Reactive alloy melt infiltration for SiC composite matrices: Mechanistic insights

Abstract

A comparative study of reactive melt infiltration using Si and Si‐Y alloys is presented to provide insight into the governing processes that control the effectiveness of the melt interaction with a carbonaceous preform and the temperature capability of the SiC matrix for ceramic matrix composites. Through experiments on two substantially different scales of capillaries in porous graphite tubes using Si and Si‐Y alloys, the current study has characterized the phenomena that play a role in the infiltration of the melt and its reaction with the preform. It is shown that (i) the interface reaction controls wetting in both large and small capillaries and the climb rate is enhanced by the presence of Y; (ii) reaction choking occurs at critical throats within the pore network, usually behind the infiltration front; and (iii) different residual silicides can form during reaction and upon cooling. A potential mechanism for SiC growth is described, and the implications for the interplay between SiC growth and infiltration are discussed.

Vapor‐mediated melt infiltration for synthesizing SiC composite matrices

“…This approach has the potential to increase the temperature capability of the matrix. [7][8][9] Previous studies on the mechanisms of Si and Si alloy infiltration in porous carbonaceous preforms identified three 1 A similar limitation arises from the use of Si as a bond coat in current generation CMCs.…”

“…issues that may limit the extent of penetration and, by extension, the thickness of composite that may be densified in this manner. [7][8][9] First, since molten Si does not naturally wet C surfaces, inward melt flow into a porous network is controlled by reactive wetting wherein the surface must be modified first to form SiC before it can be wetted by the melt. 7,10 Second, behind the leading melt front, SiC continues to form and constrains the flow, eventually choking the infiltration at the narrowest necks in the network.…”

“…[7][8][9] First, since molten Si does not naturally wet C surfaces, inward melt flow into a porous network is controlled by reactive wetting wherein the surface must be modified first to form SiC before it can be wetted by the melt. 7,10 Second, behind the leading melt front, SiC continues to form and constrains the flow, eventually choking the infiltration at the narrowest necks in the network. [7][8][9] Third, when alloys are designed to form refractory intermetallics following partial removal of the Si to form SiC, the emerging silicide may also block the melt flow, even if the pores are larger than those that would prevent choking with (pure) Si melts.…”

“…7,10 Second, behind the leading melt front, SiC continues to form and constrains the flow, eventually choking the infiltration at the narrowest necks in the network. [7][8][9] Third, when alloys are designed to form refractory intermetallics following partial removal of the Si to form SiC, the emerging silicide may also block the melt flow, even if the pores are larger than those that would prevent choking with (pure) Si melts. 11 New strategies are thus needed to improve infiltration rates, mitigate choking, and eliminate residual Si from the matrix.…”

“…A variant on the latter approach involves use of a Si alloy that captures residual Si (that remaining after consumption of C) in the form of a high‐temperature silicide. This approach has the potential to increase the temperature capability of the matrix 7–9 …”