Search

We present a method for evaluating the damage in and failure of fiber reinforced composite materials. The method is based on dissipated energy and has three major components: (i) a relationship between the strain in the material and the magnitude of the dissipated energy density, (ii) a model that relates the dissipated energy density to matrix damage and fiber failure inside the material, and, accordingly, reduces the elastic properties of the material, and (iii) a nonlinear finite element code for calculating the strain at each point in the material taking into account the degradation of the stiffness as damage occurs. The method provides the parameters of interest, namely, the regions of matrix damage and fiber failure, the load-displacement curve, the failure load, and the failure displacement. Tensile, shear, and four point bend tests were performed with S-glass/epoxy solid coupons as well as with coupons containing a central hole. The load-displacement curves, failure loads, and failure displacements were measured and compared to the results computed by the method. The data and the results of the method are in good agreement. [ABSTRACT FROM AUTHOR]

The fatigue lives of two different parts (made of short glass fiber reinforced Copolyamide) were investigated. The first one was a 150 mm by 10 mm by 4 mm coupon with a 2 mm hole at the center. The second one was an automotive gear shift link. The fiber orientations and the stresses inside these parts were calculated, respectively, by the commercial softwares C-Mold and by ABAQUS. The fatigue lives under different cyclic loads were measured; they were also estimated by a model based on the Generalized Miner’s Rule, the results of C-Mold and ABAQUS, and the S-N data generated as part of this study [1,2]. Comparisons between the measured and estimated (by the model) fatigue lives are in reasonable agreement, indicating that the procedure employed is a useful tool for estimating the fatigue lives of parts made of short fiber reinforced thermoplastics. [ABSTRACT FROM PUBLISHER]

Tests were conducted measuring the fatigue lives of 150 mm by 10 mm by 2 mm short glass fiber reinforced Copolyamide coupons under different types of cyclic loads. The fatigue lives measured in these tests were compared to those provided by the Generalized Miner’s Rule, and good agreements were found between the test and model results. [ABSTRACT FROM PUBLISHER]

The fatigue lives (i.e., the number of cycles to failure) of short fiber reinforced thermoplastic matrix composites were investigated. First, S-N curves were generated at stress ratios of R = 0 and R = -1 for short glass fiber reinforced Copolyamide coupons with 0° (30% or 50% fiber content by weight), 45° (50% fiber content) and 90° (30% or 50% fiber content) fiber orientations. Second, these S-N curves were compared to data reported in the literature for a wide range of short glass and short carbon fiber reinforced thermoplastics materials. On the basis of these comparisons, all available data were “collapsed” on two S-N curves, one for R = 0 and one for R = -1. [ABSTRACT FROM PUBLISHER]

In a previous paper, models are described for estimating the deformations and failures of masonry walls reinforced with composites when the wall is subjected to static, in-plane shear and normal loads. In this follow-up investigation, tests were performed to assess the accuracies of these models. Three types of test specimens were constructed: i) walls made of clay bricks and mortar and reinforced with chopped glass mat and epoxy resin, ii) walls made of wood bricks and reinforced with paper, and iii) beams made of wood bricks and reinforced with paper. Static, in-plane normal and shear loads were applied to each test specimen and the load versus deflection, failure load and deflection at failure were recorded. The data were compared with corresponding results generated by the models. These comparisons showed reasonable agreements between the data and the values calculated by the models. [ABSTRACT FROM PUBLISHER]

A technique was developed for simulating the resin transfer molding (RTM) process. The major feature of the technique is a computational steering system that enables the user to make changes during the simulation. Specifically, at any instance, the user can inspect the progress of the resin front. On the basis of the observed resin front position, the user can, as needed, change the port and vent locations, open and close ports and vents, adjust the inlet and exit pressures or flow rates, and reorient the mold with respect to the gravitational field. Additionally, the user can "rewind" the simulator to any previous time in the mold filling process, make any of the above changes, and then continue the simulation.The technique is augmented by a computer code which has three main components: the Simulator, the Gtaphics User Interface (GUI), and the Global Data Storage. The Simulator is a finite element code that calculates the resin flow inside the fiber preform. The GUI serves as the interface between the user and the Simulator; it provides the commands to the Simulator and displays the results. The Global Data Storage is the module that manages the exchange of data between the GUI and the Simulator.The computer code (designated as SUPERTM_3D) is suitable for simulating the resin flow inside two-dimensional as well as three-dimensional fiber preforms of arbitrary shapes. The use of this computer code is illustrated through sample problems. These problems demonstrate how (with this code) the designer can establish the port and vent locations, opening and closing sequences of ports and vents such that the fiber preform is filled completely in the shortest time with the fewest number of vents. [ABSTRACT FROM PUBLISHER]

Tests were conducted with continuous glass fiber reinforced epoxy matrix composites that had different types of fiber reinforcements. Composites reinforced with either unidirectional, triaxial or random fibers were tested. The tensile and compressive strengths and moduli, the Iosipescu shear moduli, the short beam shear strengths, the mode I critical strain energy release rates, and the damage introduced by impact were measured. The results demonstrate the influence of fiber form on these properties and provide design guides for the use of different fiber forms. [ABSTRACT FROM PUBLISHER]

Tests were performed to measure the failure loads of axially compressed glass-fiber-reinforced and graphite-fiber-reinforced composite cylinders. The data were compared with the results of a previous model, which was based on a three-dimensional stress analysis and the Tsai-Wu quadratic first-ply failure criterion. This model predicted the failure loads for glass-fiber-reinforced composites with good accuracy, but less accu rately for failure loads of graphite-epoxy composites. [ABSTRACT FROM PUBLISHER]

Tests were performed measuring the damage initiation loads and the loca tions, shapes, and sizes of delaminations in Fiberite T300/976 graphite/epoxy, Fiberite IM7/977-2 graphite-toughened epoxy, and ICI APC-2 graphite-PEEK plates subjected to transverse static loads. The data were compared to the results of the Finn-Springer model, and good agreements were found between the measured and calculated delamination lengths and widths. [ABSTRACT FROM PUBLISHER]

In-plane compressive and three-point bend tests were performed to evalu ate the strengths of composite laminates containing embedded sensors. Unidirectional, cross-ply and quasi-isotropic specimens were fabricated using Fiberite T300/976 or Fiberite APC-2 unidirectional tape. A thermocouple or a strain gauge was embedded at the center of the specimen, and the in-plane compressive and bending strengths were measured. For comparison, the strengths of specimens without sensors were also mea sured. The data showed that the sensors had little effect on the strength. [ABSTRACT FROM PUBLISHER]

Tests were performed to evaluate the effects of moisture and temperature on the com pressive and short beam shear properties of Fiberite T300/976 fabric. The maximum mois ture content was measured as a function of relative humidity and temperature. The diffusivity was determined as a function of temperature. The compressive and short beam shear strengths and moduli of dry and moisturized specimens were measured at 70, 250, and 350°F. The changes in these mechanical properties caused by drying after moisturiza tion and by subsequent remoisturization were also assessed. [ABSTRACT FROM PUBLISHER]

The effects of moisture on the compressive and interlaminar shear strengths of aramid fiber reinforced epoxy matrix composites were investigated. The compressive strengths were measured with a Celanese fixture, the interlaminar shear strengths were inferred from the results of three point bend tests. The influence of fiber surface treatment on the compressive and interlaminar shear strengths were also assessed by testing composites containing fibers with their surfaces treated and with fibers without surface treatment. [ABSTRACT FROM PUBLISHER]

Tests were conducted measuring the changes in weight of aramid fiber reinforced epoxy matrix composites during exposure to humid air. Composites containing surface treated fibers and composites containing fibers without surface treatment were tested. In addition, for the purpose of comparison, the weight changes of glass fiber reinforced epoxy com posites and pure epoxy were measured. From the weight change data the values of the maximum moisture contents and diffusivities were determined. The effects of fiber surface treatment on the moisture absorption characteristics of the aramid fiber reinforced com posite were assessed. An expression was developed for estimating the transverse diffu sivities of composites from the known resin and fiber properties when fibers absorb a sig nificant amount of moisture. [ABSTRACT FROM PUBLISHER]

Tests were performed to evaluate the effects of crystallinity on the mechanical properties of PEEK 150P polymer and APC-2 graphite/PEEK composite. The tensile strengths, ten sile moduli and fracture toughnesses of the polymer, and the mixed mode I and mode II fracture toughnesses of the composite were measured at crystallinities ranging from 27 to 43 percent by weight. The results show that, in the crystallinity range studied, the tensile properties are not affected significantly by changes in crystallinity, while the fracture toughness depends strongly on the crystallinity. A model is also described which can be used to calculate the crystallinity from the known applied cooling rate. The validity of the model was verified by tests performed with PEEK 150P polymer. [ABSTRACT FROM PUBLISHER]

Procedures are described for calculating the in-plane deformations and curvature changes of sandwich panels caused by nonuniform temperature and moisture distribu tions. The procedures apply to panels consisting of a honeycomb core covered by fiber reinforced, composite face sheets. Results are presented in tables and charts, which can be used to estimate the deformations and curvature changes of panels made of dif ferent, commonly used, composites. [ABSTRACT FROM PUBLISHER]

The effects of the environment on adhesive bonded single lap joints formed using XMC-3 and SMC-R50 composites were investigated. Tests were performed at temperatures of 23C and 93C with test coupons immersed in air, water, and 5% NaCl water mixture. The weight changes of both bonded joints (XMC-3 to SMC-R50 and SMC-R50 to SMC-R50) and unbonded materials were measured. Data were also ob tained showing the effects of moisture, temperature, and applied load on changes in weight, on creep deformation, lap shear strength, and separation modes of the joints. [ABSTRACT FROM PUBLISHER]

A model is presented for predicting the changes in the mechanical properties of organic matrix composites exposed to elevated temperatures. The model can be used to calculate changes in the tensile, compressive and shear strengths and moduli as func tions of exposure time. The model also provides the decrease in the stress ratio with time and the time of failure. [ABSTRACT FROM PUBLISHER]

A summary is given on the effects of moisture and temperature on the properties of chopped fiber reinforced sheet molding compounds. The properties surveyed include tensile strength and modulus, compression strength and modulus, shear strength and modulus, flexural strength and modulus, fatigue, creep, vibration damping, moisture absorption characteristics, and thermal expansion. [ABSTRACT FROM PUBLISHER]

A model was developed for calculating the motion of dilute suspensions of rigid, slender particles in a uniformly sheared viscous liquid, taking into ac count interactions between particles. The model was applied to the plane Couette flow problem in which particles are suspended in a liquid moving parallel to a plane with a uniform, shearing motion. The results of the model were compared to existing data, and good agreement was found between the calculated and experimental results. Using the model, results were also ob tained which illustrate the major effects of particle-particle interactions on the angular motion of the particles. [ABSTRACT FROM PUBLISHER]

Tests were conducted with the objective of evaluating the effectiveness of techniques used for repairing damaged fiber reinforced laminated composites. Effectiveness of the repair was assessed by the tensile strength of the repaired laminate. In the tests the failure loads of laminates repaired either by the scarf, the uniform lap, or the stepped lap technique were measured. Data were generated with the following parameters having been varied: type of material of the damaged laminate, type of repair material, scarf angle and number of external plies in scarf repair, length and number of repair plies in uniform and stepped lap repair, moisture content of the laminate prior to repair, moisture content of the laminate after repair, test temperature, roughness of grinding tool used in preparing the repair surface, and the temperature applied during the cure of the repair patch. The aforementioned parameters were varied over wide ranges resulting in systematic sets of data which shed light on the influence of each of these parameters on the effectiveness of the repair. [ABSTRACT FROM PUBLISHER]

Models were developed for calculating the failure loads of composite laminates repaired by scarf and uniform lap techniques. The models take into account anisotropy of each ply in the laminate and in the repair ply and non-elastic behavior of the adhesive or resin interlayer between the laminate and the repair patch. On the basis of the models two computer codes were written for generating numerical values of the failure loads. The first code is for calculating the failure loads of laminates repaired by the uniform lap technique, the second for calculating the failure loads of laminates repaired by the scarf technique. Failure loads calculated by the models, and the corresponding computer codes, were compared with data, and good agreements were found between the results of the models and the tests. [ABSTRACT FROM PUBLISHER]

A procedure is described for designing prosthetic implants made of composite materials used in central bone and combined joint and bone replacements. The primary objective was to arrive at an implant made of fiber reinforced composites which has a higher bone growth stimulus and a lower micromotion level than titanium alloy (Ti6Al4V) implants currently in use. A secondary objective was to provide a means by which the physician can select, rapidly and with ease, the proper implant for a specific patient. To achieve these objectives, a procedure was developed consisting of three major elements. The first is a performance parameter which indicates the relative merits of different implant designs and takes into account that bone growth and micromotion have different importance from a clinical point of view. This relative importance is taken into account with weighting factors, the values of which were estimated from a survey of orthopedic surgeons. The second element of the procedure is a finite element code for calculating the strains in the bone, which are the parameters of importance in evaluating the implant's performance. The third element is a strain library which stores strains precalculated for a large number of different patient characteristics and different implant types. The strains for the specific patient under treatment are obtained by interpolation of the data in this strain library. The performance of an implant is then evaluated based on these patient specific strains. An example is presented for implants made of graphite AS4/PEEK composite. The results indicate that, in most applications, composite implants have better performances than titanium alloy implants both from the point of the overall implant performance and from the points of view of bone growth and micromotion. [ABSTRACT FROM PUBLISHER]

A procedure is presented for determining the layups of unidirectional fiber reinforced composite plates (with or without cutouts) subjected to multiple in-plane and out-of-plane loadsets, such that the plate satisfies the Tsai-Wu Quadratic Strength Failure criterion, is of low weight, and has a layup which is easy to manufacture. Additionally, the desired cure cycle is generated for thermosetting matrix composite plates such that during cure the maximum allowed temperature is not exceeded, after cure the plate is fully cured and fully compacted, and the cure is achieved in the shortest time.The design of the layup for strength and weight (utilizing a finite element method for the stress analysis) is based on the Hybrid Algorithm for Laminate Optimization, modified to result in plates in which preassigned regions are of uniform thickness. The design of the layup for ease of manufacture is based on rules. The cure cycle is determined by simulating the cure process of the thermosetting matrix with a deterministic model "controlled" interactively by an expert system.A computer code SMARTOPT was developed which combines all the above features of the design process and, for specified geometry, material properties, and applied loadsets, provides the desired layup and cure cycle. Results for sample problems were generated which demonstrate the usefulness of the computer code for designing composite plates. The results also show that with proper design, plates made of composite materials can be made lighter than plates made of metals. [ABSTRACT FROM PUBLISHER]

A model is developed for determining the farfield strains in laminated composites from the strain values indicated by a microsensing device embedded inside the material. The microsensing device considered here consists of three elements: a sensor, a circuit chip, and an antenna. The sensor detects changes in the strains in the material surrounding the sensor. The circuit chip processes the sensor signals which are then transmitted, via the antenna, to a receiver located outside of the composite. The sensor and the circuit are powered by electromagnetic waves generated outside the material and transmitted to the antenna.The model proposed provides a means by which the signal emitted by the microsensing device, and picked up by a receiver, is converted to the engineering value of the farfield strain, i.e., to the strain which would exist in the material at the location of the sensor if the sensor were not present. On the basis of the model, two computer codes were written. The first code calculates the influence coefficient matrix, which is the parameter required to convert the sensor strain to the farfield strain; the second gives the farfield strain directly from the measured sensor strain.Tests were performed to verify the model and the computer codes. In these tests, a strain gauge, a rectangular steel plate, and a circular steel disk (simulating the components of a microsensing device) were embedded in a glass/epoxy beam. The sensor strains were measured by the embedded sensor; the farfield strains were measured by two surface mounted sensors. The measured farfield strains were compared to the farfield strains calculated by the present model, and good agreements were found between the measured and calculated farfield strains. [ABSTRACT FROM PUBLISHER]

A computer code was developed for simulating the resin transfer molding process. The unique feature of the code is that it is interactive such that, during the simulation process, the user can (a) change the locations of the inlet ports and exhaust vents, and (b) add, remove, open and close ports and vents, and (c) change the inlet pressure of flow rate at the inlet ports. Additionally, the user can "rewind" the simulator to any previous time in the mold filling process, make any of the above changes and then continue the simulation. The computer code has two main components, the Simulator and the Graphic User Interface (GUI). The Simulator is a finite element code which calculates the resin flow, and the temperature, degree of cure and viscosity of the resin inside the fiber preform. The GUI serves as the interface between the user and the Simulator, and is used to provide the commands to the Simulator and to display the results. Sample results are presented which illustrate the use of the computer code (designated as SUPERTM) in selecting the port and vent locations, their opening and closing sequences such that the fiber preform is filled completely in the shortest time. [ABSTRACT FROM PUBLISHER]

The drapabilities of uncured unidirectional aligned discontinuous fiber composites were investigated, the drapability being defined as the ratio of the tip deflections, in one minute, of a discontinuous to a continuous fiber reinforced composite canti-levered beam. Equations were developed for calculating the deflections of such composite beams taking into account large deflections, and elastic and viscoelastic material behavior. On the basis of the model, a computer code was written which, for given fiber and matrix properties, discontinuous fiber geometry, fiber volume fraction, and fiber orientation, provides the stresses and strains in and the displacements of the composite beam, and the drapability. Tests were performed with beams made of uncured discontinuous and continuous fiber composites. Tip deflections were measured during the tests and were compared with tip deflections calculated by the computer code. Good agreements were found between the data and the results of the present model. Using the computer code, parametric studies were also conducted to assess the effects of fiber geometry, fiber volume fraction and fiber orientation on the drapability of uncured composites, and the effects of fiber geometry on the on-axis elastic moduli of discontinuous cured fiber composites. The results of these studies indicate that while discontinuous fibers increase the ability of uncured composites to drape, such fibers reduce the longitudinal modulus of cured composites. [ABSTRACT FROM PUBLISHER]

A method is proposed to predict the future load response of structures made of composite materials. The method makes the predictions on the basis of past loadstrain data. The method was developed for structures subjected to monotonically increasing or to random cyclic tensile loads. A computer code PREDICT was written which, in real time, can provide predicted values of residual strengths and number of cycles to failure from measured data. Tests were performed to evaluate the validity of the method using specimens made of different materials, lay-ups, and geometries. The specimens were subjected to monotonically increasing or to cyclic tensile loads. During the tests, the residual strengths and number of cycles to failure were predicted, and were compared to the actual value generated in the tests. The predicted and measured values were in good agreement lending confidence to the method. [ABSTRACT FROM PUBLISHER]

The overall objective of this investigation was to study the manufacture of organic matrix composites by the tape laying process. Specifically, the effects of the processing conditions on the quality of composite parts were investigated. Models were developed which relate the applied heating, applied pressure, tape laying speed, and mold material to the thermochemical and mechanical response of the material. On the basis of these models, computer codes were written to generate numerical results. With these computer codes, parametric studies were performed which indicated the influence of the various processing parameters on the temperature, degree of cure (for thermosetting matrix composites), crystallinity (for thermoplastic matrix composites), viscosity, and stresses inside the composite. From the numerical results, the processing conditions required for manufacturing good quality parts were identified. [ABSTRACT FROM PUBLISHER]

Tests were performed to assess the effects of processing conditions and laminate layup on the change in shape (springback). Twelve inch long and one inch wide coupons made of Fiberite T300/976 graphite/epoxy were tested. Coupons of different layups were cured using different types of molds and different types of vacuum bagging arrangements. The curing was performed with different temperature cycles. Springback was measured from the change in shape of initially flat coupons. The results indicated the effects of the applied temperature, applied pressure, material of the mold, arrangement of the teflon surrounding the coupon, and laminate layup on springback. [ABSTRACT FROM PUBLISHER]

An experimental technique is described which can be used to measure simultaneously the three principal permeabilities of fiber reforms made of continuous or short fibers. The technique utilizes embedded optical fibers for detecting the position of the liquid front inside the reform. Expressions were derived for calculating the permeabilities from the three-dimensional measurement of the liquid front. Permeabilities measured by the present fiber optic and by the conventional, pressure drop and flow visualization techniques were compared. Good agreements were found between the results of these techniques. [ABSTRACT FROM PUBLISHER]

Simple models were developed for determining the vent locations and the fill time of resin transfer molds containing thin flat preforms with isotropic permeabilities. The in-plane shape of the preform is arbitrary and the preform may contain impermeable inserts. The vent locations required to avoid trapping air bubbles are determined by geometric considerations. The time required to fill the mold is calculated by treating the resin flow inside the mold as partly radial and partly channel-like flow. Tests were performed with different shape molds with and without inserts. The vent locations required to avoid air bubbles and the time required to fill the mold were determined. The experimentally established vent locations were in good agreement with the vent locations given by the model. The measured and calculated fill times also agreed well. [ABSTRACT FROM PUBLISHER]

The changes in shapes of fiber-reinforced composite beams, plates and shells affected by embedded piezoelectric actuators were investigated. An analytical method was developed to determine the voltages needed to achieve a specified desired shape. The method is formulated on the basis of mathematical models using twodimensional, linear, shallow shell theory including transverse shear effects which are important in the case of sandwich construction. The solution technique is a minimization of an error function which is a measure of the difference between the deformed shape caused by the application of voltages and the desired shape. A computationally efficient, userfriendly computer code was written which is suitable for performing the numerical calculations. The code, designated as SHAPE2, gives the voltages needed to achieve specified changes in shape. To validate the method and the computer code, results generated by the code were compared to existing analytical and experimental results. The predictions provided by the SHAPE2 code were in excellent agreement with the results of the other analyses and data. [ABSTRACT FROM PUBLISHER]

The use of embedded intrinsic Fabry-Perot sensors in measuring strains and temperature changes was explored. Tests were performed with intrinsic Fabry-Perot sensors embedded in plates made of Fiberite T300/976 graphite-epoxy and of Fiberite APC2-PEEK graphite thermoplastic unidirectional tapes. The plates were also in strumented with thermocouples and strain gauges. The plates were subjected either to three point bending at room temperature, or to a uniform temperature change without an applied external stress. The temperatures and strains measured by the Fabry-Perot sensors agreed well with the thermocouple and strain gauge data, showing that with a single sensor and a single, constant wavelength light source, either the axial strain component or the temperature change can be measured. [ABSTRACT FROM PUBLISHER]

A model was developed for embedded intrinsic and extrinsic Fabry-Perot fiber optic sensors. This model relates the strains and the temperature changes in the material surrounding the sensor to the reflected light intensity. The model consists of three parts. Submodel I relates the temperature and the strains in the material to the temperature and the strains inside the sensor. Submodel 2 relates the temperature and the strains inside the sensor to the change in sensor length and to the changes in the optical properties of the sensor. Submodel 3 relates the changes in sensor length and optical properties to the changes in the intensity of the reflected light.On the basis of the model, a computer code SENSOR1 was written which can be used to calculate the reflected light intensity for specified strains and temperature changes inside the material. The model and the code were verified by measuring changes in the light in tensity of an intrinsic Fabry-Perot sensor embedded in a graphite-epoxy composite plate subjected to loads varying with time. The changes in light intensities were measured and were also calculated by the code. The measured and calculated light intensities were found to be in good agreement. [ABSTRACT FROM PUBLISHER]

A Monte Carlo procedure was developed for optimizing symmetric fiber reinforced composite laminates such that the weight is minimum and the Tsai-Wu strength failure criterion is satisfied in each ply. The laminate may consist of several materials in cluding an idealized core, and may be subjected to several sets of combined in-plane and bending loads. The procedure yields the number of plies, the fiber orientation, and the material of each ply and the material and thickness of the core. A user friendly computer code was written for performing the numerical calculations. Laminates optimized by the code were compared to laminates resulting from existing optimization methods. These comparisons showed that the present Monte Carlo procedure is a useful and efficient tool for the design of composite laminates. [ABSTRACT FROM PUBLISHER]

Tests were conducted to validate the thermoplastic manufacturing process models described by Mantell and Springer. In the tests, APC-2 graphite/PEEK plates were fabricated both in a press and with a specially constructed tape laying apparatus. The tem perature, intimate contact and bonding inside the plates were measured. The data agreed well with the results of the model. [ABSTRACT FROM PUBLISHER]

Models were developed which simulate the processing of thermoplastic matrix composites. The models relate the applied temperature, pressure, speed, and time to the temperature, crystallinity, consolidation, interlaminar bonding, and residual stresses and strains inside the composite. The formulations follow the models proposed by Springer, Loos, and their coworkers for manufacturing plates in a press or in an autoclave, but were extended to include cylindrical geometries, variations in the applied temperatures and pressures with position and time, and rapid bonding. These extensions make the models applicable to the manufacture of plates by tape laying and to filament winding of cylinders. These models were incorporated into a user friendly computer code which can be used to generate numerical results and to select the appropriate processing conditions for processing thermoplastic composite plates 1) in an autoclave, 2) in a press, or 3) by tape laying, and 4) for filament winding thermoplastic composite cylinders. [ABSTRACT FROM PUBLISHER]

Tests were performed assessing the effectiveness of repair in restoring the mechanical properties of damaged, solid composite plates made either of Fiberite T300/976 graphite-epoxy, Fiberite IM7/977-2 graphite-toughened epoxy, or ICI APC-2 graphite-PEEK. The plate length, the layup and the amount of damage were also varied. Damage was introduced in the plates either by impacting them with a solid projectile or by applying a transverse static load. Some (75 % ) or all (100 % ) of the damaged zone was then cut out, and the plate was repaired by plugging and patching the hole. The effec tiveness of the repair was evaluated by measuring the compressive strengths of undamaged plates, damaged plates with no cutout, damaged plates with a cutout, and repaired plates. The data at an intermediate stage of repair (damaged plates with a cutout) provide informa tion on the effect of each repair step on the compressive strength. The results indicated that for the solid plates used in these tests, the repair methods used herein did not improve the compressive strength of already damaged plates. [ABSTRACT FROM PUBLISHER]

Tests were conducted to validate the model and the CHAR computer code previously developed for calculating the thermomechanical behavior of carbon-phenolic and carbon-carbon composites exposed to high temperatures. Temperature distributions in and damage of rapidly heated carbon-phenolic composites were measured. The data com pared well to the results calculated by the code. Parametric studies were also performed, the results of which yield physical insights into the ply-lift phenomenon observed in solid rocket motor exit cone linings, and provide a guide in the design of parts made of carbon- phenolic and carbon-carbon composites. [ABSTRACT FROM PUBLISHER]

The high temperature behavior of carbon-phenolic and carbon-carbon composites was investigated. Models were developed which describe the thermal and me chanical responses of such composites when rapidly heated to high temperatures. On the basis of the models, a computer code was written that provides the temperature and pressure distributions, vapor and volatile formation rates, amount of char, stress and strain, and damage in carbon-phenolic and carbon-carbon plates. [ABSTRACT FROM PUBLISHER]

Tests were performed measuring the temperature, ionic conductivity, and compaction in 16 to 200 ply thick graphite-epoxy laminates made of either Fiberite T300/976 (tape or fabric) or Hercules AS/3501-6 (tape). The data were compared to results calculated by the Loos-Springer CURE model. Good agreement was found between the model results and the data indicating that the model can be used to select the cure cycle. It is also shown that the cure cycle recommended by the manufacturer may result in ex cessive temperatures and in incomplete compaction and cannot be used to manufacture "thick" or even moderately thick composites. [ABSTRACT FROM PUBLISHER]

A rule based expert system was developed for controlling the autoclave temperature and pressure during cure of fiber reinforced thermosetting matrix composites. Rules were established which ensure that the temperatures and pressures remain within the required limits, the laminate is fully compacted and fully cured, the void content is minimum, and the residual (curing) stresses are low. The inputs to the expert system are the measured instantaneous autoclave temperature and pressure, the composite midpoint and surface temperatures, composite thickness, and ionic conductivity. These inputs are examined by preestablished rules and, on the basis of decisions dictated by the rules, the autoclave heater, cooler and pressure controllers are adjusted to the appropriate settings. An algorithm was written and hardware was developed for collecting and processing the data, for applying the rules, for making the decisions, and for controlling the autoclave. The software and hardware were integrated into a control system which was validated by simulated as well as actual cures of flat graphite-epoxy laminates ranging in thickness from 0.1 to 6.5 inches. The tests showed that the control system (designated as SECURE) performed all its required functions well, resulted in reduced cure times, and produced laminates whose mechanical properties were as good as those cured by conventional methods. [ABSTRACT FROM PUBLISHER]

A model is presented describing the pultrusion process of thermoplastic matrix composites. The model provides the temperature, crystallinity, pressure, and consolidation inside the composite along the axis of the die, and the required pulling force. On the basis of the model, a computer code was written that can be used to obtain numerical results. A sample problem is presented illustrating the type of information which can be generated by the code. [ABSTRACT FROM PUBLISHER]

Teas were conducted measuring the deformations of composite open cylindrical shells subjected to temperature changes. The cylindrical shells were made of 16 plies of Fiberite T300/934 unidirectional graphite-epoxy tape arranged as cither [02/902]2, or [902/02)2,. The measured deformations were compared to the results of a higher order shell theory. It is shown that terms representing extensional strain in the thickness direction must be considered for the analysis to represent the data accurately. [ABSTRACT FROM PUBLISHER]

A technique is described for determining the mechanical properties of thick composite laminates. According to this technique the test specimens are cut out from thin sublaminates separated from the main laminate by porous Teflon sheets. Data are presented which show the usefulness of the method in measuring tensile, compressive and short beam shear properties, void contents and fiber volumes of laminates consisting of as many as 1000 plies. [ABSTRACT FROM PUBLISHER]

Analytical and experimental studies were performed to validate the model developed by Lee and Springer for simulating the manufacturing process of filament wound composite cylinders. First, results calculated by the Lee-Springer model were com pared to results of the Calius-Springer thin cylinder model. Second, temperatures and strains calculated by the Lee-Springer model were compared to data. The data used in these comparisons were generated during the course of this investigation with cylinders made of Hercules IM-6G/HBRF-55 and Fiberite T-300/976 graphite-epoxy tows. Good agreement was found between the calculated and measured stresses and strains, indicating that the model is a useful representation of the winding and curing processes. [ABSTRACT FROM PUBLISHER]

The behavior of elliptical sublaminates created by delaminations in composite plates was investigated for plates subjected to in-plane compressive, shear, and thermal loads. The axes of the ellipse may be arbitrarily oriented with respect to the applied loads. A model was developed which provides the stresses, strains, and displacements of the sublaminate, and the loads applied to the plate at which the sublaminate buckles and at which it grows in size. A series of experiments was conducted on sandwich plates made of Fiberite T300/976 graphite-epoxy laminates bonded to an aluminum honeycomb core. Teflon film in either a circle or an ellipse was embedded in the laminate, simulating the presence of a delamination. Each plate was loaded in compression, and the load-strain history of the sublaminate, the sublaminate buckling load, and the sublaminate growth load were measured. The predictions of the model agreed reasonably well with the data. [ABSTRACT FROM PUBLISHER]

By using the Lee-Springer filament winding model temperatures, degrees of cure, viscosities, stresses, strains, fiber tensions, fiber motions, and void diameters were calculated in graphite-epoxy composite cylinders during the winding and subsequent cur ing. The results demonstrate the type of information which can be generated by the model. It is shown, in reference to these results, how the model, and the corresponding WIND- THICK code, can be used to select the appropriate process variables. [ABSTRACT FROM PUBLISHER]

Models are presented for describing the time dependent viscoelastic and viscoplastic stress-strain relationships of unidirectional fiber reinforced thermoset matrix composites exposed to elevated temperatures. For Fiberite T300/976 graphite-epoxy com posite the constants in the stress-strain relationships were determined from data obtained in four point bend tests over the range 75 to 350°F. A procedure is also described which, together with the stress-strain models developed in this study, can be used to calculate the stresses and strains in loaded multidirectional laminates at high temperatures. Strains in loaded multidirectional Fiberite T300/976 laminates were measured as a function of time at temperatures varying from 75 to 350°F. Strains calculated by the models were compared to these data, and good agreements were found between the analytical and experimental results. [ABSTRACT FROM PUBLISHER]

Models are presented for describing the stress-strain relationships and the strengths of a thermoset matrix composite exposed to elevated temperature. The models take into account the nonlinear elastic characteristics of the material. For Fiberite 976 resin and for unidirectional Fiberite T300/976 composite the constants required in the models were determined by measuring the tensile, compressive, and shear properties, and Poisson ratios at 75, 250, 300, and 350°F. The strengths, initial (stress → 0) moduli, and initial Poisson's ratios of the unidirectional composite were calculated from the known resin and fiber properties using micromechanics equations. A procedure is also described which, together with the stress-strain-strength models developed in this study, can be used to calculate the deformations and strengths of loaded multidirectional laminates at high temperatures. To evaluate the accuracies of the models and the calculation procedure deformations and strengths of multidirectional Fiberite T300/976 laminates were calculated. These results were compared to data obtained at temperatures ranging from 75 to 350°F. The calculated results agreed well with the data. [ABSTRACT FROM PUBLISHER]