Fiber Metal Matrix Composites Research Articles

Interfacial reactions in Ti–6AI–4V/σ(SM1240) fibre metal matrix composite

AbstractThe effect of thermal exposure on the stability and effectiveness of carbon and TiBx coatings on σ(SM1240) fibre in a Ti–6Al–4 V metal matrix composite was studied. The results suggest that extensive loss of carbon coating starts only after consumption of the TiBx layer and that the heat treatment atmosphere has little effect on the carbon coating loss. The activation energies for the formation of TiC in specimens heat treated in air and in vacuum are 241 and 257 kJ mol−1 respectively.MST/3046

Anisotropic Damage of Fiber-Reinforced MMC Using Overall Damage Analysis

An anisotropic damage model is proposed for fibrous metal matrix composites (MMC) with a ductile matrix. The model incorporates damage mechanics with micromechanical behavior. An overall damage tensor, M for the whole composite is used in this analysis. This formulation allows the proposed damage model to directly use the elastoplastic stiffness tensor obtained for the undamaged effective configuration. An explicit expression is obtained for the elastoplastic stiffness tensor for the damaged composite material. Numerical solutions are obtained for different types of laminate layups compared with experimental results.

FREE EDGE EFFECT ON RESIDUAL STRESSES AND DEBOND OF A COMPOSITE FIBRE/MATRIX INTERFACE

In this study, the free edge effect on the fibre matrix interface of a single fibre metal matrix composite has been investigated. It was found, using both an approximate elasticity model and finite...

Luminescence sensing of stress in Ti/Al 2O 3 fiber reinforced composites

The high temperature processing of fibrous metal matrix composites has been predicted to result in the development of large residual stresses because of a difference between the coefficients of thermal expansion (CTE) of the fibers and matrix. The R1 and R2 optical luminescence line shifts of single crystal α-Al 2O 3 fibers embedded in a titanium matrix have been measured using a fiber optic method and related to the reinforcing fiber's principal stresses. The fiber's axial stress component has been found to be more than twice that predicted by a concentric cylinder's (CTE difference) model, whilst the fiber's radial (and hoop) direction stresses are smaller than expected. We propose that the stress in these composites is affected by two presently unmodeled effects: an elevation of composite's residual stress due to the large CTE difference between the tooling used for processing of the composite and a relaxation of the transverse stress components by radial matrix cracking. The result has important consequences for the manufacture of metal matrix composites and for efforts to predict their mechanical properties.

Transverse strength of continuous fiber metal matrix composites

The composite limit flow stress for transverse loading of metal matrix composites reinforced with continuous fibers is calculated using the finite element method. The focus of this work is to compare how different models for these composites influence the resulting composite flow behavior under transverse loading. Cell models with regular square and hexagonal arrangements [1] are compared with a so-called “embedded cell model” as developed by Dietrich et al. [2,3]. In the models for regularly arranged fibers, the hexagonal arrangement is seen to result in the highest limit stress for volume fractions less than about 0.42, whereas the square arrangement of fibers loaded in the direction of nearest neighbors provides greater strengthening at higher volume fractions. When the square arrangement is loaded 45° to the nearest neighbor direction, virtually no strengthening is seen. The interference of fibers with flow paths is seen to play an important role in the strengthening mechanism for these composites with regularly arranged fibers. Additionally, the influence of matrix hardening as a strengthening mechanism in these composites is seen to increase with volume fraction due to increasing fiber interaction. This is strengthening that results over and above the strengthening due to the addition of fibers. The embedded cell model is seen to agree quite well with the hexagonal model when the matrix has perfectly plastic behavior. However, when the matrix work hardens, the embedded cell model results in a higher asymptotic reference stress than calculated with the hexagonal model.

Experimental and Finite Element Analytical Guidelines for Fabricating Continuous Fiber (SCS-6) Metal Matrix (Ti-6A1-4V) Composites via the Foil/Fiber/Foil Technique

The consolidation behavior of Ti-6A1-4V matrix SCS-6 silicon carbide fiber metal matrix composites fabricated using the foil/fiber/foil technique was studied by both experimental physical modeling and finite element analysis. The experiments were carried out at 8750C (T/Tm = 0.597 for the matrix material) for varying levels of uniaxial compressive stress and times to characterize the stages of consolidation observed prior to reaching full density. Three stages of consolidation were identified: (1) instantaneous plasticity, (2) time-dependent foil deformation prior to establishing contact between neighboring foils, and (3) time dependent consolidation to close the pores at each side of the fiber that remain after neighboring foils have contacted between the fibers. Finite element analysis using the commercial code ABAQUS was employed to model the process and to examine the role of fiber spacing and applied stress level on the time required to densify the composite. The simulation predictions for the times necessary to achieve various densities were in good agreement with the experimental observations. The FEA results were used to establish analytical relations that describe the entire densification process. These equations are quite general and can be used to characterize any foil/fiber/foil composite fabrication process since they include as variables the matrix flow properties, processing stress and temperature, foil thickness and fiber diameter. The results predicted a slowing in densification at the latter stages of pore closure. This slowing was attributed to the low stresses in the matrix near a pore. Also, as the fiber spacing decreases, the time to densify increases because the load per fiber is now smaller.

Effect of fibre surface morphology on stability of C/TiBxcoatings in Ti–6Al–4V/σ (SM1240) fibre metal matrix composite

AbstractSigmafibres (SM1240) produced by a chemical vapour deposition process using a 15 μm tungsten wire corefor SiC deposition have a duplex coating of graphitic carbon and TiBx. Nodules present on the fibre surface are attributed to the deposition of the carbon coating over soot particles present on the substrate. Both the carbon and TiBx coatings were stable in vacuum or air at temperatures up to 973 K. The nodules werefound to be sites of preferential attack by the titanium alloy matrix. The average number of nodules per fibre decreased more rapidly when the specimens were heated in air than in vacuum. It is suggested that the nodules may reduce the stability temperature of the coatings.MST/2028

Tensile Property Evaluation of Continuous Fibre MMC

A tensile test method is described for continuous fibre metal matrix composites and used to evaluate the properties of aluminium alloy reinforced with Altex fibre. The composite, which was manufactured to a high quality using liquid-metal infiltration, was shown to possess high longitudinal modulus and strength and to have a transverse strength comparable with that of unreinforced alloy.

Modeling of thermal cycling creep damage of short fiber metal matrix composites

An analytical damage model is proposed to estimate the inelastic strain accumulation in short fiber metal matrix composites subjected to a combination of mechanical and cyclic thermal loading. Based on experimental observations, the enhanced non-linear strain accumulation associated with thermal cycling is attributed to damage mechanisms at the fiber-matrix interface. The proposed approach models the composite as one consisting of both perfectly bonded and debonded short fibers where the progression of debonding is described by an evolution law.

Preparation of solid and hollow diamond fibres and the potential for diamond fibre metal matrix composites

The development of techniques to grow diamond thin films using chemical vapour deposition (CVD) is now an area of active world-wide research [1, 2] and the unique physical and chemical properties of diamond promise many potential applications in optical components, semiconducting devices and hard wear-resistant coatings [3, 4]. Previous work has focused primarily on planar silicon on molybdenum substrates, but some preliminary results have recently been reported for coatings on wires [5, 6]. This paper describes a technique for producing uniform diamond coatings on the surface of metallic wires or ceramic fibres and the production of free-standing diamond tubes. The factors that affect the quality of the diamond fibres are discussed and their potential use in reinforced composites is considered. In the present experiments, diamond-coating was carried out in a standard hot filament CVD reactor [1, 2], in which CH 4 and H2 in a ratio of 1:100 were passed into a vacuum chamber at a total flow rate of 200 standard cm 3 min -I and a pressure of about 4000 Pa. A Ta filament held at 2000 °C dissociated the gases allowing carbon to deposit on to the surface of the wires and fibres in the form of a polycrystalline diamond film at a rate of about 0.5/xmh -1. If the wire was placed parallel to, and a few millimetres from, the filament (as for planar substrates) the uniformity of the diamond-coating was limited by the thickness of the wire, since diamond grew fastest on the side of the wire facing the filament. This effect became noticeable for wires and fibres with diameter > 250/zm, placing an upper limit upon the thickness of wires or fibres that can be uniformly coated by this method of around 300/zm. Alternatively, if the wire was positioned centrally and coaxially within the coils of the filament, uniform coatings on wires and fibres with a wide range of diameters were achieved. In this case, for thicker wires or fibres (even up to a few millimetres diameter), the diameter of the filament coils was simply increased to maintain an optimum distance of about 4-5 mm between the surface of the wire and the filament. This ensured that the wire was heated to a sufficient temperature to favour diamond deposition (typically about 900 °C), and also that the

Predicting High Temperature Ultimate Strength of Continuous Fiber Metal Matrix Composites

A model to predict the high temperature ultimate strength of a continuous fiber metal matrix composite (CFMMC) has been developed. The model extends the work of Rosen by including high temperature processes such as matrix creep, fiber-matrix de bond, and the effects of randomly spaced fiber breaks which typically exist in the MMC prior to loading. A finite element model (FEM), developed in the form of a representative volume element (RVE), is used to calculate the time-dependent stress field surrounding a fiber break. Variables included in the calculation are process-related parameters such as the fiber diameter, the fiber-matrix interface strength, and interface roughness. Statistical analysis is used to infer the strength of a large composite sample from the stress analysis of a single break provided by the FEM.

The Influence of Cool-Down Temperature Histories on the Residual Stresses in Fibrous Metal-Matrix Composites

The vanishing fiber diameter model together with the thermoviscoplastic ity theory based on overstress including a recovery of state formulation is introduced. They are employed to analyze the effects of temperature rate and of annealing at constant temperature on the residual stresses at room temperature when unidirectional fibrous metal-matrix composites are cooled down from 1000°C during the manufacturing process. For the present analysis the fibers are assumed to be transversely isotropic thermoelastic and the matrix constitutive equation is isotropic thermoviscoplastic including recovery of state. All material functions and constants can depend on current temperature. Yield sur faces and loading/unloading conditions are not used in the theory in which the inelastic strain rate is solely a function of the overstress, the difference between stress and the equi librium stress, a state variable of the theory. Assumed but realistic material elastic and vis coplastic properties as a function of temperature which are close to W/9Cr-lMo steel com posite permit the computation of residual stresses. Due to the viscoplasticity of the matrix time-dependent effects such as creep and change of residual stresses with time are found. It is found that the residual stresses at room temperature change considerably with temper ature history. The matrix residual stress, upon reaching room temperature, is highest for the fastest cooling rate, but after 30 days rest the influence of cooling rate is hardly notice able since relaxation takes place. Annealing periods can reduce the residual stresses com pared to continuous cooling.

Erratum to: Failure Diagrams for Unidirectional Fiber Metal Matrix Composites

Erratum to: Failure Diagrams for Unidirectional Fiber Metal Matrix Composites

Failure diagrams for unidirectional fiber metal-matrix composites

Pertinent failure processes in unidirectional fiber metal-matrix composites (MMCs) have been identified and analyzed. The critical conditions for interface delamination in several composites are compared with the theoretical delamination diagram proposed by He and Hutchinson, in which interface delamination and fiber fracture are delineated on the basis of their relative tough-ness values. It is shown that the delamination diagram does not provide information about the extent of interface cracking or the onset of fiber bridging. An alternative failure diagram that depicts composite fracture processes, such as matrix yielding followed by fiber fracture, inter-face cracking, and fiber bridging, is proposed. Development of the composite failure diagramvia micromechanical modeling of individual mechanisms is presented together with experimental results from the literature. Good correlation between theory and experiment suggests that the composite failure diagram might be used for tailoring composite properties through control of the dominant fracture mechanism.

Modeling of the consolidation of continuous- fiber metal matrix composites via foil- fiber- foil techniques

The consolidation of metal-matrix composites (MMC) via hot isostatic pressing (HIP) of foil-fiber-foil layups has been investigated using finite element method (FEM) metal flow analysis. For this purpose, the deformation pattern for various fiber arrangements was determined using representative unit cells to describe extremes in behavior. For a given fiber architecture, the consolidation time was found to be heavily dependent on the ratio of the HIP pressure to average flow stress and the friction conditions at the matrix-fiber interface. The specific influence of material properties such as the rate sensitivity of the flow stress appears to enter only as a second order effect. The FEM solutions were used to construct HIP diagrams delineating temperature-time-pressure combinations for full composite consolidation. Laboratory trials on subscale foil-fiber coupons were used to validate the FEM predictions.

Cavity formation during tensile straining of particulate and short fibre metal matrix composites

The formation of cavities in commercially pure aluminium composites, made by both powder and casting routes and reinforced with alumina (short fibres, angular particles and spherical particles), has been monitored using periodic density measurements during tensile testing and microstructural examinations. Stable cavities always form well before final failure, usually adjacent to the reinforcement, particularly when it is elongated in the loading direction and has a relatively flat surface normal to the stress axis. Sharp corners are not favoured cavitation sites and cavities can form at spherical particles, although the incidence is somewhat less than for angular particles. Cavitation occurred earlier for higher reinforcement contents and with powder-route, as opposed to cast, material, although the void contents and composite strains at failure were similar. A simple geometrical model is proposed, allowing prediction of the failure strain as a function of the reinforcement content, aspect ratio and strain to failure of the unreinforced matrix. The data presented are in good agreement with predictions from this model.

Overall damage and elastoplastic deformation in fibrous metal matrix composites

A phenomenological constitutive model for fibrous composite materials with a ductile matrix is postulated incorporating damage mechanics with micromechanical behavior. The model is first formulated in an undamaged composite system and then transformed consistently achieved in terms of an overall damage tensor M for the whole composite. In the process of formulating this model, interesting results are obtained demonstrating the necessity of using a non-associated flow rule for plasticity in the damaged composite system together with a Hill's type yield criterion. It is also shown that using a Ziegler-Prager kinematic hardening rule for the ductile matrix leads to a general kinematic hardening rule for the composite that is a combination of a generalized Ziegler-Prager model and a Phillips-type model. Finally, an explicit expression for the elastoplastic stiffness tensor for the damaged composite is obtained.

Thermoviscoplasticity Based on Overstress Applied to the Analysis of Fibrous Metal-Matrix Composites

The vanishing fiber diameter (VFD) model together with the thermovisco plasticity theory based on overstress (TVBO) are used to analyse the thermomechanical behavior of unidirectional fibrous metal-matrix composites. Fiber and matrix can both be transversely-isotropic, thermoelastic-viscoplastic. All material constants can depend on current temperature. Yield surfaces and loading/unloading conditions are not used in the theory in which the inelastic strain rate is solely a function of the overstress, the difference between stress and the equilibrium stress, a state variable of the theory. The three- dimensional equations are derived and specialized for various simple loading cases such as isothermal uniaxial and biaxial proportional and nonproportional loadings. The predic tions of this viscoplasticity theory during loading compare favorably with results from the rate-independent plasticity theory. In addition it is capable of predicting creep, relaxation and rate sensitivity.

A Numerical Simulation of the Effects of Volume Fraction, Creep and Thermal Cycling on the Behavior of Fibrous Metal-Matrix Composites

A previously derived cyclic neutral thermoviscoplasticity theory for fi brous metal-matrix composites (Yeh and Krempl [1]), which combines the vanishing fiber diameter (VFD) model with the thermoviscoplasticity theory based on overstress (TVBO) is specialized for transversely isotropic, thermoelastic fibers and an isotropic, thermovis coplastic matrix. Numerical experiments are used to illustrate the predictive capability of the theory. Simulations include the influence of volume fraction on the monotonic and cyclic loading behavior in the fiber and transverse directions with creep holds for B/A1 and B/Ti unidirectional composites and thermal cycling of B/A1. The three-dimensional theory permits the calculation of the actual Poisson's ratio in the tests. The results compare favor ably with sparsely available experimental results.

Interfacial stresses in a continuous fibre metal matrix composite

Finite element modeling was used to study the distribution of radial interfacial stresses in a SiC/Al composite caused by both thermally induced internal strain and by subsequently applied transverse tension. In particular, numerical results are presented for a composite consisting of an aluminum alloy (6061) matrix reinforced by large-diameter silicon carbide fibers. The model predictions are in good agreement with experimentally determined stress-strain curves, with an interface debonding stress of 93 MPa implied by imposing the recorded ultimate strain on the finite element model. 7 refs.