+ All Categories
Home > Documents > COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher...

COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher...

Date post: 26-Aug-2020
Category:
Upload: others
View: 10 times
Download: 6 times
Share this document with a friend
17
COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee 37831-7274 P.O. BOX 2003, K-1225, MS7294 Halen E. Philpot Oak Ridge K-25 Site' G. Louis Powell Oak Ridge Y-12 Plant' To whom all correspondence should be addressed ABSTRACT The effects of heat damage were determined on the residual mechanical, physical, and chemical properties of lM6/3501-6 laminates, and potential nondestructive techniques to detect and assess material heat damage were evaluated. About one thousand preconditioned specimens were exposed to elevated temperatures, then cooled to room temperature and tested in compression, flexure, interlaminar shear, shore-D hardness, weight loss, and change in thickness. Specimens experienced significant and irreversible reduction in their residual properties when exposed to temperatures exceeding the material upper service temperature of this material (35OOF). The Diffuse Reflectance Infrared Fourier Transform and Laser- Pumped Fluorescence techniques were found to be capable of rapid, in-service, nondestructive detection and quantitation of heat damage in lM6/3501-6. These techniques also have the potential applicability to detect and assess heat damage effects in other polymer matrix composites. * Managed for the U.S. Department of Energy by Martin Marietta Energy Systems, Inc., under Contract No. DE- AC05-840R21400. DI%RIBUTDN OF MIS DOCUMENT IS UNLIMITED M E
Transcript
Page 1: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

COMPOSITE HEAT DAMAGE ASSESSMENT

Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory.

Oak Ridge, Tennessee 37831-7274 P.O. BOX 2003, K-1225, MS7294

Halen E. Philpot Oak Ridge K-25 Site'

G. Louis Powell Oak Ridge Y-12 Plant'

To whom all correspondence should be addressed

ABSTRACT

The effects of heat damage were determined on the residual mechanical, physical, and chemical properties of lM6/3501-6 laminates, and potential nondestructive techniques t o detect and assess material heat damage were evaluated. About one thousand preconditioned specimens were exposed t o elevated temperatures, then cooled t o room temperature and tested in compression, flexure, interlaminar shear, shore-D hardness, weight loss, and change in thickness. Specimens experienced significant and irreversible reduction in their residual properties when exposed t o temperatures exceeding the material upper service temperature of this material (35OOF). The Diffuse Reflectance Infrared Fourier Transform and Laser- Pumped Fluorescence techniques were found t o be capable of rapid, in-service, nondestructive detection and quantitation of heat damage in lM6/3501-6. These techniques also have the potential applicability to detect and assess heat damage effects in other polymer matrix composites.

* Managed for the U.S. Department of Energy by Martin Marietta Energy Systems, Inc., under Contract No. DE- AC05-840R21400.

DI%RIBUTDN OF MIS DOCUMENT IS UNLIMITED M E

Page 2: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

1. I NTRO DUCT1 ON

Organic resin systems used in the manufacture of polymer matrix composites (PMC) are susceptible t o significant and irreversible thermal oxidative degradation following subjection t o temperatures above or within a narrow window of their upper service temperatures. When exposed t o temperatures sufficiently high enough t o cause resin degradation, these materials experience a drop in the glass transition temperature that effectively lowers the upper service temperature and significantly reduces the mechanical strength properties of the composite (1,2). In addition, embrittlement and cracking of the surface cause a loss in t he impact strength of the material. Therefore, PMC structures exposed to overheat conditions can suffer irreversible and catastrophic damage in a very short time.

Typical heat damage can result from fires, lightning strikes, supersonic dashes, heating blankets, or curing ovens/autoclaves. Heat damage can also result from exposure t o hot gases from missile efflux or the ground reflected engine efflux from vertical or short-takeoff and landing aircraft. Composite structures that have the potential for heat damage include the AV-8B Harrier jumpjet (31, the newly developed V-22 Osprey, the F-I 8, and any other PMC structures that experience temperatures near or above the material upper service temperature.

Nondestructive evaluation (NDE) techniques based on ultrasonic methods, including digital pulse (broadband), tone burst (monochromatic), and ultrasonic-C scanning, have been investigated by researchers a t the Oak Ridge National Laboratory/Applied Technology Division (ORNL/ATD) and determined inadequate for detecting heat damage in lM6/350 1-6 laminates (4). The early onset of composite heat damage cannot be detected by conventional NDE techniques because the size of the damage is well below the practical detection limits for these techniques. Other techniques investigated by ORNL/ATD researchers include attenuated total reflectance, fluorescence, specular reflectance, fourier transform-raman, diffuse reflectance visible, ultraviolet-visible diffuse reflectance, Diffuse Reflectance Infrared Fourier Transform (DRIFT) and Laser-Pumped Fluorescence (LPF). The DRIFT and LPF techniques demonstrated the highest degree of promise for detecting and assessing composite heat damage (5).

2. EX PER I MENTAL

2.1 Specimen Preparation and Preconditioninq Twenty-one ply lM6/3501-6 panels were fabricated and cured in an autoclave in accordance with the recommendations of Hercules (the manufacturer of the material). The panels were then cut into about one thousand test specimens. These specimens were sized as required by the American Society for Testing and Materials (ASTM) standards. The lay-up for the specimens was [ + 4 5 O , 90°, (-45O, +45 o)4, 0°0.5]5. The testing orientation was either O o or 90°.

Half of these specimens were preconditioned in a vacuum oven at 49OC (120OF) (dry specimens). The other half were placed in a humidity chamber at 49OC (120OF) and 9 8 % relative humidity (wet specimens). Selected specimens from each group were monitored weekly to determine when the specimens were preconditioned according to ASTM D570-81. The average percent weight gain, determined on five preconditioned specimens per panel, varied from 0.64 t o 1.20%. The average percent weight loss for the dry specimens varied from 0.04 t o 0.32%.

2.2 Heat Exposure of SDecimens exposed on one face, without load, t o one of the following temperaturehime conditions:

Dry and wet preconditioned specimens were then

Page 3: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

0 177 , 232, 288,343OC (350,450, 550, and 650OF) for 60 min 0 288, 343, 399OC (550, 650,'and 75OOF) for 5 min 0 288, 343, 3990C (550, 650, and 75OOF) for 30 min 0 399OC (75OOF) for 10 min

Following heat exposure, the specimens were removed from the heating chamber and allowed to cool t o room temperature, 23OC ( 7 3 O F 1 , prior t o mechanical testing. Dry and wet room- temperature control specimens were also prepared, removed from the preconditioning ovens, and allowed t o cool before testing.

The heat source was a manually controlled, parabolic clamshell furnace equipped with twelve radiant quartz lamps. The ramp-up t o the exposure temperature was achieved as quickly as possible t o simulate a sudden one-sided heat blast from a jet engine. Thermocouples were placed beneath the second ply of the heated face of selected reference specimens and were used to control and monitor the temperature during each heat run. After the steady-state temperature was reached (which took from several seconds up t o two minutes), heating times were initiated.

2.3 Mechanical ProDertv Testing The mechanical properties of test specimens were. determined at room temDerature in accordance with ASTM standards and tested in compression, ASTM D3410-87 (IITRI fixture); flexure (4-point bend), ASTM D790-87; short- beam shear (SBS), ASTM D2344-84; and shore D, ASTM D2240-86. The actual specimen geometries and testing procedures can be found in Ref. 5. Tables 1 through 3 summarize the average O o and 90° strength data, percent strength retention, and percent coefficient of variation for the lM6/3501-6 specimens. The percent strength retention data for the dry and wet specimens was calculated based on the highest average strength having 100% strength retention.

A Shore tester was used to record shore-D hardness values on specimens t o determine the effects of heat exposure on composite hardness according t o ASTM D2240-86. To obtain an average hardness value, each specimen was tested at six different locations, at least 0.64 cm (0.25 in.) apart, on the heat-exposed side. An average shore-D hardness value was recorded for ten specimens a t each temperaturehime set point.

Figure 1 is a plot of the average shore-D hardness data and percent strength retention vs temperature/time for the dry and wet lM6/3501-6 test specimens, according to ASTM D2440- 86. The percent shore-D hardness retention for the dry and wet averages was calculated based on the highest average shore-D hardness (wet 177OC/60 min.) having 100% hardness retention.

2.4 Phvsical ProDertv Testing The percent weight loss data were determined on wet and dry preconditioned, heat-damaged IM6/3501-6 specimens. Specimens were weighed immediately before and after heat exposure in an attempt t o correlate strength degradation with weight loss as a function of temperature and time. Figure 2 summarizes the average percent weight loss for the lM6/3501-6 specimens after exposure t o various temperaturehime treatments.

Maximum thickness measurements were determined on wet and dry preconditioned heat- damaged 1M6/3501-6 specimens after exposure t o various temperaturehime treatments. Micrometer readings were taken between opposite faces of the specimens a t their thickest section. Figure 3 summarizes the average micrometer measurements.

Page 4: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

2.5 Laser-PumDed Fluorescence SDectroscoDy LPF is a powerful means for probing localized changes in the electronic structure and environment of individual functional groups and for detecting small populations of molecules in a complex matrix. For materials such a s the lM6/3507-6, the resin molecules are expected to fluoresce, but t he IM6 graphite fibers are not expected to produce substantial fluorescence.

The LPF experiments were performed on actual heat-damaged lM6/350 1-6 flexural specimens with a research laser raman spectroscopy system, using tunable krypton-ion (482.5 nm) and argon-ion lasers (488.0 nm) for excitation. For the 482.5-nm excitation experiments, each specimen was placed in a holder supported by a precision X-Y stage, and its position was optimized to provide the greatest average signal. For all experiments in these studies, average intensities were calculated over 5-nm-wide regions of t he spectrum, covering wavelengths from 510 to 535 nm. Results from these studies on selected specimens are compiled in Table 4.

2.6 Diffuse Reflectance Infrared Fourier Transform SoectroscoDy DRIFT is a method of scattering infrared (IR) light from the surface of a material, collecting the diffusely scattered light, and reducing that light t o an IR spectrum from which qualitative and quantitative chemical information can be obtained. The IR spectral information is specific to chemical functional groups such a s ketones, esters, hydroxyls, and hydrocarbons. Unlike internal reflectance techniques for obtaining similar IR spectra, DRIFT does not require mechanical contact with the material surface for spectral measurement.

Laboratory equipment for the DRIFT analyses included a 610-RAD FTS-60 Fourier transform infrared (FTIR) spectrometer that had a praying-mantis, diffuse reflectance specimen compartment accessory (Harrick Scientific Model DRA3CQ) with evacuable cells for analyzing small specimens in a controlled environment. The use of evacuable cells with the praying- mantis optics was developed jointly by Powell and Smyrl (6) with Harrick Scientific, Inc., for analyzing materials in real time at elevated temperatures in controlled gaseous environments and is essentially described by Smyrl et al. (7). DRIFT experiments performed in the evacuable cell used specimens 1.27 cm (0.5 in.) square by 0.31 8 cm (0.125 in.) thick. Every diffuse reflectance spectrum was obtained from a 2-mm-diameter (0.08-in.) area near t he center of the specimens. The inlet gas w a s regulated by a precision metering valve for which the supply g a s could be either argon or compressed air. The specimen was maintained at constant temperature, and oxidation was initiated by switching the supply g a s from argon to compressed air.

The FTlR spectrometer was also used with a barrel ellipsoid IR inspection accessory "SPECTROPUS" system for the post-mortem analysis of actual heat-damaged flexural specimens. The SPECTROPUS system (8) is a prototype system developed by G.L. Powell and Harrick Scientific, Inc., to extend diffuse reflectance measurements beyond the spectrometer specimen compartment to analyze variably sized surfaces (9). This accessory allows diffuse reflectance spectra to be obtained from almost any convex or flat surface, with no purge-time delays and with spectral quality comparable to the best capabilities of FTlR instruments. The barrel ellipsoid accessory was used to obtain DRIFT spectra from heat-damaged specimens. Both the evacuable cell/praying mantis system and the barrel ellipsoid system have excellent rejection of specularly reflected light, which w a s the main problem encountered by Cole et al. with the diffuse reflectance method (1 0).

The main features in the spectrum shown in Figure 4 are the O-H (hydroxyl) stretching band

Page 5: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

a t 3500 wave numbers (cm-'1, the C-H stretching bands near 3000 cm-' (aromatic hydrocarbons above 3000 cm'l, aliphatic hydrocarbons below 3000 cm-'1, the aliphatic ketone near 1 730 cm'l , and aromatic ring bands at 1 600 and 1 520 cm". Spectral data were reduced to a f e w simple numbers and were baseline corrected (linear between 3800 and 2000 cm-l). Peak heights and areas and the spectral collection time were normalized by using quantitative analysis software BIO-RAD QUANT32 CLASS.

3. RESULTS

3.1 Mechanical and Phvsical Prooertv Testinq On 1M6/3501-6 laminates, shore-D hardness and 0 O compressive strength properties were the least sensitive to elevated temperature exposure. Our studies indicated that shore D is a poor indicator for relating loss in mechanical strength properties t o heat damage. No noticeable decrease in average shore-D hardness values manifests itself until approximately 343OC (65OOF). By the time the material had reached tha t temperature, 66 t o 92% of its original O o and 90° interlaminar shear and flexural strength was lost.

lnterlaminar shear and flexural strength properties of t he lM6/3501-6 specimens were very sensitive to heat damage and dropped dramatically after exposure to 288OC (55OoF), losing up to 8 0 % of their original room-temperature strength. The 90 O SBS room-temperature control specimens exhibited strengths over 1.5 times tha t of t he O o SBS specimens. However, after exposure to 288OC (55O0F) and higher, t he strengths for t he 90° and Oo SBS specimens were almost equivalent. The same trend w a s also observed for t he Oo and 90° flexural specimens.

No visual or microscopic damage or increase in thickness of the lM6/3501-6 specimens w a s evident at exposure temperatures of 288 O C (55OOF) and below. Delamination, translaminar cracking, and a n increase in specimen thickness were first visually observed at exposure temperatures of 343OC (65OOF). At an exposure temperature of 343OC (650OF) for 60 min, t he thickness of the specimens w a s almost double that of specimens exposed to 288OC (55OOF) and below. The cause of this tremendous change in laminate thickness can probably be attributed to the rapid volatilization of resin degradation products generated during thermal exposure. These undesirable dimensional changes produce areas of high stress concentration and contribute to significant strength loss especially above 288 OC (55OOF).

Weight losses were evident at exposure temperatures of 232OC (45OOF) through 343OC (65OOF) and accounted for 0 to 1 0 % of the original dry weight of the laminate. Water, propenal, and other volatile organic compounds have been reported to be the major by- products generated during thermal exposure at these temperatures (1 1 ).

3.2 Laser-Pumoed Fluorescence Soectroscooy The LPF results confirm the differences between composite specimens tha t exhibit varying degrees of thermal degradation. In general, LPF intensity measurements in the 5 10- t o 535-nm region decreased in intensity a s strength decreased (Table 4). These intensity decreases were fairly significant at the early to middle s t ages of heat damage, and tended t o level off a t the advanced s tages of heat damage. Furthermore, t he results obtained using laser excitation at 482.5 and 488.0 nm provide equivalent results. Hence, t he strong 488-nm argon-ion-laser can be used in place of the much weaker 482-nm krypton-ion-laser.

3.3 Diffuse Reflectance Infrared Fourier Transform Soectroscoov The evacuable cell

Page 6: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

specimen w a s given a temperature cycle in s teps up t o 32OOC (608OF) where it w a s held'for 1 hour. An argon flow w a s used to check the temperature control system and to evaluate spectral changes tha t were related t o thermal effects rather than to oxidation. These thermally induced changes were minor. The same specimen was heated again, then isothermed at 289OC (553OF) under an argon flow and, a t time = 0 s, t he gas was switched to compressed air. The specimen was maintained a t 289OC (553OF) for 172 ks. Spectra were obtained a t a rate of 140 s per spectrum for 4 ks, after which spectral acquisition was slowed for long- term monitoring. Figure 5 shows baseline-corrected spectra in the carbonyl region indicating oxygen addition reactions. The specimen experienced rapid addition of phenones (1 650 cm") and aliphatic ketones. At longer times, the carbonyl region expanded below 1800 cm-', indicating an increase in the degree of oxidation with the formation of acids, esters, acid anhydrides, and carbonate esters. A decrease in the aromatic-ring character was also apparent from the 1 6 0 0 and 1520 cm-' bands. Figure 6 shows baseline-corrected spectra in the C-H, O-H, N-H region, where the loss of aliphatic hydrocarbons (2900 cm-') occurred in preference to aromatic hydrocarbon (3050 cm"). A similar loss w a s observed for aliphatic hydroxyls (3550 cm"), and a band developed (perhaps an amine) at 3470 cm''.

Figure 7 s h o w s the time dependence of the oxidation processes using reduced-spectral data as relative absorbance, obtained a s follows. The term "band" refers to a normalized peak height, The term "band edge" is the value of the baseline-corrected spectrum at the indicated cm" divided by the maximum value between 1720 and 1750 cm'' for t he carbonyl region (1 600 to 2000 cm") and the maximum value between 3350 and 3425 cm-' for t he hydroxyl region (3000 to 4000 cm"). The "band area" is an integral over some region of the baseline- corrected spectrum divided by the integral between 1320 and 1380 cm", a region of t he spectrum tha t is not strongly affected by oxidation. For the carbonyl region, t he integration range was 1600 to 2000 cm". For the hydrocarbon region the integration range was from 2670 to 3130 cm" with the baseline corrected for a 20-crn" average over the integration limits. For Figures 8-1 0, the band edge and band area determinations were further normalized by a factor tha t made the value for the last unoxidized spectrum unity.

The 1730-cm-' carbonyl band present in the specimen tha t w a s not heat damaged was assimilated into the ketone-phenone region of the spectrum as the phenone band appeared (Le., in the first kilosecond of oxidation). During the first 10 k s of oxidation, half t he aliphatic hydrocarbon and hydroxyls were burned away and the apparent resin thickness (1 350 cm") decreased by about 20%. Eventually, all of the aliphatic hydrocarbons and hydroxyls were removed by oxidation at approximately the same rate. The hydroxyl band edge measurement does not go to zero because of the formation of another band, probably a n amine, with a maximum value at 3470 cm-'. The carbonyl band area best monitors t he total oxidation. The 1775-cm" band edge responds more slowly and may be interpreted in a simplified manner as the formation of organic acids. The 7800-cm-' band edge grows even more slowly and probably indicates t he formation of acid anhydrides or carbonate esters (1 2,13).

The barrel ellipsoid w a s used for the analyses of actual heat-damaged, Oo lM6/3501-6 flexural specimens. This accessory w a s operated in an inverted mode so that t he specimen was simply placed on the opening containing the specimen position focal point. Two analyses were obtained for each specimen, one from each end of the specimen. These specimens were analyzed after exposure t o one of the temperature/time conditions tha t were described in Sect. 2.2, with the exception of the specimens exposed to 399OC (750OF) for either 5 min or 10 rnin. Figures 8-1 1 give the results of these analyses in terms of the flexural strength plotted a s a function of reduced-spectral parameters. The data correlated well with the exception of t he 343OC (650OF) and 399OC ( 7 5 0 O F ) data. Specimens given these extreme

Page 7: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

conditions did not show as much evidence of carbonyl addition or hydrocarbon loss a s did the specimens oxidized a t lower temperatures. The hydroxyl band edge, however, did correlate well with flexural strength data for all heat treatments (Figure 11). This indicates tha t t he oxidation process probably comprises a number of different reactions tha t involve both the addition of oxygen as various carbonyl species, which generally increase in oxygen content as oxidation progresses, and the removal of hydrocarbons by water formation. The early addition of oxygen as phenones and ketones or aldehydes does not appear to have a marked effect on the mechanical strength properties, but the carbonyl band area makes an excellent indicator tha t the specimen has been oxidized. The good correlation of t he loss of hydroxyl species with the loss of strength is best explained by the fact that t he hydroxyl species is the residual oxygen from the epoxide ring, and its loss may be directly associated with the breaking of the resin cross-links that were formed during the curing reactions. The loss of hydroxyl groups and the formation of amines, along with the observation in Ref. 11 of the formation of propenal in the gas phase, is consistent with the expulsion of the three-carbon chain tha t had contained the original epoxide ring being the main reason for loss of strength.

4. CONCLUSIONS

DRIFT and LPF possess the capability to detect heat damage in lM6/3501-6 laminates and to correlate this damage with the residual mechanical strength properties. These techniques can de tec t molecular changes of the polymer matrix resin that result from elevated temperatures. They can be used for lightweight and field portable inspection devices for rapid, in-service, nondestructive detection and quantitation of heat damage in structures fabricated using lM6/3501-6. The DRIFT and LPF techniques also offer t he potential for detecting and assessing heat damage in other resin systems including bismaleimides, polyimides, polyether ether ketone, etc. These techniques could also be used for cure monitoring, quality control, and material characterization.

5. ACKNOWLEDGEMENTS

We would like to acknowledge the Naval Aviation Depot, Cherry Point, North Carolina, and the Naval Air Systems Command, Washington, DC, for support under Interagency Agreement Number 1822-1 822-A1.

Page 8: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

6. REFERENCES

1. 2. 3. 4.

5.

6.

7. 8. 9.

10. 11. 12. 13.

J. F. Haskins, SAMPE J., 25 (21, 29 (1 989). G. A. Luoma and R. D. Rowland, J. Appl. Polym. Sci., 32, 5777 (1986). T. A. Collings and D. L. Mead, Composites, 19 (11, 61 (1988). C. J. Janke, e t al., Composite Heat Damage, ORNL/ATD-33, Martin Marietta Energy Systems, Inc., Oak Ridge Natl. Lab., May 1990. C. J. Janke, et al., Composite Heat Damage Spectroscopic Analysis, ORNL/ATD-42, Martin Marietta Energy Systems, Inc., Oak Ridge Natl. Lab., September 1990. G. L. Powell, N. R. Smyrl, and N. J. Harrick, Research and Development, 26, 88 (1 984). N. R. Smyrl, E. L. Fuller, and G. L. Powell, Appl. Spectrosc., 37, 38, (1983). G. L. Powell, e t al., Research and Development, 31, 58 (1 989). G. L. Powell, e t ai., Proceedings of the Pittsburgh Conference in New York, Applied Spectroscopy Society, pp. 1 199, (1 990). K. C. Cole, e t al., Appl. Spectrosc., 42, 761 (1988). M. A. Grayson and C. J. Wolf, J. Polym. Sci.: Polym. Chem. ed., 22, 1897 (1984). E. L. Fuller, Jr. and N. R. Smyrl, Fuel, 64, 1143 (1 985). E. 1. Fuller, Jr., et ai., Development of Diffuse Reflectance Infrared Spectroscopy Techniques and Theory for Surface Analysis, Y/DK-396, Martin Marietta Energy System, Inc., Y-12 Plant, 1985.

Page 9: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

7. BIOGRAPHIES

Christopher J. Janke received a B.S. in chemistry from Appalachian State University, Boone, North Carolina, in 1983 and an M.S. in chemistry from the University of Tennessee, Knoxville, in 1986. He was employed by Milliken and Company from 1986 t o 1988 as an engineer where he was responsible for the development of reinforcement fabrics and adhesive systems from the initial sample stages through production scale-up. Since 1988, he has been employed by Martin Marietta Energy Systems, Inc. at the Oak Ridge National Laboratory as a Development Engineer in the Engineering Technology Division. His main interests are: heat damage effects on the residual properties of composite materials; nondestructive detection techniques for composite heat damage; electron beam curing and microwave curing of polymer matrix composites; smart composite structures; and materials selection and materials replacement studies for engineered structures.

Halen E. Phiipot received a B.S. in mechanical engineering from Tennessee Technical University, Cookeville, in 1979. From 1979 t o 1985, he was employed by Union Carbide at the Oak Ridge, Tennessee, K-25 Site as an engineer in support of the gas centrifuge uranium enrichment program. From 1985 t o 1987 he was employed at the same site by Martin Marietta Energy Systems, Inc., where he was responsible for performing compressor aerodynamic work in support of the gaseous diffusion uranium enrichment program. He is currently a manager for the Mechanical Testing Laboratory a t the K-25 Site.

Dr. G. Louis Powell received a B.S. in chemistry a t Presbyterian College, Clinton, South Carolina, in 1963 and a Ph.D. in chemistry from the University of North Carolina at Chapel Hill in 1967. Since then, he has been on the Development Staff as a physical chemist for Martin Marietta Energy Systems, Inc., a t the Y-I 2 Plant in Oak Ridge, Tennessee. His main interests are kinetic spectroscopy, thermodynamics and transport phenomena and isotope effects in metal-hydrogen systems, surface analysis, and corrosion and hydrogen embrittlement in metals, metal alloys and polymeric materials. His specializations include the application of FTlR spectroscopy t o the surface analysis and inspection of materials in controlled and manufacturing environments and the correlation of chemical analyses with mechanical and metallurgical properties.

Dr. Eric A. Wachter received a B.S. in chemistry from Indiana University in 1984 and a Ph.D. in physical-analytical chemistry from the University of Wisconsin in 1988. He has been a member of the research staff of the Health and Safety Research Division a t the Oak Ridge National Laboratory since 1988 working in the area of applied spectroscopy. His main research interests include materials characterization, surface chemistry, and remote pollutant and process monitoring.

Page 10: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

90

80

70

60 Lu z n

n g 50

P O

I

0 ' 30

20

10

0

WElPRECONDlTlONlNG I DRY PRECONDITIONING cn

TIME, min 60 60 5 3060 5 3060 5 1030 lEMP.."C(OF) B(M) 177(350) 232(450) 288(550) 343(650) 399(750) @WET AND DRY STRENGM AVERAGES ARE SlATlSnCALLY SlGNlFlCANnY O l ~ A T l H E 9 5 % CONFIDENCE LEML 8

' Fig. 1. shore-D hardness vs temperatureltime.

.L c c

70

60 6 W

0

10 -

8 - a? cn' ' I- 6 -

9

v)

S

4 -

2 -

= WET PRECONDmONlNG DRY PRECONDmONlNG

0 TIME.min 60 60 5 3060 5 3060

TEMP., O C ( " F ) 177(350) 232(450) 288(550) 343(650) Fig. 2, Percent weight loss vs temperaturdtime.

Page 11: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

Table 1. Room temperature strength Owpa) test results. - EKPO- conditions: Compression SBS SBS F I e d Flexural

oc(°F)ld Drg Wet Dry Wet Dry Wet Dry Wet Dry Wet

templtime (0") (0") (90") (0") (90")

23(73)/NA 177(350)/60 232(450)/60 288(550)/5 288(550)/30 288(550)/60 343(650)/5 343(650)/30 343(6!50)/60 399(750)/5 399(750)/10 399(750)/30

308.9 318.5 284.8 302.0 281.3 221.3 282.0 128.9 31.0

129.6 83.4 28.3

301.3 295.8 275.1 271.0 264.8 183.4 233.0 142.0 25.5

152.4 73.8 19.3

42.1 42.7 60.0 68.9 42.1 40.7 64.8 61.4 39.3 40.0 60.7 51.0 37.9 33.1 61.4 44.8 28.3 22.8 31.7 29.6 22.1 145 145 16.5 10.3 7.6 17.2 23.4 2.1 2.1 0.0 0.0 23i4 165 2.1 2.1 0.0 0.0 18.6 17.2 0.0 0.0 0.0 0.0 21.4 20.7 0.0 0.0 0.0 0.0 20.0 21.4 0.0 0.0 0.0 0.0 145 26.9

3565 3675 364.0 350.3 326.8 3172 293.7 295.1 212.4 218.6 137.2 159.3 752 41.4

599.8 645.3 548.8 397.1 197.9 120.7 53.8 21.4 152 13.1 152 255

6185 601.9 495.7 m.7 240.6 1462 68.9 21.4 172 13.1 165 29.0

Table 2 Percent strength retention test results. ~ ~~~

SBS SBS Flexural Flexural Expo-

conditions: Compression

OC(OF)/min Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet

templtime . (0") (0") (90") (0") (90")

23(73)/NA 177(350)/60 232(450)/60 28 8(550)/5 288(550)/30 28 8 (550)/60 343(650)/5 343(650)/30 343(650)/60 399(750)/5 399(750)/10

97 100 100 100 92 100 98 100 93 100 100 98 100 95 100 89 100 95 100 97 90 91 94 93 94 74 90 86 85 80 95 90 91 7 7 9 4 6 5 81 80 62 65 08 88 67 52 48 43 58 60 31 39 70 61 53 34 22 24 38 43 19 24 89 7 7 2 5 18 26 34 21 11 8 11

. 41 47 5 4 0 0 6 4 3 3 10 8 5 5 0 0 5 5 2 3 41 51 0 0 0 0 6 6 2 2 26 24 0 0 0 0 6 6 2 3 9 6 0 0 0 0 4 7 4 5

Tabie 3. CoefEaent of variation (96) test results. -

conditions: Compression SBS SBS Fiexural Flexural Expo-

templtime (0") (0") (mol (0") (90")

oC(OF')/mh Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet

7 6 4 3 9 8 3 5 11 7 9 5 4 4 15 10 2 1 9 2 6 4 4 2 13 . 5 7 5 6 13 5 8 7 12 10 6 10 1 15 18 9 8 16 20 15 17 12 28 30 20

16 7 22 30 17 20 24 39 28 28 8 10 85 90 33 25 47 33 33 43

28 18 31 14 0 0 2 3 17 21 15 40 24 19 10 0 0 13 15 16 34 27 18 0 0 0 0 44 39 24 18 37 37 0 0 0 0 18 13 18 9 35 42 0 0 0 0 17 20 16 23

. ...

Page 12: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

3675 3675 363.8 356.2 350.2 350.2 326.8 317.2 317.2

295.1 295.1 294.0 218.7 218.7 2125 159.1 159.1 137.0

75.7 41.7 41.7 26.8 23.1 18.8 17.0 17.0 162 162 14.6

637 725 333 681

1900 lo60 %l 152

132 127 126 112 124 114 111 107 110

106 105 114 108 103 108 105 lo5 105 lo5 lo4

277

679 790 362 727

2100 1170 270 166 317

139 133 130 115 128 116 113 '

109 113

109 1 1 116 108 107 109 106 I05 106 106 I05

693 809 381 724

2170 1210 286 171 346

148 139 136 117 131 120 113 111 114

110 111 118 110 107 110 107 106 106 108 107

669 778 371 701

2090 1180 302 177 365

149 142 140 119 133 120 115 112 116

111 112 118 110 106 110

. 108 108 108 108 108

680 785 383 711

2130 1210 323 186 398

162 151 148 124 142 126 121 116 121

117 117 124 115 112 117 113 112 112 113 113

E E (I)- (I) UI z Y

f

= W PRECONDITIONING ES DRY PRECONDITIONING

6

5

4

3

2

1

0 TIME, min 60 60 5 3060 5 3060 5 1030

TEMP., "C("F) 23(73) 177(350) 232(450) 288(550) 343(650) 399(750) C:n 2 Mav:m..m 4h:nL-nra vc tomnnmtiirP/timP-

- _ - - - I__

-- --_

Page 13: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

I

1

0.9

0.8

0.7

$0.6

0.5

0.4

0.3

0.2

0.1

0

I d

4000 3500 3000 2500 2000 1500 1000 5 Wavenum bers

Fig. 4. DRlETspectrmn ofa specimen obtaincd b y e the bamd ellipsoid impeciionaazssmy

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0 2000 1900 1800 1700 1600

Wavenumbers 1500 1400

Fig. 5. DRIFT spectra showing the cadxmyl region of the qectnm.

Page 14: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

2-0

P'O . . . . . . .& . . . . . . . . . . . . .. . . .. . .. . . . ...A,. ,

9'0

8'0

0

Page 15: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

z P) C L

c

3i e 0

100 d 2 0 ~

0- As-cured 0- 17;rC (350'Fl- 60 min. A - 232'C C450'Fl - 60 min. o= 288'C (550'Fl - 5 min. m- 288'C C550'Fl - 30 min. T- 288'C C550'Fl - 60 min. A - 343'C (650'n - 5 min. e- 343'C C650'Fl - 30 min. v- 343'C (650'Fl - 60 min. A - 399'C C750'Fl- 30 min.

P

00

Ab A A A A

J 1 I I

0.0 0 2 0.4 0.6 I775 an'' Band Edge Intensify

0.8 01

Fig. 8. Flexural strength data vs the 1775 mi1 band edge intensity.

400-

0- As-cured 0- W C C350'R - 60 min. A = 232X (450'D - 60 min. 0- 288'C C550'R - 5 min. m= 288'C C550'Fl- 30 min. v- 288'C C550'Fl - 60 min. A - 343% (650'Fl - 5 min. e- 343'C (650'Fl - 30 min. t- 343'C (650'Fl- 60 min. A - 399% (750'Fl - 30 min.

A Y A A

A

, $ I 1 ~ ' ' " ' I " " " 4 5 2 3 1

Carbonyl Band Integral (Normalized)

Fig. 9. F I d strength data vs the normalized carbox@ band integral

Page 16: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

7 400

300

i s c m s 200 L ii3

* 100

0 L 1

2 A A

m

0= As-cured O= W C (350'Fl - 60 min. A = 232'C (450'D - 60 min. o= 288'C (550'D - 5 mln. m= 288% (550'D - 30 mln. t= 288'C (s50'n- 60 min. A = 343% (650'D - 5 min. 41 343% (650'F) - 30 min. v= 343'C (650.D - 60 mtn. A- 399'C (750'D - 3 0 min. 1 V - A w q 0 to

I " " 2.5 0 ~ ' " ' I " " ~ " " ~ " ' 2 0 0.5 1 l.5

Hydrocarbon Band Integral (Normalized)

Fig. 10. F l d strength data vs the normalized hydroabn band integral

I 1 400 t 1 1 I t : 0= As-cured : o= W C (350'0 - 60 min.

om 288% (550'D - 5 min. 300 7 m= 288.C C550'n- 30 min.

: t m 288% (550'D- 60 mtn. - z : A = 343'C C650'D - 5 mtn.

rn e= 343% C650'Fl- 30 m h : t= 343'C (650'Fl - 60 min.

i 3 0 5 L 3i

200 1 A = 399'C (750'0 - 30 min.

0 120 5 E

- F 50 1 Am 232% (450'n - 60 mfn.

e v) : e

5 4 0 * c

: 10 E

,- 0 1

Fig. 11. F I d strength data vs the 3520 mi1 hiydfoxyl band edge intensity.

Page 17: COMPOSITE HEAT DAMAGE ASSESSMENT/67531/metadc671415/...COMPOSITE HEAT DAMAGE ASSESSMENT Christopher J. Janke' and Eric A. Wachter Oak Ridge National Laboratory. Oak Ridge, Tennessee

DISCLAIMER

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or use- fulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any spe- cific commercial product, process, or service by trade name, trademark, manufac- turer, or otherwise does not necessarily constitute or imply its endorsement, recorn- mendattion. or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

- .


Recommended