Part A: applied scienc and manufacturing ELSEVIER Composites: Part A 30(1999)521-524 Crack deflection in ceramic composites and fiber coating design criteria Ronald J. Kerans", Triplicane A. Parthasarathy Air Force Research Laboratory, Materials and Manufacturing Directorate(WLMLLN), Wright-Patterson AFB, OH 45433-7817, USA Abstract It is widely understood that the remarkable toughnesses demonstrated by some ceramic composites are dependent upon deflection of matrix cracks into fiber/matrix interfacial cracks. Carbon and Bn fiber coatings are ideally suited for promotion of such crack deflection, but have the unfortunate weakness of limited oxidation resistance. The rational design of oxidation-resistant altermative fiber coatings is facilitated by consideration of the details of crack deflection, growth of interfacial cracks, and subsequent sliding along the interfacial crack surface. This work discusses some of these details and the implications regarding the design and testing of alternative fiber coating systems. c 1999 Elsevier Science Ltd. All rights reserved Keywords: Composites interfaces; Fiber coatings; B Debonding: Crack deflection 1. Introduction 2. Coating toughness requirements In the earliest and many subsequent ceramic composites, Experience with successful composites and theoretical fiber/matrix interfacial layers, serendipitously formed by treatments have led to the prevalent assumption that good degradation of the fiber, were sufficiently weak that they composite behavior requires that the debond crack tough- fractured under the influence of approaching matrix cracks ness and friction should both be quite low(see, for example, thereby protecting the fibers from the stress concentration of [5]). Reports of substantially improved properties of NICA- the matrix crack(see, for example, [1D). The resulting brid- LoN /C/SiC composites resulting from the sole processing ging and debonding behavior resulted in the ability to change of pre-treating the surface of the fiber [6, 7] have accommodate cracking and still carry load, thereby provid- motivated reexamination of these assumptions regarding ing relaxation and load transfer from over-stressed regions he upper limits of allowable values The behavior of the composite to less loaded regions[2] It is this yielding of the materials is not fundamentally different, but is suffi type of mechanism that provides high toughness; however application of such composites has been limited primarily as that they suggest new possibilities for interface coating a result of environmental degradation of the interfacial properties. Recent preliminary work using rough interface layer. It has become widely appreciated that the develop- formalism(see section below) to analyze pushout tests on ment of robust resistance to environmental degradation these composites [11] may provide a revision of the the major requirement to enable exploitation of the many perceived requirements and is summarized in this section desirable characteristics of ceramic composites [3]. This Both systems consist of pyrolytic carbon-coated ceramic equirement has inspired research into oxide coatings that grade NICaLoN fibers in a matrix of CVI SiC. In one will promote crack deflection [4]. Oxidation-resistant case, the fibers were subjected to a proprietary treatment to substitutes will differ in many ways from C and Bn. reduce oxygen levels at the surface of the fiber[8, 9].The Successful application of them will likely require a much composites made with treated fibers demonstrate more complex engineering of the coatings themselves and higher strength at the same strain-to-failure, much the overall composite system to obtain comparable results. matrix crack spacing and significantly different This will require an explicit appreciation of the details of the stress-strain behavior [6] crack deflection and sliding processe Unusual fiber pushout behavior of the high strength mate- rial has been analyzed using the rough interface formalism and a rather complex failure scenario. Perhaps the most important outcome of the work is the suggestion that the Corresponding author Fax: + 1-937-255-3007 fracture energy for debonding can be higher than often 1359-835X/99/S-see front matter O 1999 Elsevier Science Ltd. All rights reserved P:S1359-835X(98)00144-4Crack deflection in ceramic composites and fiber coating design criteria Ronald J. Kerans*, Triplicane A. Parthasarathy Air Force Research Laboratory, Materials and Manufacturing Directorate (WL/MLLN), Wright-Patterson AFB, OH 45433-7817, USA Abstract It is widely understood that the remarkable toughnesses demonstrated by some ceramic composites are dependent upon deflection of matrix cracks into fiber/matrix interfacial cracks. Carbon and BN fiber coatings are ideally suited for promotion of such crack deflection, but have the unfortunate weakness of limited oxidation resistance. The rational design of oxidation-resistant alternative fiber coatings is facilitated by consideration of the details of crack deflection, growth of interfacial cracks, and subsequent sliding along the interfacial crack surface. This work discusses some of these details and the implications regarding the design and testing of alternative fiber coating systems. q 1999 Elsevier Science Ltd. All rights reserved. Keywords: Composites interfaces; Fiber coatings; B. Debonding; Crack deflection 1. Introduction In the earliest and many subsequent ceramic composites, fiber/matrix interfacial layers, serendipitously formed by degradation of the fiber, were sufficiently weak that they fractured under the influence of approaching matrix cracks thereby protecting the fibers from the stress concentration of the matrix crack (see, for example, [1]). The resulting brid￾ging and debonding behavior resulted in the ability to accommodate cracking and still carry load, thereby provid￾ing relaxation and load transfer from over-stressed regions of the composite to less loaded regions [2]. It is this yielding type of mechanism that provides high toughness; however, application of such composites has been limited primarily as a result of environmental degradation of the interfacial layer. It has become widely appreciated that the develop￾ment of robust resistance to environmental degradation is the major requirement to enable exploitation of the many desirable characteristics of ceramic composites [3]. This requirement has inspired research into oxide coatings that will promote crack deflection [4]. Oxidation-resistant substitutes will differ in many ways from C and BN. Successful application of them will likely require a much more complex engineering of the coatings themselves and the overall composite system to obtain comparable results. This will require an explicit appreciation of the details of the crack deflection and sliding processes. 2. Coating toughness requirements Experience with successful composites and theoretical treatments have led to the prevalent assumption that good composite behavior requires that the debond crack tough￾ness and friction should both be quite low (see, for example, [5]). Reports of substantially improved properties of NICA￾LONe/C/SiC composites resulting from the sole processing change of pre-treating the surface of the fiber [6,7] have motivated reexamination of these assumptions regarding the upper limits of allowable values [8–10]. The behavior of the materials is not fundamentally different, but is suffi- ciently outside the assumed envelope of workable properties that they suggest new possibilities for interface coating properties. Recent preliminary work using rough interface formalism (see section below) to analyze pushout tests on these composites [11] may provide a revision of the perceived requirements and is summarized in this section. Both systems consist of pyrolytic carbon-coated ceramic grade NICALONe fibers in a matrix of CVI SiC. In one case, the fibers were subjected to a proprietary treatment to reduce oxygen levels at the surface of the fiber [8,9]. The composites made with treated fibers demonstrate 30% higher strength at the same strain-to-failure, much finer matrix crack spacing and significantly different tensile stress–strain behavior [6]. Unusual fiber pushout behavior of the high strength mate￾rial has been analyzed using the rough interface formalism and a rather complex failure scenario. Perhaps the most important outcome of the work is the suggestion that the fracture energy for debonding can be higher than often Composites: Part A 30 (1999) 521–524 1359-835X/99/$ – see front matter q 1999 Elsevier Science Ltd. All rights reserved. PII: S1359-835X(98)00144-4 * Corresponding author. Fax: 1 1-937-255-3007