Composite Thermal Damage Measurement with Handheld FTIR. April 9, 2013 Brian D. Flinn, Ashley Tracey, and Tucker Howie University of Washington

Transcription

1 Composite Thermal Damage Measurement with Handheld FTIR April 9, 2013 Brian D. Flinn, Ashley Tracey, and Tucker Howie University of Washington

2 Composite Thermal Damage Measurement with Handheld FTIR Motivation and Key Issues Damage detection in composites requires different techniques than metals Incipient thermal damage occurs below traditional NDE detection limits Objective Determine if handheld FTIR can detect thermal damage and guide repair Approach Characterize panels with controlled thermal damage and perform repair based on FTIR inspection 2

3 FAA Sponsored Project Information Principal Investigators & Researchers Brian Flinn (PI) Ashley Tracey (PhD student, UW-MSE) Tucker Howie (PhD student, UW-MSE FAA Technical Monitor David Galella (year 3) Paul Swindell (year 1 & 2) Industry Participation The Boeing Company (Paul Shelley, Paul Vahey) Sandia National Lab (Dennis Roach) Agilent (formerly A2 Technologies) 3

4 Background Continuation of existing project (year 3 of 3) Years 1 and 2 (A2 Technologies, Boeing and U of DE) Characterization of homogeneous thermal damage Ultrasound Short beam shear (SBS) Microscopy Handheld FTIR (ExoScan) Calibration curve for FTIR detection of thermal damage (SBS data) Mapped surface of localized thermal damage Year 3 (UW and Boeing) 3-D characterization of localized thermal damage Include contact angle and fluorescence spectroscopy FTIR guided repair of thermal damage Test repair 4

5 Thermal Damage vs. Detection Method SBS, ultrasound, and microscopic analysis of composites with thermal damage Properties degrade before detection possible need method to detect incipient thermal damage (ITD) Short Beam Shear Strength Retention vs. Temp./Time Epoxy 1 No cracks visible in micrographs Onset of crack development visible in micrographs, Damage becomes visible in C-Scans 5

6 Experimental Overview Investigate ITD of composites with various inspection techniques Characterize composite samples and panels with controlled thermal damage using various methods: Contact angle (CA) Fluorescence FTIR Can results be related to SBS values and detect thermal damage? 6

7 Materials and Process Toray 3900/T800 composites with various levels of thermal damage Provided from Year 1 & 2 research SBS samples thermally exposed in air Panels with localized thermal damage in vacuum Characterize toolside (resin rich) and sanded (resin poor) surfaces Sand surfaces with random orbital sander using 120 grit 3M Al 2 O 3 sanding pads Measurement techniques: CA, fluorescence, FTIR 7

8 Materials and Process Contact Angle Measure CAs of 1 μl sessile drops from side view 5 drops (10 CAs) per fluid Fluids: DI water, diiodomethane (DIM) Measure at 0 degrees with respect to fiber orientation Drop application: dispense drop, raise surface Camera Fiber Direction θ Side-view of drop as viewed from goniometer camera 8

9 Materials and Process Fluorescence 470 nm LED Filter Detector Probe sample Excitation Light Fluorescent Emission Sample Sample absorbs excitation light and emits light at longer wavelength than the absorbed light (fluorescence). Measure changes in intensity and wavelength at max intensity (λ MAX ) of fluorescence emission 9

10 Materials and Process FTIR Mid-IR data region: 4000 cm -1 to 650 cm -1 Diffuse reflectance sampling interface Data collection: 120 coadded scans with 8 cm -1 resolution for background and specimen To detector Mirror ExoScan FTIR IR beam from interferometer θ i = 0 θ r Substrate Reflected Beam An infrared beam path for diffuse reflectance 10

11 Year 3 Results: CA Measurements on SBS Samples Average CA (deg) F 475 F DI water, sanded DI water, toolside DI water, sanded DI water, toolside DIM, sanded DIM, toolside 120 DIM, sanded DIM, toolside 100 Average CA (deg) Time (min) Time (min) increasing thermal damage increasing thermal damage CA on sanded surface lower than toolside surface No significant correlation between SBS values and CA measurement 415, 445, 475, 505 F 11

12 Year 3 Results: Fluorescence of SBS Samples λ MAX red-shifts and intensity decreases with increasing exposure λ MAX does not monotonically relate to SBS retention No fluorescence measurable on sanded surface PL Intensity (A.U.) Wavelength (nm) As-cured 445 ºF for 30 min 445 ºF for 77 min 445 ºF for 198 min 445 ºF for 510 min SBS Strength Retetion (%) λ MAX As-cured 415 ºF 445 ºF 475 ºF 505 ºF 12

13 Year 3 Results: FTIR Verification FTIR measurements on resin rich surface of SBS consistent with previous results Oxidation peaks increase with damage Year 1 & 2 Year High Damage Absorbance No Damage Wavenumber (cm -1 )

14 Year 3 Results: FTIR on Sanded Surfaces Damage is not as clear as on toolside surface Oxidation removed by sanding Need multivariate analysis to determine differences in spectra and correlate to SBS data 0.35 Full Spectra 0.35 Oxidation Region ºF for 48 min 475 ºF for 123 min Absorbance ºF for 318 min Absorbance Wavenumber (cm -1 ) Wavenumber (cm -1 )

15 Year 3 Results: FTIR Orientation Signal varies based on sample orientation FTIR needs to be rotated during repair to match fiber orientation 0 º orientation Fiber Direction 90 º orientation Absorbance º 90 º Fiber Direction Top down schematic of scarfed surface showing how fiber orientation changes at each layer Wavenumber (cm -1 )

16 Year 1 & 2 Results: Localized Damage Hot spots created 3 temperatures 440, 465, 490 F 2 panels each 16

17 Year 1 & 2 Results: Map of Localized Damage FTIR Map of Surface Damage Blue is low damage Brown is high damage BMS F / 465F / 490F Y (inches) X (inches) Low (440 ºF for 1 hr) Medium (465 ºF for 1 hr) High (490 ºF for 1 hr) 17

18 Year 3 Results: Panel Mapping Preliminary measurements performed FTIR spectra different than resin rich surface of SBS samples Panels heated in vacuum less oxidation Changes in oxidation peaks at 1720 cm -1 still observed Oxidation peak decreases as distance from center increases y x -½ 0 ½ Absorbance wavenumber (cm-1) " 1" 2" 3" 18

19 Summary Preliminary results generated No clear correlation of ITD with contact angle No clear correlation of ITD with fluorescence Oxidation detected on resin rich surfaces Resin poor surfaces require advanced analysis techniques Ready to proceed to next stage Multivariate analysis of resin poor surfaces 3-D panel mapping 19

20 Future Work Apply multivariate analysis Surface map thermal damage (all panels) 1st set of panels- mechanical testing (SBS, Tg) 2 nd set of panels scarf repair guided by FTIR Map damage ply by ply during scarfing FTIR Correlate FTIR measurements to mechanical tests to guide repair Bonded repair followed by NDE Mechanical testing of repaired panel 20

21 Looking Forward Benefit to Aviation Improved damage detection Greater confidence in repairs Future needs Application to other composite systems Other applications of handheld FTIR Chemical damage Surface prep for bonding 21

22 End of Presentation. Thank you. 22