Search

In coordination with the Office of Safety and Mission Assurance and the respective Center Pressure System Managers (PSMs), the NASA Engineering and Safety Center (NESC) was requested to formulate a consensus draft proposal for the development of additional testing and analysis methods to establish the technical validity, and any limitation thereof, for the continued safe operation of facility non-code layered pressure vessels. The PSMs from each NASA Center were asked to participate as part of the assessment team by providing, collecting, and reviewing data regarding current operations of these vessels. This report contains the outcome of the assessment and the findings, observations, and NESC recommendations to the Agency and individual NASA Centers.

This presentation provides examples of research and development that has recently or is currently being conducted at NASA, with a special emphasis on the application of structural health monitoring (SHM) of aerospace vehicles. SHM applications on several vehicle programs are highlighted, including Space Shuttle Orbiter, International Space Station, Uninhabited Aerial Vehicles, and Expandable Launch Vehicles. Examples of current and previous work are presented in the following categories: acoustic emission impact detection, multi-parameter fiber optic strain-based sensing, wireless sensor system development, and distributed leak detection.

As a next step in the development and implementation of an on-board leak detection and localization system on the International Space Station (ISS), there is a documented need to obtain measurements of the ultrasonic background noise levels that exist within the ISS. This need is documented in the ISS Integrated Risk Management System (IRMA), Watch Item #4669. To address this, scientists and engineers from the Langley Research Center (LaRC) and the Johnson Space Center (JSC), proposed to the NASA Engineering and Safety Center (NESC) and the ISS Vehicle Office a joint assessment to develop a flight package as a Station Development Test Objective (SDTO) that would perform ultrasonic background noise measurements within the United States (US) controlled ISS structure. This document contains the results of the assessment

The Director of the NASA Engineering and Safety Center (NESC), requested an independent assessment of the anomalous gaseous hydrogen (GH2) flow incident on the Space Shuttle Program (SSP) Orbiter Vehicle (OV)-105 during the Space Transportation System (STS)-126 mission. The main propulsion system (MPS) engine #2 GH2 flow control valve (FCV) LV-57 transition from low towards high flow position without being commanded. Post-flight examination revealed that the FCV LV-57 poppet had experienced a fatigue failure that liberated a section of the poppet flange. The NESC assessment provided a peer review of the computational fluid dynamics (CFD), stress analysis, and impact testing. A probability of detection (POD) study was requested by the SSP Orbiter Project for the eddy current (EC) nondestructive evaluation (NDE) techniques that were developed to inspect the flight FCV poppets. This report contains the Appendices to the main report.

Damage due to impacts from micrometeoroids and orbital debris is one of the most significant on-orbit hazards for spacecraft. Impacts to thermal protection systems must be detected and the damage evaluated to determine if repairs are needed to allow safe re-entry. To address this issue for the International Space Station Program, Langley Research Center and Johnson Space Center technologists have been working to develop and implement advanced methods for detecting impacts and resultant leaks. LaRC funded a Small Business Innovative Research contract to Invocon, Inc. to develop special wireless sensor systems that are compact, light weight, and have long battery lifetimes to enable applications to long duration space structures. These sensor systems are known as distributed impact detection systems (DIDS). In an assessment, the NASA Engineering and Safety Center procured two prototype DIDS sensor units to evaluate their capabilities in laboratory testing and field testing in an ISS Node 1 structural test article. This document contains the findings of the assessment.

This paper outlines cryogenic Y-joint testing at Langley Research Center (LaRC) to validate the performance of optical fiber Bragg grating strain sensors for measuring strain at liquid helium temperature (-240 C). This testing also verified survivability of fiber sensors after experiencing 10 thermal cool-down, warm-up cycles and 400 limit load cycles. Graphite composite skins bonded to a honeycomb substrate in a sandwich configuration comprised the Y-joint specimens. To enable SHM of composite cryotanks for consideration to future spacecraft, a light-weight, durable monitoring technology is needed. The fiber optic distributed Bragg grating strain sensing system developed at LaRC is a viable substitute for conventional strain gauges which are not practical for SHM. This distributed sensing technology uses an Optical Frequency Domain Reflectometer (OFDR). This measurement approach has the advantage that it can measure hundreds of Bragg grating sensors per fiber and the sensors are all written at one frequency, greatly simplifying fiber manufacturing. Fiber optic strain measurements compared well to conventional strain gauge measurements obtained during these tests. These results demonstrated a high potential for a successful implementation of a SHM system incorporating LaRC's fiber optic sensing system on the composite cryotank and other future cryogenic applications.

The loss of the Space Shuttle Columbia as a result of impact damage from foam debris during ascent has led NASA to investigate the feasibility of on-board impact detection technologies. AE sensing has been utilized to monitor a wide variety of impact conditions on Space Shuttle components ranging from insulating foam and ablator materials, and ice at ascent velocities to simulated hypervelocity micrometeoroid and orbital debris impacts. Impact testing has been performed on both reinforced carbon composite leading edge materials as well as Shuttle tile materials on representative aluminum wing structures. Results of these impact tests will be presented with a focus on the acoustic emission sensor responses to these impact conditions. These tests have demonstrated the potential of employing an on-board Shuttle impact detection system. We will describe the present plans for implementation of an initial, very low frequency acoustic impact sensing system using pre-existing flight qualified hardware. The details of an accompanying flight measurement system to assess the Shuttle s acoustic background noise environment as a function of frequency will be described. The background noise assessment is being performed to optimize the frequency range of sensing for a planned future upgrade to the initial impact sensing system.

An acoustic monitoring device has at least two acoustic sensors with a triggering mechanism and a multiplexing circuit. After the occurrence of a triggering event at a sensor, the multiplexing circuit allows a recording component to record acoustic emissions at adjacent sensors. The acoustic monitoring device is attached to a solid medium to detect the occurrence of damage.

By using dual-wavelength fiber-optic Bragg gratings, a new technique has been developed for sensing both temperature and strain simultaneously in cryogenic temperature range. Two Bragg gratings with different wavelengths were inscribed at the same location in an optical fiber to form a dual-wavelength sensor. By measuring the wavelength shifts that resulted from the fiber being subjected to different temperatures and strains, the wavelength-dependent thermo-optic coefficients and photoelastic coefficients of the fiber were determined. These coefficients were used to construct the elements of the K matrix, which enables to determine inversely the strain and temperature changes by measuring the wavelength shifts of the dual-wavelength Bragg grating. In this study, measurements were made over the temperature range from room temperature down to about 10 K, addressing much of the low temperature range of cryogenic tanks. A structure transition of the optical fiber during the temperature change was found from about 70 K to 140 K. This transition caused splitting of the waveforms characterizing the Bragg gratings, and the determination of wavelength shifts was consequently complicated. Several alternatives are proposed to resolve this problem. The effectiveness and sensitivities of these measurements in different temperature ranges are discussed. The separation of two wavelengths for the dual-wavelength Bragg grating has been widened to increase the sensitivities of measurement; however, this separation can still be covered in the scanning range from single scanning laser.

This paper discusses the modeling of acoustic emissions in plate structures and their sensing by embedded or surface bonded piezoelectric sensor arrays. Three different modeling efforts for acoustic emission (AE) wave generation and propagation are discussed briefly along with their advantages and disadvantages. Continuous sensors placed at right angles on a plate are being discussed as a new approach to measure and locate the source of acoustic waves. Evolutionary novel signal processing algorithms and bio-inspired distributed sensor array systems are used on large structures and integrated aerospace vehicles for AE source localization and preliminary results are presented. These systems allow for a great reduction in the amount of data that needs to be processed and also reduce the chances of false alarms from ambient noises. It is envisioned that these biomimetic sensor arrays and signal processing techniques will be useful for both wireless and wired sensor arrays for real time health monitoring of large integrated aerospace vehicles and earth fixed civil structures. The sensor array architectures can also be used with other types of sensors and for other applications.

A new technique has been developed for sensing both temperature and strain simultaneously by using dual-wavelength fiber-optic Bragg gratings. Two Bragg gratings with different wavelengths were inscribed at the same location in an optical fiber to form a sensor. By measuring the wavelength shifts that resulted from the fiber being subjected to different temperatures and strains, the wavelength-dependent thermo-optic coefficients and photoelastic coefficients of the fiber were determined. This enables the simultaneous measurement of temperature and strain. In this study, measurements were made over the temperature range from room temperature down to about 10 K, addressing much of the low temperature range of cryogenic tanks. A structure transition of the optical fiber during the temperature change was found. This transition caused splitting of the waveforms characterizing the Bragg gratings, and the determination of wavelength shifts was consequently complicated. The effectiveness and sensitivities of these measurements in different temperature ranges are also discussed.

Acoustic emission (AE) signals created by impact sources in thin aluminum and graphite/epoxy composite plates were analyzed. Two different impact velocity regimes were studied. Low-velocity (less than 0.21 km/s) impacts were created with an airgun firing spherical steel projectiles (4.5 mm diameter). High-velocity (1.8 to 7 km/s) impacts were generated with a two-stage light-gas gun firing small cylindrical nylon projectiles (1.5 mm diameter). Both the impact velocity and impact angle were varied. The impacts did not penetrate the aluminum plates at either low or high velocities. For high-velocity impacts in composites, there were both impacts that fully penetrated the plate as well as impacts that did not. All impacts generated very large amplitude AE signals (1-5 V at the sensor), which propagated as plate (extensional and/or flexural) modes. In the low-velocity impact studies, the signal was dominated by a large flexural mode with only a small extensional mode component detected. As the impact velocity was increased within the low velocity regime, the overall amplitudes of both the extensional and flexural modes increased. In addition, a relative increase in the amplitude of high-frequency components of the flexural mode was also observed. Signals caused by high-velocity impacts that did not penetrate the plate contained both a large extensional and flexural mode component of comparable amplitudes. The signals also contained components of much higher frequency and were easily differentiated from those caused by low-velocity impacts. An interesting phenomenon was observed in that the large flexural mode component, seen in every other case, was absent from the signal when the impact particle fully penetrated through the composite plates.

Advanced, waveform based acoustic emission (AE) techniques have been developed to evaluate damage mechanisms in the testing of composite materials. This approach, more recently referred to as Modal AE, provides an enhanced capability to discriminate and eliminate noise signals from those generated by damage mechanisms. Much more precise source location can also be obtained in comparison to conventional, threshold crossing arrival time determination techniques. Two successful examples of the application of Modal AE are presented in this work. In the first, the initiation of transverse matrix cracking in cross-ply, tensile coupons was monitored. In these tests, it was documented that the same source mechanism, matrix cracking, can produce widely different AE signal amplitudes dependent on laminate stacking sequence and thickness. These results, taken together with well known propagation effects of attenuation and dispersion of AE signals in composite laminates, cast further doubt on the validity of simple amplitude or amplitude distribution analysis for AE source determination. For the second example, delamination propagation in composite ring specimens was monitored. Pressurization of these composite rings is used to simulate the stresses in a composite rocket motor case. AE signals from delamination propagation were characterized by large amplitude flexural plate mode components which have long signal durations because of the large dispersion of this mode.

Previous research on scaling effects in composite materials has demonstrated that the stress levels at first ply failure and ultimate failure of composite laminates are dependent on the size of the laminate. In particular, the thickness dimension has been shown to be the most influential parameter in strength scaling of composite coupons loaded in tension. Geometrically and constitutively scaled laminates exhibit decreasing strength with increasing specimen size, and the magnitude of the strength-size effect is a function of both material properties and laminate stacking sequence. Some of the commonly used failure criteria for composite materials such as maximum stress, maximum strain, and tensor polynomial (e.g., Tsai-Wu) cannot account for the strength-size effect. In this paper, three concepts are developed and evaluated for incorporating size dependency into failure criteria for composite materials. An experimental program of limited scope was performed to determine the first ply failure stress in scaled cross-ply laminates loaded in tension. Test specimens were fabricated of AS-4/3502 graphite-epoxy composite material with laminate stacking sequences of [0 deg n/90 deg n/o deg n]subT where n=1-6. Two experimental techniques were used to determine first ply failure, defined as a transverse matrix crack in the 90 deg ply: (1) step loading with dye penetrant x-ray of the specimen at each load interval, and (2) acoustic emission. The best correlation between first ply failure analysis and experimental data was obtained using a modified Weibull approach which incorporated the residual thermal stress and the outer ply constraint, as well as the ply thickness effect. Finally, a second set of experiments was performed to determine the tensile response and ultimate failure of the scaled cross-ply laminates. The results of these experiments indicated no influence of specimen size on tensile response or ultimate strength.

Advanced, waveform based acoustic emission (AE) techniques have been successfully used to evaluate damage mechanisms in laboratory testing of composite coupons. An example is presented in which the initiation of transverse matrix cracking was monitored. In these tests, broad band, high fidelity acoustic sensors were used to detect signals which were then digitized and stored for analysis. Analysis techniques were based on plate mode wave propagation characteristics. This approach, more recently referred to as Modal AE, provides an enhanced capability to discriminate and eliminate noise signals from those generated by damage mechanisms. This technique also allows much more precise source location than conventional, threshold crossing arrival time determination techniques. To apply Modal AE concepts to the interpretation of AE on larger composite specimens or structures, the effects of modal wave propagation over larger distances and through structural complexities must be well characterized and understood. To demonstrate these effects, measurements of the far field, peak amplitude attenuation of the extensional and flexural plate mode components of broad band simulated AE signals in large composite panels are discussed. These measurements demonstrated that the flexural mode attenuation is dominated by dispersion effects. Thus, it is significantly affected by the thickness of the composite plate. Furthermore, the flexural mode attenuation can be significantly larger than that of the extensional mode even though its peak amplitude consists of much lower frequency components.

Advanced, waveform based acoustic emission (AE) techniques have been previously used to evaluate damage progression in laboratory tests of composite coupons. In these tests, broad band, high fidelity acoustic sensors were used to detect signals which were then digitized and stored for analysis. Analysis techniques were based on plate mode wave propagation characteristics. This approach, more recently referred to as Modal AE, provides an enhanced capability to discriminate and eliminate noise signals from those generated by damage mechanisms. This technique also allows much more precise source location than conventional, threshold crossing arrival time determination techniques. To apply Modal AE concepts to the interpretation of AE on larger composite structures, the effects of wave propagation over larger distances and through structural complexities must be well characterized and understood. In this research, measurements were made of the attenuation of the extensional and flexural plate mode components of broad band simulated AE signals in large composite panels. As these materials have applications in a cryogenic environment, the effects of cryogenic insulation on the attenuation of plate mode AE signals were also documented.

An ultrasonic wave is applied to an anisotropic sample material in an initial direction and the intensity of the ultrasonic wave is measured on an opposite surface of the sample material by two adjacent receiving points located in an array of receiving points. A ratio is determined between the measured intensities of two adjacent receiving points, the ratio being indicative of an angle of flux deviation from the initial direction caused by an unknown applied load. This determined ratio is then compared to a plurality of ratios of a similarly tested, similar anisotropic reference material under a plurality of respective, known load conditions, whereby the load applied to the particular anisotropic sample material is determined. A related method is disclosed for determining the fiber orientation from known loads and a determined flux shift.

Composite materials are being more widely used today by aerospace, automotive, sports equipment, and a number of other commercial industries because of their advantages over conventional metals. Composites have a high strength-to-weight ratio and can be constructed to meet specific design needs. Composite structures are already in use in secondary parts of the Douglas MD-11 and are planned to be used in the new MD-12X. Plans also exist for their use in primary and secondary structures on the Boeing 777. Douglas proposed MD-XX may also incorporate composite materials into primary structures such as the wings and tail. Use of composites in these structures offers weight savings, corrosion resistance, and improved aerodynamics. Additionally, composites have been used to repair cracks in many B-1Bs where traditional repair techniques were not very effective. Plans have also been made to reinforce all of the remaining B-1s with composite materials. Verification of the structural integrity of composite components is needed to insure safe operation of these aerospace vehicles. One aspect of the use of these composites is their response to fatigue. To track this progression of fatigue in aerospace structures, a convenient method to nondestructively monitor this damage needs to be developed. Traditional NDE techniques used on metals are not easily adaptable to composites due to the inhomogeneous and anisotropic nature of these materials. Finding an effective means of nondestructively monitoring fatigue damage is extremely important to the safety and reliability of such structures. Lamb waves offer one method of evaluating these composite materials. As a material is fatigued, the modulus degrades. Since the Lamb wave velocity can be related to the modulus of the material, an effective tool can be developed to monitor fatigue damage in composites by measuring the velocity of these waves. In this work, preliminary studies have been conducted which monitor fatigue damage in composite samples using strain gage measurements as well as Lamb wave velocity measurements. A description of the test samples is followed by the results of two different measurements of Lamb wave velocity. The first technique is a contact measurement done at a single frequency, while the second involves an immersion study of Lamb waves in which dispersion curves are obtained. The results of the Lamb wave monitoring of fatigue damage is compared to the damage progression measured by strain gages. The final section discusses the results and conclusions.

In new method of determining load on specimen, direction of stress wave propagation measured by array of acoustic sensors attached to specimen. Responses of sensors analyzed by strain-energy-density calculations to calculate load. Change in anisotropy of elasticity caused by applied load and corresponding induced strain. Is means to monitor applied load corresponding to measurable shift in direction of propagation of stress wave. Monitoring device based on this method made integral part of airframe in which unidirectional section of structure, such as spar cap or longeron, carries primary load.

An ultrasonic wave is applied to an anisotropic sample material in an initial direction and an angle of flux deviation of the ultrasonic wave front is measured from this initial direction. This flux deviation angle is induced by the unknown applied load. The flux shift is determined between this flux deviation angle and a previously determined angle of flux deviation of an ultrasonic wave applied to a similar anisotropic reference material under an initial known load condition. This determined flux shift is then compared to a plurality of flux shifts of a similarly tested, similar anisotropic reference material under a plurality of respective, known load conditions, whereby the load applied to the particular anisotropic sample material is determined. A related method is disclosed for determining the fiber orientation from known loads and a determined flux shift.

A method for determining the angle of impact of an object on a thin-walled structure which determines the angle of impact through analysis of the acoustic waves which result when an object impacts a structure is presented. Transducers are placed on and in the surface of the structure which sense the wave caused in the structure by impact. The waves are recorded and saved for analysis. For source motion normal to the surface, the antisymmetric mode has a large amplitude while that of the symmetric mode is very small. As the source angle increases with respect to the surface normal, the symmetric mode amplitude increases while the antisymmetric mode amplitude decreases. Thus, the angle of impact is determined by measuring the relative amplitudes of these two lowest order modes.

In isotropic materials, the direction of the energy flux (energy per unit time per unit area) of an ultrasonic plane wave is always along the same direction as the normal to the wave front. In anisotropic materials, however, this is true only along symmetry directions. Along other directions, the energy flux of the wave deviates from the intended direction of propagation. This phenomenon is known as energy flux deviation and is illustrated. The direction of the energy flux is dependent on the elastic coefficients of the material. This effect has been demonstrated in many anisotropic crystalline materials. In transparent quartz crystals, Schlieren photographs have been obtained which allow visualization of the ultrasonic waves and the energy flux deviation. The energy flux deviation in graphite/epoxy (gr/ep) composite materials can be quite large because of their high anisotropy. The flux deviation angle has been calculated for unidirectional gr/ep composites as a function of both fiber orientation and fiber volume content. Experimental measurements have also been made in unidirectional composites. It has been further demonstrated that changes in composite materials which alter the elastic properties such as moisture absorption by the matrix or fiber degradation, can be detected nondestructively by measurements of the energy flux shift. In this research, the effects of nonlinear elasticity on energy flux deviation in unidirectional gr/ep composites were studied. Because of elastic nonlinearity, the angle of the energy flux deviation was shown to be a function of applied stress. This shift in flux deviation was modeled using acoustoelastic theory and the previously measured second and third order elastic stiffness coefficients for T300/5208 gr/ep. Two conditions of applied uniaxial stress were considered. In the first case, the direction of applied uniaxial stress was along the fiber axis (x3) while in the second case it was perpendicular to the fiber axis along the laminate stacking direction (x1).

Simulated acoustic emission signals were induced in a thin-walled graphite/epoxy tube by means of lead breaks (Hsu-Nielsen source). The tube is of similar material and layup to be used by NASA in fabricating the struts of Space Station Freedom. The resulting waveforms were detected by broad band ultrasonic transducers and digitized. Measurements of the velocities of the extensional and flexural modes were made for propagation directions along the tube axis (0 degrees), around the tube circumference (90 degrees) and at an angle of 45 degrees. These velocities were found to be in agreement with classical plate theory.

A recently developed ultrasonic techique which has been successful in monitoring the integrity of advanced structures is described. The technique is based on a two-transducer contact type arrangement that can be used to determine the dispersion curves of guided waves generated within the specimen. The phase velocity of antisymmetric plate waves is determined through spectral analysis of signals recorded from a lead break source on the surface of graphite epoxy laminates. The influence of all five stiffness constants on the dispersion curves of both symmetric and antisymmetric plate waves is theoretically investigated for propagation at 0, 45, and 90 deg to the fibers. All but c12 were found to have a strong influence on the dispersion curves in this frequency range. The fit with the theoretically predicted dispersion curves was excellent for propagation parallel and perpendicular to the fibers in the unidirectional specimen and good for the other cases considered.

Acoustic emission was interpreted as modes of vibration in plates. Classical plate theory was used to predict dispersion curves for the two fundamental modes and to calculate the shapes of flexural waveforms produced by vertical step function loading. There was good agreement between theoretical and experimental results for aluminum. Composite materials required the use of a higher order plate theory (Reissner-Mindlin) to get good agreement with the measured velocities. Four composite plates with different laminate stacking sequences were studied. The dispersion curves were determined from phase spectra of the time dependent waveforms. Plate modes were shown to be useful for determining the direction of source motion. Aluminum plates were loaded by breaking a pencil lead against their surface. By machining slots at angles to the plane of a plate, the direction in which the force acted was varied. Changing the source motion direction produced regular variations in the waveforms. To demonstrate applicability beyond simple plates, waveforms produced by lead breaks on a thin walled composite tube were also shown to be interpretable as plate modes. The tube design was based on the type of struts proposed for Space Station Freedom's trussed structures.

Lead breaks (Hsu-Neilsen source) were used to generate simulated acoustic emission signals in an aluminum plate at angles of 0, 30, 60, and 90 degrees with respect to the plane of the plate. This was accomplished by breaking the lead on slots cut into the plate at the respective angles. The out-of-plane and in-plane displacement components of the resulting signals were detected by broad band transducers and digitized. Analysis of the waveforms showed them to consist of the extensional and flexural plate modes. The amplitude of both components of the two modes was dependent on the source orientation angle. This suggests that plate wave analysis may be used to determine the source orientation of acoustic emission sources.

The normalized change in ultrasonic "natural" velocity as a function of stress and temperature was measured in a unidirectional laminate of T300/5208 graphite/epoxy composite using a pulsed phase locked loop ultrasonic interferometer. These measurements were used together with the linear (second order) elastic moduli to calculate some of the nonlinear (third order) moduli of this material.

Ultrasonic waves suffer energy flux deviation in graphite/epoxy because of the large anisotropy. The angle of deviation is a function of the elastic coefficients. For nonlinear solids, these coefficients and thus the angle of deviation is a function of stress. Acoustoelastic theory was used to model the effect of stress on flux deviation for unidirectional T300/5208 using previously measured elastic coefficients. Computations were made for uniaxial stress along the x3 axis (fiber axis) and the x1 for waves propagating in the x1x3 plane. These results predict a shift as large as three degrees for the quasi-transverse wave. The shift in energy flux offers a new nondestructive technique of evaluating stress in composites.

Ultrasonic waves suffer energy flux deviation in graphite/epoxy because of the large anisotropy. The angle of deviation is a function of the elastic coefficients. For nonlinear solids, these coefficients and thus the angle of deviation is a function of stress. Acoustoelastic theory was used to model the effect of stress on flux deviation for unidirectional T300/5208 using previously measured elastic coefficients. Computations were made for uniaxial stress along the x3 axis fiber axis) and the x1 axis for waves propagating in the x1x3 plane. These results predict a shift as large as three degrees for the quasi-transverse wave. The shift in energy flux offers new nondestructive technique of evaluating stress in composites.

Measurements of stress-induced velocity changes for propagation along the direction of applied stress in graphite/epoxy composites are presented. A linear relation between normalized velocity shift and stress is demonstrated for propagation and stress direction perpendicular to the fiber direction. The ratio of the SAC to the elastic compliance for this loading direction was assessed and found to have a value similar to that of numerous other materials which have very different linear elastic properties. Measurements with stress and propagation along the fibers yielded unusual behavior; the curves were very nonlinear and even shifted direction at higher loads.

The theoretical treatment of linear and nonlinear elasticity in a unidirectionally fiber reinforced composite as well as measurements for a unidirectional graphite/epoxy composite (T300/5208) are presented. Linear elastic properties were measured by both ultrasonic and strain gage measurements. The nonlinear properties were determined by measuring changes in ultrasonic natural phase velocity with a pulsed phase locked loop interferometer as a function of stress and temperature. These measurements provide the basis for further investigations into the relationship between nonlinear elastic properties and other important properties such as strength and fiber-matrix interfacial stength in graphite/epoxy composites.