Monitoring of a full-scale wing fatigue test - RNLAF F-16 aircraft, 4200 FH -

Transcription

1 Monitoring of a full-scale wing fatigue test - RNLAF F-16 aircraft, 4200 FH - Jaap Heida and Jason Hwang Aerospace Vehicles Division Nationaal Lucht- en Ruimtevaartlaboratorium National Aerospace Laboratory NLR

2 National Aerospace Laboratory NLR Amsterdam Flevoland Staff of 650 Turnover ~ 75 MEuro/yr 2

3 Contents Wing fatigue test Load monitoring with optical FBG s Structural health monitoring (SHM) Comparative vacuum monitoring (CVM) Acoustic emission (AE) Conclusions 3

4 Objectives wing fatigue test Determine whether the wing contains damage not accounted for in the DADTA of LM Generate data for assessment of current inspection programme Crack growth curves for critical locations Establish the most likely failure scenario, including an estimate of the technical end of life Goal for fatigue test: cover two design lifetimes (16000 FH) Decide on purchase of new F-16 wing sets, or planning of modification programme Evaluate load monitoring and SHM techniques 4

5 - lay-out of 23 hydraulic actuators - Load spectrum derived from measured flight profiles of RNLAF fleet (blocks of 500 FH or 412 flights) Bovenaanzicht Echter: Vijzels grijpen op onderkant vleugel aan!! Actuator 1-10: aerodynamic and inertia loads on wing surface (via bonded rubber load pads) 18 Actuator 11-16: loads from stores and launchers Actuator 17-23: LEF (5x) and flaperon loads (2x) Only vertical loads included BL BL 41.5 Top View However: Actuators act on lower wing skin z x y FS 294 FS 309 FS 325 FS 341 FS 357 FS 374 5

6 6

7 Maximum/minimum load +9g/-3g, deflection at wing tip ~70 cm 7

8 Flight hours F-16 wing fatigue test - test progress - Fatigue test duration ~ ½ year, no failure of wing FH 1st crack visually detected 1st crack NDI detected 8

9 Contents Objectives wing fatigue test Load monitoring with optical FBG s Objective: evaluate optical FBG s as load monitoring technique over a long-term fatigue loading environment Structural health monitoring Conclusions 9

10 Fibre optic Bragg grating sensor Core Cladding Coating λ B λ B = 2n.Λ 10

11 - FBG load monitoring - Two FBG interrogator systems Fiberpro - IFIS100 (Korea) Wavelength range nm Total of 3 channels, plus reference channel Maximum sampling rate: 200 Hz Technobis - Deminsys Ultra (Netherlands) Wavelength range nm Total of 4 channels Maximum sampling rate: 20 khz 11

12 - FBG load monitoring on wing upper skin - Purpose: Compare FBG with conventional SG Load monitoring on spots with limited access Endurance test of the FBG sensors Fiberpro 2 fibres 7+8 FBG s Technobis 1 fibre 4 FBG s 3 optical fibres with 19 FBG s polyamide cladding, M-bond 200 or X60 adhesive 12

13 - FBG load monitoring on wing upper skin - 13

14 Strain [µε] Strain [µε] F-16 wing fatigue test - FBG load monitoring during Strain Surveys - Monitoring of different load cases during Strain Surveys Stepwise increase and subsequent decrease of static loading Fiberpro interrogator: linear correlation between the FBG s (-) and strain gauges (-) Strain transfer factor (STF): compensating for strain transfer loss between specimen and glass core SG FBG SG FBG Time [s] Load [%] 14

15 - FBG load monitoring during Strain Surveys - Downward bending SG x FBG corrected with STF STF = 1.31, independent of strain level and loading condition 15

16 - FBG findings - Strain measurements, after STF correction Average absolute error 1.3% 97% of all FBG strain measurements within ± 5% of the SG measurements Overall conclusions Application of FBG sensors for load monitoring is feasible FBG s can withstand a long term fatigue load environment 16

17 Contents Objectives wing fatigue test Load monitoring with optical FBG s Structural health monitoring (SHM) Comparative vacuum monitoring (CVM) Acoustic emission (AE) Conclusions oppervlaktescheur surface crack 17

18 Comparative vacuum monitoring (CVM) Vacuum technique developed by SMS Ltd. (Australia) cross-section dwarsdoorsnede top bovenaanzicht view sensor vliegtuigcomponent to naar SIM8 oppervlaktescheur surface crack polymer sensor self-adhesive 18

19 p [Pa] CVM reading delta P [Pa] Comparative vacuum monitoring (CVM) p Kvac SIM CVM lab kit with 0 - reference vacuum source (Kvac-5) - sensitive flow meter (SIM-8) CVM recording Pa 11:37:48 11:37:57 11:38:05 11:38:14 11:38:23 1 Time Time [hrs:min:sec] 19

20 Comparative vacuum monitoring (CVM) - measurements on FSW panels - FSW panel with residual stresses Central hole with starter notch CVM sensor on blade stiffener Fatigue test, R = 0.1, Fmax = 80 kn threshold 100 Pa 20

21 - CVM application on lower wing attach fittings - Periodic monitoring with 9 CVM intercept sensors 21

22 - CVM application on lower wing attach fittings - CVM sensor CVM sensor FWD OUTBD 22

23 - CVM findings on lower wing attach fittings - Good performance of the CVM sensors (no bonding problems) No fatigue cracks detected with CVM, but: no cracks occurred at the locations under CVM monitoring No false calls during the complete fatigue test 23

24 Contents Objectives wing fatigue test Load monitoring with optical FBG s Structural health monitoring (SHM) Comparative vacuum monitoring (CVM) Acoustic emission (AE) Conclusions 24

25 Acoustic emission (AE) AE: generation of high-frequency transient elastic waves by the rapid release of strain energy from a localised source within a material under stress AE are generated when defects initiate or grow 25

26 Acoustic emission (AE) AE monitoring: PE sensors in the passive mode Real time and on-line detection of defect initiation and defect growth Registration of e.g. time histories, amplitude histogram, correlation plots F-16 wing test 16-channel SAMOS 24 system of PAC 16 resonant PE sensors (150 khz) Continuous monitoring of 5 critical locations on the lower and upper skin AE parameters 26

27 - AE application on lower wing skin WAF 16W262 AE group 1 LEF #2 AE group 2 Y X 27

28 - AE application on lower wing attach fitting FS OUTBD FWD 28

29 - AE application on upper wing skin -. Wing tip rib AE group 3 5 Cut-out #5 AE group 4 4 Cut-out #6 AE group 5 3 Y 29 X

30 - AE application on upper wing skin cut-out #6 - FWD OUTBD 30

31 - calibration with lead-pencil breaks - OUTBD AFT Event location plot Requirement for localisation: 3 Hits/Event M 1 2 AE group 1 WAF 16W262 31

32 - AE observations first phase of test - Calibration of sensors O.K. (lead-pencil breaks) Good performance of sensors (after initial bonding problems) But: High AE activity combined with high amplitude of the hits External noise (e.g. from hydraulic actuators) Internal noise (e.g. from surface rubbing at fastener locations) 32

33 - AE event location plots - AE groups 3-5, test file FAT2-8, 24 hrs (flights ), 3 Hits/Event AE group 3 AE group 4 AE group 5 WAF 16W262 Cut-out #5 Cut-out # events events events 33

34 - change in AE test setup - Rearrangement of AE sensor groups Sensors of group 3 (wing tip rib) divided over the 3 groups with 3 sensors 4 AE groups with 4 sensors each Event Location Setup: 4 Hits/Event Graphical filters for AE event location plot Correlation factor: 0.85 CF 1.0 Source amplitude: 60 db SA 90 db 34

35 - AE application on upper wing skin - 35

36 - AE application on upper wing skin cut-out #6 - FWD OUTBD 36

37 - AE event location plots, influence of filtering - AE group 5, cut-out #6, test file FAT4-10, 24 hrs (FH ) OUTBD AFT no cracks yet unfiltered data ( events) filtered data (28538 events) Q: Can the location plots yield relevant info about crack initiation and growth? 37

38 - analysis of AE data with confirmed crack history - AE group 5, cut-out #6, first crack detection crack length 2 mm 38

39 - second crack detection- AE group 5, cutout #6 first crack length 3 mm second crack length 2 mm 39

40 - final inspection - AE group 5, cutout #6 first crack length 4 mm third crack length 7 mm second crack length 2 mm fourth crack length 5 mm 40

41 - AE event location plots - AE group 5, cut-out #6, filtered data (CF , SAF db) FH OUTBD FH FWD 41

42 - crack indications- AE group 1 WAF 16W262 42

43 - final inspection - AE group 1, WAF 16W L 1-3L 2-1L 2-2L 1-1L crack lengths: 1-1L 4 mm 1-2L 13 mm 1-3L 17 mm 2-1L 10 mm 2-2L 14 mm OUTBD FWD 43

44 - AE event location plots - AE group 1, WAF 16W262, filtered data (CF , SA db) FH OUTBD FH FWD 44

45 - AE findings - Good performance of AE sensors througout the test High AE activity drastic filtering of AE data necessary Some trends observed but the AE data could not be reliably related to initiation/growth of fatigue cracks in the areas monitored Most AE data probably non-relevant and caused by e.g. frictional noise from fastener locations and other surface rubbing areas 45

46 Conclusions Fatigue test successful, no purchase of new wings FBG load monitoring: Linear correlation between the FBG s and conventional strain gauges was obtained using a specific strain transfer factor (STF) CVM monitoring: No cracks detected (but no cracks occurred at the locations under periodic monitoring). Furthermore, no false calls during the complete test AE monitoring: Registration of a lot of AE activity (also after drastic filtering of the AE data) but the AE data could not be reliably related to the initiation and growth of fatigue cracks in the areas monitored 46

47 Questions? 47

48 - AE measurements, further analysis - AE group 5, cutout #6, influence of filter settings (File FAT 10-11, FH ) CF , SAF db CF , SAF db 48

49 - Linkage of all 24-hr AE data files - Total data file size 21.4 GB for AE measurements with 4 AE groups (104 days) Maximum size for Replay with AE equipment: 2 GB Employment of different data file filters on 24-hrs data files Exclude hits with Energy of V.s Exclude hits with Average Frequency of khz Resulting data file size < 1 GB 49

50 - Events versus fatigue test blocks - AE group 1 WAF AE group 5 cut-out #6 50