.37%) was achieved with an atmosphere created by pulling vacuum at room temperature, and then maintaining a static atmosphere during sintering. For optimally sintered specimens exposed to these atmospheres, lower Vickers hardness (15.03-18.35 GPa) were measured compared to solid-state sintered SiC, but very high Vickers indentation fracture toughness (2.92-7.85 MPa⋅m1/2) were obtained. This is associated with the relatively weak grain boundary phase deflecting/branching propagating cracks. 

This work then investigated the sintering of SiC with lower additive concentrations: 1-4 wt% of AlN and 0-2 wt% of Y2O3, using a flow-through He atmosphere, with the compacts immersed in a pure AlN powder bed. Relative densities were inferior to the previous study; it increased with increasing Y2O3 content. In the absence of Y2O3, AlN acted as a grain growth inhibitor, and points toward the potential merit of a Prochazka composition with AlN additions.   

A 2-D computer model of sintering was constructed using MATLAB. Green microstructures were represented in a 2-D view. The filled circles representing particles were generated with random number generator and a fall-and-roll algorithm. The sintering process was simulated with sequential algorithms of the initial, intermediate, and final stages of sintering. Each stage with controlling factors that could be input depicts microstructures that would result under differing conditions. The simulation depicted particle neck formation, particle re-shaping, pore elimination (densification), and grain growth, forming microstructures generally consistent with those observed after sintering.