International Permeability Benchmark B. Laine, S.V. Lomov, P. Henrat, Ch. Binetruy, P. Ermanni, V. Michaud, F. Trochu, H. Sol, A. Long, M. Wietgrefe, P. Beauchene, A. Endruweit, F. Gommer, E. Ruiz, S. Comas-Cardona, S. Hasanovic, G. Morren, L. Bizet, J. Bréard downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm
Plan of the presentation • Introduction • Why, Who and using What ? • Experimental setups and associated problems • 1D, 2D • Saturated, Non saturated • Results • Way forward • Time schedule Direction - Conférence • Toward Normalization ? 2 downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm
Introduction • Why doing an international permeability benchmark? • The permeability tensor of fibrous reinforcements is the main “material” input parameter for macroscopic flow simulation • To reduce the existing “scattering” of permeability measurements between setups (labs) for both saturated and partially saturated measurements � to understand why we get such dispersions and produce recommendations in order to reduce them • To understand the limitations of each type of measurement • 1D (in plane or transverse), 2D, saturated, partially saturated and their combinations � To initiate a normalization process Direction - Conférence • To do so, we need to have the right “countries” around the table but also to be aware of the industrial needs 3 downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm
Introduction University of Delaware Advani , Suresh US University of Twente Akkerman , Remko Netherlands • Who is involved? Mines de Douai Binetruy , Christophe France 18 Universities and Institutes • Université du Havre Bréard , Joel France from 11 countries EADS IW Chatel , Sylvain France 2 industrial end users INEGI-UMEC Correira , Nuno Portugal • Polytechnique Zurich Ermanni , Paolo Suisse 1 industrial providing the • Hexcel Reinforcments Henrat , Patrick France reinforcements: HEXCEL Lightweight Structures BV Labordus , Maarten Netherlands ONERA Laine , Bertrand France KU Leuven Lomov , Stepan V. Belgium University of Nottingham Long , Andrew UK Lulea University Lundstrom , Staffan Sweden Polytechnique Lausane Michaud , Véronique Switzerland UCONN Parnas , Richard US Free University of Brussels Sol , Hugo Belgium PCCL Talvensaari , Henna Austria Direction - Conférence Polytechnique Montréal Trochu , François Canada Mines de St Etienne Vautrin , Alain France DLR-Airbus Wietgrefe , Mathias Germany TU Clausthal Ziegmann , Gerhard Germany 4 downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm
Introduction • What are the investigated materials Balanced Carbon • twill 2x2 (G986) Unbalanced Glass • twill 2x2 (G1113) Carbon non-crimped fabric • Direction - Conférence 45/-45 (NC2) 5 downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm
Experimental setups • 1D Dry or partially saturated = position of the flow front as function of time • Saturated = Mass flow as function of the pressure drop • • 2D Dry or partially saturated = position of the main axes of the pseudo- • ellipse as function of time Saturated = mass flow as function of the pressure drop along (at • least) two directions Continuous = Compression of a saturated medium (force - • displacement curve gives K) • Transverse Continuous (idem but with flow in the transverse direction) • Direction - Conférence 1D saturated = mass flow as function of pressure drop • 1D Dry or partially saturated = position of the flow front as function of • time 6 downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm
Experimental setups • List of parameters which should be recorder during Benchmark for all measurements Mould: nature, thickness, deflection, temperature, position of pressure • sensors Fluid test: nature, behavior (viscosity vs strain rate…), temperature, • viscosity as function of T , density of the fluid Reinforcement: volume fraction of fiber, number of plies, lay-up, • contact angle with fluid test Sketch of setup, experimental procedure, type of experiment • (saturated …), raw data (pressure …), data processing, size of injection hole if any, way of determining the position of flow front if relevant, determination of main directions o flfow Direction - Conférence 7 downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm
Results • Permeability of G986 (all results): from 4.10 -11 to 6.10 -10 !! Permeability of G986 1,00E-09 Permeability (m²) Kxx Kyy 1,00E-10 Direction - Conférence 1,00E-11 0 2 4 6 8 10 12 Lab or Setup 8 downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm
Results • Permeability of G986 as function of Vf and experimental setup: ~Kdry < Ksat, difficult to said anything else! • Permeability of G986 as function of Vf and experimental setup 1,00E-09 Kxx 2D sat Kyy 2D sat Kxx 2D dry Kyy 2D dry Permeability (m²) Kxx 1D sat 1,00E-10 Kyy 1D sat Kxx 1D dry Kyy 1D dry Kxx 2D Direction - Conférence continuous 1,00E-11 45 46 47 48 49 50 51 52 53 54 Vf of fibre 9 downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm
Results • Permeability of G986 A lot of dispersion! • At first sight, no trend in terms of difference between 1D, 2D, • saturated and partially saturated… � These trends are probably hidden by a bigger issue! When a same lab performs both saturated and unsaturated • experiments, the saturated permeability is always higher Anisotropy ratio (Kyy/Kxx) varies between 1.3 and 2.6 • Direction - Conférence 10 downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm
Results • Permeability of G1113 (all results): Kxx: from 4.10 -11 to 4.10 -10 • Kyy: from 1.10 -11 to 6.10 -11 • Permeability of G1113 1,00E-09 Permeability (m²) Kxx Kyy 1,00E-10 Direction - Conférence 1,00E-11 0 1 2 3 4 5 6 7 8 9 10 Lab or Setup 11 downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm
Results • Permeability of G1113 as function of Vf and setups A little bit less dispersion than for G986, but no trend… • Permeability of G1113 as function of Vf and experimental setup 1,00E-09 Kxx 2D sat Kyy 2D sat Kxx 2D dry Kyy 2D dry Kxx 1D sat Kyy 1D sat Permeability (m²) Kxx 1D dry Kyy 1D dry 1,00E-10 Direction - Conférence 1,00E-11 44 45 46 47 48 49 50 51 52 53 54 Vf of fiber 12 downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm
Results • Permeability of G1113 A lot of dispersion, even if a little bit less than for G986 • Again, at first sight, no trend in terms of difference between 1D, 2D, • saturated and partially saturated… When a same lab performs both saturated and unsaturated • experiments, the saturated permeability is always higher High dispersion of anisotropy ratio (Kyy/Kxx), which varies between • 1.7 and 4.1 Direction - Conférence 13 downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm
Results • Permeability of NC2 as function of Vf and experimental setups Few points, but dispersion is low so far! • Probably less local geometrical variation in a NCF than woven • Permeability of NC2 as function of Vf and experimental setups 1E-10 Kxx 2D sat Kyy 2D sat Kxx 2D dry Kyy 2D dry Permeability (m²) Direction - Conférence 1E-11 47 47,5 48 48,5 49 49,5 50 50,5 51 Vf of fiber 14 downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm
Results • High dispersion for woven fabrics • Still many results to come (less than half of the partners have send something) • So far, we almost always received enough information on the way the measurements were performed • Some Kzz values are already available • It is difficult to extract trends from these data, but we are working on them. The more results we have, the easier it will be to deduce anything… Direction - Conférence 15 downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm
Way forward • Please send your results ( cf website: permeability.onera.fr) !! • Some complementary tests/information will be asked to all partners in order to understand what is happening • First “draft” analysis of the dispersion will be proposed before end of the year to partners who have already send something… • Again, we need to fully understand were does this dispersion comes from in order to initiate a normalization procedure • Thank you for your help, we look forward to receive more results before the end of this year: this benchmark is the only Direction - Conférence way to close this topic of permeability dispersion! 16 downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm
Stereolithographic specimen K, 10 -9 mm 2 7 6 5 4 3 Direction - Conférence 2 Gerd Morren’s poster 1 Kx Ky 0 FPCM-9 Montreal - July 17 downloaded from: http://www.mtm.kuleuven.ac.be/Research/C2/poly/index.htm 2008
Recommend
More recommend
