EXPERIMENTAL DETERMINATION OF KEY PARAMETERS FOR MODELLING THE TENSILE AND COMPRESSIVE FATIGUE BEHAVIOUR OF NOTCHED GRP LAMINATES Bill Broughton, Mike Gower, Maria Lodeiro, Gordon Pilkington and Richard M. Shaw 5th International Conference on Composites Testing and Model Simulation, EPFL, Lausanne, 2011 Content Introduction Test Programme Constant Amplitude Cyclic Fatigue Tension-Tension Compression-Compression Tension-Compression Multiple Step T-T Block Loading Concluding Remarks Introduction Aims and Rationale: Ensuring the long-term structural integrity and safety of composite structures throughout in-service lifetime Develop and validate fatigue test methods for composites Identify and evaluate key parameters for modelling tensile and compressive fatigue behaviour of FRPs Test Programme E-glass/913 (Hexcel Composites) Quasi-isotropic (QI) lay-up [45°/0°/-45°/90°]4S Open-hole tension (OHT) Open-hole compression (OHC) Quasi-static loading Constant amplitude cyclic loading (f = 5 Hz) Tension-tension (OHT): R = 0.1 and 0.5 Compression-compression (OHC): R = 10 Tension-compression (OHC): R = -1 Stress: 80, 70, 55, 40 and 25% UTS/UCS Strain measurement DIC, FBGs, strain gauges, extensometry Open-Hole (Notched) Tension Tension-Tension Fatigue Unnotched Exx (GPa): 21.9 ± 0.4, xx: 0.31 ± 0.01 Strength (MPa): 484 ± 18 Open-Hole Tension (OHT) Exx (GPa): 20.6 ± 0.3 Strength (MPa): 347 ± 5 Embedded Fibre Bragg Gratings FBG 50 150 50 =6 x 18 7.5 End-tab 50 36 50 y 125 Strain Gauges and FBGs Multiple-Plexed FBGs Centre positions of FBG’s 50mm 50mm 3 gratings at different centre wavelengths ~1540, 1550 and 1560nm 50 mm centre-to-centre spacing We would need a total of 6 fibres. Depending on the success of the FBG work, t of the project will be written up as either an NPL measurement note or as a sectio report detailing all the outcomes of the project tasks plus a possible journal/confe Although you are not permitted to sell the fibres, any work published would a City University’s help with the provision of the fibres. 38 mm end-to-end spacing 12 mm grating length 215 mm buffer free region centred on middle grating Length – 660 mm Core – glass, 9 m diameter Coating - 125 m diameter (acrylate re-coated) Cladding – glass, 125 m diameter Quasi-Static Strain Measurements 250 Applied Stress (MPa) 200 150 100 Strain Gauge Extensometer Fibre Bragg Grating Digital Image Correlation 50 0 0 5000 10000 Strain () 15000 20000 Quasi-Static Loading DIC xx Strain Maps 40.3 kN LaVision® DIC System Single megapixel (1280 x 1024 pixel) video camera Image recording frequency: 1 Hz LaVision® Strainmaster software Data capture/analysis 42.5 kN Quasi-Static Loading xx Strain Across Specimen Mid-length 0 kN 3 5.40 kN 11.80 kN Increasing Load 2.5 17.67 kN 20.17 kN 2 26.28 kN xx (%) 32.03 kN 35.70 kN 1.5 1 0.5 0 0 3 6 9 12 15 18 21 24 27 -0.5 Distance across specimen width (mm) 30 33 36 T-T Cyclic Fatigue Fatigue Damage (55% UTS) 5,000 20,000 30,000 Nf = 27,979 ± 9,142 cycles T-T Cyclic Fatigue Pulse Thermography (55% UTS) 0 5,000 10,000 15,000 20,000 25,000 30,000 T-T Cyclic Fatigue Normalised Stress-Cycle (S-N) Curves Normalised Stress (MAX/UTS) 1.0 0.8 MAX/UTS = 1 - 0.09 logNf 0.6 0.4 MAX/UTS = 1 - 0.10 logNf 0.2 R = 0.1 R = 0.5 0.0 0 10 1 10 2 10 3 10 4 10 5 10 Number of Cycles to Failure (Nf) 6 10 7 10 Normalised Residual Stiffness (E/E0) T-T Cyclic Fatigue Residual Stiffness (40 % UTS) I 1.0 II 0.8 III 0.6 0.4 0.2 Extensometer Fibre Bragg Grating 0.0 0 200000 400000 600000 Number of Cycles (N) 800000 1.0 1.0 0.8 0.8 Normalised Stiffness (E/E0) Normalised Stiffness (E/E 0) T-T Cyclic Fatigue Residual Stiffness 0.6 70% UTS 0.4 0.2 0.6 40% UTS 0.4 0.2 0.0 0.0 0 200 400 600 Number of Cycles (N) 800 1000 0 1x10 5 2x10 5 3x10 5 4x10 5 5x10 5 Number of Cycles (N) 6x10 5 7x10 5 8x10 5 T-T Cyclic Fatigue Residual Strength (55% UTS) 350 255 ± 6 MPa Residual Strength (MPa) 300 250 200 150 100 50 Experimental Boltzmann distribution 0 0 5000 10000 15000 20000 25000 30000 Number of Cycles (N) Monotonic decrease in stiffness is not accompanied by decrease in residual strength during fatigue life T-T Cyclic Fatigue xx Strain Distribution vs. Loading Cycles Fatigue: 44% UTS (f = 5 Hz, R = 0.1) Static load for measurements: 20 kN T-T Cyclic Fatigue Strain Distributions vs. Loading Cycles yy xy xx (%) T-T Cyclic Fatigue xx Strain Across Specimen Mid-length 10 9 8 7 6 5 4 3 2 1 0 -1 10,000 cycles 30,000 cycles 40,000 cycles 50,000 cycles 60,000 cycles 70,000 cycles 80,000 cycles Increasing Cycles 0 6 12 18 24 30 Distance across specimen width (mm) Fatigue: 44% UTS (f = 5 Hz, R = 0.1) Static load for measurements: 20 kN 36 T-T Cyclic Fatigue Maximum xx Strain at Hole Perimeter 8 Maximum Strain xx (%) 7 -5 xx = 0.97 + 8.35 x 10 N 6 5 4 3 2 Experimental Linear fit 1 0 0.0 4 2.0x10 4 4.0x10 4 6.0x10 Number of Cycles (N) 4 8.0x10 T-T Cyclic Fatigue Global xx Strain Values Stress (% UTS) Stress (MPa) Initial Strain (%) Final Strain (%) Nf (cycles) mean max mean max mean max R = 0.1 40 55 70 76.2 104.8 133.4 138.6 190.6 242.5 0.363 0.534 0.698 0.658 0.949 1.375 0.663 0.835 1.081 1.100 1.421 2.117 822804 26976 706 R = 0.5 40 55 70 103.8 142.7 181.7 138.6 190.6 242.5 0.412 0.570 0.763 0.554 0.765 1.017 0.753 1.063 1.331 1.004 1.373 1.775 5993898 62093 2262 f max i max i mean mean max max mean E0 E0 1 R 2 OHT QI Laminate (T-T Cyclic Fatigue) Maximum Failure Strain fmax 3.0 f max 2.5 max of k log10 Nf Ef f Maximum Failure Strain ( max (%)) 3.5 2.0 1.5 1.0 0.5 0.0 0 10 Experimental Linear fit 10 1 10 2 10 3 10 4 10 5 Number of Cycles to Failure (N f) 10 6 10 7 T-T Cyclic Fatigue Hysteretic Heating Effects (40% UTS) 80 o Surface Temperature ( C) 70 60 50 40 30 20 0 200000 400000 600000 Number of Cycles (N) 800000 T-T Cyclic Fatigue Maximum Surface Temperature (ºC) Test Condition (% UTS) Initial Final Ultimate failure R = 0.1 40 55 55 (1 Hz) 70 33 46 30 23 46 78 34 65 87 104 41 68 R = 0.5 40 25 33 Measured at hole perimeter Frequency is 5 Hz (unless otherwise specified) T-T Cyclic Fatigue (55 %UTS) Normalised Residual Fatigue Stiffness Normalised Residual Stiffness (E/E0) 1.0 0.8 0.6 0.4 0.2 Experimental Linear fit 0.0 20 40 60 80 100 o Surface Temperature ( C) 120 OHT QI Laminate (T-T Cyclic Fatigue) Normalised Residual Fatigue Stiffness Normalised Residual Stiffness (E/E0) 1.0 E 1 AT E0 0.8 0.6 0.4 70% UTS 55% UTS 40% UTS Linear fit 0.2 0.0 20 40 60 80 100 o Surface Temperature ( C) 120 Open-Hole (Notched) Compression Compression-Compression Unnotched SCxx (MPa): 617 ± 19 Open-Hole Compression (OHC) Strength (MPa): 346 ± 54 C-C Cyclic Fatigue Damage/Failure C-C Cyclic Fatigue Normalised S-N Curve Normalised Stress (MAX/UTS) 1.0 MAX/UTS = 1 - 0.07 logNf 0.8 0.6 MAX/UTS = 0.54 + 0.57/(1 + Nf/74.42) 0.31 0.4 Experimental Linear fit Sigmoidal fit 0.2 0.0 0 10 1 10 2 10 3 10 4 10 5 10 Number of Cycles to Failure (Nf) 6 10 7 10 C-C Cyclic Fatigue xx Strain Across Specimen Mid-length 0 6 12 18 24 30 36 0.00 xx (%) -0.50 -1.00 -1.50 Increasing Cycles 10,000 cycles 20,000 cycles 30,000 cycles 40,000 cycles 50,000 cycles 60,000 cycles -2.00 -2.50 Distance across specimen width (mm) Fatigue: 61% UCS (f = 5 Hz, R = 10) Static load for measurements: -25 kN C-C Cyclic Fatigue Maximum xx Strain at Hole Perimeter Maximum Compressive Strain xx (%) 2.0 1.5 1.0 -5 xx = 0.97 + 1.46 x 10 N 0.5 Experimental Linear fit 0.0 4 1x10 4 2x10 4 3x10 4 4x10 Number of Cycles (N) 4 5x10 4 6x10 C-C Cyclic Fatigue Hysteretic Heating Effects (5 Hz) Applied Stress MAX/UTS Surface Temperature (°C) 60% 41 65% 54 70% 59 70%* 45 * Unnotched Open-Hole (Notched) Compression Tension-Compression Open-Hole Compression (OHC) Strength (MPa): 346 ± 54 Requirements Rigid test frame and well aligned grips Max. Bending Strains: < 8% (C) and < 3% (T) T-C Cyclic Fatigue Normalised S-N Curve Normalised Stress (MAX/UTS) 1.0 0.8 0.6 0.4 run out MAX/UTS = 1 - 0.12 logNf 0.2 Experimental Linear fit 0.0 0 10 1 10 2 10 3 10 4 10 5 10 Number of Cycles to Failure (Nf) 6 10 7 10 T-C Cyclic Fatigue xx Strain Across Specimen Mid-length 5.00 10,000 cycles xx (%) 4.50 20,000 cycles 4.00 30,000 cycles 3.50 40,000 cycles 50,000 cycles 3.00 60,000 cycles Increasing Cycles 2.50 2.00 1.50 1.00 0.50 0.00 0 6 12 18 24 Distance across specimen width (mm) Fatigue: 61% UTS/UCS (f = 5 Hz, R = -1) Static load for measurements: 15 kN 30 36 T-C Cyclic Fatigue Maximum xx Strain at Hole Perimeter Maximum Tensile Strain xx (%) 4 3 -5 xx = 0.41 + 5.37 x 10 N 2 1 Experimental Linear fit 0 4 1x10 4 2x10 4 3x10 4 4x10 Number of Cycles (N) 4 5x10 4 6x10 T-C Cyclic Fatigue Fully Reversed Loading S-N Response Normalised Stress ( max/ULT) 1.0 0.8 0.6 0.4 0.2 0.0 0 10 Experimental Predicted 10 1 10 2 10 3 10 4 10 5 Number of Cycles to Failure (N f) 10 6 10 7 Multiple-Step T-T Block Loading Applied Stress QI E-glass/913 laminate OHT: Tension-tension Ni = 1,000 cycles 40% 25%, 55% 25%, 55% 40% UTS 50% 40% 25% UTS (repeated) Time T-T Block Loading Global xx Strain Values (R = 0.1) Stress (% UTS) 40-2 5 25 40 55-2 5 25 40 55-40 25 40 55-40-25 25 40 55 Stress (MPa) mean 47.6 76.2 47.6 104.8 76.2 104.8 47.6 76.2 104.8 max 86.6 138.6 86.6 190.6 138.6 190.6 86.6 138.6 190.6 Initial Strain (%) Final Strain (%) mean mean 0.230 0.368 0.244 0.538 0.385 0.530 0.238 0.381 0.524 max 0.418 0.669 0.444 0.978 0.700 0.963 0.433 0.693 0.953 0.353 0. 581 0.515 0.940 0.591 0.812 0.303 0.485 0.666 max Nf (cycles) 0.642 1.065 1980585 990000 990584 0.684 1.548 74796 37000 37796 1.047 1.478 82569 41000 41569 0.681 0.994 1.574 51564 17000 17000 17564 T-T Cyclic Fatigue Global Strain Values f max f max i max i mean max mean E0 E0 max of k log10 Nf Ef mean max 1 R 2 T-T Block Loading (55%40%25% UTS) Surface Temperature 80 55% UTS o Surface Temperature ( C) 100 40% UTS 60 25% UTS 40 20 0 0 10000 20000 30000 40000 Number of Cycles (N) 50000 60000 Concluding Remarks Alignment and rigidity of loading chain is critical for compressioncompression and tension-compression tests DIC suitable for monitoring local and global strains Providing critical information on changes in strain distribution around the hole of notched laminates due to damage formation/growth incurred through either increasing load or number of loading cycles Optical fibres (FBGs) suitable for monitoring fatigue performance – superior fatigue performance compared with strain gauges Longitudinal strain and stiffness along with surface temperature – indication of level of remnant life of notched components Possible to estimate fatigue life for fully reversible and block loading conditions from T-T and C-C cyclic data Acknowledgements The work was supported by United Kingdom Department for Business, Innovation and Skills (National Measurement Office), as part of the Materials 2007 Programme. The authors would also like to thank: Hexcel Composites Limited Dr F Surre and Dr T Venugopalan - City University London