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.     

Effect-of Processing Varcables on Interfacial Properties of an SiG-Fiber-Reinforced Reaction-Bonded Si3N4 Matrix Composite

Fiber/matrix interfacial debonding and frictional sliding stresses were evaluated by single-fiber pushout tests on unidirectional continuous silicon-carbide-fiber-reinforced, reaction-bonded silicon nitride matrix composites. The debonding and maximum pushout loads required to overcome interfacial friction were obtained from load–displacement plots of pushout tests. Interfacial debonding and frictional sliding stresses were evaluated for composites with various fiber contents and fiber surface conditions (coated and uncoated), and after matrix densification by hot isostatic pressing (HIPing). For as-fabricated composites, both debonding and frictional sliding stresses decreased with increasing fiber content. The HIPed composites, however, exhibited higher interfacial debonding and frictional sliding stresses than those of the as-fabricated composites. These results were related to variations in axial and transverse residual stresses on fibers in the composites. A shear-lag model developed for a partially debonded composite, including full residual stress field, was employed to analyze the nonlinear dependence of maximum pushout load on embedded fiber length for as-fabricated and HIPed composites. Interfacial friction coefficients of 0.1–0.16 fitted the experimental data well. The extremely high debonding stress observed in uncoated fibers is believed to be due to strong chemical bonding between fiber and matrix.     

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.     

Analysis of a pull-out test with real specimen geometry. Part II: the effect of meniscus

This paper continues our study on the platelet model of the pull-out specimen, in which the matrix droplet shape is approximated by a set of thin parallel disks with the diameters varying along the embedded fiber. Using this model, the fiber tensile stress and the interfacial shear stress profiles were calculated for real-shaped matrix droplets, including menisci (wetting cones) on the fibers, taking into account residual thermal stresses and interfacial friction. Then, these profiles were used to numerically simulate the processes of crack initiation and propagation in the pull-out test and to obtain theoretical force-displacement curves for specimens with different embedded lengths and wetting cone angles. Our simulations showed that the interfacial crack in real-shaped droplets initiated at very small (practically zero) force applied to the fiber, in contrast to the popular ‘equivalent cylinder’ approximation. As a result, the equivalent cylinder approach underestimated the interfacial shear strength (IFSS) value determined from the pull-out test and at the same time overestimated the interfacial frictional stress; the smaller was the wetting cone angle, the greater the difference. We also investigated the effects of the embedded fiber length and interfacial frictional stress in debonded areas on the calculated IFSS. The simulated force–displacement curves for the real-shaped droplets showed better agreement with experimental curves than those plotted using the equivalent cylinder approach.     

Effect of Interface Roughness on Fiber Push-Out Stress

A theoretical analysis is given for single-fiber push-out with a rough interface whose asperity amplitude can be expressed by a Fourier series that has a good convergence. Based on a three-cylinder model for the single-fiber push-out test, the solutions of the fiber frictional push-out stress and the relative displacement of the fiber to the matrix at the fiber-matrix interface are obtained. A new asperity wear model is introduced in the analysis to simulate the degradation of the asperity amplitude during the fiber push-out process. The interfacial roughness between the fiber and the matrix is found to have a pronounced influence on the fiber sliding behavior.     

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.     

Theoretical Analysis of the Fiber Pullout and Pushout Tests

The fiber pullout and pushout tests have been analyzed to predict the load-displacement behavior in terms of fiber/matrix interface parameters. The effects of residual axial strain in the fiber and fiber surface topography were included. The residual axial strain was found to be a significant parameter. It is shown that the interface failure can be progressive or catastrophic. In the case of a progressive failure of the interface, the load-displacement curve is nonlinear. The portion of the curve from above the first nonlinearity to near the peak load can be predicted in terms of parameters of the interface, viz., the friction coefficient, the radial stress at the interface, the fracture toughness of the interface, and the residual axial strain in the fiber. Values for these parameters can be obtained from a single loaddeflection curve. The peak load and load drop, which are usually reported, are found not to be directly relatable to any interface property, since the length of the last portion of the fiber to debond is influenced by end effects and hence not easily predicted. However, for data which describe the peak load as a function of initial embedded length, that factor can be eliminated and the data reduced to yield the relevant interface parameters. In pullout, the peak and friction loads saturate with large specimen thickness. Catastrophic failure is favored when the debond initiation load is high or when residual stress is low. Finally, a methodology to extract interface parameters from experimental data is suggested.     

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.     

Theory and Analysis of Fracture Energy in Fiber Reinforced Composites

A micro-mechanics model is developed to analyze the stress distributions and fracture energies associated with crack propagation and fiber pull-out in reinforced composites. The stress and work mechanisms of interfacial debonding, fiber deformation, and the frictional work of fiber pull-out are considered as semi-independent contributions to fracture toughness. The theoretical expressions of Cottrell for frictional work W_F and Outwater and Murphy for fiber deformational work W_D are obtained as special relations in a general relation for the total work W_T = W_s + W_F + W_D where W_s defines the matrix shear work for interfacial debonding of fiber and matrix. Three dimensional diagrams of fracture energies W_T, W_s, or W_r versus interfacial shear bond strength λ₀ and frictional shear stress λ_f identify regions of optimized fracture energy. The influence of environmental degradation of bond strength upon fracture energy is analyzed in terms of the theory.     

Control of Interfacial Properties through Fiber Coatings: Monazite Coatings in Oxide–Oxide Composites

Fiber pushout tests were used to quantify the effects of fiber coating thickness on the mechanical properties of two model composite systems: a monazite-coated (LaPO₄-coated) alumina (Al₂O₃) fiber in an Al₂O₃ matrix and a LaPO₄-coated yttrium aluminum garnet (YAG) fiber in an Al₂O₃ matrix. Interface properties were quantified using the Liang and Hutchinson (LH) pushout model and mechanistically rationalized by considering the change in residual thermal stresses with changes in the coating thickness. Measures of the pure Mode II interfacial fracture energy, the coefficient of friction, and a radial clamping pressure are extracted by fitting the LH equations to the experimental results. Using the approach that has been developed herein, a methodology is available for measuring the interfacial properties, predicting the effect of coating thickness, and selecting the coating thickness to     

含脱粘界面纤维增强复合材料应力传递的理论分析

取具有脱粘界面的连续纤维增强复合材料的特征体积单元为研究对象,在常规剪滞模型的基础上通过引入摩擦力概念,并考虑横向泊松效应及基体径向力作用的影响,得到了纤维、基体的轴向应力及界面剪应力沿纤维方向的解析表达式。结果表明:本文所用的改进剪滞模型能较准确地反映各相介质沿纤维方向的应力分布特征,特别是较清晰地描述了脱粘界面的应力渐变以及界面粘结与脱粘临界处出现的界面剪应力跳跃现象,取得了与有限元解较为一致的结果。     

Design Guidelines for In-Plane Mechanical Properties of SiC Fiber-Reinforced Melt-Infiltrated SiC Composites

In-plane tensile stress–strain, tensile creep, and after-creep retained tensile properties of melt-infiltrated SiC–SiC composites reinforced with different fiber types were evaluated with an emphasis on obtaining simple or first-order microstructural design guidelines for these in-plane mechanical properties. Using the minimatrix approach to model stress–strain behavior and the results of this study, three basic general design criteria for stress and strain limits are formulated, namely a design stress limit , a design total strain limit , and an after-creep design retained strength limit . It is shown that these criteria can be useful for designing components for high-temperature applications.     

Profiling solid volume fraction in a conical bed of dry micrometric particles: Measurements and numerical implementations

In order to develop predictive process models in conical fluidized beds, there is an ongoing search for experimental methods and simulation tools to measure and model hydrodynamics parameters. Accordingly, experiments carried out in a conical fluidized bed containing micrometric TiO₂ particles with a wide particle size distribution. An optical fiber technique was employed to determine the effect of particles loading on the solid volume fraction. The axial and radial profiles of solid volume fraction have then been determined to evaluate the sensitivity of different models, including Syamlal-O?Brien (1988), Arastoopour et al., (1990) and Gidaspow (1994) drag models. The Eulerian-Eulerian model with frictional stress models and three different boundary conditions (BCs), consisting of no-slip, partial-slip and free-slip have been used in the numerical simulations. The Gidaspow model with the partial-slip BC gives the best agreement with experimental data for different particle loadings.     

Peak force as function of the embedded length in pull-out and microbond tests: effect of specimen geometry

We have derived the equations which explicitly express the peak force, F _max, and the apparent interfacial shear strength, τ _app, measured in the pull-out and microbond tests, as functions of the embedded length. Three types of test geometries were considered: (1) a fiber embedded in a cylindrical block of the matrix material; (2) microbond test with spherical matrix droplets; and (3) pull-out test in which the matrix droplet had the shape of a hemisphere. Our equations include the local interfacial shear strength (IFSS), τ _d, and the frictional interfacial stress, τ _f, as parameters; the effect of specimen geometry appeared in the form of dependency of the effective fiber volume fraction on the embedded length. The values of τ _d and τ _f were determined by fitting our theoretical curves to experimental F _max (l _e) plots by using the least squares method. Our analysis showed how the local IFSS and the frictional interfacial stress affected the measured F _max and τ _app values. In particular, it was revealed that intervals of embedded lengths could exist in which frictional interfacial stress had no effect on F _max and τ _app, even if the τ _f value was high. We also derived an equation relating the scatter in the interfacial strength parameters (τ _d and τ _f) to the scatter in τ _app, which is experimentally measurable, and proposed a procedure to determine the standard deviations of τ _d and τ _f from experimental pull-out and/or microbond test data.     

Stress Transfer in Single Fiber/ Resin Tensile Tests

Microscale (25 mm gauge length) “dogbone” resin specimens with single carbon fibers embedded through the length of the specimen have been studied as a method for determining the fiber-resin interphase strength. The specimens are pulled in tension until the fiber fragments to a critical length, l_c. Evidence is presented here, based primarily on the relaxation of stress birefringence around the fiber fragment, that this test may not be an unambiguous measure of fiber-resin adhesion. Data obtained for various production lots of AS-4, AS-6, and IM-6 fibers indicate an increase in l_cd with laminate tensile strength. Although there is theoretical justification for this correlation, it requires that the interphase shear strength is relatively constant.

In those instances where interfacial adhesion was expected to be low, i.e., surface contamination or unsurface treated fiber, there was a significant increase in l_c/d and usually a distinct difference in stress birefringence compared to “good” adhesion. However, the distinction in stress birefringence was not always clear cut.     

Influence of Interfacial Shear Stress on First-Matrix Cracking Stress in Ceramic-Matrix Composites

The first-matrix cracking stress and fiber-matrix interfacial shear stress were measured in zircon-matrix composites uniaxially reinforced with either uncoated or BN-coated silicon carbide filaments to study the role of intentional changes in interfacial shear stress on first-matrix cracking stress. The first-matrix cracking stress was measured by mechanical tests performed in either tension or flexure, and the filament-matrix interfacial shear stress was measured by a fiber pushout test. The first-matrix cracking stress was independent of the measured interfacial shear stress and did not conform to the predictions of a number of energy-based micromechanics models. In contrast, the first-matrix cracking stress showed a good correlation with the first-matrix cracking strain, which is hypothesized to be a more realistic criterion for first-matrix cracking in this class of filament-reinforced ceramic-matrix composites.     

Model of Oxidation‐Induced Fiber Fracture in SiC/SiC Composites

Stress rupture of SiC/SiC composites at intermediate temperatures in oxidizing environments is the result of a series of internal chemical and thermomechanical processes that lead to premature, localized fiber fracture. This article presents analytical models for two potentially critical steps in this process. The first involves the generation of tensile stresses in the fibers due to SiO₂ scale formation (following removal of fiber coatings) and the associated reduction in the applied stress required for fiber fracture. The second occurs once the gaps produced by coating removal are filled with oxide and subsequent oxidation occurs subject to the constraints imposed by the matrix crack faces. In this domain, the failure model is couched in terms of the stress intensification within the fibers caused by constrained oxidation. The models incorporate the combined kinetic effects of oxide growth and viscous flow. The competing effects of increased oxidation rate and accelerated stress relaxation with increasing temperature on fiber stress feature prominently in the results. The results suggest that, in dry air environments, the highest risk of fiber fracture occurs at temperatures in the range 840°C–940°C. In this range, the oxide scales grow at appreciable rates yet the resulting growth stresses cannot be mitigated sufficiently rapidly by viscous flow.     

Effective Fiber Properties to Incorporate Coating Thermoelastic Effects in Fiber/Matrix Composite Models

A method to incorporate the thermoelastic effects of fiber coatings into models of fiber/matrix composites was deter-mined. The coated fiber was replaced by an "effective" transversely isotropic fiber so that the properties of this effective fiber could be used in composite models. This ap-proach was used to determine the magnitude of errors re-sulting from neglect of the coatings in modeling fiber/matrix debonding and sliding and in reduction of data from real composites. The effects of carbon and BN coatings in the Nicalon/SiC system were found to be significant. It was found that significant errors could be expected from fitting models to experimental data if the compliance and coeffi-cient of thermal expansion of the coatings were ignored, even when the coatings were thin. Wide use of the approach required revision of composite models to allow inclusion of a transversely isotropic fiber. Such a revision was de-rived for a popular model of matrix cracking stress, and significant effects again were found to result from neglect of coatings.