Engineering Composite Reinforcements

Transcription

1 Facing up to the Challenges of Natural Fibres as Potential Engineering Composite Reinforcements Jim Thomason Composites September th 2013 Leuven, Belgium

2 Jim why are you so down on Natural Fibre? Natural Fibre Composites Some personal NF History Some of the NFC Challenges NF Anisotropy NF cross section Some Conclusions

3 Natural Fibre some Personal History Half Day Symposium on Natural Fibre Composites

4 ECCM-8, Naples 1998 Prof Verpoest also attended Held standard team meeting

5 ECCM-8 NF Symposium The Natural Fibre Conundrum NF green, cheap, great properties can replace glass fibre Hey Lets make some NF composites and replace glass fibre Hmmm my NF composites are nowhere near what I predicted

6 Natural Fibre some Personal History Potential Advantages of Natural Fibres Potentially low cost Low density (1.45 g/cc vs 2.6 g/cc for GF) Very green image Incinerable - thermally recyclable (with no net increase in CO2 balance) Modulus range exceeds that of E-glass Non-abrasive - low wear of processing equipment No skin irritation problems during handling Identified Disadvantages of Natural Fibres Fibre properties dependent on level of processing (high properties require more fibre processing = cost penalty) Properties dependent on seasonal conditions High levels of water adsorption and poor dimensional stability Low strength compared to E-glass Anisotropic structure - low transverse properties = poor flex & compression performance Composites generally require higher fibre loading resulting in high processing viscosities. Surface treatments and polymer coupling agents required for best composite properties Odour problem after composite processing Potentially Bioactive

7 Natural Fibre some Personal History Owens Corning NFC Project Based on Long Fibre PP Process Technology 12 mm pultruded pellets, 20-50% NF-PP Pilot Plant capability 500kg/day plan for first production plant in India New Sizings developed - some patented for Natural Fibre and Regenerated Cellulose Fibre (Rayon) Multiple Demonstrator Parts Moulded and Tested Huge automotive OEM, Tier 1 and Tier 2 interest

8 Natural Fibre Composite - Demonstrators INDUSTRY LEADER Organizer / Moulder OC, MIG Plastics JCI Part Method Part Wt. (kg) Buick - Door handle Inj. Mould 0.14 pull Jeep Grand Cherokee - Door Inner Inj. Mould 0.91 Mayco Plastics Delphi, Proto Plastics SEG Kunststoff technik Pelzer, Clion Gmbh Chrysler - Air deflector GM Instrument panel retainer Audi A2 - Fender stiffener DCX PT Cruiser - Underbody shield Inj. Mould 0.31 Inj. Mould 2.6kg Inj. Mould 0.45 Extrusion Compress 1.35

9 Typical Properties of PP Based Compounds INDUSTRY LEADER 30% 30% 30% 30% Talc Jute-A Jute-B Glass* Modulus (GPa) Tensile (MPa) Flex Str (MPa) N Izod (J/m) Un Izod (J/m) HDT ( C) Density

10 Relative Performance (%) Comparison PP Composite Performance INDUSTRY LEADER Modulus Strength Notched Unnotched Short Glass Modulus Strength Notched Unnotched Long Jute Relative Performance (%) Modulus Strength Notched Unnotched Modulus Strength Notched Unnotched Fibre Content (% weight) Fibre Content (% weight) Long Glass Long Rayon

11 Comparison Composite Cost/Performance INDUSTRY LEADER $ / kg Composite per MPa Tensile Strength 0.03 $ / kg Composite per GPa Tensile Modulus 0.02 Strength 20% NF 40% NF 20% Talc 40% Talc 20% GF 40%GF $ / kg Natural Fiber Input % NF 40% NF 20% Talc 40% Talc % GF 40%GF $ / kg Natural Fiber Input Modulus $ / kg Composite per ft-lb Unnotched Izod $ / kg Composite per ft-lb Notched Izod Unnotched Impact Jute 20% NF 40% NF 20% Talc 40% Talc 20% GF 40%GF % NF $ / kg Natural 40% NF Fiber Input % GF 40%GF Talc> $ / kg Natural Fiber Input Notched Impact

12 Natural Fibre some Personal History Owens Corning NFC Project Project shelved 2002, Natural Fibre Composites are not Performance-Cost competitive with existing materials

13 Some Philosophy If you know your enemies and know yourself, you can win a hundred battles without a single loss. If you only know yourself, but not your opponent, you may win or may lose. If you know neither yourself nor your enemy, you will always endanger yourself. The Art of War, Sun Tzu

14 Some INDUSTRY LEADER of the Challenges of Working with NF Fibre natural variability Fibre highly anisotropic Low transverse and shear reinforcement performance Fibres mostly non-circular Fibre lumen = composite voids Fibre cross section non-uniform along length Fibre diameter often much larger than man-made fibres High moisture content in fibres at ambient RH processing issues Temperature sensitivity in particular odour issues in processing Many forms of NF not suitable for use in standard industry processes Poor (often negative) performance in composites Composite fibre content measurement? Moisture sensitivity in composite Bio-activity (rotting,fungus, mould) Fibre-matrix interaction - poor Fogging/Emission issues in Automotive applications

15 Some typical fibre properties are shown in the Table below Why Natural Fibre Composites? Sisal Jute Flax Glass Modulus (GPa) Strength (GPa) >1.5 Density Specific Modulus So some natural fibre may have the potential to replace glass fibres??? E = V E + V C η 0 η L f f m E m

16 Typical Specs for Automotive Application INDUSTRY LEADER There are very few applications where only modulus is required!!! A typical automotive spec sheet will need Melt Flow Rate (ISO 1133, ASTM D1238) Glass Fibre Content (ISO 3451/1) Density (ISO 1183, ASTM D792) Tensile Strength (ISO R527, ASTM D638M) Flexural Modulus (ISO 178, ASTM D790M) Shear Modulus (ASTM D4065) Impact Strength, Izod (ISO 180, ASTM D256) Heat Deflection temperature (ISO 73, ASTM D648) Heat Aging Performance (ISO 188, ASTM D573) Flammability (ISO 3795) Fogging (FLTM BO ) Mould Shrinkage (ISO 2577) Coeff. of Linear Thermal Expansion (ASTM D696)

17 Modulus (GPa) Comparison Predicted Composite Modulus For injection moulded long fibre polypropylene Glass Fibre NF 20 GPa(Sisal) NF 40 GPa (Jute) NF 60 GPa (Flax) Remember comparison on weight content (i.e. specific fibre properties) means NO weight saving advantage! Fibre Content (% weight)

18 Actual Modulus Injection Moulded Jute-PP INDUSTRY LEADER Modulus (GPa) ASTM GFPP ASTM Bar Plaque Flow Plaque Cross Flow Glass Jute Fibre Content (% weight)

19 Thermoelastic Anisotropy of Flax and Sisal Fibres INDUSTRY LEADER Goal Quantify anisotropy of Flax & Sisal fibres Full thermoelastic characterisation Measure UD fibre-epoxy laminates E(θ,T), G 12,ν 12, ν 21,α(θ,T) Epoxy matrix E m (T),ν m, α m (T) Laminate fibre volume fraction? Flax & Sisal fibre E 1f (fibre cross section?) Calculate E 1f (T), E 2f (T), G 12f (T), ν 12f (T), α 1f (T), α 2f (T)

20 Composite DMA Results Log Storage Modulus (Pa) Sisal 0 Sisal 10 Sisal 20 Sisal 30 Sisal 50 Sisal 80 Resin Temperature ( C)

21 Anisotropy of Fibre Modulus 100 Modulus (GPa) Flax E1 Sisal E1 Flax G12 Flax E2 Sisal E2 Sisal G Temperature ( C)

22 Thermal Strain (mm/m) Composite Thermal Strain Epoxy Flax 90 Flax 65 Flax 45 Flax 25 Flax Temperature ( C)

23 CLTE mm/m C Fibre Expansion Coefficients Fibre Transverse Sisal Transverse Flax Transverse Sisal Axial Flax Axial Fibre Axial Temperature ( C)

24 Summary INDUSTRY LEADER Thermo-Mechanical Properties NF Glass Flax Sisal Longitudinal Modulus (GPa) Transverse Modulus (GPa) Shear Modulus (GPa) Axial LCTE (µm/m. o C) Transverse LCTE (µm/m. o C)

25 Single Fibre Testing 1 mm Fibre Stress = Load/Area = P/A f (= 4P/πD f 2???)

26 Single Fibre Cross Section Area A f in single fibre testing is almost universally evaluated from D f using a transverse image of fibre and assumption of circular cross-section Is this acceptable for Natural Fibres??

27 Single Fibre CSA Measurements 1. Single fibre diameter determined by averaging 4 transverse measurements at 2 mm intervals 2. Fibres embedded, cut and polished 3. true cross sectional area determined at approximately the same position on fibre 4. Sample ground down 2 mm and polished 5. Steps 3-4 repeated 10x

28 Natural Fibre CSA Evaluation Diameter CSA (mm 2 ) y=2.55x Flax Sisal Y X y=1.97x y=x Measured CSA (mm 2 )

29 Single Fibre Modulus 1/Modulus (10 6 /GPa) Sisal Flax y=11.4x E y=10.7x+33.0 * f = 1 E f + C A L Sisal, 1000/33=30 GPa Flax, 1000/14.1=71 GPa f Fibre CSA/Gauge Length (mm)

30 Natural Fibre CSA Evaluation Diameter CSA/True CSA Flax Sisal Average Diameter (mm)

31 Natural Fibre CSA Evaluation Diameter method significantly overestimates CSA Underestimates single fibre modulus and strength Magnitude of error is diameter dependent

32 Effect CSA on Single Fibre Properties INDUSTRY LEADER CSA method Diameter Actual Flax Strength (MPa) Sisal Strength (MPa) Flax Modulus (GPa) Sisal Modulus (GPa) 20 30

33 Apparent Modulus (GPa) Effect of Diameter CSA on Apparent NF Modulus Assume a diameter independent modulus Flax, E1f=71.0 GPa Sisal, E1f=30.5 GPa Average Diameter (mm)

34 Simple Model of NF CSA Diameter Errors INDUSTRY LEADER 10µm Flax 20µm NF non-circular simplest model is oval X-section Sisal

35 Simple Model of NF CSA Diameter Errors INDUSTRY LEADER Due to NF natural twist the oval cross section will be viewed differently at different positions along the fibre Transverse view from microscope

36 Parameteric Ellipse Analysis y True CSA = 0.25πAB B " Diameter"CSA = 0.25πD 2 φ t X max x A X(t),Y(t) X(t) = 0.5ACos(t)Cos( φ) 0.5BSin(t)Sin( φ) fibre diameter D Can solve for X max for any φ and then average over φ=0-90 for different A:B ratios

37 5 CSA Ratio from Ellipse Analysis CSA Ratio (D 2 /AB) A/B 5 A/B 3 A/B 2 A/B 1 Av CSA Ratio Ellipse Major Axis Orientation Angle

38 Natural Fibre CSA Evaluation 'Diameter"CSA/True CSA Flax measured Sisal measured Average Fibre "Diameter" (mm)

39 Natural Fibre CSA Evaluation 'Diameter"CSA/True CSA Lines of fixed CSA and varying ellipse A:B ratio Flax measured Sisal measured Flax thin Flax average 0 Flax thick Sisal thin Sisal average Sisal thick Average Fibre "Diameter" (mm)

40 Abaca Other Fibres Coir Kenaf Jute

41 Abaca Other Fibres Ellipse A:B Coir Kenaf Jute Similar issues probable in CSA estimation from fibre diameter

42 Natural Fibre CSA Evaluation Average CSA (mm 2 ) Diameter CSA / True CSA Sisal Flax Jute Hemp Kenaf Abaca Coir

43 Summary INDUSTRY LEADER Thermo-Mechanical Properties NF Longitudinal Modulus (GPa) 75 Glass Flax Sisal 61.5 (71.0) 24.9 (30.5) Transverse Modulus (GPa) Shear Modulus (GPa) Axial LCTE (µm/m. o C) Transverse LCTE (µm/m. o C)

44 What does this anisotropy mean for the reinforcement performance of natural fibres? E = η η V E + V 0 C L f Comparison NF and GF often assumes isotropic fibre f m E m Hence simple Krenchel analysis for η 0 η = cos 4 ( θ) 0 NF is more like an orthotropic composite material Apply laminate theory to model reinforcement performance

45 E = x ε x ε y γ xy Engineering Stiffness, Off-axis Orthotropic Lamina = σ ε x x S S S S S S ε = Sσ xy xy set σ xy = { σx 0 0} S S S σ 0 0 x and for all θ, S11 = S cos θ + (2S + S )sin θcos θ + S sin The terms S 11, etc., are found from S= 33 hence ε = x S11 1 E ν E E σ ν E 1 E x = x 4 21 θ S G

46 Offaxis Stiffness Contribution of Anisotropic Fibre Fibre Modulus Contribution Flax Krenchel Flax "Laminate" Sisal Krenchel Sisal "Laminate" Off-axis Angle ( )

47 Offaxis Stiffness Contribution of Anisotropic Fibre Fibre Modulus Contribution Integration of curves gives an average orientation Flax Krenchel factor for RoM random Flax in-plane "Laminate" GMT Krenchel Sisal ηkrenchel 0 =0.375 Laminate Sisal η "Laminate" 0 = Off-axis Angle ( )

48 Comparison Predicted Composite Modulus INDUSTRY LEADER For Randon Inplane moulded long fibre polypropylene GMT Modulus (GPa) Glass Fibre NF 20 GPa(Sisal) NF 40 GPa (Jute) NF 60 GPa (Flax) 0 Krenchel Fibre Content (% weight)

49 Comparison Predicted Composite Modulus INDUSTRY LEADER For Randon Inplane moulded long fibre polypropylene GMT Modulus (GPa) Glass Fibre NF 20 GPa (Sisal) NF 40 GPa (Jute) NF 60 GPa (Flax) 0 Laminate Fibre Content (% weight)

50 NF Anisotropy Challenge Fibre Modulus Contribution OK - Lets just make Unidirectional Composites? Actual contribution 40-50% less than expected Flax Krenchel at only 10 - how unidirectional Flax "Laminate" and non-wavy Sisal can you Krenchel make your UD NF composites Sisal "Laminate"? Off-axis Angle ( )

51 Conclusions (1) GLOBAL Estimation VISION of natural fibre cross section area via the diameter method leads to significant overestimation of CSA. results in significant underestimation of mechanical properties obtained by single fibre testing. also contributes significantly to the variability observed in the measurement of natural fibres properties. since the magnitude of the CSA error is diameter dependent single fibre properties will appear to be diameter dependent.

52 Conclusions (2) Flax and Sisal fibres exhibit very high levels of mechanical and thermomechanical anisotropy. Ignoring natural fibre anisotropy and using only the axial modulus of natural fibres in estimating their composite reinforcing ability will significantly overestimate their potential in any off-axis composite loading scenario.

53 Announcement Sustainable Composites In August 2013 the Advanced Composites Group at the University of Strathclyde filed its first patent application in the area of Glass Fibre Recovery covering cost effective, industrially applicable, treatments to regenerate the strength of thermally recycled glass fibres.