Search

Impact damage tolerance is a frequently listed design requirement for composites hardware. The effect of impact damage and open hole size on laminate compressive strength was studied on sandwich beam specimens which combine CFRP-GFRP hybrid skins and a syntactic foam core. Three test specimen configurations have been investigated for this study. The first two were sandwich beams which were loaded in pure bending (by four point flexure). One series had a skin damaged by impact, and the second series had a circular hole machined through one of the skins. The reduction of compressive strength with increasing damage (hole) size was compared. Additionally a third series of uniaxially loaded open hole compression coupons were tested to generate baseline data for comparison with both series of sandwich beams.

The thermoviscoelastic characterization of Kevlar 49/Fiberite 7714A epoxy composite lamina and the development of a numerical procedure to predict the viscoelastic response of any general laminate constructed from the same material were studied. The four orthotropic material properties, S sub 11, S sub 12, S sub 22, and S sub 66, were characterized by 20 minute static creep tests on unidirectional (0) sub 8, (10) sub 8, and (90) sub 16 lamina specimens. The Time-Temperature Superposition-Principle (TTSP) was used successfully to accelerate the characterization process. A nonlinear constitutive model was developed to describe the stress dependent viscoelastic response for each of the material properties. A numerical procedure to predict long term laminate properties from lamina properties (obtained experimentally) was developed. Numerical instabilities and time constraints associated with viscoelastic numerical techniques were discussed and solved. The numerical procedure was incorporated into a user friendly microcomputer program called Viscoelastic Composite Analysis Program (VCAP), which is available for IBM PC type computers. The program was designed for ease of use. The final phase involved testing actual laminates constructed from the characterized material, Kevlar/epoxy, at various temperatures and load level for 4 to 5 weeks. These results were compared with the VCAP program predictions to verify the testing procedure and to check the numerical procedure used in the program. The actual tests and predictions agreed for all test cases which included 1, 2, 3, and 4 fiber direction laminates.

A methodology is described for the durability analysis of polymer matrix composites, based on nonlinear viscoelasticity theory. The durability analysis is performed on the basis of a certain number of tests carried out on limited and, if possible, short time scale by the use of accelerating factors. The method was applied to thermomatrix composites under uniaxial and biaxial loadings, showing satisfactory agreement between the life-time predictions and the published data on real-time behavior.

A numerical procedure based on classical lamination theory has accurately predicted time-dependent creep in laminated composites with only two fiber directions. Predictions for laminates with more than two fiber orientations have been accurate as short times, but have fallen well below the longer-time experimental results obtained on narrow (12 mm) specimens. Further testing has helped establish that relaxation of the fiber trusses along the free edges of narrow specimens is much more significant for time-dependent behavior than under static conditions.

In response to the tremendous growth in the development of advanced materials, such as fiber-reinforced plastic (FRP) composite materials, a new numerical method is developed to analyze and predict the time-dependent properties of these materials. Basic concepts in viscoelasticity, laminated composites, and previous viscoelastic numerical methods are presented. A stable numerical method, called the nonlinear differential equation method (NDEM), is developed to calculate the in-plane stresses and strains over any time period for a general laminate constructed from nonlinear viscoelastic orthotropic plies. The method is implemented in an in-plane stress analysis computer program, called VCAP, to demonstrate its usefulness and to verify its accuracy. A number of actual experimental test results performed on Kevlar/epoxy composite laminates are compared to predictions calculated from the numerical method.

Numerical solution methods for viscoelastic orthotropic materials, specifically fiber reinforced composite materials, are examined. The methods include classical lamination theory using time increments, direction solution of the Volterra Integral, Zienkiewicz's linear Prony series method, and a new method called Nonlinear Differential Equation Method (NDEM) which uses a nonlinear Prony series. The criteria used for comparison of the various methods include the stability of the solution technique, time step size stability, computer solution time length, and computer memory storage. The Volterra Integral allowed the implementation of higher order solution techniques but had difficulties solving singular and weakly singular compliance function. The Zienkiewicz solution technique, which requires the viscoelastic response to be modeled by a Prony series, works well for linear viscoelastic isotropic materials and small time steps. The new method, NDEM, uses a modified Prony series which allows nonlinear stress effects to be included and can be used with orthotropic nonlinear viscoelastic materials. The NDEM technique is shown to be accurate and stable for both linear and nonlinear conditions with minimal computer time.

The impact of flight error in measured viscoelastic parameters on subsequent long-term viscoelastic predictions is numerically evaluated using the Schapery nonlinear viscoelastic model. Of the seven Schapery parameters, the results indicated that long-term predictions were most sensitive to errors in the power law parameter n. Although errors in the other parameters were significant as well, errors in n dominated all other factors at long times. The process of selecting an appropriate short-term test cycle so as to insure an accurate long-term prediction was considered, and a short-term test cycle was selected using material properties typical for T300/5208 graphite-epoxy at 149 C. The process of selection is described, and its individual steps are itemized.

A discussion of current standard durability prediction methodology for bonded joints is given. The influence of time-dependent adhesive properties on lap shear stress distributions and interfacial stresses is presented. It is suggested that these effects may be important in durability considerations. Test geometries from which better durability information can be obtained are outlined and a general mechanics approach for durability prediction is suggested.

The literature for creep to failure cumulative damage laws are reviewed. Creep to failure tests performed on polycarbonate and polysulfone under single and two step loadings are discussed. A cumulative damage law or modified time fraction rule is developed using a power law for transient creep response as the starting point. Experimental results are approximated well by the new rule. Damage and failure mechanisms associated with the two materials are suggested.

Fiber reinforced plastics were considered for many structural applications in automotive, aerospace and other industries. A major concern was and remains the failure modes associated with the polymer matrix which serves to bind the fibers together and transfer the load through connections, from fiber to fiber and ply to ply. An accelerated characterization procedure for prediction of delayed failures was developed. This method utilizes time-temperature-stress-moisture superposition principles in conjunction with laminated plate theory. Because failures are inherently nonlinear, the testing and analytic modeling for both moduli and strength is based upon nonlinear viscoelastic concepts.

A viscoelastic response scheme for the accelerated characterization of polymer-based composite laminates in applied to T300/5208 graphite/epoxy. The response of uni-directional specimens is modeled. The transient component of the viscoelastic creep compliance is assumed to follow a power law approximation. A recursive relationship is developed, based upon the Schapery single-integral equation, which allows approximation of a continuous time-varying uniaxial load using discrete steps in stress. The viscoelastic response of T300/5208 to transverse normal and shear stresses is determined unsing 90 deg and 10 deg off-axis tensile specimens. In each case the seven viscoelastic material parameters required in the analysis are determined experimentally using short-term creep and creep recovery tests. It is shown that an accurate measure of the power law exponent is crucial for accurate long-term prediction. A short term test cycle selection procedure is proposed, which should provide useful guidelines for the evaluation of other viscoelastic materials.

Existing strain gage technologies as applied to orthotropic composite materials are reviewed. The bonding procedures, transverse sensitivity effects, errors due to gage misalignment, and temperature compensation methods are addressed. Numerical examples are included where appropriate. It is shown that the orthotropic behavior of composites can result in experimental error which would not be expected based on practical experience with isotropic materials. In certain cases, the transverse sensitivity of strain gages and/or slight gage misalignment can result in strain measurement errors.

The utilization of adhesive bonding for composite structures is briefly assessed. The need for a method to determine damage initiation and propagation for such joints is outlined. Methods currently in use to analyze both adhesive joints and fiber reinforced plastics is mentioned and it is indicated that all methods require the input of the mechanical properties of the polymeric adhesive and composite matrix material. The mechanical properties of polymers are indicated to be viscoelastic and sensitive to environmental effects. A method to analytically characterize environmentally dependent linear and nonlinear viscoelastic properties is given. It is indicated that the methodology can be used to extrapolate short term data to long term design lifetimes. That is, the method can be used for long term durability predictions. Experimental results for near adhesive resins, polymers used as composite matrices and unidirectional composite laminates is given. The data is fitted well with the analytical durability methodology. Finally, suggestions are outlined for the development of an analytical methodology for the durability predictions of adhesively bonded composite structures.

Creep to failure tests performed on polycarbonate and polysulfone under single and two step loadings are discussed. A cumulative damage law or modified time fraction rule is developed using a power law for transient creep response as the starting point. Experimental results are approximated well by the new rule. Damage and failure mechanisms associated with the two materials are suggested.

Possible treatments of the nonlinear viscoelastic behavior of materials are reviewed. A thermodynamic based approach, developed by Schapery, is discussed and used to interpret the nonlinear viscoelastic response of a graphite epoxy laminate, T300/934. Test data to verify the analysis for Fiberite 934 neat resin as well as transverse and shear properties of the unidirectional T300/934 composited are presented. Long time creep characteristics as a function of stress level and temperature are generated. Favorable comparisons between the traditional, graphical, and the current analytical approaches are shown. A free energy based rupture criterion is proposed as a way to estimate the life that remains in a structure at any time.

An experimental method used to measure the in-plane normal-axial strains produced within a single-lap joint is described in which a resistance-foil strain gage is embedded within the joint prior to curing of the adhesive. Nominal dimensions of the titanium Ti-6-4 adherends were 0.13 x 2.5 x 12.7 cm and an overlap of 2.5 cm was used. The joint was bonded with FM-300 structural adhesive. The average ultimate shear strength of the gaged specimens and control specimens was 16.1 MPa and 14.1 MPa. A significant advantage of the proposed method is that strains internal to the joint are measured, rather than strains at an external edge. The presence of the gage was found to be not detrimental to bond performance.

Nonlinear viscoelastic and thermoelastic characterization procedures were applied to a rubber toughened adhesive that is commercially available as FM-300. Long time (accelerated testing) predictions on the basis of stress-time-superposition and time-temperature-superposition were compared with actual long term data. Good verification was obtained.

A nonlinear viscoelastic model based on the procedure of Findley et al., (1948) has been used to accurately represent the creep of several graphite/epoxy composites. Applying this approach to unidirectional 0, 90 and 10 deg off-axis tensile specimens, the viscoelastic response of a lamina can be characterized. The resulting lamina model has been useful for representing the behavior of a lamina in a numerical procedure to predict creep and delayed failures of general laminates. Also, independently, the Findley procedure has been used to characterize the nonlinear viscoelastic behavior of general laminates.

Measurements of the nonliner viscoelastic behavior of two adhesives, FM-73 and FM-300, are presented and discussed. Analytical methods to quantify the measurements are given and fitted into a framework of an accelerated testing and analysis procedure. The single integral model used is shown to function well and is analogous to a time-temperature stress-superposition procedure (TTSSP). Advantages and disadvantages of the creep power law method used in this study are given.

The current paper describes the utilization of a thermodynamic based analytical nonlinear viscoelastic approach to represent lamina properties. Test data to verify the analysis for both transverse and shear properties of a T300/934 composite are presented. Master curves as a function of stress level and temperature are generated. Favorable comparisons between the traditional graphical and the current analytical approaches are shown.

Lifetime prediction of matrix resin composites is an important problem because of their viscoelastic character. If the Time-Temperature-Stress-Superposition principle is valid, results from short time tests at high stress levels can give the necessary information for the prediction of long time behavior. At such stress levels the viscoelastic response is nonlinear. In this contribution the basis of a computer routine is presented starting from a thermodynamical formulation given by Schapery. This method gives in an accurate way the different material parameters in that nonlinear constitutive equation for different stresses and temperatures.

A numerical procedure is described for predicting the viscoelastic response of general laminates. A nonlinear compliance model is used to predict the creep response of the individual laminae. A biaxial delayed failure model predicts ply failure. The numerical procedure, based on lamination theory, increases by increments through time to predict creep compliance and delayed failures in laminates. Numerical stability problems and experimental verification are discussed. Although the program has been quite successful in predicting creep of general laminates, the assumptions associated with lamination theory have resulted in erroneous bounds on the predicted material response. Delayed failure predictions have been conservative. Several improvements are suggested to increase the accuracy of the procedure.

Primarily because of the polymeric matrix, graphite/epoxy composites are viscoelastic materials which exhibit creep and delayed failures. Guided by the time-temperature superposition principle, the authors are developing accelerated characterization techniques to predict the long term compliance behavior and creep rupture strength of composite materials based on short term tests at elevated temperatures. The effect of a post-cure cycle on the compliance and creep rupture strength of composite materials is discussed. The Zhurkov type failure law is applied to experimental creep rupture data of a typical laminate and is shown to correlate the data. A beneficial mechanical strengthening effect is described which significantly affects the failure strengths of specimens which have been under creep loading for a period of time. This aging effect is reversible if the specimen is allowed to recover before being loaded to failure. Other environmental aspects of composite materials are also reviewed.

Uniaxial tensile creep and recovery data from polycarbonate at six temperatures and six stress levels are analyzed for nonlinear viscoelastic constitutive modeling. A theory to account for combined effects of two or more accelerating factors is presented.

A method for the experimental and analytical accelerated characterization of long term viscoelastic properties is presented. The time-temperature-stress superposition principle (TTSSP) is shown to serve as the basis for determining long term compliance data from short term creep test results. Nonlinear viscoelastic theories are discussed to provide the framework for the mathematically modeling of such a process. A time dependent Tsai-Hill-Zhurkov failure theory is used to determine long term failure properties from short term data. Compliance and failure data are incorporated in an incremental lamination theory to make long term laminate predictions. Comparisons are made between theory and experiment.