Interfacial Bonding and Friction in Silicon Carbide [Filament]-Reinforced Ceramic- and Glass-Matrix Composites

Interfacial shear strength and interfacial sliding friction stress were assessed in unidirectional SiC-filament-reinforced reaction-bonded silicon nitride (RBSN) and borosilicate glass composites and 0/90 cross-ply reinforced borosilicate glass composite using a fiber pushout test technique. The interface debonding load and the maximum sliding friction load were measured for varying lengths of the embedded fibers by continuously monitoring the load during debonding and pushout of single fibers in finite-thickness specimens. The dependences of the debonding load and the maximum sliding friction load on the initial embedded lengths of the fibers were in agreement with nonlinear shear-lag models. An iterative regression procedure was used to evaluate the interfacial properties, shear debond strength (T _d ), and sliding friction stress (T _f ), from the embedded fiber length dependences of the debonding load and the maximum frictional sliding load, respectively. The shear-lag model and the analysis of sliding friction permit explicity evaluation of a coefficient of sliding friction (μ) and a residual compressive stress on the interface (σ₀). The cross-ply composite showed a significantly higher coefficient of interfacial friction as compared to the unidirectional composites.     

Influence of Residual Stresses, Interface Roughness, and Fiber Coatings on Interfacial Properties in Ceramic Composites

Fiber pushout tests are performed on zircon-matrix composites especially fabricated with a variety of silicon carbide reinforcing fibers and fiber coatings in order to create samples with different interfacial properties, surface roughness, and possibly in different states of residual stress to demonstrate their role on the interfacial and mechanical properties of fiber-reinforced composites. The data obtained from fiber pushout tests are analyzed using linear, shear-lag, and progressive debonding models to extract important interfacial properties, residual stresses, and surface roughness. The nature and magnitude of residual stresses in composites are independently characterized by measuring the coefficient of thermal expansion of the fiber, the matrix, and the composite for comparison with similar values measured using the fiber pushout tests. These results are then compared for self-consistency among different ways of analyzing data and with independently measured and calculated values. The results have shown that independent and complementary methods of data acquisition and analysis are required to fully understand interfacial properties in ceramic composites. In particular, independent measures of the coefficient of thermal expansion, residual stresses, and surface roughness are required to confidently interpret interfacial properties obtained by different analytical approaches and then relate them to the overall mechanical response of composites. It is also shown that composites with optimum mechanical response can be created by suitably engineering the interface using multiple fiber coatings.     

Carbon-Coated-Glass-Fiber-Reinforced Cement Composites: I, Fiber Pushout and Interfacial Properties

Interfacial mechanical properties of carbon-coated-S-glass-fiber-reinforced cement were characterized by a fiber pushout technique. The pushout experiments were conducted on model composites, where the S-glass monofilaments with and without carbon coating were unidirectionally embedded in ordinary portland cement. Interfacial properties, including bonding strength, frictional stress, residual stress, and fracture energy, were extracted from the previously developed progressive debonding model. The composite with a carbon interface exhibited a weaker interfacial bonding strength and frictional stress than did the composite without a carbon interface. The interfacial fracture energy of the composite with a carbon interface was 7.9 J/m², as compared to 47.6 J/m² for the composite without a carbon interface. The composite with the carbon interface exhibited a smaller residual clamping stress (18 MPa), in comparison to that for the composite without a carbon interface (69 MPa). Scanning electron microscopy observations indicated that the filament without a carbon coating was significantly attacked by the alkaline environment and was strongly bonded onto the matrix, whereas the filament with a carbon coating remained intact under the same curing conditions. These studies suggest that carbon coating provides the glass fiber with significantly improved corrosion resistance to alkali in the cement environment.     

Residual Stresses in Silicon Carbide–Zircon Composites from Thermal Expansion Measurements and Fiber Pushout Tests

Coefficients of thermal expansion (CTE) in the axial direction of two types of SiC fibers, monolithic zircon, monolithic SiC, and several SiC_f-zircon composites were measured in the temperature range of 50 to 1380C. The measured CTE values of composites were compared with values predicted by the rule-of-mixtures approach, and a small difference in measured and calculated values was ascribed to the nature of interfacial bonding and assumptions implicit in the rule-of-mixtures approach. Fiber pushout tests were performed on these composites and the residual stresses were extracted from the analysis of the load–displacement plots in terms of the shear-lag and progressive debonding models. The radial and axial residual stresses arising from the mismatch in CTE were calculated and compared with values obtained from the fiber pushout tests. The fiber pushout tests in general produced lower values of the residual stresses, but the residual stresses obtained using shear-lag analysis were in good agreement with the calculated values based on the CTE mismatch in composites with lower values of the interfacial shear stress. The influence of anisotropic fiber expansion in the radial and axial directions on the radial and axial residual stresses in composites were also examined.     

Interfacial Sliding Friction in Silicon Carbide–Borosilicate Glass Composites: A Comparison of Pullout and Pushout Tests

Interfacial sliding friction stress (τ_f) was assessed using both pushout and pullout tests on SiC-borosilicate glass composite specimens. Single-filament composite specimens were fabricated by heating to 950°C in argon borosilicate glass rods with fine-diameter (250-μm) capillary in which SiC filaments were inserted. The composite specimens prepared in this manner showed only frictional bonding. The maximum frictional sliding loads for pushout and the initial frictional sliding loads in pullout were measured as functions of the embedded length of the filament in the glass rods. The nonlinear variations of the frictional loads were analyzed using shear-lag models that include corrections for the effects of Poisson expansion or contraction on the sliding friction stress. It is shown that under identical conditions of composite fabrication the two tests give nearly identical properties for the interfaces. Pushout tests on hotpressed bulk composite specimens, however, showed both chemical bonding and a higher sliding friction stress relative to the single-filament capillary specimens. The presence of compressive residual stress on the filaments was independently confirmed by evidence of stress-induced birefringence.     

Evaluation of Porous ZrO2-SiO2 and Monazite Coatings Using NextelTM 720-Fiber-Reinforced Blackglas™ Minicomposites

A procedure was developed to fabricate oxide-fiber-reinforced minicomposites with a dense matrix and evaluate two oxidation-resistant interface coatings, porous oxide (zirconia-silica mixture) and monazite. The coatings were evaluated using Nextel^TM 720-fiber-reinforced Blackglas^TM-matrix minicomposites. Boron nitride (BN) coated and uncoated fibers were used as controls for comparison. The evaluation was based on ultimate failure strengths, fractography, and fiber pushin tests. All the composites that used fiber coatings had ultimate strengths significantly better than the control that used uncoated fibers. In addition, porous-oxide-coated fibers were found to be similar to BN-coated fibers in strength, fractography, and fiber pushin behavior. Monazite-coated fibers resulted in similar ultimate strengths but showed no appreciable fiber pullout. Fiber pushin tests showed that monazite debonds readily but frictional resistance is higher than for BN or porous oxide fiber coatings.     

Use of a Crack-Bridging Single-Fiber Pullout Test to Study Steel Fiber/Cementitious Matrix Composites

The fracture process of steel fiber/cementitious matrix composites has been studied using a single-fiber pullout test that permits detailed measurements of the load-crack opening displacement relationship during fiber debonding and unloading. Using a suitable analytical model, the interfacial fracture energy and interfacial sliding friction have been calculated for composites incorporating steel fibers with cement paste or mortar matrices. Comparison of theoretical debonding curves with the experimental data show that the model accurately represents the fiber debonding process, except for a decrease in interfacial sliding friction due to wear of matrix asperities at the interface. Differences between the calculated interfacial properties of several specimens are associated with changes in the interfacial microstructure.     

Interfacial Mechanical Characterization of Nicalon SiC Fiber/Alumina-Based Composites

Interfacial mechanical properties of both Nicalon SiC/aluminum borate and Nicalon SiC/aluminum phosphate with various fiber coatings and heat treatments were evaluated using a commercially-available indenter to induce fiber sliding during load cycling experiments. Varying degrees of sliding due to different coating materials were found. The interfacial characteristics including the shear, the residual axial fiber, and debond stresses were estimated by matching the experimental stress-displacement curves with curves predicted from an existing model. The elastic modulus and hardness of the interphase/interface in ceramic matrix composites were also evaluated. These results provided important insights into the ultimate mechanical performance of fiber-reinforced ceramic-matrix composites.     

Interfacial Mechanical Characterization of Nicalon SiC Fiber/Alumina-Based Composites

Interfacial mechanical properties of both Nicalon SiC/aluminum borate and Nicalon SiC/aluminum phosphate with various fiber coatings and heat treatments were evaluated using a commercially-available indenter to induce fiber sliding during load cycling experiments. Varying degrees of sliding due to different coating materials were found. The interfacial characteristics including the shear, the residual axial fiber, and debond stresses were estimated by matching the experimental stress-displacement curves with curves predicted from an existing model. The elastic modulus and hardness of the interphase/interface in ceramic matrix composites were also evaluated. These results provided important insights into the ultimate mechanical performance of fiber-reinforced ceramic-matrix composites.     

Predicted Effects of Interfacial Roughness on the Behavior of Selected Ceramic Composites

Potential effects of interfacial roughness in ceramic composites were studied using a model that included the progressively increasing contribution of roughness with relative fiber/matrix displacement during debonding of the fiber/matrix interface. A parametric approach was used to study interfacial roughness in conjunction with other parameters such as the strength, radius, and volume fraction of the fiber. The progressive roughness contribution during initial fiber/matrix sliding caused a high effective coefficient of friction, as well as an increased clamping stress, which led to rapidly changing friction with increasing debond length. Calculated effects implied a potentially significant contribution to the behavior of real composite systems and the necessity for explicit consideration in the interpretation of experimental data to understand composite behavior correctly. In a tension test, the Poisson's contraction of the fiber may negate the effects of roughness, allowing an "effective constant shear stress" (tau) approximation. This was evaluated using a piecewise linear approximation to the progressive roughness model in an analysis of composite stress-strain behavior; for the Nicalon/SiC system, the effective tau value was lower than the values that would be obtained from fiber pushout tests and/or matrix crack spacings.     

Interface Properties in a Porous-Matrix Oxide Composite

This study focuses on the interfacial properties of a family of porous matrix oxide composites with uncoated fibers. Measurements of debond energy and sliding stress are made using a modified version of the established fiber push-in test. Modifications include the following: (i) use of a sphero-conical indenter (not a sharp-tipped one) to produce only elastic deformation of the fibers, and (ii) analysis of the loop width (instead of absolute displacements) to ascertain interface properties. The method obviates the need for indentation tests on reference (non-sliding) fibers. It also mitigates the problems associated with the elastic deformation of the surrounding matrix. The measured debond toughnesses (about 0.05 J/m²) are about two orders of magnitude lower than the fiber toughness. This ensures that debonding will occur when a matrix crack impinges on a fiber. Additionally, the sliding stresses are in the same range as those reported for C-coated Nicalon fibers in glass–ceramic matrices (about 5 MPa). The latter results are qualitatively consistent with the observed damage tolerance in these two seemingly disparate systems, as manifested in the degree of fiber pullout as well as the notch sensitivity of tensile strength.     

Analysis of Fiber Frictional Sliding in Fiber Bundle Pushout Test

A simple theoretical model is developed to analyze the fiber frictional sliding resistance in a fiber bundle pushout test. The effect of the radial constraint imposed by the neighboring fibers on the stress transfer and frictional pushout stress is included in this analysis. Comparisons of theoretical results of this study and those of two existing fiber pushout models (i.e., single-fiber pushout and three–cylinder) are also presented. For SiC–RBSN and SiC–glass composites with short embedded fiber lengths less than 1 mm, there is little difference between all these models. However, for larger embedded fiber lengths, the present model gives the highest frictional pushout stress caused by the more realistic radial constraint condition used in the analysis.     

Effect of Interfacial Roughness Parameters on the Fiber Pushout Behavior of a Model Composite

The effect of interfacial roughness on the frictional sliding in composites has been studied using fiber pushout and pushback tests on a model composite of Plexiglas rods in an epoxy matrix. Different extents of roughness were introduced on the Plexiglas rods and the resulting roughness profiles measured. The roughness profiles were characterized using six different roughness parameters. An attempt was made to find a correlation between the sliding resistance and the selected roughness parameters. A parameter defined as the maximum coefficient in the Fourier transform of the roughness profile was found to yield the best correlation. If the roughness introduced is periodic, then the pushout traces exhibit periodic dips, but the magnitude of this periodic dip is significantly smaller than the seating drop obtained from pushback tests.     

Measurement of Interfacial Shear Strength in SiC-Fiber/Si3N4 Composites

An indentation method for measuring shar strength in brittle matrix composites was applied to SiC-fiber/Si₃N₄-matrix samples. Three methods were used to manufacture the composites: reaction bonding of a Si/SiC preform, hot-pressing, and nitrogen-overpressure sintering. An indentation technique developed by Marshall for thin specimens was used to measure the shear strength of the interface and the interfacial friction stresses. This was done by inverting the sample after the initial push through and retesting the pushed fibers. SEM observations showed that the shear strength was determined by the degree of reaction between the fiber and the matrix unless the fiber was pushed out of its (well-bonded) sheath.     

Effect of a Boron Nitride Interphase That Debonds between the Interphase and the Matrix in SiC/SiC Composites

Typically, the debonding and sliding interface enabling fiber pullout for SiC-fiber-reinforced SiC-matrix composites with BN-based interphases occurs between the fiber and the interphase. Recently, composites have been fabricated where interface debonding and sliding occur between the BN interphase and the matrix. This results in two major improvements in mechanical properties. First, significantly higher failure strains were attained due to the lower interfacial shear strength with no loss in ultimate strength properties of the composites. Second, significantly longer stress-rupture times at higher stresses were observed in air at 815°3C. In addition, no loss in mechanical properties was observed for composites that did not possess a thin carbon layer between the fiber and the interphase when subjected to burner-rig exposure. Two primary factors were hypothesized for the occurrence of debonding and sliding between the BN interphase and the SiC matrix: a weaker interface at the BN/matrix interface than the fiber/BN interface and a residual tensile/shear stress-state at the BN/matrix interface of melt-infiltrated composites. Also, the occurrence of outside debonding was believed to occur during composite fabrication, i.e., on cooldown after molten silicon infiltration.     

A new method for study of the fiber-matrix interface in composites: single fiber pull-out from a microcomposite

—A new method, single fiber pull-out from a microcomposite (SFPOM), was developed to study the fiber/matrix interface in composites. By pulling a fiber out of a seven-fiber microcomposite, the SFPOM test provides the real feeling of a fiber pulled out of an environment similar to that in a real composite. Interfacial shear strength decreased as the fiber volume fraction increased in the fiber-matrix system tested in the experiment. Three factors were suggested to be responsible for the phenomenon: (1) poor bonding between fibers when close to each other; (2) shear stress concentration in the matrix between neighboring fibers; and (3) possible change in matrix properties, thus altering the failure mechanism from interfacial debonding to a mixture of interfacial debonding and matrix fracture.     

Friction and Wear Properties of Si3N4/Carbon Fiber Composites with Aligned Microstructure

The friction and wear properties of silicon nitride/carbon fiber composites have been assessed and compared with monolithic Si₃N₄. Three different types of composites have been produced; one in which both the Si₃N₄ grains and the carbon fibers were aligned, one in which only the fibers had alignment, and a third where both the grains and fibers had random orientation. The friction coefficients of all of the composites, following running in, were around 0.2–0.3, typically less than one-third of that of the monolithic material. However there was no significant difference in friction coefficient between the three different types of composite. The specific wear rates of all the materials decreased with sliding distance and those of the composites were lower than the monolithic material. Among the composites, the wear rate of the one with aligned fibers in a randomly oriented Si₃N₄ matrix showed no dependence on sliding direction relative to the fiber alignment, and the specific wear rates of these samples were similar to that of the randomly oriented fiber composite, indicating little effect of fiber alignment alone on the wear properties under the present testing conditions. However, the specific wear rate of the composite with both fiber and grain alignment showed directional dependence. Grain cracking was observed perpendicular to the sliding direction, and the Spara specimen, in which the sliding direction was parallel to the Si₃N₄ grain alignment, showed higher wear rates than the Sperp and N samples of this composite. Such cracks are perpendicular to the major axis of the grains in the Spara sample and are thought to lead to easier removal of grains following their cracking under the tensile stresses induced particularly during the running in period.     

Fiber–Matrix Interfacial Characteristics in a Fiber-Reinforced Ceramic-Matrix Composite

Fiber—matrix interfacial shear stress and mechanical properties in zirconmatrix composites uniaxially re-inforced with either uncoated or BN-coated silicon carbide filaments were measured in the as-fabricated state and after a thermal treatment at 1300°C for 100 h. All the tested composites showed higher fiber—matrix interfacial debond stress than frictional stress and toughened-composite behavior. Mechanical properties of composites reinforced with uncoated filaments were unaffected by an annealing treatment at 1300°C for 100 h. In contrast, an identical annealing treatment given to a composite reinforced with BN-coated filaments led to a decrease in first matrix cracking stress and ultimate strength. These observations were related to the changes in the measured interfacial shear stress values as a result of the annealing treatment.     

Multiple Cracking of Unidirectional and Cross-PlyCeramic Matrix Composites

This paper examines the multiple cracking behavior of unidirectional and cross-ply ceramic matrix composites. For unidirectional composites, a model of concentric cylinders with finite crack spacing and debonding length is introduced. Stresses in the fiber and matrix are found and then applied to predict the composite moduli. Using an energy balance method, critical stresses for matrix cracking initiation are predicted. Effects of interfacial shear stress, debonding length and bonding energy on the critical stress are studied. All the three composite systems examined show that the critical stress for the completely debonded case is lower than that for the perfectly bonded case. For cross-ply composites, an extensive study has been made for the transverse cracking in 90° plies and the matrix cracking in 0° plies. One transverse cracking and four matrix cracking modes are studied, and closed-form solutions of the critical stresses are obtained. The results indicate that the case of combined matrix and transverse crackings with associated fiber/matrix interfacial sliding in the 0° plies gives the lowest critical stress for matrix cracking. The theoretical predictions are compared with experimental data of SiC/CAS cross-ply composites; both results demonstrated that an increase in the transverse ply thickness reduces the critical stress for matrix cracking in the longitudinal plies. The effects of fiber volume fraction and fiber modulus on the critical stress have been quantified. Thermal residual stresses are included in the analysis.     

Temperature Dependence of Interfacial Shear Strength in SiC-Fiber-Reinforced Reaction-Bonded Silicon Nitride

The interfacial shear strength of AVCO SCS-6 SiC-fiber-reinforced reaction-bonded Si₃N₄ (RBSN) composites was studied as a function of temperature. Fiber "push-through" experiments were conducted with a diamond indenter and a high-temperature microhardness tester. The interfacial shear strength was variable and depended mostly on interfacial bonding at low temperatures (5 to 18 MPa at room temperature) and frictional forces at high temperatures (12 to 32 MPa at 1300°C). The frictional component is attributed to the surface roughness of the fibers. The interfacial shear strength increased with temperature, because of the relief of residual stresses arising from the thermal expansion mismatch between fiber and matrix. Because of the composite nature of these fibers, a number of interfaces were tested in each experiment. The interface which debonded and slid was not always the same. Interfacial fracture took place either between the two outermost carbon layers of the SCS-6 fibers, or between the SiC core and the innermost of the two outer carbon layers. The outermost carbon layer of the fiber always stayed bonded to the Si₃N₄ matrix.