Composites Part A Applied Science and Manufacturing

Published by Elsevier
Print ISSN: 1359-835X
A numerical method was used to study the interaction between a crack and the filler phase in a particle-reinforced polymer composite. The simulation was achieved by implementing a progressive damage-and-failure material model and element-removal technique through finite element analysis, providing a framework for the quantitative prediction of the deformation and fracture response of the composite. The effect of an interphase on composite toughness was also studied. Results show that a thin and high strength interphase results in efficient stress transfer between particle and matrix and causes the crack to deflect and propagate within the matrix. Alternatively, a thick and low strength interphase results in crack propagation within the interphase layer, and crack blunting. Further analysis of the effect of volume fraction and particle-particle interactions on fracture toughness as well as prediction of the fracture toughness can also be achieved within this framework.
This study examines the effect of stitching on the impact performance of a class of graphite/epoxy cross-ply laminates with the aim of investigating the ability of through-thickness reinforcement to improve the delamination resistance of laminates.Unstitched and stitched rectangular specimens (65 mm × 87.5 mm) were simply supported by a steel plate having a rectangular opening 45 mm × 67.5 mm in size and impacted at the center with energies ranging between 1 and 13 J. Stitched and unstitched laminates revealed similar structural performances in terms of force versus displacement response, energy absorption and residual indentation depth. It was also observed that whereas stitching does not appear capable of preventing the initiation and spread of delaminations, it induces a clear reduction of damage area when stitches bridge delaminations sufficiently developed in length.
Interfacial reactions in SCS-6 SiC/Ti–25Al–10Nb–3V–1Mo composites processed by fibre coating with matrix material, hot isostatic pressing and thermal treating for simulation of service conditions were studied by analytical transmission electron microscopy. In the as-processed specimen three reaction layers were observed. Adjacent to C coating of the SCS-6 fibre no mixture of TiC and Ti5Si3 was found reported in literature [Rhodes, C. G., Mat. Res. Soc. Symp. Proc., 1992, 273, 17; Baumann, S. F., Brindley, P. K. and Smith, S. D., Metall. Trans., 1990, 21A, 1559; Smith, P. R., Rhodes, C. G. and Revelos, W. C., in Interfaces in Metal–Ceramics Composites, eds R. Y. Lin, R. J. Arsenault, G. P. Martins and S. G. Fishman, TMS Press, New York, 1989, pp. 35–58; Badini, C., Terraris, M. and Marchetti, F., J. Mater. Sci., 1994, 2]. Instead, two sublayers were determined. Sublayer 1 close to C coating was identified as (Ti,V)C and sublayer 2 as (Ti,V,Nb)5Si3. The second layer is composed of larger equiaxed grains of (Ti,Nb)C and separated from matrix by a third layer of (Ti,Nb)5(Si,Al)3. Small amounts of (Ti,Nb)3(Si,Al) and (Ti,Nb)3(Al,Si)C were also identified. In the specimens heat treated at 700°C and 800° for up to 3000h, three, four or five reaction layers were found. In all specimens the reaction products of the first two layers are the same. The additional layers consist of the (Ti,Nb)3(Al,Si)C, the (Ti,Nb)3(Si,Al) or the (Ti,Nb)5(Si,Al)3 phase. These phases are arranged in different sequences. At the treatment temperature of 700°C the thickening of the interfacial reaction zone is mainly due to the growth of (Ti,Nb)3(Al,Si)C and (Ti,Nb)3(Si,Al). The growth of these phases is probably responsible for the slow decrease of mechanical properties of the composites during heat treatment. A kinetic analysis indicates that the growth of the reaction zone is a diffusion controlled process. The activation energy was determined to be 368kJ/mol for the total reaction zone, 302kJ/mol for (Ti,Nb)3(Al,Si)C and 181kJ/mol for (Ti,Nb)3(Si,Al) phase. The experimental results and a crystallographic analysis indicate that (Ti,Nb)3(Si,Al) is not a diffusion barrier for the carbon and titanium diffusion.
This paper is a review of low-velocity impact responses of composite materials. First the term ‘low-velocity impact’ is defined and major impact-induced damage modes are described from onset of damage through to final failure. Then, the effects of the composite's constituents on impact properties are discussed and post-impact performance is assessed in terms of residual strength.
Current and future potential applications for three-dimensional (3D) fibre reinforced polymer composites made by the textile processes of weaving, braiding, stitching and knitting are reviewed. 3D textile composites have a vast range of properties that are superior to traditional 2D laminates, however to date these properties have not been exploited for many applications. The scientific, technical and economic issues impeding the more widespread use of 3D textile composites are identified. Structures that have been made to demonstrate the possible uses of 3D composites are described, and these include applications in aircraft, marine craft, automobiles, civil infrastructure and medical prosthesis.
Charge transport properties, such as the temperature dependent dc conductivity and the frequency dependent conductance, of polymer matrix–metal particles composites, are investigated in the present study. Dc and ac conductivity is examined with varying parameters the filler content, temperature and the frequency in the case of ac field. The examined systems, though they are characterized as dielectrics, exhibit considerable conductivity, which alters by several orders of magnitude with temperature and frequency. The temperature and frequency dependence of conductivity gives evidence for the charge carriers transport mechanism via the occurred agreement of experimental results with the employed hopping models (variable range hopping model and random free-energy barrier model).
Shape memory polymers (SMPs) are a kind of very important smart polymers. In order to improve the properties or obtain new functions of SMPs, SMP composites and blends are prepared. We thoroughly examine the research in SMP composites and blends achieved by numerous research groups around the world. The preparation of SMPs composites and blends is mainly for five aims: (1) to improve shape recovery stress and mechanical properties; (2) to decrease shape recovery induction time by increasing thermal conductivity; (3) to create new polymer/polymer blends with shape-memory effect (SME); (4) to tune switch temperature, mechanical properties, and biomedical properties of SMPs; (5) to fabricate shape memory materials sensitive to electricity, magnetic, light and moisture. The trend of SMP composite development is discussed. SMP composites and blends exhibit novel properties that are different from the conventional SMPs and thus can be utilized in various applications.
Poly(ethylene 2,6-naphthalate) (PEN) nanocomposites reinforced with a very small quantity of carbon nanotube (CNT) were prepared by melt compounding using a twin-screw extruder. Morphological observations revealed that the modified CNT was uniformly dispersed in the PEN matrix and increased interfacial adhesion between the nanotubes and the polymer matrix, as compared to the untreated CNT. Furthermore, a very small quantity of the modified CNT can improve the mechanical and thermal properties of the PEN nanocomposites. This study also demonstrates that the non-isothermal crystallization behaviors of the PEN nanocomposites are strongly dependent on the presence of the modified CNT and cooling rate. The variations of the nucleation activity and activation energy for crystallization reflected the enhancement of crystallization of the PEN nanocomposites induced by the modified CNT. Combined Avrami and Ozawa analysis was found to be effective in describing the non-isothermal crystallization of the PEN nanocomposites in the presence of the modified CNT.
The interfacial behaviour of a 2.5D woven SiC/C/SiC composite under creep conditions has been studied by means of fibre push down tests performed with a nanoindentation system. For two samples tested in creep at 1000 and 1100 °C the interface was initially cohesioned whereas for a sample tested at 1300 °C the interface is partially decohesioned. Moreover, the interface of the sample tested at 1100 °C differs from that of the sample tested at 1000 °C because only sudden debond is observed. Fractographic and microstructural analyses have been performed in each specimen in order to define possible correlations between measured parameters such as fibre and bundle pullout, oxidation, temperature, etc. and interfacial mechanical properties.
The 2XXX series aluminium alloys reinforced with intermetallics present a special behaviour due to the reaction between matrix and reinforcement. This reaction forms an interphase that influences the mechanical and chemical behaviour of the composite, reducing the capability of the material to improve its properties after heat treatment. In this work, an approach is made to the study of this interphase, using particulate intermetallics with the same chemical and stoichiometric composition but obtained in different ways: mechanical alloying and gas atomising. A microstructural study was carried out by SEM, including qualitative analysis, showing the chemical gradient formed at the interphase both as-extruded and after T6 treatment. The mechanical behaviour of the interphase is studied through nanoindentation that allows the determination of hardness and Young modulus. Finally, all these properties are correlated with a fractrographic study of the fracture mechanisms. A harder interphase is formed for the mechanically alloyed system promoting a transgranular cleavage fracture micromechanisms, while intergranular cleavage fracture is found in atomised intermetallic containing composites.
The wear behaviour of varying size and weight fraction of particles up to 30 wt% Al2O3 particles reinforced 2024 Aluminium alloy Metal Matrix Composites (MMCs) fabricated by a vortex method was investigated in a pin on disc abrasion test apparatus against different SiC abrasives at room conditions. Wear tests performed under the load of 2 N against SiC abrasive papers of 20 (600 grit), 46 (320 grit) and 60 μm (240 grit). The effects of sliding distance, Al2O3 particle content and size, and abrasive grit sizes on the abrasive wear properties of the composites have been evaluated. The main wear mechanisms were identified using a scanning electron microscope. The results showed that Al2O3 particles reinforcement improved the abrasion resistance against all the abrasives used, and the abrasive wear resistance decreased with an increase in the sliding distance and the abrasive grit size. The wear resistance of the composites was considerably bigger than that of the aluminium alloy and increased with increasing Al2O3 particles content and size.
The effect of Friction Stir Processing (FSP) on the mechanical properties of 2618 aluminium alloy reinforced with 20% of alumina particles aluminium alloy has been studied in the present paper. The material was processed into the form of sheets of 7 mm thickness after T6 treatment and was tested in tension and fatigue at room temperature.Tensile tests were also performed at higher temperatures and different strain rates in the nugget zone, in order to analyse the superplastic properties of the recrystallized material and to observe the differences with the parent materials as a function of the strong grain refinement due to the Friction Stir Process. The high temperature behaviour of the material was studied, in longitudinal direction, by means of tensile tests in the temperature and strain rate ranges of 400–500 °C and 10−3–10−1 s−1, respectively.Fracture surfaces of the deformed fatigue test specimens were comprehensively examined in a scanning electron microscope equipped with field emission gun to determine the macroscopic fracture mode and characterize the fine-scale topography and microscopic mechanisms governing fatigue fracture.The mechanisms governing fatigue life, cyclic deformation and fracture characteristics are analysed in function of magnitude of applied stress, intrinsic micro structural evolution and material deformation behaviour.
Metal matrix composites reinforced with Al2O3 particles combine the matrix properties with those of the ceramic reinforcement, leading to higher stiffness and superior thermal stability with respect to the corresponding unreinforced alloys. However, their wide application as structural materials needs proper development of a suitable joining processes. The present work describes the results obtained from microstructural (optical and scanning electron microscopy) and mechanical evaluation (hardness, tensile and low-cycle fatigue tests) of an aluminium alloy (AA6061) matrix composite reinforced with 20 vol.% fraction of Al2O3 particles (W6A20A), welded using the friction stir welding process. The mechanical response of the FSW composite was compared with that of the base material and the results were discussed in the light of microstructural modifications induced by the FSW process on the aluminium alloy matrix and on the ceramic reinforcement. The FSW reduced the size of both particles reinforcement and aluminium grains and also led to overaging of the matrix alloys due to the frictional heating during welding. The FSW specimens, tested without any post-weld heat treatment or surface modification showed lower tensile strength and higher elongation to failure respect to the base material. The low-cycle fatigue life of the FSW composite was always lower than that of the base material, mainly at the lower strain-amplitude value. The cyclic stress response curves of the FSW composite showed evidence of progressive hardening to failure, at all cyclic strain-amplitudes, while the base material showed a progressive softening.
This article presents a model to simulate the RIFT process for complex geometries in advance. Compaction and permeability experiments were carried out for two types of preform. A significant difference between dry and wet preform compaction was observed. The model takes both into account and by doing so, the general assumption, that the RIFT process can be modeled as a quasi-static process, becomes invalid. Therefore, a fully transient model is proposed, including the preform compaction flux. Experiments were carried out to validate the model. It was found that using the wet and dry preform properties leads to a good prediction of the height distributions, flow front positions and filling times.
The aim of this study is to evaluate the possibility of using the linear friction welding (LFW) technique to produce sound joints on a 2124Al/25 vol.%SiCp composite. The MMC joints were subjected to microstructural and mechanical characterization, including hardness, tensile and fatigue tests, without any post-weld heat treatment. The microstructural analyses showed substantially defect-free joints, with a uniform particle distribution in the central zone and a relevant plastic flow of the aluminium matrix alloy. The hardness decrease in the welded zone was approximately 10% in respect to the base material. The joint efficiency was higher than 80%, both in respect to the ultimate tensile strength and fatigue strength at 107 cycles. S–N probability curves were calculated using the maximum likelihood method. Generally, the fracture occurred in the Thermo-Mechanically Affected Zone (TMAZ), with a relevant reduction in the elongation to failure.
Thermomechanical response of a cross-ply SCS-6/Timetal-21S composite subjected to a generic hypersonic flight profile with the temperature ranging from -130 C to 816 C was evaluated experimentally and analytically. A two dimensional micromechanical anlaysis, VISCOPLY, was used to predict the stress-strain response of the laminate and of the constituents in each ply during thermomechanical loading conditions. In the analysis, the fiber was modeled as elastic with transverse orthotropic and temperature dependent properties and the matrix was modeled using a thermoviscoplastic constitutive relation. The fiber transverse modulus was reduced in the analysis to simulate fiber-matrix interface failure. Reasonable agreement was found between measured and predicted laminate stress-strain response when fiber-matrix debonding was modeled.
The micromechanics of reinforcement of a model composite consisting of continuous high-modulus fibre embedded in epoxy resin has been investigated as a function of fibre sizing. The composite was subjected to incremental tensile loading up to full fragmentation, while the stress in the fibre was monitored at each level of applied strain with the new technique of remote laser Raman microscopy. The two systems exhibited differences in the residual stress field with the unsized fibre being in compression. The average stress in the fibre increased linearly with applied matrix strain up to first fracture. After fracture, the stress in the fibre was found to build from the tips of the fibre breaks, reaching a maximum value at the middle of each fragment. The shape of the stress transfer profiles indicated minor differences between the two systems at moderate strains. At high strains, the stress transfer profiles of the two systems were distinctly different possibly owing to the presence of two different interfacial failure modes in the two types of model composites. The maximum interfacial shear stress for both systems was of the order of 40 MPa with the sized system exhibiting slightly better adhesion. SEM examination of the fracture surfaces revealed clear interfacial failure for the unsized system whereas the sized system indicated areas of good adhesion.
The aim of the present investigation is to study the mechanical properties of as-cast and heat-treated ZA-27 zinc-aluminium alloy/graphite particulate composites containing graphite particles of size 90–150 μm and of content ranging from 0 to 5% by weight. The vortex method of production was employed, in which the graphite particles were poured into the vortex created by stirring the molten metal by means of a mechanical agitator. Heat treatment was performed at 280°C for durations of 1, 2, 3 and 4 h, respectively. The results of this study revealed that as graphite composition was increased, there were significant increases in the ductility, ultimate tensile strength (UTS) and compressive strength of the composite, accompanied by a tremendous drop in the hardness of the material. Heat treatment was found to have a similar effect. An attempt is made in the paper to furnish explanations for these phenomena.
2D triaxially braided carbon fiber composites were subjected to off-axis compression under static and low velocity impact loading. Three types of specimens with axial fiber tows at angles of 30°, 60°, and 75° to the direction of loading were studied under uniaxial compression loading using a modified compression loading fixture that has anti-buckling guides. Static tests were carried out using a hydraulically activated MTS loading frame, while, low velocity impact tests were conducted using a drop tower facility. A three parameter homogenized orthotropic plasticity model was used to characterize the inelastic response and the constants associated with this characterization were determined uniquely. Results from this characterization were subsequently used to analyze the rate dependent interfacial effects in these materials. Two types of composites made from two different resins, but with the same textile architecture were used for this study. The effect of the resin properties on the unusual rate effects exhibited by the composite is seen to be captured well with the three parameter orthotropic plasticity model adopted.
Textile preforming is the stitching, cutting, and assembling of reinforcement textiles to enhance mechanical properties or optimize the RTM-tool loading. The stitching of the reinforcing textile has direct influence on the permeability of the preform. In this paper the influence on permeability of two different stitching patterns with five different seam distances is described. The two-dimensional permeability has been determined continuously in a matched metal tool incorporating capacitive sensors. Beforehand, the glass twill weave textile has been thoroughly evaluated to determine the permeability behavior of the textile without stitching in dependence on the fiber volume fraction and the cavity height. The paper reveals the significant influence of the stitching seam distance and the stitching pattern on the permeability values K1 and K2, the orientation angle of the flow front ellipse, and the anisotropy of the preform for two different fiber volume contents.
A comparison of the as-fatigued and re-consolidated properties have been made between notched quasi-isotropic [0/45/90/−45]2S and cross-ply [0/90]4S AS4/PEEK laminates. For the former, the ±45° plies tend to constrain longitudinal damage development so that damage growth primarily occurred in the transverse direction, causing more widespread damage. This led to prominent mechanical properties degradation, shorter fatigue lives and lower residual strengths. For cross-ply laminates, quick and extensive longitudinal crack tangential to the hole and the corresponding 90° fiber shear off brought about effective stress concentration alleviation. This discouraged further damage development. Hence, their fatigue lives exceeded one million cycles even at high cyclic stress levels and their residual strengths were significantly higher than their virgin strength. On the other hand, the re-consolidation process removed most of the defects that alleviated the stress concentration and thus decreased the strengths. Detailed study of the residual strength changes and damage development history revealed that the residual as-fatigued and re-consolidated strengths were governed by the competition between local structural decay and its resulting stress concentration alleviation.
A phenylethynylphthalic anhydride terminated poly(etherimide) oligomer with a molecular weight of Mn=3000 g mole−1 has been synthesized for use in high-performance composite and adhesive applications. This reactive poly(etherimide) exhibits excellent thermo-oxidative stability and solvent resistance upon curing at high temperatures. Towpreg has been produced by coating unsized, G30-500, 12k carbon fiber tow with this oligomer in a powder form. A minimal dry powder prepregging technique was employed for rapid processing to circumvent the problems associated with high viscosity and to perform the operation in an organic solvent-free environment. This towpreg has been manufactured into flat composite laminates via manual lay-up and hot press consolidation and cure. The Mode I and Mode II fracture toughness of these composite panels has been investigated via the double cantilever beam (DCB) and end-notch flexure (ENF) test methods. Fracture surface and microstructure features are correlated with the measured mechanical properties.
Accurate cure kinetic model is crucial for correctly identifying the amount of heat generated for composite process simulation. This paper presents a new technique for identifying cure kinetics parameters for Hercules AS4/3502 prepreg by normalizing the DSC data. The cure kinetics is based on an autocatalytic model for the proposed method, which uses dynamic and isothermal DSC data to determine its parameters. Existing models are also used to determine kinetic parameters but rendered inadequate because of the material's temperature dependent final degree of cure. The model predictions determined from the new technique showed good agreement to both isothermal and dynamic DSC data. The final degree of cure was also in good agreement with experimental data.
A dynamic crash loading experiment is performed on polypropylene foam which is used in composite sandwich structures for safety applications. Several interrupted shocks are conducted, in between which, microtomographic acquisitions are made showing the change of the sample during its compression. These data can help construct and validate predictive models, although, because this material is multi-scale (constitutive beads at the mesoscopic scale are made of microscopic closed cells), image processing is required to extract useful quantitative measurements. Such processing is described here, so as to determine a representative volume for each bead of the sample, to associate values such as bead density to each bead and to each stage of the compression. This can help build a predictive model at the mesoscopic scale.
In this paper, a 3D macro/micro finite element analysis (FEA) modeling approach and a 3D macro/micro analytical modeling approach are proposed for predicting the failure strengths of 3D orthogonal woven CFRP composites. These approaches include two different scale levels, macro- and micro-level. At the macro-level, a relatively coarse structural model is used to study the overall response of the structure. At the micro-level, the laminate block microstructure is modeled in detail for investigating the failure mechanisms of 3D orthogonal woven CFRP composites. The FEA and analytical models developed previously [Tan P, Tong L, Steven GP. Modeling approaches for 3D orthogonal woven composites, Journal of Reinforced Plastics and Composites, 1998:17;545–577] are employed to predict the mechanical properties of 3D orthogonal woven CFRP composites. All models presented in this paper are validated by comparing the relevant predictions with the experimental results, which were reported earlier in Part I of the paper [Tan P, Tong L, Steven GP. Behavior of 3D orthogonal woven CFRP composites. Part I. Experimental investigation, Composites, Part A: Applied Science and Manufacturing, 2000:31;259–71]. The comparison shows that there is a good agreement for the mechanical properties. An acceptable agreement exists for the failure strength in the x or stuffer yarn direction even though the FEA model gives a lower bound and the analytical model gives an upper bound. However, for the failure strength in the y or filler yarn direction, the difference between the predicted and experimental results is significant due to primarily ignoring of the waviness of filler yarn in the models. A curved beam model, which considers the waviness of the filler yarn, will be presented in Part III of the paper.
The tensile fatigue properties of a specific type of 3D orthogonal woven composite that contained different volume contents of through-thickness reinforcing fibres (called z-binders) was studied in this paper. The 3D woven composite contained z-binder contents of 0.3, 0.5 or 1.1% by volume. The fatigue life (S–N curve) and residual fatigue strength (after one million load cycles) of the 3D woven composites was lower than a 2D woven composite. Furthermore, the fatigue life and strength of the 3D woven composite decreased with increasing z-binder content. The fatigue performance was degraded due to the development of fatigue-induced damage caused by the insertion of the z-binders. Plastic shear yielding and cracking of the polymer matrix surrounding the z-binders was found to initiate fatigue damage in the 3D woven composite. Mode I interlaminar fracture tests were also performed on the 3D woven composites to demonstrate the large improvement to the delamination toughness gained by increasing the z-binder content.
A comprehensive experimental study was performed to determine the strength of several co-cured and adhesively bonded joints of composite panels reinforced with non-crimp 3D orthogonal woven E-glass fiber fabrics. Various single-lap and double butt-strap joints were fabricated using co-curing and adhesive bonding and tested in uniaxial in-plane tension. The co-cured joints included special stepped 3D woven preforms, stitched and stapled joints. The joint strength appears to have the lowest values for the co-cured single-lap joints, intermediate values for the co-cured double butt-strap joints and the highest values for the adhesively bonded double butt-strap joints. Stitching and stapling dry preforms resulted in significant increase of the co-cured single-lap joint strength. In the range of strap thicknesses studied for the bonded joints, thinner straps provide a higher joint break force than thicker ones. Tapering strap ends to as small angle as possible was found to be the most effective method of increasing break force of double butt-strap bonded joints.
Braiding is a relatively less explored textile process for producing composite preforms. Biaxial braids can be produced as hoses and subsequently be draped over different three-dimensional surfaces. However, triaxial braids are relatively stable structures and should be produced to the desired shape during the braiding process. This is achieved by over-braiding on mandrels that either form part of the finished composite or removed before the moulding process. Triaxial braided composites have superior mechanical properties due to fibre orientations along three directions.Geometry of a braided structure depends on the number of yarn carriers, rotational speed of the carriers, take-up speed and the effective perimeter of the cross-section of a mandrel. In the present work, a VRML based geometrical visualisation tool has been developed to simulate a braid structure on any predefined mandrel geometry, and using a predefined yarn cross-section. Braid angles, cover factors and yarn volume fractions can be computed from these simulations. A triaxial braiding machine has been developed with an independent servo control of the carrier movement and the take-up mechanism; geometrical simulation is used as an input to the control system to continuously vary the braid structure along the length of a mandrel. Flexible tooling is important for rapid product development. A flexible mandrel has been developed that can be mechanically adjusted to change the cross-section and the taper. This system enables rapid development of braided preforms.
The uniaxial tensile properties of 4-step 3D braided E-Glass/epoxy composites at quasi-static (0.001/s) and high strain rates (up to 2800/s) were tested with MTS and split Hopkinson tension bar (SHTB), respectively. From the stress–strain curves of the composites at various strain rates, it is shown that the 3D braided composite is a kind of the rate-sensitive material. The uniaxial tensile stiffness and failure stress increase with the increase of the strain rate, while the failure strain decreases. The fractograph of 3D braided composites depicts the 3D braided composites fail in a more brittle mode in tension at high strain rates. More experiments should be conducted to establish the constitutive equation of 3D braided composites at various strain rates.
Tailored fibre placement (TFP) preforms made of carbon fibre were 3D reinforced with aramid, polybenzoxazol (PBO), polyethylene and polyester fibres and vacuum injected with epoxy resin. The effects of stitch distribution and stitching process parameters on Mode I interlaminar fracture toughness were analysed using a statistical approach. Stitch distribution had a minor effect but 3D thread tension had to be carefully chosen to gain optimum mechanical properties. PBO fibre provided the most improvement in fracture toughness. 3D reinforcing with aramid fibre reduced tensile and flexural properties by 3-8%. Low velocity impact damage in TFP was larger than in fabric but smaller than in tape laminates. Compression-after-impact strength was partly increased by 3D reinforcing in some circumstances but no improvement was found under other conditions.
This paper describes a modified system to predict the properties, in particular the areal density and z-axis fibre content, of a 3D woven preform. Previously a model used had incorporated an idealised tow path to describe the placement of a warp tow within the fabric. The idealised tow path was found to provide some correlation between predicted and actual values particularly for integrated type structures. It was realised however that the idealised yarn model did not truly reflect the actual tow path for an interlinked type structure. Hence a modified model still using a lenticular cross-section, but incorporating a more realistic path for the through-the-thickness part of the binder is put forward as an alternative. Following the procedure used in previous work, fabrics were produced and tested. Predictions made using the modified-model were shown to have a closer correlation than previously in terms of both areal density and percentage z-axis fibre prediction.
A method is described for preparing composite foam using expandable PAN-based microspheres reinforced with continuous fibers. Composite foams were produced by mixing expanded and non-expanded microspheres in select proportions, packing the dry microspheres into a fibrous preform in a closed mold, and heating the assembly to expand and weld the microspheres and fibers together. The composite foams exhibited mechanical performance and formability that surpassed the unreinforced foams. The tensile modulus and strength were increased by 750 and 400% respectively, and showed enhanced resistance to crack propagation compared with unreinforced foam samples. The improvement in compression properties was modest by comparison (<10%). Fiber performs were comprised of 3D, stochastically arranged long fibers, and typical fiber loadings were ∼8 wt%. Long fibers were deeply anchored in the foam and bridged crack wakes, resisting crack growth and delaying catastrophic failure of the foams during tensile tests.
This Part 2 paper presents results of comparative experimental study of progressive damage in 2D and 3D woven glass/epoxy composites under in-plane tensile loading. As Part 1, this Part 2 work is focused on the comparison of in-plane tensile properties of two non-crimp single-ply 3D orthogonal weave E-glass fibre composites on one side and a laminate reinforced with four plies of E-glass plain weave on the other. The damage investigation methodology combines mechanical testing with acoustic emission registration (that provides damage initiation thresholds), progressive cracks observation on transparent samples, full-field surface strain mapping and cracks observation on micrographs, altogether enabling for a thorough characterisation of the local micro- and meso-damage modes of the studied composites. The obtained results demonstrate that the non-crimp 3D orthogonal woven composites have significantly higher in-plane strengths, failure strains and damage initiation thresholds than their 2D woven laminated counterpart. The growth of transverse cracks in the yarns of 3D composites is delayed, and they are less prone to a yarn–matrix interfacial crack formation and propagation. Delaminations developing between the plies of plain weave fabric in the laminate at certain load level never appear in the 3D woven single-ply composites.
Composites fabricated by VARTM technology with the use of single-ply non-crimp 3D orthogonal woven preforms 3WEAVE® find fast growing research interest and industrial applications. It is now well understood and appreciated that this type of advanced composites provides efficient delamination suppression, enhanced damage tolerance, and superior impact, ballistic and blast performance characteristics over 2D fabric laminates. At the same time, this type of composites, having practically straight in-plane fibers, show significantly better in-plane stiffness and strength properties than respective properties of a “conventional” type 3D interlock weave composites. One primarily important question, which has not been addressed yet, is how the in-plane elastic and strength characteristics of this type of composites compare with respective in-plane properties of “equivalent” laminates made of 2D woven fabrics. This 2-part paper presents a comprehensive experimental study of the comparison of in-plane tensile properties of two single-ply non-crimp 3D orthogonal weave E-glass fiber composites on one side and a laminate reinforced with four plies of plain weave E-glass fabric on the other. Results obtained from mechanical testing are supplemented by acoustic emission data providing damage initiation thresholds, progressive cracks observation, full-field surface strain mapping and cracks observation on micrographs. The obtained results demonstrate that the studied 3D non-crimp orthogonal woven composites have considerably higher in-plane ultimate failure stresses and strains, as well as damage initiation strain thresholds than their 2D woven laminated composite counterpart. Part 1 presents the description of materials used, experimental techniques applied, principal results and their mutual comparisons for the three tested composites. Part 2 describes in detail the experimentally observed effects and trends with the main focus on the progressive damage: detailed results of AE registration, full-field strain measurements and progressive damage observations, highlighting peculiarities of local damage patterns and explaining the succession of local damage events, which leads to the differences in strength values between 2D and 3D composites.
A shell/3D modeling technique was developed for which a local three-dimensional solid finite element model is used only in the immediate vicinity of the delamination front. The goal was to combine the accuracy of the full three-dimensional solution with the computational efficiency of a plate or shell finite element model. Multi-point constraints provided a kinematically compatible interface between the local three-dimensional model and the global structural model which has been meshed with plate or shell finite elements. Double cantilever beam (DCB), end notched flexure (ENF), and single leg bending (SLB) specimens were modeled using the shell/3D technique to study the feasibility for pure mode I (DCB), mode II (ENF) and mixed mode I/II (SLB) cases. Mixed mode strain energy release rate distributions were computed across the width of the specimens using the virtual crack closure technique. Specimens with a unidirectional layup and with a multidirectional layup where the delamination is located between two non-zero degree plies were simulated. For a local three-dimensional model, extending to a minimum of about three specimen thicknesses on either side of the delamination front, the results were in good agreement with mixed mode strain energy release rates obtained from computations where the entire specimen had been modeled with solid elements. For large built-up composite structures modeled with plate elements, the shell/3D modeling technique offers a great potential for reducing the model size, since only a relatively small section in the vicinity of the delamination front needs to be modeled with solid elements.
Much less research has been done on failure characteristics of composites under transverse shear, especially for 3D textile composites. This work is an attempt to this need. General characteristics of 3D composites related to the present study are first discussed. Three types of 3D woven carbon/epoxy composites were made with identical internal yarn structures but different external loop patterns. For comparison purposes, a unidirectional carbon/epoxy composite with the same numbers of axial fibers and a monolithic epoxy material were also made to reveal the role of transverse yarns in resisting the shear. To apply the transverse shear, a special fixture was used to clamp and cut the specimen using two cutters. With the fixture, no notch on the specimen is needed, and thus the interlacing loops on the surface remain intact before the test. The gap between the cutters was varied to examine its influence on the failure behavior. Damage in fibers is most intensive within the cutting zone. Microscopic observations on the induced damage were carried out. Two failure modes in axial yarns are prevailing: shear fracture and tensile rupture. Matrix cracking leading to the loss of the shear rigidity is responsible for the tensile rupture of the axial yarns. The transverse shear resulted in complex but intriguing damage modes. The loop pattern, gap length, and cutting position are the crucial influencing factors to the damage modes, maximum load, and the maximum shear displacement to failure.
This paper is aimed at elucidating the processing-property relationships of composites through experimental characterization. The fabrication processes involve three primary steps: the pultrusion of 1 mm diameter unidirectional rods, the formation of three-dimensional preforms incorporating the rods, and the impregnation of resin into the preforms. The use of the rods is intended to address crimp problems often seen in textile composites. A modified two-step set-up to incorporate the rods has been developed. A series of fabrics with varying braiding fibers, pitch lengths and braider sizes has been made to investigate their respective influence on processing and damage behavior. In comparison with conventional textile composites, the axial yarns are straight and are packed in a rather dense and orderly manner. The composite moduli have been analysed by using the fabric geometry model and compared with the experimental results. The material characterization has been carried out on the basis of flexure, short-beam and compression tests. Damage configurations and accumulation for each loading case have been examined. Unique features include pull-in and push-out of the rods, which are the dominant modes in the short-beam tests. Buckling of the rods is the major damage in the flexure and compression tests. The buckled rods form kink bands analogous to fiber microbuckling in compressed unidirectional composites. The kink bands were found to propagate along the interlacing loops on the surface. These unique damage characteristics associated with the use of the rods are discussed in detail.
Measurements of the internal geometry of a carbon fiber non-crimp 3D orthogonal woven composite are presented, including: waviness of the yarns, cross sections of the yarns, dimensions of the yarn cross sections, and local fiber volume fraction. The measured waviness of warp and fill yarns are well below 0.1%, which shows that the fabric termed here “non-crimp” has nearly straight in-plane fibers as-produced, and this feature is maintained after going through all steps of fabric handling and composite manufacturing. The variability of dimensions of the yarns is in the range of 4–8% for warp and fill directions, while the variability of the yarn spacing is in the range of 3–4%. These variability parameters are lower than respective ranges of variability of the yarn waviness and the cross-sectional dimensions in typical carbon 2D weave and 3D interlock weave composites, which are also illustrated in this work for comparison.
A comprehensive drape model has been developed to deal with a range of 3D surfaces, from simple open surfaces to closed tubular sections with 3D bends. Existing drape algorithms, developed mainly for broadcloth composites, cannot cope with closed sections. These algorithms consider the woven fabric as a network of linkages with pin joints and perform kinematic mapping by solving a set of sphere-intersection equations. This method of kinematic drape assumes only in-plane shear deformation and hence cannot be readily applied to a number of 3D shapes, involving other modes of deformation. In the present work, a kinematic mapping algorithm was implemented at first and subsequently modified to drape two-layer tapered preforms to open surfaces. Following this work, a more general algorithm was developed to drape closed preforms on bent tubes, which the authors believe to be the first such attempt.
A comparison of substantial published data for 3D woven, stitched and pinned composites quantifies the advantages and disadvantages of these different types of through-thickness reinforcement for in-plane mechanical properties. Stitching or 3D weaving can either improve or degrade the tension, compression, flexure and interlaminar shear properties, usually by less than 20%. Furthermore, the property changes are not strongly influenced by the volume content or diameter of the through-thickness reinforcement for these two processes. One implication of this result is that high levels of through-thickness reinforcement can be incorporated where needed to achieve high impact damage resistance. In contrast, pinning always degrades in-plane properties and fatigue performance, to a degree that increases monotonically with the volume content and diameter of the pins. Property trends are interpreted where possible in terms of known failure mechanisms and expectations from modelling. Some major gaps in data and mechanistic understanding are identified, with specific suggestions for new standards for recording data and new types of experiments.
Tensile tests are reported for some graphite/epoxy composites with three-dimensional woven interlock reinforcement. Composite failure consists of the accumulation of discrete tow rupture events distributed over a band of damage typically 10–20 mm wide. Load—displacement data for gauges spanning the band indicate work of fracture values ranging from 0.4 to 1.1 MJ m−2. Most of these unusually high vales derives from the ability of the composite to sustain loads near peak load (≈1 GPa) for displacements significantly beyond those at which tows have all failed. The key mechanism is very strong friction or lockup that couples sliding, broken tows to the surrounding composite. Lockup is the product of the geometrical irregularity of nominally straight tows and clamping compressive stresses generated by the through-thickness reinforcement. Lesser contributions to the work of fracture arise from plastic straightening of tows prior to their rupture and the relatively easy but prolonged pull-out of tows following failure of the lockup mechanism.
The mechanical behavior and failure mechanisms of 3D orthogonal woven carbon fiber reinforced plastic (CFRP) composite panels were investigated. The 3D fiber architectures were measured and visualized in a micrograph form. The quasi-static tensile coupon tests were performed to measure the Young's modulus, Poisson's ratio, tensile fracture strength, and failure strains in both stuffer and filler yarn directions. The average Young's modulus in the filler yarn direction was found to be higher than that in the stuffer yarn direction, and the average failure strain in the filler yarn direction was found to be lower than that in the stuffer yarn direction. A scanning electron microscope was used to study the fracture surfaces of the specimen.
In Liquid Composite Molding (LCM), the fiber preform is placed in a mold and resin is injected through a gate to fill the empty spaces within the mold. LCM processes are modeled as resin flow through fibrous porous media in which if one knows the permeability values, one can determine the arrival times of the resin at any location. A mold with radial injection with 192 flow arrival detection sensors along 16 radial lines are mounted flush with the top and the bottom mold surface with an additional sensor on the top surface opposite to the injection hole of the bottom surface. The inverse problem addressed here is from the recorded resin arrival times at sensor locations, how accurately can one determine the permeability of the preform? The proposed method uses correlation between the experimentally recorded resin arrival times and 3D flow simulation of the experiment. The optimization routine varies the permeability components in the simulation to achieve the best possible match with the experimental arrival times at all the sensor locations. Currently, all in-plane permeability components and through-the thickness permeability are characterized from a single experiment with the potential to evaluate the cross-thickness off-diagonal terms as well. In addition, the technique also demonstrates how one can reduce the variation in through thickness permeability by the use of a distribution media at the injection hole to avoid blockage of the inlet by the fiber tows. The optimization routine sequentially optimizes the values of the individual components of the permeability tensor using golden search method, and then repeats the entire sequence until the best match is found. The validation and sensitivity of this method is explored and it has been shown that this technique is promising for characterization of all permeability components from a single radial flow experiment.
Tensile tests were performed on glass reinforced polymer (GRP) composites with three-dimensional (3D) orthogonal, normal layered interlock, and offset layered interlock woven fibre architectures. The mechanical properties and failure mechanisms under tensile loading were similar for the three composites. Cracks formed at low strains within the resin-rich channels between the fibre tows and around the through-thickness binder yarns in the composites, although this damage did not alter the tensile properties. At higher applied tensile stresses the elastic modulus was reduced by 20–30% due to inelastic tow straightening and cracking around the most heavily crimped in-plane tows. Further softening occurred at higher strains by inelastic straightening of all the tows. Composite failure occurred within a localised region and the discrete tow rupture events that have caused tow lock-up and pullout mechanisms in other 3D woven composites were not observed.
The load transfer between fibre and matrix in a metal matrix composite (MMC) depends on the properties and conditions of the fibre/matrix interfacial region. The objective of this investigation is to gain a better understanding of the stresses generated within a continuously reinforced MMC, particularly at this interface. Finite element analysis is used to investigate the effect of thermal and transverse mechanical loading on the SiC/Ti–6Al–4V composite system. The effect on the stress field of a carbon coating on the SiC fibres is also investigated. The results indicate that the interfacial region affects the stress distribution, with the presence of the carbon coating significantly altering the stress profiles generated. It is also found that the residual stresses generated as a result of cooling down the composite from processing temperature, has a marked effect on the stress profile and the behaviour of the composite when subsequent mechanical loading is applied.
Dielectric spectroscopy was applied in the present work for the first time to polymeric composite materials containing transcrystallinity, wherein the dielectric properties of pure nylon 6,6 and of aramid fibre-reinforced nylon 6,6 microcomposites were examined over wide frequency and temperature ranges. The temperature behaviour of the dielectric losses of the materials indicated three polarization processes, related to either local or collective molecular mechanisms of motion. In addition, interfacial polarization of the Maxwell–Wagner–Sillars type was observed. The dielectric response was found to be sensitive to the presence of transcrystallinity in the microcomposites. It was found that the activation energy of the α, β and γ relaxations exhibits typical variations in the presence of reinforcement and transcrystallinity. The dielectric strength, calculated from fitting of the relaxation spectra to the Havriljak–Negami empirical term for all the relaxation processes, was found to be very sensitive to the morphology of the systems. The specific values at 1 MHz of the dielectric constant, dielectric loss and dielectric loss tangent as a function of temperature for the transcrystalline layer were retrieved from the composites data using the rule-of-mixtures. A comparison was conducted between the values of the transcrystalline layer and those of the bulk matrix to determine the effect of the transcrystalline layer on the dielectric properties.
The properties of calcined and hydrous kaolin filled nylon 6,6 composites have been investigated with respect to particle size and surface treatment with an amino-silane coupling agent (A1100). The mechanical properties of untreated kaolin filled nylon 6,6 composites are kaolin type (hydrous and calcined) independent. However, A1100 surface treatment leads to a distinction between calcined and hydrous kaolin composites, promoting adhesion only with calcined kaolin. Treated calcined kaolin filled nylon 6,6 exhibits improved strength and elongation properties, which are influenced by the silane loading level relative to the filler surface area. At high silane loading levels a plasticing effect was observed with impact strength decreasing. The wet strength of A1100 treated calcined kaolin composites is almost double that of the untreated condition. As A1100 does not promote adhesion with hydrous kaolin wet and dry strength properties are relatively unaffected, with the mode of failure being interfacial. Evidence of interfacial chemical bonding was found only in A1100 treated calcined kaolin. It is thought that the interaction is controlled by an acid base interaction between surface free primary amines and carboxylic acid polymer end groups with amide formation at the interface. The distinction between treated calcined and hydrous kaolin composites is due to the availability of surface free amines. Calcined kaolin has a relatively neutral surface on which the amines are not inhibited in their interaction with nylon functional groups. Hydrous kaolin has an acidic surface with bronsted activity that protonates surface amines inhibiting their interaction with nylon 6,6.
Three-dimensional fibre orientation (1-2 plane in Fig. 1) by computer tomography of the central part of the plate at the mid-plane through the thickness (core layer).
This paper investigates the anisotropic behaviour of mechanical properties of a short glass fibre reinforced polyamide 6.6 (PA66-GF35) under quasi-static loading. For this purpose tensile tests were carried out on dog-bone specimens, machined out from injection moulded plates 80 × 80 mm, of three different thicknesses t (1–3 mm) at eight different orientation angles. The tests were performed at room temperature as well as at 130 °C. Material elastic constants were estimated from fitting experimental tensile moduli according to the theory of elasticity for orthotropic materials. A fit on geometrical tensile strengths with the Tsai–Hill failure criterion provided instead the material strength parameters. Both specimen thickness and temperature appear to have a strong influence on mechanical properties and degree of anisotropy.
Multiaxial stress tests were performed on filament wound glass fibre reinforced epoxy tubulars and the first failure modes observed and the stress and strain failure envelopes recorded. The tubes were produced by filament winding technique. Specimens were tested under various ratios of axial stress and circumferential stress in a MTS testing system modified to apply the internal pressure. Fourteen stress ratios were tested (hoop stress: axial stress) ranging from 0:1 (pure axial tension), 1:1, 2:1, constrained ends (εA = 0, ≈3.5:1), 4:1, 4.5:1, 5:1, 7:1, 12:1, 1:0 (pure internal pressure), 7:−1, 2:−1, 1:−1 (pure shear) and 0:−1 (pure axial compression). Five specific failure modes were observed: (1) axial failure along a dominant helical crack under axial tension dominated stress ratios. (2) Weepage at the stress ratio of 2:1 and constrained end condition. (3) Local leakage via fluid jets in the range of 4:1 to 5:1. (4) Burst under stress ratios dominated by hoop stress. (5) Axial collapse under compressive axial loads. The resulting failure strain failure envelope was fit using the maximum strain criteria with necessary modifications in the regions of leakage and axial compression. Both of these ranges were found to be fit by linear segments.
The microstructures of the matrix and particularly of the interfacial zone related to titanium alloy coated SiCCVD filaments processed by a fast liquid route, were investigated through SEM, HREM, EPMA and AES analyses. The carbon transfer from the pyrolytic protection of SiC filaments towards the titanium alloy coating gives rise to the formation of a continuous titanium carbide interphase and a significant volume fraction of precipitated titanium carbide. This precipitated phase, resulting from the liquid Ti/C coating interaction, can be partly lessened by means of a suitable heat treatment, which is expected to generate a deviation of the non-stoichiometry of the carbide. The observed microstructures enable a description of the mechanisms involved in the new liquid route processing method used for coating carbon protected SiC filaments.
Top-cited authors
I. Verpoest
  • KU Leuven
K. L. Pickering
  • The University of Waikato
A.P. Mouritz
  • RMIT University
Lawrence T Drzal
  • Michigan State University
Stepan Lomov
  • KU Leuven