14. deutsches LS-DYNA Forum 2016 Integration of Single Cells of Lithium Ion Traction Battery in Crash Simulation

Transcription

1 14. deutsches LS-DYNA Forum 2016 Integration of Single Cells of Lithium Ion Traction Battery in Crash Simulation Bamberg, 10. October 2016 Dipl.-Ing. Michael Funcke Forschungsgesellschaft Kraftfahrwesen Aachen mbh Slide No. 1

2 Motivation [SAH12a] [SAH12b] [TES15] Low energy density of Lithium-ion cells (compared to conv. fuels) Heavy storage systems Energy storage + kg Integration into vehicle body Dimensioning If single cells are not part of crash simulations Damage of cells not tracked Manufacturers use conservative simulation approaches Oversizing [MIK11] Public interest in safety of electric vehicles Potential hazard (e.g. thermal runaway) More weight through crash-proof battery integration Slide No. 2

3 System Vehicle Cell Vehicle Research Approach and Methodology FEM Simulation Mechanical Characterisation Input parameters E-vehicle structure Battery systems Identification of relevant crash load cases Evaluation of GIDAS database Impact velocity, direction, object Evaluation of cell deformation Simplified cell simulation model Determination of a design load case Derivation of cell load cases Test parameters Cell simulation Adaptation to cell tests Mechanical behaviour Qualitative behaviour Cell simulation Forcedisplacement curves model Full vehicle simulation Detailed cell model Derivation of a system crash scenario Test parameters Cell tests Mechanical behaviour Voltage drop Thermal Runaway System crash simulation Detailed cell modell Mechanical behaviour Comparison System crash testing Mechanical behaviour Validated cell simulation model Slide No. 3

4 Battery structure E-vehicle structure Toyota Yaris Load cases Crash data vehicle front vehicle tail Input Parameters Input paramet ers Max. Intrusion depth of 2% of all vehicle collisions Intrusion probability x [%] [FUN14] Side impact ECE-R95 Euro NCAP Euro NCAP pole crash Battery casing Battery Models: Prismatic Cylindrical Prismatic Cylindrical Pouch cell Pouch 50 km/h km/h Pole (Ø = 254 mm) Slide No. 4

5 Evaluation of energy storage deformation Set up of simulation models for the energy storages and integration into the overall vehicle model Use of simplified cell models: Linear-elastic material behaviour Representing outer geometry Simulation of aforementioned load cases Deformation of vehicle body and energy storage within pole impact very localised Variation of pole position Evaluation of energy storage deformation Low / no deformation at barrier impact Pole impact is the more critical load case Following consideration of pole impact at 50 km/h Slide No Impact points for pole z Pole axis x Critical area Cylindrical cell Prismatic cell Pouch cell

6 Impact position Intrusion [mm] Intrusion [mm] Cell deformation Cell selection Pfahlaufprall - Anprallpunkt 3 z x A1 B1 C1 Front A2 B2 C2 A3 B3 C3 pole Rear A4 Deformation of energy storage housing (FE simulation) B4 C4 A5 B5 C5 B3 C A3 Pole impact - Impact point 3 A3 B3 C Time [ms] A3 B3 C3 max. 123 mm max. 73 mm max. 108 mm Load Zeit [ms] Pouch cell [Impact point 4] Cylindrical Prismatic Pouch cell high low high Cell deformation (FE simulation) low low high Tolerance of the cell to deformation low low high Internal cell security mechanisms available available unavailable Massive cell housing / protection against sharp objects available available unavailable Potential risk of cell chemistry medium (LiFePO4) medium (LiFePO4) high (NMC) Slide No. 6 [KOR13]

7 Derivation of load cases on cell level Top view Rear view Overall load on vehicle level Superimposed load cases Side wall Cover sheet Force Abutment Compression Frame Frame Sill inside 150 mm 75 mm Cell bending curve Force Bending Frame Abutment Cover sheet Punch through force l Cover Deckblech sheet Frame l 1 = 100 mm l 2 = 170 mm Slide No. 7

8 Group 2: Load considering the assembly conditions Group 1: generic load cases Derivation of load cases on cell level Compression 47 mm Bending Punch through short Punch through long Abutment Cell Cell edge Surface impactor Single frame Frame 120 mm 240 mm Punch through short 120 mm 240 mm Abutment Bending single cell Impaktor 75 mm Mounting cell 150 mm edge 120 mm mm mm 47 mm l 2 =100 mm Bending couple of cells 120 mm 240 mm 12 mm l 1 =170 mm Mounting cell edge Slide No. 8

9 Set-up of cell simulation model pouch cell (example) cell edge electrolyte layer Al layer cell closure FE model Simulation approach: Five layers of solid elements representing the electrolyte Six layers of shell elements representing the electrodes No connection between solid and shell elements Slipping between layers possible Cell edge representing the surrounding clamping area Connected to solid elements (tied contact) Cell closure connecting the outer shell layers Failure (Mat_add_erosion) Part_composite (three layers) Time step equivalent to full vehicle simulation Slide No. 9

10 Uniaxial crushing Three point bending Punch through short Punch through long Force [kn] Force [kn] Force [kn] Built-up of cell simulation model Execution of cell tests Derivation of target range for the simulation curves Sensitivity analysis to determine the parameter influences of the simulation model Component 1. Monotonization and normalization Raw data testing res. testing curve Displacement [mm] Parameter Solid element layer Stress-strain curve Interaction cell edge and clamping Friction coefficient Forming the arithmetic mean Test 1 Test 2 Test 3 Test 4 Arithmetic mean Displacement [mm] Influence on force-displacementcurve of the load case Successive determination of the model parameters through comparison of testing and simulation Uniaxial crushing three point bending punch through short punch through long Derivation of rating curves Arithmetic mean Upper boundary Lower boundary Displacement [mm] Parameter n.a. Low influence No influence High influence Slide No. 10

11 Force Kraft [kn] [kn] Built-up of cell simulation model 3,0 2,0 1,0 0,0 Upper Obere boundary Grenzwertkurve Bruchbereich Cracking of frames Rahmen Lower Untere boundary Grenzwertkurve Simulationsmodell model Displacement Weg [mm] [mm] 1. crack: Bruch: Bruch: crack: Bruch: crack: unterer lower three drei frames Rahmen oberer upper Rahmen frame Rahmen frame side Elektrodenseite with electric Side w/o abgewandte electric connectors Seite Side abgewandte w/o electric connectors Seite connectors Built-up and validation of the frame model Simulation of group 2 load cases Cell simulation model validated for quasistatic load cases Screws Steel profiles 4x Frames 3x Cells Slide No. 11

12 Validation of the simulation model and discussion of results Integration of cell model in full vehicle simulation model Examination of the cell deformation within the load case pole impact x y Load Area I Area II Increasing deformation Area III Slide No. 12 High deformation within Area I Stiffness of area II influences deformation within area I Deformation within area III very low Validation on system or vehicle level necessary No vehicle available for crash testing Derivation of a system load case, which considers the loads acting on full vehicle level diameter Impact velocity Absorbed kinetic energy, leading to impactor mass

13 Intrusion [mm] Validation of the simulation model and discussion of results Sidewall failure Frame rupture External shortcut [AUT14] t = 0 ms Connectors Pole Sidewall failure Frame rupture t = 0 ms Module (Frames + cells) t = 10 ms t = 10 ms t = 15 ms t = 15 ms t = 25 ms x y t = 20 ms t = 29 ms t = 29 ms Testing Simulation Time [s] Slide No. 13

14 Summary Within this research a simulation approach for pouch cells applicable to crash simulations on full vehicle level was investigated A simplified cell model was used to derive load cases on cell level from full vehicle simulations Test of theses load cases were carried out and the results were used for a stepwise model built-up Final validation by a system crash test showed a good correlation between simulation and testing Generation of a cell model applicable for full vehicle simulations was successful As long as the mechanical cell loads correspond to those used for the built-up process the generated cell model is applicable for various storage system layouts and positions within the vehicle Slide No. 14

15 Contact Thank you for your Attention! Dipl.-Ing. Michael Funcke fka Forschungsgesellschaft Kraftfahrwesen mbh Aachen Steinbachstr Aachen Germany Phone Fax funcke@fka.de Internet Slide No. 15

16 Sources [AUT14] AUTOLIV OSTLER Project Results Autoliv Sverige AB, Vårgårda, Sweden, 2014 [FUN14] FUNCKE, M., SCHÄFER, S., WOHLECKER, R., DUFAUT, D., STURK, D., VAVALIDIS, K. D Evaluation report of active and passive protection solutions Institut für Kraftfahrzeuge Aachen, RWTH Aachen University, Aachen, 2014 [KOR13] KORTHAUER, R. Handbuch Lithium-Ionen-Batterien Springer-Verlag, Berlin, Heidelberg, 2013 [MIK11] MIKOLAJCZAK, C., KAHN, M., WHITE, K., LONG, R.T. Lithium-Ion Batteries Hazard and Use Assessment Springer-Verlag, New York, Heidelberg, Dordrecht, London, 2011 [SAH12a] SAHRAEI, E., HILL, R., WIERZBICKI, T. Calibration and finite element simulation of pouch lithium-ion batteries for mechanical integrity Journal of Power Sources, Heft 201, S (2012) Slide No. 16

17 Sources [SAH12c] SAHRAEI, E., CAMPBELL, J., WIERZBICKI, T. Modeling and short circuit detection of Li-ion cells under mechanical abuse conditions Journal of Power Sources, Heft 220, S (2012 [TES15] TESLA MOTORS INC. Herstellerseite abgerufen am Tesla Motors Inc., Palo Alto, Kalifornien, USA, 2015 Slide No. 17