Acknowledgent:
The paper was first published at theConference on NDE applied to Composite Fabrication,
organized by NTIAC and McDonnell Douglas Aerospace and took place in St. Louis Missouri in Oct '94.
See also all 54 NDE Abstracts of the '94 and '96 Conference Proceedings.
|
Figure 1. Examples of parameters and properties that can be nondestructively monitored for process control of organic matrix composites. | |
| THERMOSET | T HERMOPLASTIC |
| Cure (Polymerization) | Crystallinity |
| Temperature | MeIt Stability |
| Exotherm | MeIt Temperature |
| Viscosity | Viscosity |
| Modulus | Fiber/Resin Ratio |
| Tg Shifts | Density |
| Fiber/Resin Ratio | Porosity |
| Density | Degradation of Resin |
| Porosity | Delaminations |
| Fiber Orientation | Consolidation |
| Delaminations | Fiber Orientation |
| Overall Material Integrity | Overall Material Integrity |
| Others | Others |
The critical process step requiring monitoring is the final consolidation process for both thermoset and thermoplastic composites. Examples of the parameters and properties that can be nondestructively monitored for process control of organic matrix composites are listed in Figure 1. The two parameters currently routinely monitored are the temperature and the pressure applied to the filaments, tows, or laminates. Commonly this is done with conventional thermocouples and pressure gauges. The only other device fairly routinely used to monitor the cure process is a dielectric sensor mounted between two composite plys in autoclave curing which measures the ionic conductivity.
| Figure 2. Potential devices for process monitoring of organic matrix composite matrials | |
| THERMOSET | THERMOPLASTIC |
| Thermocouple | Acoustic waveguide |
| Acoustic Waveguide | Optical fibers |
| Optical Fibers | Thermal |
| Capacitance | Ultrasonics |
| Conductivity | Magnetostrictive wire |
| Dielectric | |
| Thermal | |
| FT-IR | |
| Ultrasonics | |
| Magnetostrictive wire | |
Figure 2 lists the potential devices for process monitoring of organic matrix composites. At the present time, no technique exists to optimally monitor the mechanical properties during the curing or melt process, although the mechanical properties are the most important factor for overall integrity of composite structures. Ultrasonic tests can be designed and optimized for measurement of the elastic wave speed and attenuation of the wave propagating through the composite during the consolidation process. The wave speed is a function of the elastic modulus, viscosity and density of the composite, while the attenuation is a monitor of the change in viscosity of the resin, fiber compaction, resin-fiber ratio and porosity. Figure 3 compares NDE techniques available for process monitoring of organic
composite materials.
Figure 4 shows the results of an experiment where an ultrasonic waveguide embedded in the midplane of 16-ply graphite/epoxy panel during the prepreg lay-up process was used for in-process cure monitoring.
|
Figure 3. Comparison of NDE techniques for process monitoring of composite materials. | |||||
| SENSORS | PARAMETER | INSTALLATION | RUGGEDNESS | IMPLEMENTATION COST | REMARKS |
| CONVENTIONAL | Strain Temperature Pressure | Surface bonded | Moderate | Moderate | Difficult to use in production |
| EMBEDDED |
Dynamic-strain acoustic signals cure... | In-situtd | Potentialy high | Potentialy low | Emerging technology |
| ULTRASONICS | Cure, modulus density | Surface or non contact | Potentialy high | Potentialy low | Emerging remote sensor technology |
It is also important to note that the character of the piezoelectric transducer itself exerts a major influence on the components of the ultrasonic signal, since conventional transducers do not respond as a simple vibrating piston and have their own frequency, amplitude, and directional response. In addition, they "ring" at their resonance frequency and it is extremely difficult to distinguish between the amplitude excursions caused by this "ringing" and the amplitude variations actually characteristic of the ultrasonic signal. They are also limited in their frequency response and, since they are in contact with the surface of the material to be tested, they can load this surface and thereby modify the ultrasonic wave itself.
EMAT's, on the other hand, have been successfully used for material defect characterization particularly in metal bars, tubes, pipes, and plates. One major problem with EMAT's is that the efficiency of ultrasound and detection rapidly decreases with lift-off distance between the EMAT's face and the surface of the test object. They can obviously only be used for examination of electrically conducting materials. Because of the physical processes involved they are much better detectors than generators of ultrasound.
Air-coupled ultrasonic systems have been recently developed and currently research is underway to optimize them for practical non-contact ultrasonic applications. These systems are relatively similar to conventional contact ones and, therefore, when optimized may play an important role in modem nondestructive evaluation of composites. They may also be incorporated into hybrid systems using lasers for ultrasound generation and air-coupled transducers for efficient ultrasound detection. Furthermore, gases more dense than air may be used to improve the acoustical impedance match between transducer and test structure. An air-coupled c-scan system has been used for imaging of defects in wooden substrates for panel paintings and is currently being optimized for applications in the lumber industry. Figure 6 shows a schematic portrayal of an air-coupled ultrasound system for online monitoring of prepreg tows. Recently, such a system has been successfully used to characterize graphite/epoxy prepreg materials.
Three different mechanisms account for the generation of ultrasonic waves by the impact of pulsed laser beams, namely radiation pressure, ablation, and themmoelasticity. Radiation pressure, which is caused by momentum transfer from the incident electromagnetic pulse, is the least efficient of the three mechanisms and is therefore of little importance for practical applications. At the other extreme, when a laser pulse possessing sufficient power density strikes the surface of a material the electromagnetic radiation is absorbed in a very thin layer of the material and vaporizes it. The amplitude of the ultrasonic wave generated in the material by this ablation process can often be increased by placing a coating of a material possessing different themmal properties on the surface of the test object. Although the surface of the test object is slightly damaged when ablation occurs, in certain cases the amount of damage is acceptable when such a generation process is the only way to obtain ultrasonic waves of sufficient amplitude in a non-contact manner. The themmoelastic process consists of absorption of a laser pulse possessing moderate energy in a finite depth of the material under investigation such that themmal expansion causes a volume change and consequently an elastic wave. Thus, the themmoelastic process is the only process which is truly nondestructive and still capable of generating an ultrasonic wave of sufficient amplitude for nondestructive evaluation purposes.
Laser detection of ultrasonic signals is usually performed using an interferometer system to detect the minute displacements or velocity of an object surface as the ultrasonic waves impinges upon or propagates along it. A beam of laser light is reflected from the object surface back into the interferometer system so that changes in the optical phase or Doppler shifted frequency can be detected. Detection of ultrasonic signals can be recorded with very broad frequency bandwidths in excess of 100 MHz, if necessary, thereby resulting in very high fidelity recording of actual ultrasonic surface displacements. Fundamental calibration based upon the known wavelength of the laser light means that ultrasonic displacements amplitudes can be measured with a high degree of accuracy. Consequently, absolute attenuation measurements can be made independent of couplant materials, transducer resonances, and other variations inherent in conventional piezoelectric transducer systems. In fact, laser interferometer systems have been designed specifically to facilitate highly accurate attenuation measurements of surface and plate waves.
Although they possess high fidelity and are highly accurate, laser interferometric receivers are the weak link in laser-based ultrasonic systems. Although laser generator can be a relatively efficient source of ultrasonic signals, a laser interferometric receiver is somewhat less sensitive than its piezoelectric counterpart. For this reason several schemes have been explored to enhance the overall laser ultrasonic system sensitivity. Direct methods for enhancing sensitivity involve the use of very bright (and often quite expensive) lasers for both the generating and receiving sources. Preliminary results have shown great promise for other, more sophisticated methods such as laser generation of narrow-band, tone burst signals, and the use of spatial arrays. Figure 7 illustrates the application of a laser based ultrasonic system for monitoring the mechanical properties of several layers of a composite laminate. A pulsed laser is used to generate broad frequency ultrasonic pulse at the surface of the composite. A dual beam laser interferometer is used to measure the velocity and attenuation of the waves as they pass beneath the two probing beams of the interferometer. This technique permits direct unmodified measurement of velocity and absolute attenuation as a function of frequency and, as a result, monitoring of mechanical properties as a function of depth into the composite
|
A completely new laser generation system has been designed and constructed to generate discrete frequency ultrasound rather than the broad frequency band ultrasound generated by conventional laser pulses. For the non-contact hybrid laser generation/EMAT detection ultrasonic system, an optimized EMAT system has been designed and constructed to detect the laser generated ultrasonic pulses. For the laser generation/laser detection ultrasonic system a specially designed laser interferometer is used which has a flat frequency response with virtually no phase shift over the bandpass range of interest, namely, l0 kHz to l0 Mhz. This interferometer provides absolute amplitude calibration at the beginning and end of a test as well as instantaneous relative calibration at the moment of measurement. Since, in some experiments, the test specimen may move or change reflectivity, the latter feature is critical in assuring accurate measurements. For the gas-coupled ultrasonic system, an innovative system is being designed and optimized for practical applications.
Conventional pulsed laser systems used to generate ultrasonic waves are high power pulsed lasers operated at reduced power to eliminate damage to the test structure. By their very nature such a single laser pulse generates a broad band of frequencies. Unfortunately, the signal-to-noise ratio of a laser interferometer is inversely proportional to the frequency bandwidth. In order to increase the laser detector signal-to-noise ratio it is desirable to have a laser capable of generating discrete narrow band frequencies similar to what is accomplished with conventional narrow band piezoelectric transducers. Figure 8 is a schematic drawing of a new ten cavity laser system which is the second generation version of special laser systems developed in the Johns Hopkins University Center for Nondestructive Evaluation for generation of a discrete frequency chain of ultrasonic pulses. This multi-element laser is composed of ten Q-switched Nd:YAG cavities with a common power supply and timing electronics. The timing circuit permits the cavities to be fired over a large range of delays. Turning mirrors are used to direct the infrared light pulses through a cylindrical lens to the test structure.
Figure 9
shows the optical signal output from the ten cavity laser system and its power spectrum for a 2 MHz firing rate.
Since, as presently designed, lasers systems are far better generators of ultrasound in materials than detectors, while EMAT's are far better ultrasound detectors than generators, researchers in CNDE have used the ten cavity Nd:YAG laser system to generate narrow-band ultrasound and a meander line EMAT as a detector,
Figure l0 to generate and detect vertically polarized, angle beam shear waves. For cost of construction, compactness, and operational capability this system is the superior non-contact ultrasound system developed up to the present time.
Figure 11 is a schematic of a proposed in-process laser ultrasonic system for monitoring composite tape laying with on-the-fly curing.
Figure 12 shows a schematic of a proposed system which uses a pulsed laser to generate ultrasonic waves at the surface of a tow, tape, or laminate
and a laser beam optical interferometer for detection of the reflected waves. The pulsed laser beams are delivered to the composite surface through a linear array of optical fibers and the reflected probe laser beams are detected with a second linear array of optical fibers. An optical multiplexer permits a single pulsed laser (or the set of ten multi-element lasers) described above, to be used for generation and a single continuous wave laser to be used for detection. Depending on the width of the optical fiber arrays the entire width of a tape being placed could be monitored extremely rapidly without mechanical scanning. In addition, by the use of different optical fiber arrays, different widths of composite and composites of different shapes could be monitored and inspected using the same laser generation and detection systems.
For more information see: NDT in Aerospace - UTonline 11/97
| | UTonline | |