Skip Menu

As fiber-reinforced polymer (FRP) composites find application in highway bridge structures, methodologies for describing their long-term performance under service loading will be a necessity for designers. The designer of FRP bridge structures will be faced with out-of-plane damage and delamination at ply interfaces. The damage most often occurs between hybrid plys and dominates the lifetime response of a thick-section FRP structure. In this paper a model is developed to address these issues. The methodology employs the quadratic delamination initiation failure criteria, in conjunction with a delamination growth law to describe the out-of-plane damage under bending. These phenomena are combined with the critical element residual strength life prediction tool to determine remaining bending stiffness and moment capacity of a pultruded and hybrid FRP 20.3-cm-deep structural shape. The model successfully describes the onset of delamination prior to fiber failure and suggests that out-of-plane failure controls the life of the structure.

Pinhole formation in proton exchange membranes (PEM) may be caused by a process of flaw formation and crack propagation within membranes exposed to cyclic hygrothermal loading. Fracture mechanics can be used to characterize the propagation process, which is thought to occur in a slow, time-dependent manner under cyclic loading conditions, and believed to be associated with limited plasticity. The intrinsic fracture energy has been used to characterize the fracture resistance of polymeric material with limited viscoelastic and plastic dissipation, and has been found to be associated with long-term durability of polymeric materials. Insight into this limiting value of fracture energy may be useful in characterizing the durability of proton exchange membranes, including the formation of pinhole defects. In an effort to collect fracture data with limited plasticity, a knife slit test was adapted to measure fracture energies of PEMs, resulting in fracture energies that were two orders of magnitude smaller than those obtained with other fracture test methods. The presence of a sharp knife blade reduces crack tip plasticity, providing fracture energies that may be more representative of the intrinsic fracture energies of the thin membranes. Three commercial PEMs were tested to evaluate their fracture energies (G(c)) at temperatures ranging from 40 to 90 degrees C and humidity levels varying from dry to 90% relative humidity (RH). Experiments were also conducted with membrane specimens immersed in water at various temperatures. The time temperature moisture superposition principle was applied to generate fracture energy master curves plotted as a function of reduced cutting rate based on the humidity and temperature conditions of the tests. The shift with respect to temperature and humidity suggests that the slitting process is viscoelastic in nature. Also such shifts were found to be consistent with those obtained from constitutive tests such as stress relaxation. The fracture energy is more sensitive to temperature than on humidity. The master curves converge at the lowest reduced cutting rates, suggesting similar intrinsic fracture energies; but diverge at higher reduced cutting rates to significantly different fracture energies. Although the relationship between G, and ultimate mechanical durability has not been established, the test method may hold promise for investigating and comparing membrane resistance to failure in fuel cell environments. (C) 2009 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 48: 333-343, 2010

Pinhole defects that form in proton exchange membranes (PEMs) due to the cyclic hygrothermal stresses induced during the operation of a fuel cell and cause gas crossover may be interpreted as a result of crack formation and propagation. The goal of this study is to employ a fracture test to approach the intrinsic fracture energy of a perflourosulfonic acid proton exchange membrane. The intrinsic fracture energy has been used to characterize the fracture resistance of polymeric materials with minimal plastic dissipation and the in absence of viscous dissipation, and has been associated with the long-term durability of polymeric materials where subcritical crack growth Occurs under slow time-dependent or cyclic loading conditions. Insights into this limiting value of fracture resistance may offer insights into the durability of PEMs, including the formation of pinhole defects. in order to achieve this goal, a knife slit test which significantly reduces the plastic deformation during the test by limiting the plastic zone size with a sharp blade is conducted. Additionally, double edge notched tension tests and trouser tear tests are conducted to obtain the essential work of fracture and tear energy, respectively. It has been found that although the fracture energy obtained with the knife slit test is still several times larger than the intrinsic fracture energy of regular polymer materials, it is several orders of magnitude lower than those obtained with the other two methods, where process-dependent viscous and plastic dissipation dominate over the intrinsic material property.

The widely used yttria-stabilized zirconia (YSZ) electrolyte is subjected to thermal and external stresses under operation, so that the enhancement of the mechanical properties is an important issue in planar solid oxide fuel cells. Fracture strengths of 8 mol % YSZ electrolytes as 100Ã--100 mm squire plates, 23 mm disks, and 17 mm disks were evaluated using plate tensile, ball-on-ring, and pressure-on-ring testing methodologies, respectively. Finite element analysis (FEA) was validated and used to calculate the stress distribution and peak stress for the biaxial strength tests. A Weibull analysis was carried out on the test/FEA-predicted peak stresses, and Weibull strength, modulus, and material scale parameters were found for each test methodology. The methodologies were compared and evaluated based on the results of the Weibull analysis; the pressure-on-ring test is preferred for brittle thin-film fracture strength testing. ©2005 American Society of Mechanical Engineers

The structural response of continuous fiber, polymer matrix composites to fire exposure is presently of interest to the construction and marine industries. This paper addresses experimentally measured lifetime of E-glass vinyl ester composite laminates subjected to combined centric compression and one-sided simulated fire exposure. Under such conditions, these laminates (having a nearly quasi-isotropic stacking sequence) support a 10 MPa compressive stress under low heat fluxes (20-30 kW/m2) for approximately 102 s. Thermally modified micromechanics and laminate mechanics are successfully used to describe the observed life times when limited to thermal reversible effects. In these cases, the glass transition temperature controls the life the composite.

A theoretical linear-elastic model is used to study the effect of embedded sensors and actuators on a composite's compression strength. An existing analytical model is employed to predict the local undulations in the plies which result from the presence of the embedded devices. Using these results in conjunction with a micromechanical model for compression strength, predictions are made for the reduction in this property due to sensors or actuators. Comparisons are made with experimental data for graphite epoxy systems, and the model is shown to accurately predict the trends in such data. Parametric studies are performed to investigate the effect of embedded sensor/actuator diameter, laminate thickness, laminate stiffness, and laminate shear strength on the compression strength of composites containing these embedded sensors/actuators. It is shown that degradations in a composite's compression strength is as severe for sensors/actuators embedded at 30-degrees as those embedded at 90-degrees to the structural fibers.

The objective of this work is to investigate the crack growth properties of a natural rubber vulcanizate. This work utilizes a modified double cantilever beam (DCB) for measuring the crack propagation in natural rubber. Results of the crack growth in the plane strain dominated DCB are presented and compared to results from plane stress single edge notch test specimens. It is found that the crack growth rate in natural rubber under plane strain conditions is one hundred times faster than under plane stress. The failure surface of the DCB is examined for cavitation occurring within the rubber due to high hydrostatic pressure. No evidence of cavitation was found. (C) 2002 Elsevier Science Ltd. All rights reserved.

The objective of this study was to examine the effects of hygrothermal aging on the durability of a graphite/epoxy woven composite material system. The study was part of a larger project in which the objective was to evaluate and model the effects of moisture, temperature, and combined hygrothermal conditions on the strength and life of a graphite/epoxy woven composite material system. The results presented here represent an extension of the work by Patel and Case (Int. J. Fatigue 22 (2000) 809).The hygrothermal aging consisted of cyclical temperature and moisture variations which were meant to simulate mission conditions for an advanced subsonic aircraft. Durability studies were carried out on the aged material system in the form of fatigue and residual strength testing under humid and elevated temperature environments. Damage mechanisms and failure modes were determined through fatigue testing, residual strength testing, and nondestructive evaluation.Changes in physical appearance, thermal analysis results, fracture surfaces, and moisture diffusion behavior all supported the idea that the material was affected by the aging process. However, experimental testing also showed that the initial and residual tensile properties of the aged material were virtually unaffected by the imposed environmental aging (as compared to unaged material testing results), except when tested at elevated temperature. At elevated temperature, both the dynamic stiffness and residual strength were noticeably reduced from that at room temperature. (C) 2002 Elsevier Science Ltd. All rights reserved.

The objective of this effort was to evaluate and model the effects of moisture, temperature and combined hygrothermal conditions on the strength and life of a graphite/epoxy woven composite material system. The material system specified is a candidate for use in an advanced subsonic aircraft engine, and the imposed environmental conditions were considered to be representative of engine service conditions. Fatigue and residual strength data showed that initial and residual tensile properties and fatigue life of the material were only minimally affected by any of the imposed environmental conditions for the fatigue stress levels considered in this study. Based on this data, it was shown a residual-strength-based life prediction approach could be used to model strength and life with reasonable results. Fatigue damage progression and accumulation were found to be dependent on testing environment, suggesting that more adverse effects of environment on strength and life may be manifested for other types of loading (i.e., off-axis loading). (C) 2000 Elsevier Science Ltd. All rights reserved.

This work addresses issues of long-term durability of hydrogen-air proton exchange membrane fuel cells (PEMFCs) under cyclic current loading conditions, simulating the real road driving conditions for automotives. The same type of membrane-electrode assembly (MEA) was also aged under constant current mode as a control and the results were compared with those of the cyclically aged MEA. Both MEAs were characterized for cell polarizations, impedance spectra, Tafel plots, hydrogen crossover rates as well as electrochemical active surface areas at intervals of 100 h of aging. It was demonstrated that hydrogen crossover increased dramatically after 500 h of current cycling due to pinhole formation and was the most dominant degradation source. The fuel cell approached the end of its useful lifetime after 1000 h of operation. On the other hand, the hydrogen crossover rate remained approximately constant for the MEA under constant current operation. Mass transport limitations were identified as the major source of decreased performance during the constant current operation. This decrease in performance was partially reversible when cathode flooding was resolved by setting the cell at lower current densities. At the end, a phenomenological durability model was established successfully to describe the aging processes and cell performance at different time nodes.

The focus of this research was to investigate the effect of thermal degradation upon the mechanical properties of a natural rubber compound. We examined both the quasi-static and dynamic mechanical properties of a natural rubber vulcanizate which had been subjected to isothermal, anaerobic aging. The thermal aging was conducted between the temperatures of 80 degreesC and 120 degreesC for times ranging from 3 to 24 days. The effect of thermal degradation was measured using the changes in the crosslink distribution of the vulcanizates as functions of time at temperature. A master curve relationship between the crosslink distribution of the vulcanizates due to thermal degradation and the static and dynamic mechanical properties has been developed. It was found that the both the quasi-static and dynamic mechanical properties correlated with the percentage of poly and monosulfidic crosslinks, where in general higher levels of polysulfidic crosslink gave rise to the highest mechanical properties.

The ability to determine the state of strain within a plane is of great importance to the health monitoring of composite materials and structures. In this preliminary study, fiber optic strain sensors are arranged in a rosette configuration to study their effectiveness in monitoring arbitrary strain states within a neat resin and a graphite/epoxy laminate. The extrinsic Fabry Perot Fiber Optic Strain Rosette Sensor (FP-FOSRS) is very capable of measuring axial (0-degrees) and transverse (90-degrees) strains. However, discrepancies are noted between strain measured by the embedded Fabry Perot Sensors and externally mounted strain gages at 45-degrees. Further, it is believed that the compressive strength of FP-FOSRS embedded composites is most affected by the sensors' presence. A compressive strength reduction of 300% is measured. This reduction is compared with the result predicted by a micromechanical model.

Interphase strengths of carbon reinforced vinyl ester composites are investigated by a, microindentation. technique as a function of moisture exposure and fatigue cycles. This data will provide the basis for understanding the mechanics of stress transfer in the composite when subjected to similar environmental and mechanical loading conditions. These interphase properties will be incorporated into a micromechanical model for the tensile strength of the composites. Initial experimental results show a 30% and 27% reduction in interfacial shear strength, IFSS, in samples sized with a standard epoxy material due to fatigue and moisture degradation, respectively. Residual tensile strengths of the bulk composites reduced 30% due to fatigue and 16% due to moisture. The micromechanical modeling showed that these large changes in IFSS resulted in very little change in the predicted tensile strength, suggesting that other damage mechanisms may play more important roles in the tensile strength reduction due to moisture and fatigue. The over-prediction of the tensile strength of unaged material indicates the need for a tensile strength model that incorporates damage present upon processing in virgin material.

The goal of this effort was to investigate the effect of temperature and environment (particularly temperature cycling) on the behavior of polymer composite materials typical of that used in infrastructure applications as well as in aerospace applications. To this end, freeze-thaw tests were conducted on a glass/vinyl ester composite using a DSC to verify the presences of freezable water within the composite material. Additionally, freeze-thaw conditioning of composites was conducted according to the protocol specified in ASTM C 666. Tensile test data encompassing strength, stiffness, and strain-to-failure for "as-received", moisture saturated, and freeze-thaw conditioned material is presented as well as saturation moisture uptake data. To examine the role of temperature cycling on an aerospace composite, durability studies were carried out on a woven (5 harness satin) graphite/epoxy system targeted at applications in areas of the aircraft industry where operating environment is of concern. Fatigue and residual strength testing was carried out on the composite material in question with the objective of developing a residual strength based life prediction model in which the effects of environment are considered. In this case, the approach called for the evaluation of tensile strength degradation during fatigue cycling in specified environments. The results could then be combined to predict the fatigue life of the material under periodically changing environments. The study was duplicated with material previously aged (unstressed) under hygrothermal cycling to study the effects of environment alone on the composite.

In this paper we present an analytical model to predict the compression strength of a composite containing coated, circular fibers. The theoretical development is based on the assumption that shear failure in the matrix governs the compression strength of a unidirectional composite material. The stress concentrations in the matrix are dependent upon the geometry of the fiber and the interphase or coating region around the fiber. Analytical results suggest that an optimum interphase might exist to maximize compression strain to failure. However, in general stiffer coatings appear to provide larger compression strength than do compliant coatings. To evaluate the analytical model, composites containing glass fibers coated with 5 different materials were manufactured. Tests on these samples support the trends predicted by the model and the strong dependence of compression strength predictions on coating properties.

Temperature and humidity fluctuations in operating fuel cells impose significant biaxial stresses in the constrained proton exchange membranes (PEMs) of a fuel cell stack. The strength of the PEM, and its ability to withstand cyclic environment-induced stresses, plays an important role in membrane integrity and consequently, fuel cell durability. In this study, a pressure loaded blister test is used to characterize the biaxial strength of Gore-Select(R) series 57 over a range of times and temperatures. Hencky's classical solution for a pressurized circular membrane is used to estimate biaxial strength values from burst pressure measurements. A hereditary integral is employed to construct the linear viscoelastic analog to Hencky's linear elastic exact solution. Biaxial strength master curves are constructed using traditional time-temperature superposition principle techniques and the associated temperature shift factors show good agreement with shift factors obtained from constitutive (stress relaxation) and fracture (knife slit) tests of the material. (C) 2009 Elsevier B. V. All rights reserved.

Composites reinforced with ultra-high molecular weight polyethylene (UHMWPE) fibers are known to perform well under ballistic impact, while being lightweight. In this study, the response of Spectra Shield (a proprietary UHMWPE fiber composite manufactured by Honeywell) is examined. The aim of this study is to determine the modes of failure of this material subjected to dynamic hard-body impact. Dynamic impact tests are performed with a single stage gas gun on 0, 902s panels of Spectra Shield II SR-3124 using hemispherically capped and flat cylindrical impactors (16.7 g mass) with incident velocities between 50 m/s and 300 m/s. Panel failure modes are compared through the range of velocities tested. Damage progression during the impact event is observed using high speed photography. Additionally, quasi-static punch tests are performed and the results compared with those from the dynamic event.

In this paper, we examine the fatigue behavior of a Nextel 610 reinforced alumina-yttria composite at room and elevated temperatures. Our main interest is in the prediction of elevated temperature fatigue behavior from simpler and/or shorter duration tests. We consider a residual strength based approach which involves a stepwise integration of the material stress-rupture behavior. This analysis is found to accurately predict the elevated temperature fatigue behavior.

A 5-harness satin woven graphite/epoxy composite was studied in both the as-fabricated and a hydrothermally aged condition. The fatigue damage mechanisms for both unaged and aged specimens were characterized using dynamic stiffness loss monitoring during fatigue loading,. The major damage mechanism,, of transverse yarn cracking, inter-yarn debonding, and delamination also were modeled using simple mechanistic models. It as possible to use the dynamic stiffness loss curves along with the simple damage mechanism models to characterize the evolution of transverse yarn cracking and delamination during fatigue, for both the as-fabricated and the aged specimens at room temperature (RT) and at 121 degreesC (250 degreesF). The present analysis predicted the trends in the matrix crack evolution and the delamination growth reasonably well. Based on the experiments and the analysis, the unaged material tested at RT and had the lowest rate of increase in the crack density with fatigue cycles. The unaged material tested at 121 degreesC had the highest rate of crack density increase with cycles. Crack density for the 12,000 hour hygrothermally aged material increased at a rate of 95% high er than the unaged material tested at RT. Delaminations were computed to initiate at higher fatigue cycles and also reach higher delamination levels with decreasing cyclic stress levels. For the same stress level, elevated temperature and aping led to more rapidly increasing and larger delamination lengths. The hydrothermally aged material tested at 121 degreesC had a very low threshold for delamination, which initiated below 1.000 cycles. The aged material tested at 121 degreesC also had the largest delamination length, The models developed in this paper provide a simple means to derive information about fatigue damage mechanisms (which are often difficult to characterize), using stiffness loss measurements, which are quite easy to make.

In this study, three commercially available proton exchange membranes (PEMs) are biaxially tested using pressure-loaded blisters to characterize their resistance to gas leakage under either static (creep) or cyclic fatigue loading. The pressurizing medium, air, is directly used for leak detection. These tests are believed to be more relevant to fuel cell applications than quasi-static uniaxial tensile-to-rupture tests because of the use of biaxial cyclic and sustained loading and the use of gas leakage as the failure criterion. They also have advantages over relative humidity cycling test, in which a bare PEM or catalyst coated membrane is clamped with gas diffusion media and flow field plates and subjected to cyclic changes in relative humidity, because of the flexibility in allowing controlled mechanical loading and accelerated testing. Nafion (R) NRE-211 membranes are tested at three different temperatures and the time-temperature superposition principle is used to construct stress-lifetime master curve. Tested at 90 degrees C, 2%RH extruded Ion Power (R) N111-IP membranes have a longer lifetime than Gore (TM)-Select (R) 57 and Nafion (R) NRE-211 membranes. (C) 2009 Elsevier B.V. All rights reserved.

Many of the loads applied to engineering components during their service life are not constant with time; in many instances, loads are applied in a dynamic or repeated fashion. In the transportation industry, for example, it is easy to find cases of repeated loading resulting from such situations as an automobile driving down a bumpy road or an airplane flying through turbulent air. Other common examples are associated with rotating or reciprocating parts, including automobile drive shafts and piston connecting rods in combustion engines.

Wind loads on buildings, traffic over bridges, and the pressurization/depressurization of pressure vessels are other common examples of time-dependent degradation. It was noticed in the 1840's that many materials fail abruptly, and with little or no change in properties prior to that failure, when a mechanical load is applied cyclically over a long period of time. This phenomenon eventually became known as "fatigue." It is difficult to find a way to characterize such material behavior systematically, but the engineering approach usually adopted is the idealization of the dynamic or repeated loading as a cyclic loading with a constant amplitude. The susceptibility or resistance of a given material to this cyclic loading is typically discussed in terms of the number of cycles to failure, i.e., the life, N, of the material under a given cyclic stress amplitude, S, and is generally represented on an "S-N curve" as we will see in more detail later.

Although fatigue behavior occurs over time (especially over long periods of time and under the application of many cycles of loading), the phenomenon is not time-dependent; fatigue is a cycle-dependent phenomenon. The cycles of load application may be applied quickly over a long period of time, or discontinuously (with periods of no loading interspersed), with no change in the number of cycles to failure (the fatigue life). As an engineering definition, the number of cycles to failure is the defining feature of this phenomenon. Of course, as we will discuss in later chapters, the material behavior may not be strictly independent of time in many practical situations. It is possible for fatigue and creep (as well as stress rupture) to occur at the same time, for example. In such a case, the physical damage mechanisms that cause creep and fatigue may or may not be independent; some mechanisms may contribute to both cycle-dependent and time-dependent changes in properties and performance. As we will see, there may be significant interactions between time-dependent and cycle-dependent mechanisms. Any attempt at lifetime estimation for design purposes with composite materials must include each of these effects (as appropriate). It is our goal in the sections that follow to highlight the influences of these effects, and to suggest a methodology that may be employed to analyze the composite materials subjected to fatigue loads.

In the context of our discussion of the physical mechanisms and the behavior that underlies the subject of durability and damage tolerance, the subject of fatigue introduces another important topic, namely, the statistical nature of the problem. Like most specific topics raised in this chapter, we will revisit this subject many times in subsequent chapters because it is especially important. In fact, in many cases, the statistical nature of the micro-events that control remaining strength and life is a canon of the philosophy and modeling that describe and explain the resulting properties and performance. At this point, we will only mention a few salient features to support our conclusions about physical behavior. The statistical variation of the characteristics of composites is generally significantly larger than that observed for homogeneous materials. Moreover, the statistical distribution of life for a composite has a large spread. Finally, the statistical spread of the strength for a composite is, typically, less than the spread of the fibers. The details of this subject are complex for composites, and the specific nature of these details is important. For example, there is a statistical distribution of strength from point-to-point in any homogeneous material (including the fibers--especially due to surface defects and their large surface-to-volume ratio), but that does not explain the statistical behavior of composite materials (which are inhomogeneous) in which many damage events occur.

Indeed, these facts bring us to two of the most important points of discussion in this chapter. First, the local fracture of a constituent in a composite rarely causes the global failure of the composite. Constituents typically fail many times, from point-to- point, and well-designed composites usually do not fail as a result of any single flaw or single flaw growth. Second, composites typically fail as a result of the statistical accumulation of micro-defects, rather than as a result of the statistical distribution of defects or flaws. In order to demonstrate these points, and to substantiate their consequences, we will explore several of the physical mechanisms that contribute to the defect distributions referred to previously. To do so, we will first seek to classify composite materials according to the type of reinforcement, as well as to the types of constituent materials. After doing so, we will examine the behavior of composites in these classifications in fatigue loading conditions.

In this paper, an analysis is developed to predict the stress redistribution in the presence of single and multiple fractured fibers in a unidirectional composite material. This analysis includes the effects of constituent properties, fiber volume fraction, and crack size on the strain concentrations experienced by the adjacent fibers. These effects are not included in other predictions such as shear lag or those of Hedgepeth and Van Dyke. In addition, the predictions are compared with direct experimental measurements obtained from model composite tests.

When a proton exchange membrane (PEM) based fuel cell is placed in service, hygrothermal stresses develop within the membrane and vary widely with internal operating environment. These hygrothermal stresses associated with hygral contraction and expansion at the operating conditions are believed to be critical in membrane mechanical integrity and durability. Understanding and accurately modeling the viscoelastic constitutive properties of a PEM is important for making hygrothermal stress predictions in the cyclic temperature and humidity environment of operating fuel cells. The tensile stress relaxation moduli of a commercially available PEM, Gore-Select (R) 57, were obtained over a range of humidities and temperatures. These tests were performed using TA Instruments 2980 and Q800 dynamic mechanical analyzers (DMA), which are capable of applying specified tensile loading conditions on small membrane samples at a given temperature. A special humidity chamber was built in the form of a cup that encloses tension clamps of the DMA. The chamber was inserted in the heating furnace of the DMA and connected to a gas humidification unit by means of plastic tubing through a slot in the chamber. Stress relaxation data over a temperature range of 40-90 degrees C and relative humidity range of 30-90% were obtained. Thermal and hygral master curves were constructed using thermal and hygral shift factors and were used to form a hygrothermal master curve using the time temperature moisture superposition principle. The master curve was also constructed independently using just one shift factor. The hygrothermal master curve was fitted with a 10-term Prony series for use in finite element software. The hygrothermal master curve was then validated using longer term tests. The relaxation modulus from longer term data matches well with the hygrothermal master curve. The long term test showed a plateau at longer times, suggesting an equilibrium modulus.

As civil engineers seek to improve buildings and bridges, they are turning to composite materials for some structural components. Thus the long-term life of these materials in damp highly acidic conditions (from concrete pore solutions) is critical. This article presents a study of moisture ingression and damage in a pultruded compression element under exposure to pore solution. The goal was to experimentally find the orthotropic diffusion coefficients for pore solution in the composite material and to evaluate the damage in the composite using SEM. A method for calculating 3D diffusion coefficients based on weight measurements by selectively sealing some surfaces against moisture ingression was successfully employed. This research extends the application of the 3D diffusion solution developed by Pierron et al. [3] to selectively sealed specimens of fixed dimension. The Arrhenius equation was then used to model the diffusion coefficients with respect to temperature. For the first time, sequential SEM images were performed in the same location before and after specimen exposure in an attempt to identify damage development separately from initial damage. These images showed little if any change in specimens exposed at room temperature over the first 19 days: however, characteristically different damage was notable at elevated temperature and for a specimen exposed for 650 days.

A life prediction method for ceramic matrix composites is developed. This model is based upon damage mechanics concepts included in the framework of the critical element model. One unique feature of the model is its ability to include general variations of temperature and applied loads as functions of time. A detailed description of application of the model to elevated temperature fatigue is examined. In addition, a validation example is presented that includes the combined effects of rupture and fatigue. The comparison with the experimental data is shown to be good, although the result does appear to be dominated by the rupture effect.

This paper describes the development of a ceramic matrix composite component design tool that integrates the life prediction methodology, MRLife, with a finite element code CSTEM. The life prediction methodology is based on damage mechanics using residual strength as the damage metric and the critical element concept. Previous work has shown that such an approach is able to give accurate predictions of fatigue lives under combined mechanical and thermal loads. The development of the design tool based on this technique involves modifying the finite element program to account for damage in the component and the linking of the lifing code to it. At the component level, damage results in the redistribution of stresses while at the element level the same damage causes changes in the local stress state and changes the remaining strength in the component to propagate failure. The use of the design tool is described by means of a case study involving the life prediction of a composite notched plate at high temperature and cold grips under an axial fatigue load at high temperature.

This paper presents the results to date of an ongoing study on the effects of temperature and mechanical fatigue on the residual strength of a five-harness satin weave reinforced polymeric composite. Unaged specimens underwent mechanical fatigue testing at room temperature and at 120 degrees C. Tests were interrupted at two different stages of fatigue life to monitor damage and to obtain stiffness changes and residual strengths. Initial material strength and stiffness were measured to be approximately 738 MPa and 65.6 GPa, respectively, at both temperatures. As a result of fatigue loading (R = 0.1) at 455 MPa and 10 Hz, the strengths remained unchanged, but the residual stiffness decreased at both temperatures, particularly at 120 degrees C. Data collected thus far show that temperature has a significant impact on fatigue life.

The present paper investigates the concept of evolution of strength in fibrous composite systems. In particular, the conceptual foundations of the 'critical-element method,' developed by the authors, are defined and systematically combined to construct a philosophy for the prediction of remaining strength and life of composite materials in the presence of combined mechanical, thermal, and environmental applied conditions that may produce concomitant fatigue, creep, and stress rupture. This philosophy, developed and applied over a fifteen year period, has not been presented in collected form in previous publications. Since Dr. Pagano was a contributor to some of the micromechanical concepts used in the philosophy, the present volume was judged to be an appropriate place for a statement of the philosophy. (C) 2000 Elsevier Science Ltd. All rights reserved.

In this chapter, we describe our efforts to develop many of the principal elements required for a failure prediction of composite materials subjected to fire exposure in conjunction with structural compressive loads. This work is a combination of thermal modeling and prediction of mechanical response. Our ultimate goal is to combine computational fire modeling with structural modeling using the finite element method.

To support the thermal modeling, we begin with the characterization of thermal properties--particularly thermal conductivity--using the HotDisk transient plane source technique for both uncharred and fully charred Vetrotex 324/Derakane 510A composite samples. Results are compared with those collected by Lattimer and Ouellette.

Our structural modeling is a combination of experimental characterization supported with analytical modeling, and has been focused on compression-controlled failures. Because compression failures are believed to be influenced by the shear response of the composite, extensive characterization of the nonlinear, thermo-viscoelastic behavior of Vetrotex 324/Derakane 510A - 40 has been performed using tensile loading of [±45]2S laminates. Results from these tests were used in a modified version of the Budiansky and Fleck compression strength model to predict the initial compression strength of warps-aligned laminates. In addition, the shear creep compliance was used in the modified Budiansky and Fleck model to predict iso-thermal compression creep rupture behavior. Good agreement was observed in comparison with the experimental data. In particular, master curves of compression creep rupture failure were developed using the shift factors obtained from the shear compliance tests, suggesting that the correct mechanism was being modeled.

To simulate actual fire exposure in conjunction with mechanical loading, a one-sided heat flux combined with compression loading technique has been developed. In this test, a quartz IR heater is used to apply a controlled heat flux while constant compressive loads are applied. Micro-kinking failures are observed--consist with the failure mode modeled using the modified Budiansky and Fleck model.

Finally, we describe our efforts to connect a computational fire simulation tool--the Fire Dynamics Simulator--with a structural finite element analysis. We have used the results of an FDS simulation as thermal boundary conditions in a transient thermal model. The thermal boundary conditions are input as prescribed surface temperatures or heat fluxes on a FE solid model surface.

In this paper, the problem of a penny-shaped crack in the center of multiple concentric cylinders is considered. By making appropriate choices for the stress functions in each of the constituents and making the standard assumptions of linear elasticity, it is possible to reduce the problem to the solution of a Fredholm integral equation of the second I:ind which may be solved numerically. This solutionis used in conjunction with a geometry approximation to model the stress concentrations due to a broken fiber in a unidirectional composite material. The results suggest that models of such a shear lag may be missing important features of the stress state including surrounding broken fibers in composite materials. Copyright (C) 1996 Elsevier Science Ltd.

It is now well known that Fick's Law is frequently inadequate for describing moisture diffusion in polymers and polymer composites. Non-Fickian or anomalous diffusion is likely to occur when a polymer composite laminate is subjected to external stresses that could give rise to internal damage in the form of matrix cracks. As a result, it is necessary to take into account the combined effects of temperature, stress (or strain), and damage in the construction of such a model, In this paper, a modeling methodology based on irreversible thermodynamics applied within the framework of composite macro-mechanics is extended to the case of a bi-axially damaged laminate. The model allows characterization of non-Fickian diffusion coefficients as well as moisture saturation level from moisture weight gain data for laminates with pre-existing and non-evolving damage. A symmetric damage tensor based on continuum damage mechanics is incorporated in this model by invoking the principle of invariance with respect to coordinate transformations. For tractability, the diffusion governing equations are simplified for the special case of a laminate with biaxially oriented matrix cracks that is subjected to a uniaxial tensile stress. The final equations obtained from this derivation indicate that both effective diffusivity and maximum saturation level for this particular case can be expressed as quadratic functions of crack density. Comparisons with test data for a bi-axially damaged AS4/PR500 graphite/ epoxy woven composite are provided for model verifications. (C) 2001 Elsevier Science Ltd. All rights reserved.

One scheme for reliability-based design that is growing in popularity for civil and naval applications is the load and resistance factor design (LRFD). Our goal in this research is the development of a simulation to predict the remaining strength of structural composites subjected to variable fatigue loading and environmental exposure. The results of this simulation can then be used in LRFD to determine appropriate material "knockdown" factors for use in engineering design applications. The work so far focuses on modeling the response of the material to fatigue damage only. The proposed Monte-Carlo style simulation combines a phenomenological residual strength based life prediction model for composites materials with an empirical stiffness based damage accumulation model. This model is demonstrated using data for two glass reinforced vinyl-ester polymer composite systems. The analysis of these results has led to new insight into how the changes in mechanical properties of these materials exposed to fatigue loading can be modeled.

In this manuscript we use a concentric cylinder model to determine the stresses in a composite material subjected to transverse loading. After selecting a failure criterion, it is possible for us to determine a fiber coating which optimizes the strain to failure of the composite material. Experimental results are presented for a model composite system and the results are compared to the analytical predictions. The model is shown to give reasonable agreement with the experimental data.

A key need for the commercial acceptance and viability of composite structural material systems is a methodology for reliability and lifetime predictions. In this paper, we present a philosophy in which the residual strength of the composite material is calculated as a function of applied loading conditions, environment, and resulting failure mode. Two specific example applications are considered: the assessed and verified performance of fiber wound composite tubes and pultruded shapes. We attempt to demonstrate the understanding of the controlling damage mechanisms and the degradation processes that contribute to the loss in stiffness and strength under simulated service conditions.

Structural composites are increasingly being utilized in many large naval and civil structures where it is vital that their long-term performance be predictable and their variability definable over the life of the structure. However, these properties may be influenced by the degree of cure of the resin, particularly for room-temperature-cured systems. Thus, this investigation defines the postcure effects on E-glass/vinyl-ester fiber-reinforced polymer (FRP) composites manufactured using the vacuum-assisted resin transfer molding (VARTM) method, which are typical of those used by the US Navy for ship structures. The composites are differentiated by varying levels of postcure temperature and duration, and examined for the effects of advancing cure at various points in the time after postcure. Pseudo-quasi-isotropic [0/+45/90/-45/0]s and angle ply laminate [±45]2s samples from each level of postcure are examined at 1, 10, 30, 100, and 300 days after postcure in order to track strength, stiffness, failure strain, creep, and fatigue performance as functions of time. In parallel, the matrix polymer is inspected using FTIR (Fourier transform infrared spectroscopy) to directly assess the degree of conversion. Dynamic mechanical analysis and shrinkage measurements are also undertaken to assess the Tg and the amount of shrinkage undergone during post-curing, as well as the advancing of the level of cure during the prescribed aging time. Results suggest that the degree of conversion is limited to 80% for the vinyl-ester oligomer and 90-95% for styrene following a postcure of 93°C. It is observed that after 300 days of ambient storage the nonpostcured samples approach the degree of conversion exhibited by those postcured at 93°C, as measured by FTIR. Resin dominated quasi-static properties are greatly affected by the degree of cure, whereas fiber dominated properties are not. Where the degree of cure is comparatively low, viscoelastic properties cause greater changes in creep response as well as influencing fatigue performance. ©2006 American Society of Mechanical Engineers

Numerous studies report the effects of temperature on the stiffness and strength of polymer matrix composites (PMC's). Due to the complexity of the relaxation phenomenon in the matrix, these studies are mainly qualitative. In the present paper models were developed that can explicitly relate the mechanical response of the composite to temperature. These models are related to the microstructure of the constituents, and therefore can be applied to any polymer matrix composite. The possibility of using these models was illustrated by the study of the mechanical behavior of carbon fiber AS4/polyphenylene sulfide (PPS) at elevated temperatures. Part I of this paper relates to the modeling of the temperature-dependent composite tensile properties (stiffness and strength). Parts II and III focus on the life prediction of AS4/PPS undergoing static and fatigue end-loaded bending at elevated temperatures.

This paper addresses proposed improvements to the carbon/vinyl ester interface via the use of engineering thermoplastic polymers as sizing agents. Two thermoplastic polymers, K-30 poly(vinylpyrrolidone) (PVP) and a modified polyhydroxyether of Bisphenol A (PhenoxyTM) were used as sizings in a controlled study. These sizings were applied to surface treated but unsized AS-4 carbon fibers using a lab-scale sizing line. The standard and commercially available G sized AS-4 fiber, was also used for comparison. Predictions of observed tensile and compression strengths of unidirectional, sized carbon fiber-vinyl ester composites fabricated using the pultrusion process were assessed using micromechanical models that take into consideration the influence of the different interfaces. The micromechanical models require inputs on either the interfacial shear strength or the global laminate shear response. The influence of both processing induced differences (misalignment of fibers and fiber damage) as well as the differences in interphase properties (shear response) have been shown to play an important role in describing the observed differences in the interphase varied composites.

We seek a method by which the residual strengths of composites subjected to combined degradation mechanisms may be calculated. Three different approaches are considered. First, the inputs to the residual strength analysis are purely phenomenological. Second, we make use of a micromechanics-based simulation to combine the effects of two specified damage mechanisms. Finally, we develop a hybrid approach in which the results from the micromechanics simulations for individual mechanisms are used as inputs to the residual strength model as if they had been obtained from experimental data. The residual strength model is then used to predict the combined behavior, and the results are compared with those obtained directly from the simulations. For the two mechanisms considered (slow crack growth and asperity creep), the results for the detailed micromechanical simulations were accurately represented by the residual strength-based model.

This paper presents a critical review of research progress in modelling the structural response of polymer matrix composites exposed to fire. Models for analysing the thermal, chemical, physical, and failure processes that control the structural responses of laminates and sandwich composite materials in fire are reviewed. Models for calculating the residual structural properties of composites following fire are also described. Progress towards validation of the models by experimental characterisation of the structural properties of composites during and following fire is assessed. Deficiencies in the fire structural models are identified in the paper, which provide the focus for future research in the field.

The need exists for a failure time schema to predict composite component lifetimes from quasi-static to dynamic strain rates. Models need to incorporate temperature, loading rate, and the time to failure of the composite components. An approach to describe the effects of strain rate and temperature on the mechanical properties of an E-glass/vinyl ester composite system is presented and verified with experimental data. A representation of the time to failure of the composite was constructed using the Monkmann-Grant equation, relating the applied steady state strain rate and the time to failure. Strain rates of 10(-6) to 1.6 s(-1) were achieved and experiments were conducted at room temperature, 80 and 120 degreesC. It was found that the composite system exhibited nearly a constant strain to failure over the applied rates and temperatures. The test data verified an approach to a failure time estimation scheme for steady state strain rate utilizing the Monkmann-Grant approach.

tert-Butylphenyl terminated disulfonated poly(arylene ether sulfone) copolymers with controlled molecular weights (Mn), 20-50 kg molâˆ'1, were successfully prepared by direct copolymerization of the two activated halides, biphenol and the endcapper, 4-tert-butylphenol. The high molecular weight copolymer (molecular weight over 80 kg molâˆ'1) was also synthesized with 1:1 stoichiometry without an endcapping reagent. The chemical compositions and the molecular weights of the endcapped copolymers were characterized by their 1H NMR spectra utilizing the 18 unique protons at the chain ends. Modified intrinsic viscosity measurements in 0.05 M LiBr/NMP solution further correlated well with NMR results. Combining the endcapping chemistry with proton NMR end group analysis and intrinsic viscosity measurements, one can demonstrate a powerful tool for characterizing molecular weight of sulfonated poly(arylene ether sulfone) random copolymers. This enables one to further investigate the influence of molecular weight on several critical parameters important for proton exchange membranes, including water uptake, in-plane protonic conductivity and selected mechanical properties. These are briefly discussed herein and will be more fully described in subsequent publications.

The tensile stress-strain behavior of Nafion 117 and sulfonated poly(arylene ether sulfone) copolymer (BPSH35) membranes were explored with respect to the effects of the strain rate, counterion type, molecular weight, and presence of inorganic fillers. The yielding properties of the two films were most affected by the change in the strain rate. The stress-strain curves of Nafion films in acid and salt forms exhibited larger deviations at strains above the yield strain. As the molecular weight of the BPSH35 samples increased, the elongation at break improved significantly. Enhanced mechanical properties were observed for the composite membrane of BPSH35 and zirconium phenylphosphonate (2% w/w) in comparison with its matrix BPSH35 film. The stress-relaxation behavior of Nafion and BPSH35 membranes was measured at different strain levels and different strain rates. Master curves were constructed in terms of plots of the stress-relaxation modulus and time on a double-logarithm scale. A three-dimensional bundle-cluster model was proposed to interpret these observations, combining the concepts of elongated polymer aggregates, proton-conduction channels, and states of water. The rationale focused on the polymer bundle rotation/interphase chain readjustment before yielding and polymer aggregate disentanglements and reorientation after yielding. © 2006 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 44: 1453-1465, 2006

This paper discusses the development of a novel test fixture for applying a uniform radial stress to cylindrical composite laminates in such a way as to eliminate bending moments and axial loads. This fixture requires the use of easily fabricated poly-tetrafluoroethylene (PTFE) ring seals. These seals, which fit just inside the composite sample, are designed to be internally pressurized in the fixture and expand outwards, applying a uniform radial load to the sample. The results of mechanics analyses of the PTFE ring seal and composite sample are presented, which estimate the radial stress applied to the composite sample as a fraction of the applied internal hydraulic pressure. Further, results are presented from a prolonged hygrothermal aging study, in which the E-glass/epoxy, cylindrically wound composite samples are aged under varying hygrothermal conditions (ambient dry; and 30, 40, and 50 degrees C submerged in water). At various points during the aging, a subset of samples are removed and destructively tested to measure the reduction in strength from the initial values. Annular specimens are tested using the fixture, and hoop tensile strengths are recorded. In addition, samples cut axially from the cylindrical composite tubes are used to evaluate tensile and compressive axial strengths. Arrhenius fits are used to estimate activation energies for the strength loss in the axial (compression) and hoop (tension) directions.