T F A R D A S I

  • Slides: 118
Download presentation
T F A R D A S I S THI NT E M U

T F A R D A S I S THI NT E M U C DO tions: i d n o c wing o l l o f e ess, r h t d d h t i a w me & se it a u rsion n n e a y v c n d u a e o t p Y upda , com e e h t m u a o n nd y our e y s e n m a c d sen So I TC S. L ) e t n u o o h • Please with il, (p t a r m a e. p e , l y title ailab e an v t a a l s u e c r m co or ci r e t l as it be a o not d e s a k. • Ple c a b d e e ! t s or f n o i t consen c m. e o r c r. o c c t s , l s ian@ stion t e a g g o u D s any t Ian e c t a a t i n c o e r c I app lease p , s n o i st Any que Livermore Software Technology Corporation All Rights Reserved For questions or comments please call LSTC at (925)-449 -2500 Copyright © 2004 1

Airbag Deployment Modeling Tutorial with LS-DYNA ALE Fluid-Structure Interaction Approach May 25 th, 2004

Airbag Deployment Modeling Tutorial with LS-DYNA ALE Fluid-Structure Interaction Approach May 25 th, 2004 Ian Do & Dilip Bhalsod Livermore Software Technology Corporation All Rights Reserved For questions or comments please call LSTC at (925)-449 -2500 Copyright © 2004 2

Fluid-Structure Interaction Modeling with LS-DYNA OUTLINE I. Introduction II. TANK TEST [a] Control Volume

Fluid-Structure Interaction Modeling with LS-DYNA OUTLINE I. Introduction II. TANK TEST [a] Control Volume [b] ALE [c] ALE without coupling III. FLAT AIRBAG [a] Control Volume [b] ALE - using traditional method [c] ALE - using *AIRBAG_ALE (newly developed) [d] ALE - using traditional method – moving mesh IV. Some NEW ALE developments V. Some general remarks on ALE modeling - Summary Livermore Software Technology Corporation 3

Fluid-Structure Interaction Modeling with LS-DYNA ACKNOWLEDGEMENT This work is a result of the collaboration

Fluid-Structure Interaction Modeling with LS-DYNA ACKNOWLEDGEMENT This work is a result of the collaboration with and learning from my good friends at LSTC. Among them are Lars Olovsson, Jason Wang, Dilip Bhalsod, Hao Chen, Mhamed Souli, Lee Bindeman, Morten Jensen, Jim Day, Khanh Bui, Todd Slavik and Mike Burger. Dilip created most example models included. Mike generated the geometry for some of the models. Jason and Hao are credited with the most difficult work of correcting and improving the algorithm continually. Last but not least is the contribution by Lars who pioneers this area of fluid-structure interaction research. His creativity, pleasant, kind nature and generosity to share his knowledge has made this task much easier and fun for all of us. This tutorial is more their contribution, though any mistakes are mine. So please give me feedback so I can continually improve it. Livermore Software Technology Corporation 4

Fluid-Structure Interaction Modeling with LS-DYNA CAUTION ALL MODELS AND ASSOCIATED DATA PRESENTED ARE FICTITIOUS

Fluid-Structure Interaction Modeling with LS-DYNA CAUTION ALL MODELS AND ASSOCIATED DATA PRESENTED ARE FICTITIOUS THEY ARE SHOWN TO FACILITATE USERS’ UNDERSTANDING ONLY. Livermore Software Technology Corporation 5

Fluid-Structure Interaction Modeling with LS-DYNA FOREWORDS This guide is created to show some basic

Fluid-Structure Interaction Modeling with LS-DYNA FOREWORDS This guide is created to show some basic steps to modeling airbag deployment using LS-DYNA Arbitrary Largrangian-Eulerian (ALE ) fluid-structure-interaction (FSI) capabilities. It is assumed that the users are already somewhat familiar with LS-DYNA ALE theory, method and command usage. For more details on the KEYWORDs used, please consult our ALE tutorial and/or latest Keyword manual. LS-DYNA ALE Airbag deployment modeling has been going through extensive evolution. The most robust and newly developed features will be presented. In the last section, some experimental/developing features will be introduced. As of the time of this writing, we are testing a new approach to simplify airbag modeling: start with ALE, then switch to CV method at a switch time defined by the user (to save CPU time). Livermore Software Technology Corporation 6

Fluid-Structure Interaction Modeling with LS-DYNA [I] INTRODUCTION Livermore Software Technology Corporation 7

Fluid-Structure Interaction Modeling with LS-DYNA [I] INTRODUCTION Livermore Software Technology Corporation 7

Fluid-Structure Interaction Modeling with LS-DYNA [I] INTRODUCTION MOTIVATION Our Control Volume (CV) approach to

Fluid-Structure Interaction Modeling with LS-DYNA [I] INTRODUCTION MOTIVATION Our Control Volume (CV) approach to modeling the inflation process of an airbag has been well developed. But out-of-position occupant tests requires the most accurate prediction possible of the pressure field inside the airbag so that precise airbag unfolding dynamics can be predicted. Control Volume (CV) method (using the *AIRBAG_ cards in LSDYNA) assumes a uniform pressure distribution inside the airbag. Such that all segments of the airbag “feels” the same internal pressure at the same time throughout the whole process. This causes the segments at the tips of the folds to be inflated early. Thus the uniform pressure assumption may open the bag too fast. Hence there is room for improvement on the airbag deployment prediction by CV method. Livermore Software Technology Corporation 8

Fluid-Structure Interaction Modeling with LS-DYNA [I] INTRODUCTION MOTIVATION (cont. ) Arbitrary Lagrangian-Eulerian (ALE) method

Fluid-Structure Interaction Modeling with LS-DYNA [I] INTRODUCTION MOTIVATION (cont. ) Arbitrary Lagrangian-Eulerian (ALE) method allows for arbitrary variation in the pressure field inside the airbag to be accounted for. This variation in pressure field will affect the way each airbag layer is unfolded as the gas flow reaches its fold. The fluid-structure interaction (FSI) and contact during the initial unfolding is a complex process. Basic physical principles and educated estimations have been employed to capture as close as possible the real physics of FSI. Many engineering assumptions must be made. The users must recognize that they are not perfect, and they are only our best “educated” hypotheses at the moment. Only the successful validations of their results will merit their application. Livermore Software Technology Corporation 9

Fluid-Structure Interaction Modeling with LS-DYNA [I] INTRODUCTION CAUTION! The users must be vigilant and

Fluid-Structure Interaction Modeling with LS-DYNA [I] INTRODUCTION CAUTION! The users must be vigilant and always: (a) Use the best material & inlet data available with consistent -system. (b) Validate the models against benchmark data. unit It is highly recommended that the users familiarized themselves with the basic assumptions made. Hence have a clear idea of the limitations of the approach employed. This guide will NOT cover theories used. Where appropriate, the limitations and assumptions will be stated. Livermore Software Technology Corporation 10

Fluid-Structure Interaction Modeling with LS-DYNA [I] INTRODUCTION BACKGROUND INFORMATION Airbag deployment modeling requires 2

Fluid-Structure Interaction Modeling with LS-DYNA [I] INTRODUCTION BACKGROUND INFORMATION Airbag deployment modeling requires 2 basic pieces of information: - The inflator INFLATING POTENTIAL (or power). - The airbag geometry or construction. Typically, the inflator INFLATING POTENTIAL is derived from an inflator tank test where the inflator is discharged into a receiving tank. The main piece of data obtained is usually the “downstream” tank pressure. This cannot be used directly as the input condition for computational model. Calculation must be done to compute additional inflow information, i. e. mass flow rate, , and stagnation temperature of the gas, , and if possible, velocity, . Livermore Software Technology Corporation 11

Fluid-Structure Interaction Modeling with LS-DYNA [I] INTRODUCTION BACKGROUND INFORMATION The airbag construction is typically

Fluid-Structure Interaction Modeling with LS-DYNA [I] INTRODUCTION BACKGROUND INFORMATION The airbag construction is typically readily available from existing purely Lagrangian Control-Volume models. Thus, providing the correct INFLATING POTENTIAL, or the inlet initial and boundary conditions, is critical to computational modeling. The inlet information, and must be provided by the inflator or automotive manufacturer. The can be estimated (more on this later). Schematically this process may be represented … Livermore Software Technology Corporation 12

Fluid-Structure Interaction Modeling with LS-DYNA [I] INTRODUCTION Inflator’s INFLATING POTENTIAL= ? TANK TEST modeling

Fluid-Structure Interaction Modeling with LS-DYNA [I] INTRODUCTION Inflator’s INFLATING POTENTIAL= ? TANK TEST modeling Inflator Manufacturer Perform tank test Car Manufacturer , ? Inflator Code Control-Volume Code (ISP, etc. ) LS-DYNA Tank Test Model [1] Control-Volume & [2] ALE method , … Species Info, ? , … check and Estimated Reproducing means Validating the Inflating Potential of a specific inflator Livermore Software Technology Corporation 13

Fluid-Structure Interaction Modeling with LS-DYNA [II] TANK TEST Livermore Software Technology Corporation 14

Fluid-Structure Interaction Modeling with LS-DYNA [II] TANK TEST Livermore Software Technology Corporation 14

Fluid-Structure Interaction Modeling with LS-DYNA TANK TEST INFLATOR & TANK A schematic of a

Fluid-Structure Interaction Modeling with LS-DYNA TANK TEST INFLATOR & TANK A schematic of a generic hybrid inflator and tank are shown (not to scale). The output of the inflator becomes the input of the tank test. In general, it is difficult to directly measure the , and inflator Gas 1 Mixing chamber orifices Gas 2 NOT to scale! Inflator is much smaller than tank Tank Livermore Software Technology Corporation 15

Fluid-Structure Interaction Modeling with LS-DYNA TANK TEST INFLATOR & TANK TEST - The tank

Fluid-Structure Interaction Modeling with LS-DYNA TANK TEST INFLATOR & TANK TEST - The tank is usually initially filled with a resident gas. - This gas is then mixed with the incoming inflator gas(es). - The tank pressure history, , is recorded. INFLATOR MANUFACTURER After in-house processing, the inflator manufacturer usually provide , , and species information to the automotive manufacturer. AUTOMOTIVE MANUFACTURER Usually perform preliminary calculation using , using Control Volume method (ISP code, or LS-DYNA *AIRBAG_ option) to reproduce to validate the inflator power characteristics. Livermore Software Technology Corporation 16

Fluid-Structure Interaction Modeling with LS-DYNA [IIa] CONTROL VOLUME TANK TEST MODELING Livermore Software Technology

Fluid-Structure Interaction Modeling with LS-DYNA [IIa] CONTROL VOLUME TANK TEST MODELING Livermore Software Technology Corporation 17

Fluid-Structure Interaction Modeling with LS-DYNA TANK TEST WITH CONTROL VOLUME METHOD GEOMETRY: A cylindrical

Fluid-Structure Interaction Modeling with LS-DYNA TANK TEST WITH CONTROL VOLUME METHOD GEOMETRY: A cylindrical tank (vol~100 lit~0. 1 E 9 mm 3). Its shell normals should point uniformly outward for CV. (A dummy inflator ring is created for locating the point sources for later ALE application, not used here. ) S 1 The tank is modeled by a Lagrangian shell structure Cut-off view S 3 A dummy inflator ring may be defined for locating the point sources later (no interaction with ring). Livermore Software Technology Corporation 18

Fluid-Structure Interaction Modeling with LS-DYNA TANK TEST WITH CONTROL VOLUME METHOD MODEL SUMMARY: Lagrangian

Fluid-Structure Interaction Modeling with LS-DYNA TANK TEST WITH CONTROL VOLUME METHOD MODEL SUMMARY: Lagrangian model requires simply: the Lagangian structure(s) or tank and the inflating potential information for the CV method. The input files are: 100 llstc_cv 3. k 100 llstc_cvctl 3. k 100 ltanks 1_normout. k 100 lrings 3. k 100 lmdot. Tstagvel. k = = = Main control volume (CV) input file *AIRBAG_HYBRID CV definitions Lagrangian shell tank geometry/mesh Lagrangian dummy inflator cylinder shell Load curves for inlet information: mdot(t)=LCID 1033; Tstag(t)=LCID 1011 The parts are: S 1 = Lagrangian shell tank S 3 = Lagrangian dummy inflator ring for locating point sources Livermore Software Technology Corporation 19

Fluid-Structure Interaction Modeling with LS-DYNA TANK TEST WITH CONTROL VOLUME METHOD KEYWORD DEFINITION FOR

Fluid-Structure Interaction Modeling with LS-DYNA TANK TEST WITH CONTROL VOLUME METHOD KEYWORD DEFINITION FOR TANK PART: To keep it simple, a rigid tank is used (it does not have to be rigid). Ø A shell structure modeling the tank maybe defined as shown below. Ø Assuming that gas 1 occupies the tank initially. Ø Inflator gas 2 is then injected into the tank. Ø Assuming perfect adiabatic mixing between the 2 ideal gases. S 1=tank *PART rigid tank : thick = 0. 4 mm $ PID SECID MID 1 1 1 *SECTION_SHELL $ SECID ELFORM SHRF 1 5 0. 0000000 $ T 1 T 2 T 3 0. 4000000 $ B 1 B 2 B 3 0. 0 *MAT_RIGID $ MID RO E 1 7. 8500 -06 2. 0000000 $ CMO CON 1 CON 2 1. 0000000 7. 0000000 $ A 1 A 2 A 3 0. 0000000 EOSID 0 HGID 0 GRAV 0 ADPOPT 0 NIP PROPT QR/IRID 1. 0000000 0. 0000000 T 4 NLOC 0. 4000000 0. 0000000 B 4 B 5 B 6 0. 0 ICOMP 1 B 7 TMID 0 B 8 PR N COUPLE M 0. 3000000 0. 0000000 This constrains the tank motion V 1 V 2 V 3 0. 0000000 Livermore Software Technology Corporation 20

Fluid-Structure Interaction Modeling with LS-DYNA TANK TEST WITH CONTROL VOLUME METHOD KEYWORD DEFINITION FOR

Fluid-Structure Interaction Modeling with LS-DYNA TANK TEST WITH CONTROL VOLUME METHOD KEYWORD DEFINITION FOR CV METHOD INFLATION: The inflator gas energy input acts to raise the pressure inside the tank. Ø and curves for the inlet gas are provided. Ø The heat capacity, Cp (per-mole unit), and molecular weight, MW, for each of the 2 gases must be defined. Ø Typically the atmospheric T, P & rho of air define the background. Ambient background *AIRBAG_HYBRID $ sidtyp 1 1 air $ atmost atmosp 2 lines defined per gas Gas 1 Gas 2 $ $ $ rbid vsca 0 0. 0 atmosd gc 298. 15001. 01325 e-41. 17913 e-9 8. 314400 c 23 lcc 23 a 23 lca 23 0. 000000 0 0 opt pvent ngas 0 0. 0 2 lcidm lcidt not used mw 0 0 0. 0288479 fmass 0. 0 lcidm lcidt not used mw 1033 1011 0. 022560 fmass 0. 0 psca 0. 0 cc 1. 000000 cp 23 0. 0 vini 0. 0 mwd 0. 000000 spsf 0. 0 lcp 23 0 ap 23 0. 0 lcap 23 2 gases will mixain the tank initm b Given 1. 000000 29. 049852 initm 0. 0 & Livermore Software Technology Corporation a 36. 23388 0. 0 c 0. 0 b 0. 0 curves from CV data INFLATING POTENTIAL 21

Fluid-Structure Interaction Modeling with LS-DYNA TANK TEST WITH CONTROL VOLUME METHOD CV TANK PRESSURE

Fluid-Structure Interaction Modeling with LS-DYNA TANK TEST WITH CONTROL VOLUME METHOD CV TANK PRESSURE RESULT: The pressure inside the tank is driven by the mass flow rate and gas temperature curves. Typically this should always be the 1 st test model as a “sanity check” of the inlet data ( & ) such that this computed pressure is close to the tank pressure data measured. Livermore Software Technology Corporation 22

Fluid-Structure Interaction Modeling with LS-DYNA [IIb] ALE TANK TEST MODELING WITH COUPLING Livermore Software

Fluid-Structure Interaction Modeling with LS-DYNA [IIb] ALE TANK TEST MODELING WITH COUPLING Livermore Software Technology Corporation 23

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING MODEL SUMMARY: ALE

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING MODEL SUMMARY: ALE model requires (1) Lagangian structure(s) ( tank) to interact with (2) ALE/Eulerian “fluid(s)” : air outside, initial resident gas inside, and inflator gas to be injected in. Tank definition is the same as before. The input files are: 100 llstc 1. k 100 llstcalectl. k 100 lfluidh 4. k 100 ltanks 1. k 100 lrings 3. k 100 lmdot. Tstagvel. k = Main input file = ALE part definitions and ALE controls = ALE geometry/mesh definition = Lagrangian shell tank geometry/mesh = Lagrangian dummy inflator cylinder shell = Load curves for inlet information: LCID 1011=Tstag(t); LCID 1022=vel(t); LCID 1033=mdot(t) The parts are: S 1 S 3 H 4 H 5 H 6 = = = Lagrangian shell tank Lagrangian dummy inflator ring for locating point AMMG 1 = background ALE mesh initially 100% filled AMMG 2 = initial gas AMMG 2, gas 2, inside the tank AMMG 2 = injected inflator gases AMMG 2, gas 3, (no sources with AMMG 1, gas 1 (no mesh defined) The Lagrangian part definitions from the CV model may be used directly. Livermore Software Technology Corporation 24

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING GEOMETRY: Using the

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING GEOMETRY: Using the same cylindrical tank (vol~100 lit~0. 1 E 9 mm 3) as CV model & dummy inflator ring for locating the point sources for ALE application. Cut-off view S 1= tank shell model ( normals pointing inward toward the gas it will couple to) H 4=AMMG 1=back ground mesh taking on outside air properties (initially defined ALE mesh) S 13=dummy inflator ring Livermore Software Technology Corporation 25

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING DIGRESSION-CLARIFICATIONS OF A

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING DIGRESSION-CLARIFICATIONS OF A FEW BASIC CONCEPTS: [1] A geometrical mesh at t=0 is defined with the *NODE & *ELEMENT_SOLID cards where each solid (or hexahedron) element is associated with a part ID (PID). [2] An ALE part can generally be defined with the *PART, *SECTION_SOLID, *MAT_, *EOS_, *HOURGLASS cards. A part can be defined with or without an association with a mesh definition. [3] An ALE multi-material group (AMMG), defined by an *ALE_MULTI-MATERIAL_GROUP card, refers to a physical H 1 fluid. mesh Consider an example where 1 mesh space is initially filled with air, and we want to inject 1 hot gas into it: H 2 - A geometrical mesh is defined. (no mesh) - A part is defined for air (H 1). This part is Livermore Software Technology Corporation associated with the mesh. 26 - A 2 nd part is defined for hot gas (H 2).

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING CLARIFICATIONS OF A

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING CLARIFICATIONS OF A FEW BASIC CONCEPTS(cont. ): There are 2 ways to define the property definitions for the air and gas: [Case A] Using traditional method (trivial) The gases do not mix thermodynamically. Their interfaces may be distinguished from each H 1 other. H 1=AMMG 1 *MAT_NULL *EOS_IDEAL_GAS *MAT_NULL H 2=AMMG 2 *EOS_IDEAL_GAS mesh H 2 (no mesh) [Case B] Using an improved method The gases do mix thermodynamically. Also, during advection, the mixture’s dissipated kinetic energy is converted into internal energy. This is a more accurate energy consideration for gas mixture. They share the same AMMG ID. H 1=AMMG 1 *MAT_GAS_MIXTURE *INITIAL_GAS _MIXTURE *MAT_GAS_MIXTURE H 2=AMMG 1 This says mesh 1 may contain material 1=air different materials defined in the same card Livermore Software Technology Corporation 27

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING CLARIFICATIONS OF A

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING CLARIFICATIONS OF A FEW BASIC CONCEPTS (cont. ): *PART Sample H 1 = background ALE air mesh = air initially fills H 1 mesh domain. $ PID SECID MID EOSID HGID GRAV ADPOPT Input for 1 1 1 0 0 *SECTION_SOLID [Case B]: 1 11 0 Note the order of the gases and the assumption of same velocity and T of all inflowing gases. NO FLOW for gas 1 TMID *MAT_GAS_MIXTURE $ MID IECONSFLG GC 1 0 0 $ Cv 1_mas Cv 2_mas Cv 3_mas Cv 4_mas Cv 5_mas Cv 6_mas Cv 7_mas Cv 8_mas 718. 782891 1237. 5623 $ Cp 1_mas Cp 2_mas Cp 3_mas Cp 4_mas Cp 5_mas Cp 6_mas Cp 7_mas Cp 8_mas 1007. 00058 1606. 1117 *INITIAL_GAS_MIXTURE $ SID STYPE MMGID T 0 1 1 1 298. 15 $ RHO 1 RHO 2 RHO 3 RHO 4 RHO 5 RHO 6 RHO 7 RHO 8 1. 17913 E-9 *HOURGLASS $ HGID IHQ QM IBQ Q 1 Q 2 QB QW 1 1 1. 00 e-05 *PART H 2 = hot gas injected, to be mixed with initial resident air in the mesh. $ PID SECID MID EOSID HGID GRAV ADPOPT TMID 2 2 1 0 0 *SECTION_POINT_SOURCE_MIXTURE $ SECID LCIDT NOTUSED LCIDVEL NIDLC 001 NIDLC 002 NIDLC 003 2 1011 0 1022 $ LCMDOT 1 LCMDOT 2 LCMDOT 3 LCMDOT 4 LCMDOT 5 LCMDOT 6 LCMDOT 7 LCMDOT 8 0 1033 $ NODEID VECID ORIFAREA This indicates that H 1 mesh may be filled with air (gas 1) Materials that mix share same *MGM card This pumps in the hot gas (gas 2) Livermore Software Technology Corporation 28

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING BACK TO THE

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING BACK TO THE CURRENT MODEL GEOMETRY: H 4 mesh or part is defined initially as gas 1 (or air) outside the tank. The *INITIAL_VOLUME_FRACTION_GEOMETRY card then initializes the space inside the tank with gas 2 (AMMG 2, H 5). At the start of Cut-off view the transient phase, the S 1= tank inflator gas 3 (AMMG 2, H 6) is injected into the H 4=AMMG 1=back tank. This gas 3 mixes ground outside air with the resident gas 2 in (initially defined ALE mesh) side the tank. Coupling: H 5=AMMG 2=initial 1. Search for spatial gas inside the tank coupling between S 1 and H 6=AMMG 2=inflator H 4. gas injected in 2. Couple S 1 to AMMG 2, physical fluid inside. S 13=dummy inflator ring Livermore Software Technology Corporation 29

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING H 4 =

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING H 4 = initial ALE mesh (for background air outside the tank) We can use {*MAT_GAS_MIXTURE (MGM) & *INITIAL_GAS_MIXTURE (IGM)} by itself even when the gas outside is not mixing with the inside gases If {MGM&IGM} is used, this first IGM is required to specify that part H 4 (mesh) may be occupied by this AMMG 1. gas prop *PART H 4 = background ALE air mesh = initial air in H 4 mesh domain, inside & outside $ PID SECID MID EOSID HGID GRAV ADPOPT TMID 4 4 4 0 0 *SECTION_SOLID 4 11 0 *MAT_GAS_MIXTURE $ MID IECONSFLG GC 4 0 0 $ Cv 1_mas Cv 2_mas Cv 3_mas Cv 4_mas Cv 5_mas Cv 6_mas Cv 7_mas Cv 8_mas 718. 782891 $ Cp 1_mas Cp 2_mas Cp 3_mas Cp 4_mas Cp 5_mas Cp 6_mas Cp 7_mas Cp 8_mas 1007. 00058 This allows AMMG 1 to be present in mesh H 4 *INITIAL_GAS_MIXTURE $ SID STYPE MMGID T 0 4 1 1 298. 15 $ RHO 1 RHO 2 RHO 3 RHO 4 RHO 5 RHO 6 RHO 7 RHO 8 1. 17913 E-9 *HOURGLASS $ HGID IHQ QM IBQ Q 1 Q 2 QB QW 4 1 1. 00 e-05 Livermore Software Technology Corporation Ref. Condition information 30

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING Alternate definition for

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING Alternate definition for H 4 = initial ALE mesh (for background air) Alternately, we can uses {*MAT_NULL (MN) & *EOS_IDEAL_GAS (EIG)} to define the outside air (only ALE part with initial mesh) as shown below. This should give similar result as using {MGM&IGM}. Difference? {MGM&IGM} allows for thermodynamic mixing of, for example, a gas that initially occupies a PID or mesh with another gas injected into it. *PART gas prop H 4 = background ALE air mesh = initial air in H 4 mesh domain, inside & outside $ PID SECID MID EOSID HGID GRAV ADPOPT TMID 4 4 4 0 0 *SECTION_SOLID 4 11 0 *MAT_NULL $ MID RO PC MU TEROD CEROD YMBEAM PRBEAM 41. 17913 E-9 0. 0 1. 844 E-11 0. 0 *EOS_IDEAL_GAS $ EOSID Cv Cp C 1 C 2 T 0 V 0 4718. 7828911007. 00058 0. 0 298. 15 1. 0 *HOURGLASS $ HGID IHQ QM IBQ Q 1 Q 2 QB QW 4 1 1. 00 e-05 Livermore Software Technology Corporation Ref. Condition information 31

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING KEYWORD DEFINITIONS: H

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING KEYWORD DEFINITIONS: H 5 = initial tank resident gas 2 (AMMG 2, no initial mesh defined). H 6 = injected inflator gas 3 (AMMG 2, no initial mesh defined). Both H 5 & H 6 can be defined with 1 {MGM&IGM} set so the inflator gas can mix with resident gas & be treated as 1 homogeneous fluid = AMMG 2 (while the air outside the tank has its own set of {MGM&IGM} cards). This MGM card will contain thermal properties of 2 gases. Another IGM card (next page) is used to specify that part H 4 (background mesh) may also be occupied by AMMG 2. This the portion of the resident gas inside the tank at t=0. Livermore Software Technology Corporation 32

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING H 5 =

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING H 5 = resident gas inside the tank at t=0 The material model for the gas mixture inside the tank is shared by the 2 gases to be mixed in the process. MGM = heat capacity information for the 2 gases inside the tank. IGM = simply states that AMMG 2 can also flow into mesh/part H 4. H 5 = initial resident gas Materials that mix share same *MGM gas 2 prop *PART H 5 = initial gas inside the tank (no associated initial mesh) $ PID SECID MID EOSID HGID GRAV 5 5 5 0 *SECTION_SOLID 5 11 0 *MAT_GAS_MIXTURE $ MID IECONSFLG GC 5 0 0 $ Cv 1_mas Cv 2_mas Cv 3_mas Cv 4_mas Cv 5_mas Cv 6_mas 718. 7828911237. 56228 $ Cp 1_mas Cp 2_mas Cp 3_mas Cp 4_mas Cp 5_mas Cp 6_mas 1007. 00058 1606. 1117 *INITIAL_GAS_MIXTURE $ SID STYPE MMGID T 0 4 1 2 298. 15 $ RHO 1 RHO 2 RHO 3 RHO 4 RHO 5 RHO 6 1. 17913 E-9 *HOURGLASS $ HGID IHQ QM IBQ Q 1 Q 2 5 1 1. 00 e-05 gas 3 prop ADPOPT 0 TMID Cv 7_mas Cv 8_mas Cp 7_mas Cp 8_mas Ref. Condition information RHO 7 RHO 8 QB QW Livermore Software Technology Corporation 33

Fluid-Structure Interaction Modeling with LS-DYNA Using ALE Coupling to Model Airbag Inflation Process H

Fluid-Structure Interaction Modeling with LS-DYNA Using ALE Coupling to Model Airbag Inflation Process H 6 = inflator gas CV inlet “data” & Flow direction may be assigned for each of the point source Inflator gas shares the *MAT_GAS_MIXTURE A 1 st estimate of the inlet velocity may be half of the sound speed *PART H 6 = ALE point sources (no initial mesh) $ PID SECID MID EOSID 6 6 5 0 *HOURGLASS $ HGID IHQ QM IBQ 6 1 1. 00 e-05 *SECTION_POINT_SOURCE_MIXTURE $ SECID LCIDT NOTUSED LCIDVEL 6 1011 0 1022 $ LCMDOT 1 LCMDOT 2 LCMDOT 3 LCMDOT 4 0 1033 $ NODEID VECID ORIFAREA 4311 3. 141600. . . 4365 17 3. 141600 *DEFINE_VECTOR $ VID XT YT ZT 4311 0. 0 20. 0. . . 17 0. 0 20. 0 HGID 6 GRAV 0 ADPOPT 0 TMID Q 1 Q 2 QB QW NIDLC 001 NIDLC 002 NIDLC 003 LCMDOT 5 LCMDOT 6 LCMDOT 7 LCMDOT 8 Orifice areas for each point source XH 1. 0 YH 0. 0 ZH 20. 0 0. 809000 -0. 588000 20. 0 Inflator gas is the 2 nd gas in the mixture. The LCMODT 2 definitions corresponds to the Cv 2_mas & Cp 2_mas under MID 5, *MAT_GAS_MIXTURE card Livermore Software Technology Corporation 34

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING 1. ALE CONTROL

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING 1. ALE CONTROL & INTERFACE TRACKING & VOLUME FILLING 2. METH=1 1 st order advection method is used. 3. PREF= 1. 01325 e-4 ambient pressure of 1 atm is defined. 4. Two physical fluid interfaces are track: AMMG 1 (H 4)=outside air; 5. AMMG 2 (H 5&H 6)=inside gases. *CONTROL_ALE 6. Volume fraction filling is used to fillbfac the space inside the tank with $ dct nadv meth afac cfac dfac efac 0 -1. 000000 0. 0 7. $ AMMG 2 end– 1 thisaafac is 1 done as a vlimit preprocessing step. start vfact ebc pref nsidebc 0. 0 01. 01325 e-4 *SET_PART_LIST 56 5 6 *ALE_MULTI-MATERIAL_GROUP $ SID IDTYPE 4 1 56 0 st *INITIAL_VOLUME_FRACTION_GEOMETRY $FPID/PSID FIDTYPE INIAMMGID 4 1 1 $ CONTTYPE FILLOPT FFLUIDID : CONTTYP=1=PSID or PID = fill inside S 1 w/AMMG 2 nd 1 0 2 $ SETID SETTYPE NORMDIR : SETTYPE=0=PSID or 1=PID 1 1 1 , fill mesh H 4 with AMMG 1 2 , fill inside of tank with AMMG 2 Livermore Software Technology Corporation 35

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING CONTROL: There are

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING CONTROL: There are 2 aspects of coupling: 1. Search for spatial overlap(s) between S 1 and H 4 geometry/meshes. 2. Couple S 1 to AMMG 2, inside physical fluids. The 1 st searches for geometric intersections between Lagrangian surfaces and ALE element space. The 2 nd defines which “physical fluid” to be coupled to the chosen Lagrangian surface. $ initial gas inside the tank and the inflator gas injected (no init mesh) AMMG 2 Use load curve for penalty factor *SET_MULTI-MATERIAL_GROUP_LIST 2 3 X 3 coupling points/Lagr. . segment 2 $ SSTYP = SLAVE elm type: 0=PSID=part-set-id; 1=PID=part-id; 2=segm-set-id *CONSTRAINED_LAGRANGE_IN_SOLID $ SLAVE MASTER SSTYP MSTYP NQUAD CTYPE DIREC MCOUP 1 4 1 1 3 4 2 -2 $ START END PFAC FRIC FRCMIN NORM ISEGNORM XDAMP 0 0 -4 0 0. 3 0 $ CQ HMIN HMAX ILEAKSTIFF VLK_PLCID IVENT IBLOCK Couple 0 0 0 2 0 0 $ pressure vs. penetration (mm) for penalty stiffness calculation when *DEFINE_CURVE compressed 4 0. 000 E-00 0. 000 E+00 0. 00000 Penalty coupling 1. 0000 4. 0 e-04 Livermore Software Technology Corporation Leakage control=type 2 36

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING TANK PRESSURE RESULTs:

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITH COUPLING TANK PRESSURE RESULTs: The pressure inside the tank ALE_mgm CV DATA ALE_ideal_gas NOTE: ALE result with MGM&IGM coincides with ALE result with MN & EIG Livermore Software Technology Corporation 37

Fluid-Structure Interaction Modeling with LS-DYNA [IIc] ALE TANK TEST MODELING WITHOUT COUPLING Livermore Software

Fluid-Structure Interaction Modeling with LS-DYNA [IIc] ALE TANK TEST MODELING WITHOUT COUPLING Livermore Software Technology Corporation 38

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITHOUT COUPLING TO ELIMINATE COUPLING

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITHOUT COUPLING TO ELIMINATE COUPLING EFFECTS A simple cylindrical model is used (vol~100 lit~0. 1 E 9 mm 3). An ALE mesh is defined to model the tank space. A dummy inflator ring is created just for locating the point sources for ALE application. The tank is modeled by an ALE mesh. H 2 Free- surface nodes are constrained from any motion to simulate tank wall condition no coupling Cut-off view A dummy inflator ring may be defined for locating the point sources. Livermore Software Technology Corporation 39

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITHOUT COUPLING KEYWORD DEFINITION :

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITHOUT COUPLING KEYWORD DEFINITION : Global nodal constraints are defined via *BOUNDARY_SPC_SET: NSID 11 = lateral face of cylinder = fixed X-Y NSID 12 = bottom face = fixed Z NSID 13 = top face = fixed Z NSID 14 = top and bottom rims = fixed X-Y-Z *BOUNDARY_SPC_SET $ NID/NSID CID 11 0 12 0 13 0 14 0 DOFX 1 0 0 1 DOFY 1 0 0 1 DOFZ 0 1 1 1 NSID 13 NSID 11 DOFRX 0 0 DOFRY 0 0 DOFRZ 0 0 NSID 14 NSID 12 Livermore Software Technology Corporation 40

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITHOUT COUPLING KEYWORD DEFINITION :

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITHOUT COUPLING KEYWORD DEFINITION : An ALE mesh spanning the tank volume/space is needed (H 2=AMMG 1 = background resident gas 1). *MAT_GAS_MIXTURE defines the Cv & Cp of the ideal gases to be mixed. *INITIAL_GAS_MIXTURE (IGM) allows this mesh to be filled with gas 1 (by specifying RHO 1). H 2: background air mesh Defined for the 2 gases that will mix together. NOTE that Cv & Cp have per-mass unit here! Unlike the permole unit defined in an analogous *AIRBAG_HYBRID card. *PART H 2 = background mesh = initial gas inside the tank $ PID SECID MID EOSID HGID GRAV ADPOPT TMID 2 2 2 0 0 *SECTION_SOLID multi-mat 2 11 0 *MAT_GAS_MIXTURE $ MID IECONSFLG GC 2 0 0 $ Cv 1_mas Cv 2_mas Cv 3_mas Cv 4_mas Cv 5_mas Cv 6_mas Cv 7_mas Cv 8_mas 718. 7828911237. 56228 $ Cp 1_mas Cp 2_mas Cp 3_mas Cp 4_mas Cp 5_mas Cp 6_mas Cp 7_mas Cp 8_mas 1007. 00058 1606. 1117 $ the rho 1 -rho 8 below refer to the gases 1 -8 define in MAT_GAS_MIXT_ID=2 *HOURGLASS $ HGID IHQ QM IBQ Q 1 Q 2 QB QW 2 1 1. 00 e-05 *INITIAL_GAS_MIXTURE This card allows MMGID 1 to flow into part/mesh 2 $ SID STYPE MMGID T 0 2 1 1 298. 15 $ RHO 1 RHO 2 RHO 3 RHO 4 RHO 5 RHO 6 RHO 7 RHO 8 1. 17913 E-9 Livermore Software Technology Corporation 41

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITHOUT COUPLING KEYWORD DEFINITION :

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITHOUT COUPLING KEYWORD DEFINITION : Injected gas (H 6) Cv & Cp shown in *MAT_GAS_MIXTURE above. *SECTION_POINT_SOURCE_MIXTURE defines the inflating potential: & curves are from CV model. curve may be estimated. The point sources have their NIDs, Vector IDs & Areas defined. This gas mixes with resident gas 1 (thus same *MAT_ card). *PART Using the same *MAT_GAS_ card as H 2 2 nd material. H 6 = ALE point sources $ PID SECID MID 6 6 2 *HOURGLASS $ HGID IHQ QM 6 1 1. 00 e-05 *SECTION_POINT_SOURCE_MIXTURE $ SECID LCIDT NOTUSED 6 1011 0 $ LCMDOT 01 LCMDOT 02 LCMDOT 03 0 1033 $ NID VID AREA 4311 3. 141600. . . *DEFINE_VECTOR $ VID XT YT 4311 0. 0. . . H 6: Inflator gas 2 EOSID 0 HGID 6 GRAV 0 ADPOPT 0 TMID IBQ Q 1 Q 2 QB QW LCIDVEL NIDLCOOR 1 NIDLCOOR 2 NIDLCOOR 3 1022 LCMDOT 04 LCMDOT 05 LCMDOT 06 LCMDOT 07 LCMDOT 08 ZT 20. 0 XH 1. 0 YH 0. 0 ZH 20. 0 Livermore Software Technology Corporation 42

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITHOUT COUPLING KEYWORD DEFINITION :

Fluid-Structure Interaction Modeling with LS-DYNA ALE TANK TEST MODELING WITHOUT COUPLING KEYWORD DEFINITION : Global ALE control, *CONTROL_ALE, defines 1 st order advection & an ambient p (pref=1 atm). *ALE_MULTI-MATERIAL_GROUP defines the material interfaces to be tracked: H 2 = AMMG 1; H 6 = AMMG 2. Ambient pressure Advection method *CONTROL_ALE $ dct nadv meth afac 0 1 1 -1. 000000 $ start end aafac vfact 0. 0 *ALE_MULTI-MATERIAL_GROUP $ SID IDTYPE 2 1 6 1 bfac 0. 0 vlimit 0. 0 cfac dfac 0. 0 ebc pref 01. 01325 e-4 efac 0. 0 nsidebc Since the fluid mesh “is” the tank space and the surface nodes are constrained, there is no coupling required. Livermore Software Technology Corporation 43

Fluid-Structure Interaction Modeling with LS-DYNA [IIIa] CONTROL VOLUME FLAT AIRBAG MODEL Livermore Software Technology

Fluid-Structure Interaction Modeling with LS-DYNA [IIIa] CONTROL VOLUME FLAT AIRBAG MODEL Livermore Software Technology Corporation 44

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL SUMMARY: Lagrangian model requires

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL SUMMARY: Lagrangian model requires simply: the Lagangian structures (airbag shell parts) and the inflating potential information for the CV method. The inlet data allows for an estimation of a uniform pressure field inside the bag via a lumped-parameter control-volume analysis. This same pressure is applied on all bag segments causing inflation. The input files are: demo 8 cv 1. k = Main control volume (CV) input file demo 8 cvhybridporos 2. k = *AIRBAG_HYBRID CV definitions mdot(t)=LCID 1; Tstag(t)=LCID 2 The parts are: S 1 S 2 S 3 S 4 S 5 S 6 S 9 = = = = Lagrangian Lagrangian shell shell bottom half of upper half of inflator rigid airbag tether airbag 2 vent airbag rim box patches Livermore Software Technology Corporation 45

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL DEPLOYMENT: CONTROL VOLUME Lagangian

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL DEPLOYMENT: CONTROL VOLUME Lagangian airbag deployment sequence Uniform Inflation Across the airbag NOTE: CV method applies uniform P over all Lagrangian surfaces The bag can open up a little too early. Livermore Software Technology Corporation 46

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL GEOMETRY: Blanking out the

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL GEOMETRY: Blanking out the top half of an inflated airbag shell structure This is a very simple representation of a typical airbag structure. An enclosed volume is inflated. S 1 = airbag lower half. S 2 = airbag upper half (blanked out) S 9 = 2 vent patches Backing platform (created by the *RIGIDWALL_PLANAR card) Livermore Software Technology Corporation S 5 & S 6 = 2 tethers restraining the top portion of the airbag Zoom-in in a later slide … 47

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL AIRBAG STRUCTURE SET-UP, *MAT_FABRIC

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL AIRBAG STRUCTURE SET-UP, *MAT_FABRIC The Lagrangian shell parts making up the airbag structure consist of PID’s: 1, 2, 5, 6, 9. For each of these parts, the definition consists of 3 cards typically looking like the following: Shear correction factor Shear modulus Material Axes option Vector for AOPT=3 *PART material type # 34 (fabric) $ PID SECID MID 1 1 1 *SECTION_SHELL $ SID ELFORM SHRF 1 5 0. 0000000 $ T 1 T 2 T 3 4. 00000 -4 $ B 1 B 2 B 3 0. 0 *MAT_FABRIC $ MID RO EA 1 8. 76000 -7. 3000000 $ GAB GBC GCA. 0400000 $ AOPT FLC FAC 3. 0000000 1. 000 -17420 $. 0000000 $ V 1 V 2 V 3 1. 0000000 # of integration points EOSID 0 HGID 0 GRAV 0 ADPOPT 0 NIP PROPT QR/IRID 4. 0000000 0. 0000000 T 4 NLOC 4. 00000 -4 0. 0000000 B 4 B 5 B 6 0. 0 ICOMP 1 Young modulus EB. 2000000 CSE 1. 0000000 ELA EC. 3000000 EL. 0600000 LNRC PRBA. 2000000 PRL. 3500000 FORM A 1 1. 0000000 D 1. 0000000 A 2. 0000000 D 2. 0000000 A 3. 0000000 D 3. 0000000 TMID 0 B 7 B 8 Poisson ratio PRCA. 2000000 LRATIO. 1000000 PRCB. 2000000 DAMP. 2000000 Material angles at integration points Porosity LCID (overwritten by *AIRBAG_HYBRID parameters) BETA Livermore Software Technology Corporation 48

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL *MAT_FABRIC POROSITY DEFINITION FLC

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL *MAT_FABRIC POROSITY DEFINITION FLC FAC = CV porous flow coefficient applies to porous fabric (If. LT. 0 defines a LCID for FLC as a function of time). = CV porous fabric area coefficient (If. LT. 0 defines a LCID for FAC versus absolute pressure). Different Applications of FLC & FAC: 1) CV method: These *MAT_FABRIC definitions can be overwritten by the POROUS flow definitions under *AIRBAG_HYBRID card (CP 23, LCCP 23, …) 2) TRADITIONAL ALE method: These *MAT_ FABRIC porous flow definitions are NOT used. 3) ALE method with *AIRBAG_ALE : These *MAT_ FABRIC definitions may be used to define POROUS flow definitions if the *AIRBAG_ALE card is used. Livermore Software Technology Corporation 49

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL INTERNAL INFLATOR GEOMETRY: Airbag

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL INTERNAL INFLATOR GEOMETRY: Airbag internal support structures (tethers) made up of 2 shell parts. S 5 S 6 Rigid inflator box rim (fixed) Rigid inflator box containing priming gas block (fixed) The tether interaction with the gas and the airbag needs careful consideration for precision modeling. S 3 S 4 Contact among the Lagrangian shell structures: *CONTACT_AIRBAG_SINGLE_SURFACE, Slave=PSID 1: 1 PID’s S 1, S 2, S 3, S 4, S 5, S 6, S 9; Master=None (selfcontact). Livermore Software Technology Corporation 50

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL INFLATOR BOX SET-UP: The

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL INFLATOR BOX SET-UP: The inflator box is made up of 2 rigid material parts (S 3 & S 4) at the bottom of and connected to the airbag. Shear correction factor Shell thickness @ its 4 nodes Center-of-mass constraint = This fixes the inflator box in the global coordinates. # of integration points across thickness *PART rigid inflator box rim : thick = 0. 4 mm 3 3 3 0 0 0 *SECTION_SHELL $ SID ELFORM SHRF NIP PROPT QR/IRID ICOMP 3 5 0. 0000000 4. 0000000 0. 0000000 1 $ T 1 T 2 T 3 T 4 NLOC 0. 4000000 0. 0000000 $ B 1 B 2 B 3 B 4 B 5 B 6 B 7 B 8 0. 0 *MAT_RIGID $ MID RO E PR N COUPLE M 3 7. 8500 -06 2. 0000000 0. 3000000 0. 0000000 $ CMO CON 1 CON 2 1. 0000000 7. 0000000 $ A 1 A 2 A 3 V 1 V 2 V 3 0. 0000000 $======================================== *PART rigid inflator box side walls and bottom: thick = 0. 4 mm $ PID SECID MID EOSID HGID GRAV ADPOPT TMID 4 3 3 0 0 0 Livermore Software Technology Corporation 51

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL CV METHOD - *AIRBAG_HYBRID

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL CV METHOD - *AIRBAG_HYBRID The inflator gas energy input causes a uniform P to inflate the bag. Ø and curves for the inlet gas are provided. Ø The heat capacity, Cp (per-mole unit), and molecular weight, MW, for each of the 2 gases must be defined. Ø Typically the atmospheric T, P & rho of air define the background. Ambient background air 2 lines defined per gas Gas 1 Gas 2 *AIRBAG_HYBRID $ sidtyp rbid vsca psca vini mwd spsf 2 1 0 0. 0000000 $ atmost atmosp atmosd gc cc 294. 40000 1. 01300 -4 1. 19950 -9 8. 3140+0 1 0. 0000000 $ c 23 lcc 23 a 23 lca 23 cp 23 lcp 23 ap 23 lcap 23 0. 0000000 0 $ opt pvent ngas 7 0 2 2 gases will mix in the tank $ lcidm lcidt not used mw initm a b c 0 0 0. 0000000 0. 028970 1. 0 29. 100000 0. 0000000 $ fmass 0. 0 $ lcidm lcidt not used mw initm a b c 1 2 0. 0000000 0. 023500 0. 0 28. 00000 0. 000000 $ fmass 0. 0 Given mass flow rate and gas temperature curves Given Livermore Software Technology Corporation INFLATING POTENTIAL 52

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL CV METHOD - *AIRBAG_HYBRID

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL CV METHOD - *AIRBAG_HYBRID VENTING & POROSITY: VENTING (one or more physical holes in the fabric for gas bleeding) C 23 = vent orifice coefficient applies to vent hole (set this to 0 if LCC 23 is defined). LCC 23 = load curve of C 23(t) as a function of time. A 23 = vent orifice areas (set this to 0 if LCA 23 is defined). LCA 23 = load curve of A 23 area as a function of absolute pressure. POROSITY can overwrite the definitions in *MAT_FABRIC card. CP 23 = porous flow orifice coefficient applies to porous fabric (set this to 0 if LCCP 23 is defined). LCCP 23 = load curve of CP 23(t) as a function of time. AP 23 = porous fabric area - if defined, it will overrides LCPA 23. (set this to 0 if LCPA 23 is defined). LCPA 23 = load curve of AP 23 area as a function of absolute pressure. Livermore Software Technology Corporation 53

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL CV METHOD - POROSITY:

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL CV METHOD - POROSITY: POROSITY alternate definitions via the *MAT_FABRIC card. If in the *AIRBAG_HYBRID card we define: OPT = 7 = fabric porosity is ON without BLOCKAGE OPT = 8 = fabric porosity is ON with BLOCKAGE Then in the *MAT_FABRIC card we must define: FLC = If porous flow fabric leakage coefficient If < 0 a LCID of fabric porous flow coefficient VS. t FAC = If porous flow fabric area coefficient If < 0 a LCID of fabric porous area coefficient VS. P (P = absolute upstream pressure) ELA = If effective leakage area for blocked fabric If < 0 a LCID of fabric Livermore Software Technology Corporation porous area coefficient VS. t 54

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL AIRBAG SELF-CONTACT DEFINITION: Typically,

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL AIRBAG SELF-CONTACT DEFINITION: Typically, airbag self-contact is modeled by considering all parts that interact (sometimes including the inflator box and tethers). For airbag *CONTACT_AIRBAG_SINGLE_SURFACE is usually used with: SBOPT = 2 (or 3 for warped segment checking) DEPTH = 3 (Complex folded airbag contact will require more sophisticated options) $ self contact for PID 1, 2, 3, 4, 5, 6, 9: all airbag parts $ SSTYP= SLAVE set types: 0=SGSID ; 1=SHSID ; 2=PSID ; 3=PID ; 4=NSID ; 6=PSID *CONTACT_AIRBAG_SINGLE_SURFACE $ SSID MSID SSTYP MSTYP SBOXID MBOXID SPR MPR 1 0 2 0 0 0 $ FS FD DC VC VDC PENCHK BT DT 0. 5000000 0. 0000000 0 0. 000000 $ SFS SFM SST MST SFMT FSF VSF 0. 0000000 0. 5000000 0. 0000000 $ SOFT SOFSCL LCIDAB MAXPAR SBOPT DEPTH BSORT FRCFRQ 2 0. 0000000 3 5 0 *SET_PART_LIST 1 1 2 3 4 5 6 9 Livermore Software Technology Corporation 55

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL CONTROL VOLUME Lagrangian airbag

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL CONTROL VOLUME Lagrangian airbag deployment movie. Note the uniform upward inflation from center to edge of airbag. Livermore Software Technology Corporation 56

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL DIGRESSION: CV MULTIPLE-CHAMBERS AIRBAG

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL DIGRESSION: CV MULTIPLE-CHAMBERS AIRBAG Consider having two connecting airbags which are inflated by different sources of gases We want to allow “interactions” or flow to pass back and forth between the 2 bags. Gas 3 Gas 4 Gas 5 Bag 1 air 1 Gas 6 Bag 2 air 2 Each *AIRBAG_HYBRID card must define the maximum number of gases present in the whole system (6 gases total) The inflation of each bag must be defined by an *AIRBAG_ card (*AIRBAG_HYBRID_ID is shown as an example) and - The *AIRBAG_INTERACTION card is defined to allow flows to go back and forth between the 2 airbags. The “*AIRBAG_#_ID” option must Livermore Software Technology Corporation be used Example … 57

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL AIRBAG 1: Background air

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL AIRBAG 1: Background air 1 for bag 1 Background air 2 input as dummy Gases 3 & 4 flow into bag 1 Gases 5 & 6 are input as dummy *AIRBAG_HYBRID_ID 1 $ sidtyp 1 1 $ atmost atmosp 294. 40000 1. 01300 -4 $ c 23 lcc 23 0. 0000000 0 $ opt pvent 7 0 $ air 1 lcm lcidt 0 0 $ fmass 0. 0 $ air 2 lcm lcidt 0 0 $ fmass 0. 0 $ gas 3 lcm lcidt 3 31 $ fmass 0. 0 $ gas 4 lcm lcidt 4 41 $ fmass 0. 0 $ gas 5 lcm lcidt 0 0 $ fmass 0. 0 $ gas 6 lcm lcidt 0 0 $ fmass 0. 0 rbid vsca psca vini mwd spsf 0 0. 0000000 atmosd gc cc 1. 19950 -9 8. 3140+0 1 0. 0000000 a 23 lca 23 cp 23 lcp 23 ap 23 lcap 23 0. 0000000 0 ngas 6 not used mw initm a b c 0. 0000000 0. 028970 1. 0 29. 100000 0. 0000000 not used 0. 0000000 mw 0. 028970 initm a b c 0. 0 29. 100000 0. 0000000 not used 0. 0000000 mw 0. 023500 initm 0. 0 a 28. 00000 b 0. 000000 c 0. 000000 not used 0. 0000000 mw 0. 028970 initm a b c 0. 0 29. 100000 0. 0000000 Livermore Software Technology Corporation 58

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL AIRBAG 2: Background air

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL AIRBAG 2: Background air 1 input as dummy Background air 2 for bag 2 Gases 3 & 4 are input as dummy Gases 5 & 6 flow into bag 2 *AIRBAG_HYBRID_ID 2 $ sidtyp 2 1 $ atmost atmosp 294. 40000 1. 01300 -4 $ c 23 lcc 23 0. 0000000 0 $ opt pvent 7 0 $ air 1 lcm lcidt 0 0 $ fmass 0. 0 $ air 2 lcm lcidt 0 0 $ fmass 0. 0 $ gas 3 lcm lcidt 0 0 $ fmass 0. 0 $ gas 4 lcm lcidt 0 0 $ fmass 0. 0 $ gas 5 lcm lcidt 5 51 $ fmass 0. 0 $ gas 6 lcm lcidt 6 61 $ fmass 0. 0 rbid vsca psca vini mwd spsf 0 0. 0000000 atmosd gc cc 1. 19950 -9 8. 3140+0 1 0. 0000000 a 23 lca 23 cp 23 lcp 23 ap 23 lcap 23 0. 0000000 0 ngas 6 not used mw initm a b c 0. 0000000 0. 028970 0. 0 29. 100000 0. 0000000 not used 0. 0000000 mw 0. 028970 not used mw 0. 0000000 0. 023500 initm a b c 1. 0 29. 100000 0. 0000000 initm 0. 0 a 28. 00000 b 0. 000000 c 0. 000000 not used 0. 0000000 mw 0. 023500 initm 0. 0 a 28. 00000 b 0. 000000 c 0. 000000 not used 0. 0000000 mw 0. 028970 initm a b c 0. 0 29. 100000 0. 0000000 Livermore Software Technology Corporation 59

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL AIRBAG 1 – AIRBAG

Fluid-Structure Interaction Modeling with LS-DYNA CONTROL VOLUME FLAT AIRBAG MODEL AIRBAG 1 – AIRBAG 2 interaction: Two connecting airbags defined previously now can interact via this card. (Notice how the total number of gases in the system must be defined on each *AIRBAG card). AB 1 = 1 st airbag ID from 1 st “*AIRBAG_*_ID” card. AB 2 = 2 nd airbag ID from 2 nd “*AIRBAG_*_ID” card. AREA = Communication area between the 2 bags. SF = Flow scale factor between the 2 bags. PID = PID defining the area between the 2 bag. mdot = mdot(d. P) LCID = flow rate as a function of d. P. If this is defined, AREA, SF &PID are ignored. IFLOW = Flow direction allowed. *AIRBAG_INTERACTION $ AB 1 AB 2 2 1 AREA 1. 5 SF 0. 0 PID 0 mdot_LCID 0 IFLOW 0. 000 Livermore Software Technology Corporation 60

Fluid-Structure Interaction Modeling with LS-DYNA [IIIb] ALE FSI FLAT AIRBAG MODEL Livermore Software Technology

Fluid-Structure Interaction Modeling with LS-DYNA [IIIb] ALE FSI FLAT AIRBAG MODEL Livermore Software Technology Corporation 61

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL SUMMARY: In general, an

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL SUMMARY: In general, an ALE model requires the same input for the airbag as a Lagrangian model, except: - no *AIRBAG_ card. - with ALE part definitions. - with Lagrangian-to-ALE coupling definition. The input files are: demo 8 flat. k = Main ALE model input file demo 8_ale. k = ALE part and coupling definitions mdot(t)=LCID 1; Tstag(t)=LCID 2 The parts are: S 1 & S 2 S 9 S 3 & S 4 S 5 & S 6 S 100000 S 100001 H 50000 H 50003 = = = = main airbag bot&top : thick=0. 4 mm SECID=1 MID=1 airbag vents : thick=0. 4 mm SECID=1 MID=1 rigid inflator can rim&base : thick=0. 4 mm SECID=3 MID=3 airbag tether 1&2 : thick=2. 0 mm SECID=5 MID=1 rigid dummy inflator ring defining orifices and vectors dummy rigid wall at bottom of airbag (by *RIGIDWALL_PLANAR). ALE surrounding air block = initial mesh. ALE inflator gas (*SECTION_POINT_SOURCE_MIXTURE vectors. ) Livermore Software Technology Corporation 62

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL SUMMARY: [Fluids=Air and gas

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL SUMMARY: [Fluids=Air and gas parts] + [Structures = Airbag and inflator parts]: Air mesh covering the space which the airbag will expand into. (H 50000) Fixed backing platforms defined by Airbag (PIDs: S 1, S 2, S 5, S 6, S 9) *RIGIDWALL_PLANAR Blanking out the outside of the bag for internal view … Livermore Software Technology Corporation 63

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL PHYSICAL SET-UP • One

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL PHYSICAL SET-UP • One ALE air mesh (H 50000) define the fluid initial mesh. • H 50000 air mesh provides room for the bag to expand into. • One ALE inflator gas (H 50003=no initial mesh). • The Lagrangian shell structures make up the airbag. • A rigid wall to limit horizontal motion of the bag. Deployed Airbag: Lower portion=S 1, Upper portion=S 2, 2 Tethers= S 5, S 6 Air mesh covering the space which the airbag will expand into (PID=H 50000). Vent holes (PID=S 9) Inflow Inflator gas (PID=H 50003). The bottom of the fluid mesh should have at least 2 or 3 layers of ALE elements beyond the Lagrangian structure Livermore Software Technology Corporation 64

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL DETAILED BOTTOM VIEW: Top

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL DETAILED BOTTOM VIEW: Top of airbag (S 2) Bottom of airbag (S 1) Air block (H 50000) H 50000 Vent holes (S 9) Rigid inflator rim (S 3) Airbag compartment opening Tethers (S 5 & S 6) Dummy orifice locator ring, dynamically inactive. (S 100000) Rigid inflator box (S 4) Livermore Software Technology Corporation 65

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL ALE AIR PART DEFINITION:

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL ALE AIR PART DEFINITION: Air at 298. 15 K and 1 atm pressure is defined via the *EOS_ card. This is the only fluid part that initially has a mesh defined. There are more than one way to define the air properties. [Method 1] *PART H 50000 = ALE background air block with initial mesh, ave T~289. 15 K $ PID SECID MID EOSID HGID GRAV ADPOPT 50000 50000 0 0 *SECTION_SOLID 50000 11 *MAT_NULL $ MID RHO PC MU TEROD CEROD YM 50000 1. 2906 E-9 -1. 0 E-05 0. 0 *EOS_IDEAL_GAS $ EOSID Cv Cp C 1 C 2 T 0 V 0 50000 718. 7829 1007. 0006 0. 0 298. 15 1. 0 *HOURGLASS $ HGID IHQ QM IBQ Q 1 Q 2 QB 50000 1 1. 00 e-04 TMID Reference rho PR Initial T Thermal properties QW or. . . Livermore Software Technology Corporation 66

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL ALE AIR PART DEFINITION

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL ALE AIR PART DEFINITION (cont. ): [Method 2] Even if this part contains only air “initially”, it can be defined with a *MAT_GAS_MIXTURE card. However, with that card, the user must also “initialize” or specify which AMMG can exist in this PID or mesh This requires a *INITIAL_GAS_MIXTURE card *PART H 50000 = ALE background air block with initial mesh, ave T~289. 15 K $ PID SECID MID EOSID HGID GRAV ADPOPT 50000 50000 0 0 *SECTION_SOLID 50000 11 *MAT_GAS_MIXTURE 50000 $ CV 1 CV 2 CV 3 CV 4 CV 5 CV 6 CV 7 718. 78289 $ CP 1 CP 2 CP 3 CP 4 CP 5 CP 6 CP 7 1007. 00058 *INITIAL_GAS_MIXTURE $ SID STYPE AMMGID T 0 STYPE: 0=PSID, 1=PID 50000 1 1 298. 15 $ RHO 1 RHO 2 RHO 3 RHO 4 RHO 5 RHO 6 RHO 7 0 1. 1504 E-9 *HOURGLASS $ HGID IHQ QM IBQ Q 1 Q 2 QB 50000 1 1. 00 e-04 Thermal properties AMMG 1 Allow this PID or mesh to contain AMMG 1 Initial T TMID CV 8 CP 8 RHO 8 QW Reference rho Livermore Software Technology Corporation 67

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL ALE INFLATOR GAS DEFINITION:

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL ALE INFLATOR GAS DEFINITION: Inflator gas (no initial mesh) with properties at an average T ~ 650 K Inflator gas point sources: - location = NID - direction = VID - orifice area Thermal properties *PART H 50003 = ALE inflator gas from point sources, @ ave T~650 K $ PID SECID MID EOSID HGID GRAV ADPOPT TMID 50003 0 0 *SECTION_POINT_SOURCE_MIXTURE $ SECID LCIDT NOTUSED LCIDVEL NIDLCOOR 1 NIDLCOOR 2 NIDLCOOR 3 50003 2 0 3 $ LCMDOT 01 LCMDOT 02 LCMDOT 03 LCMDOT 04 LCMDOT 05 LCMDOT 06 LCMDOT 07 LCMDOT 08 1 $ NID VID AREA may be initially estimated ~ 100019 1 13. 500. . . half of sound speed (more on 100018 8 13. 500 this later) *MAT_GAS_MIXTURE 50003 $ CV 1 CV 2 CV 3 CV 4 CV 5 CV 6 CV 7 CV 8 1223. 36946 $ CP 1 CP 2 CP 3 CP 4 CP 5 CP 6 CP 7 CP 8 1591. 91888 *HOURGLASS $ HGID IHQ QM IBQ Q 1 Q 2 QB QW 50003 1 1. 00 e-05 *DEFINE_VECTOR $lco or a 1 a 2 a 3 v 1 v 2 v 3 1 0. 0 -24. 50000 21. 21320 -24. 50000. . . 8 0. 0 -24. 500001. 0000 e-06 30. 00000 -24. 50000 Livermore Software Technology Corporation 68

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL NOTES ON *SECTION_POINT_SOURCE_MIXTURE CARD:

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL NOTES ON *SECTION_POINT_SOURCE_MIXTURE CARD: *SECTION_POINT_SOURCE_MIXTURE (SPSM) card basically defines the information about the inflator gas being injected into the airbag. It defines: [1] The total inflow information for all point sources: (a) information from CV input: & (b) estimated inlet gas flow velocity [2] Geometrical information for each point sources: (a) NID = node id defining location of the point (b) VID = vector ID defining the direction of flow (c) AREA = orifice area at each point source In some cases, gas speed may be roughly estimated ~ 100 -200 m/s. Our code conserves energy input via enthalpy thus speed is not critical (used for momentum accounting): Livermore Software Technology Corporation 69

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL ALE GLOBAL CONTROL: 1.

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL ALE GLOBAL CONTROL: 1. - Some 1 st-pass settings for global ALE computation controls under *CONTROL_ALE: (DCT is used for 2 D only DCT=0) 2. METH = 1 st order advection 3. NADV = 1 = Perform 1 ALE calculation per 1 Lagrangian step 4. PAMB = ambient pressure surrounding the global model 5. - *ALE_MULTI-MATERIAL_GROUP defines the fluid interface 6. reconstructions for each physical fluid: 7. H 50003 = 1 st ALE multi-material group (AMMG 1=gas)= 1 st line 8. H 50000 = 2 nd ALE multi-material group (AMMG 2=air)= 2 nd line *CONTROL_ALE $ DCT NADV METH 0 1 1 $ START END AAFAC 0. 0 *ALE_MULTI-MATERIAL_GROUP $ SID IDTYPE 50003 1 50000 1 AFAC -1. 0 VFACT 0. 0 BFAC 0. 0 PEQUIL 0 CFAC DFAC 0. 0 EBC PAMBIENT 01. 01325 e-4 EFAC EBCXSID Livermore Software Technology Corporation 70

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL FSI COUPLING DEFINITION: -

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL FSI COUPLING DEFINITION: - Use *SET_MULTI-MATERIAL_GROUP_LIST to select which AMMG or ALE fluid to couple to the Lagrangian structures under *CONSTRAINED_LAGRANGE_IN_SOLID (CLIS) CLIS card: SETID = 234 points to the 1 st AMMG = inflator gas *SET_MULTI-MATERIAL_GROUP_LIST 234 1 $ 9 are vent holes <== couple with the vent holes too (mesh too coarse) *SET_PART_LIST 123 1 2 3 4 5 6 9 - Select the Lagrangian parts for coupling with the ALE fluid: SETID = 123 points to Lagrangian shell parts 1 -6, 9 Livermore Software Technology Corporation 71

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL SOME NOTES ON LAGRANGIAN

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL SOME NOTES ON LAGRANGIAN PARTS SELECTION: - The Lagrangian shell structure simulating the airbag undergoing selfcontact is modeled with a *CONTACT card. The part-set for contact should contain all fabric parts and may include the compartment. Vent-hole patch may be omitted from this set, i. e. S 9 in this case. However, including S 9 probably will have minimal effect. This model shows 1 simple part-selection choice. The user will have to make her own judgment as to which fabric parts to include. - Lagrangian parts selected for FSI coupling with the ALE fluid, in the *CONSTRAINED_LAGRANGE_IN_SOLID (CLIS) CLIS card, should include all shell parts in touch with the inflator gas, including the Livermore Software Technology Corporation compartment. If using purely traditional FSI for venting 72

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL CLIS FSI COUPLING DEFINITION:

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL CLIS FSI COUPLING DEFINITION: Consider 2 basic coupling considerations: (1) Spatial search for overlaps between Lagrangian and ALE elements. This is defined by the SLAVE and MASTER SETIDs in CLIS. (2) Explicitly specifying which “ALE fluid” to couple to via MCOUP Typical coupling parameter settings for airbag FSI: NQUAD = 2 if Lagrangian and ALE element sizes are similar (if not, refer to the ALE tutorial for more details). CTYPE = 4 for flat bag, and 6 for complexly folded bag. DIREC = 2 (if CTYPE=4). DIREC is dictated by CTYPE=6. *CONSTRAINED_LAGRANGE_IN_SOLID ILEAK =SLAVE 0 = turn off leakage control. NQUAD set CTYPE to 1 if leaks, then 2. $ MASTER SSTYP MSTYP DIREC MCOUP $ $ 123 START 0 CQ 0 50000 END 0 HMIN 0 0 PFAC -9010 HMAX 0 1 FRIC 0 ILEAK 1 2 FRCMIN 0. 3 PLEAK 0 4 NORM 0 LCIDPOR -17420 2 NORMTYP -234 XDAMP NVENT IBLOCK Livermore Software Technology Corporation 73

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL Lagrangian ALE = coupling

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL Lagrangian ALE = coupling points Too many coupling points = BAD! S 1 NQUAD=1 (not enough) H 2 overlap H 3 overlap le leak Possib H 1 age CLIS FSI COUPLING NQUAD: Main idea = to have at least 1 coupling point per each Lagrangian-ALE overlap section. If NQUAD=1 for S 1 versus H 1&H 2 leakage may occur. For S 2 versus H 2&H 3, if NQUAD=2 (2 X 2 coupling points defined per Lagrangian segment), this may be adequate coupling NQUAD depends on the Lagrangian-ALE relative-mesh-resolution & overlap. S 2 NQUAD=2 (OK) Livermore Software Technology Corporation 74

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL CLIS FSI COUPLING PENALTY

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL CLIS FSI COUPLING PENALTY & FABRIC POROUS FLOW: PFAC usage: (1) If then it is used as a coupling-force scale factor (2) If PFAC= -I (I=positive integer), then LCID “I” contains x=fluid penetration and y= absolute coupling pressure. This curve usually can contain 2 points: {0, 0} and {max_penetration_allowed, max_absolute_coupling_pressure_observed}. This Pcoup may be estimated from fluid pressure of an ALE run using default PFAC. $ coupling pressure difference vs. penetration for penalty stiffness calculation *DEFINE_CURVE LCIDPOR = load curve defining relative_porous_fluid_velocity (y) VS. 9010 0. 000 E-00 0. 000 E+00 0. 00000 coupling_pressure_difference (x=Pin – Pout) 1. 0000 4. 0 e-04 $ porous velocity vs. delta_P data curve for porous flow thru fabrics *DEFINE_CURVE 17420 1. 000 E-06 1. 000 E+00 1. 000 0 0. 6895 0. 00000. . . 61. 7085 2. 34932 Livermore Software Technology Corporation 75

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL CLIS FSI – 2

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL CLIS FSI – 2 VENT MODELING APPROACHES: There are two methods for modeling venting with the ALE method: (1) Traditional ALE FSI method in which the “vent holes” are modeled by some Lagrangian shell part(s) but they are not included in CLIS. So they are not coupled to the gas allow the gas to pass through. (2) Specifying the number of vents via NVENT: Each vent requires 1 line of vent geometry definition input (shown below): NVENT = number of venting holes. *CONSTRAINED_LAGRANGE_IN_SOLID $ SLAVE MASTER SSTYP MSTYP NQUAD CTYPE DIREC MCOUP IBLOK = flag to turn ON|OFF blockage 123 50000 0 1 2 4 2 consideration -234 $ START END PFAC FRIC FRCMIN NORMTYP XDAMP for vents. 0 0 -9010 0 0. 3 0 $ CQ 0 $ SID 123 1 *SET_SEGMENT 123 69 HMIN 0 STYPE 2 1 HMAX 0 VENTSF 0. 35 0. 0 70 0. 0 77 ILEAK PLEAK LCIDPOR NVENT 1 0 -17420 2 (STYPE: 0=PSID; =1=PID; =2=SGSID) 0. 0 76 0. 0 IBLOCK 1 0. 0 Livermore Software Technology Corporation 76

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL CLIS FSI – 2

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL CLIS FSI – 2 VENT MODELING APPROACHES: (1) Traditional ALE FSI method relies on CLIS coupling action. It requires fine ALE mesh such that there must be at least 5 ALE elements spanning the width of a Lagrangian vent hole. It may be more accurate, but with higher computational cost. (2) Specifying the number of vents via NVENT: Each vent area is subjected to a pressure difference across it this pressure ratio is *CONSTRAINED_LAGRANGE_IN_SOLID $ SLAVE MASTER SSTYP MSTYP NQUAD CTYPE DIREC MCOUP used flow 2 equation to estimate 123 50000 in an 0 isentropic 1 4 2 -234 the flow. $ START END PFAC FRIC FRCMIN NORMTYP XDAMP A 0 0 -9010 0 0. 3 0 $ CQ HMIN HMAX ILEAK PLEAK LCIDPOR NVENT IBLOCK “discharge may be used with each 1 vent 0 0 0 coefficient” 1 0 -17420 2 $ SID STYPE VENTSF (STYPE: 0=PSID; =1=PID; =2=SGSID) hole defined. 123 2 0. 35 1 *SET_SEGMENT 123 69 1 0. 35 0. 0 70 0. 0 77 0. 0 76 0. 0 0. 0 Livermore Software Technology Corporation 77

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL DEPLOYMENT: Traditional ALE method:

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL DEPLOYMENT: Traditional ALE method: airbag deployment (fluid mesh not shown) Non-uniform Inflation Center gets inflated first Livermore Software Technology Corporation 78

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL Traditional ALE airbag model

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL Traditional ALE airbag model deployment movie (top-side view) with “FRINGE ISO” animation option under LSPREPOST. Livermore Software Technology Corporation 79

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL Traditional ALE airbag model

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL Traditional ALE airbag model deployment movie (cut-off view) Livermore Software Technology Corporation 80

Fluid-Structure Interaction Modeling with LS-DYNA [IIIc] CURRENT DEVELOPMENT ALE FSI *AIRBAG_ALE Livermore Software Technology

Fluid-Structure Interaction Modeling with LS-DYNA [IIIc] CURRENT DEVELOPMENT ALE FSI *AIRBAG_ALE Livermore Software Technology Corporation 81

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE SUMMARY: The *AIRBAG_ALE command was

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE SUMMARY: The *AIRBAG_ALE command was developed to: 1. (1) Simplify the input for ALE airbag modeling. 2. (2) Allow a switch from ALE to CV method to save computation time. 3. (3) Optionally allow an automatic ALE mesh to be generated for a 4. spatial region enveloping the airbag *PART deployment space. *AIRBAG_ALE one card replaces (Internally LSDYNA simply translates the input from *AIRBAG_ALE into that of the ALE, coupling & CV card definitions) Air *SECTION_SOLID *MAT_GAS_MIXTURE *INITIAL_GAS_MIXTURE Gas *PART *SECTION_POINT_SOURCE_MIXTURE *MAT_GAS_MIXTURE Coupling CV *CONSTRAINED_LAGRANGE_IN_SOLID *AIRBAG_HYBRID Livermore Software Technology Corporation 82

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL SUMMARY: This model demonstrates one way

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL SUMMARY: This model demonstrates one way to use of the currently-developing *AIRBAG_ALE command. This command allows the modeling of an airbag deployment in two stages: 1 st with the ALE FSI method, and subsequently with the classical CONTROL VOLUME method. This model is converted directly from the previous ALE model presented (demo 8 flat. k+demo 8_ale. k) into a new format. Compare them! INPUT FILE: demo 8 flataba 2. k = testing *AIRBAG_ALE (Courtesy of Dilip), kg-mm-ms-K MODEL ORGANIZATION (same Lagrangian construction as before): S 1&S 2 = S 9 = S 3&S 4 = S 5&S 6 = S 100000 S 100001 main airbag bot&top : thick=0. 4 mm SECID=1 MID=1 airbag vents : thick=0. 4 mm SECID=1 MID=1 rigid inflator can rim&base : thick=0. 4 mm SECID=3 MID=3 airbag teather 1&2 : thick=2. 0 mm SECID=5 MID=1 = dummy inflator ring defining 8 inflator orifices and vectors = (via *RIGID_PLANAR) rigid dummy platform on bot of airbag. Livermore Software Technology Corporation 83

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL WHAT WE NEED TO KNOW main

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL WHAT WE NEED TO KNOW main features of *AIRBAG_ALE? (2) Allow airbag modeling with ALE and CV methods at different time during the deployment thus allow a switch from ALE to CV method at a user-chosen switch time Need to collect all information to construct the input for the *AIRBAG_HYBRID card and corresponding ALE definitions. no de 1 (3) Optionally allow an automatic ALE mesh to be generated for a spatial region enveloping the airbag deployment space z In this model we only need to define the Lagrangian construction node 3 of the airbag LSDYNA will define the ALE mesh x automatically for the user HOW BIG? node 2 The user y must supply 4 points in space defining a node 0 “box” region enveloping the space spanned Livermore Software Technology Corporation by the deploying airbag. This is defined 84

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL THE OVERALL *AIRBAG_ALE DEFINITION: This command

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL THE OVERALL *AIRBAG_ALE DEFINITION: This command typically consists of 9 main cards (plus complementary cards for special options when necessary): *AIRBAG_HYBRID *CLIS VENTING FLOW ALE MESH CREATION *AIRBAG_HYBRID PTSRC + SWITCH t *AIRBAG_HYBRID GAS PROPERTIES SPSM POINT SOURCE PROPERTIES $ SIDTYP: 0=SGSID, 1=PSID; VNTTYP: 0=PSID, 1=PID, 2=SGSID; vol=60 lit=60 E 6 mm^3 *AIRBAG_ALE $1 SIDTYP RBID VSCA PSCA VINI MWD SPSF 1 1 0 0. 0 $2 ATMT ATMP=====| GC CC INFLATVOL=========| 298. 151. 01325 E-4 0. 0 8. 3144 1. 0 $3 NQUAD CTYPE PFAC FRIC FRCMIN NORMTYP ILEAK PLEAK 4 4 -11 0. 0 0. 3 1 2 0. 1 $4 IVNTSET IVNTTYP IBLOCK VENTCOEF=========|=========| 1 2 1 0. 80 $5 NEXPIDA NEYPIDG NEZ MOVERN ZOOM=========|=====| 45 44 32 0 0 $5 B X 0 Y 0 Z 0 X 1 Y 1 Z 1=========| -446. 688 -439. 19 -73. 5 $5 C X 2 Y 2 Z 2 X 3 Y 3 Z 3=========| -446. 688 439. 19 -73. 5 -446. 688 -439. 19 559. 833 $6 SWTIME=====| HG NAIR NGAS NORIF LCVEL LCT 40. 0000 0. 000 1. 0 E-5 1 1 8 3 2 $7 AIR =========| MWAIR INITM AIRA AIRB AIRC 0. 0288479 1. 0 27. 055 0. 0 $8 GAS LCM=========| MWGAS=====| GASA GASB GASC 1 0. 0225600 0. 0 35. 91369 0. 0 $9 NODEID VECTID AREA=========|=========|=====| 100019 1 13. 500. . . 100018 8 13. 500 Livermore Software Technology Corporation 85

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 1 & 2: These

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 1 & 2: These 2 cards are defined exactly the same way as cards 1 & 2 of *AIRBAG_HYBRID. Basically the Lagrangian PSID/PID/SGSID making up the airbag and some reference/ambient data are defined. One addition to card 2 is INFLATVOL (inflator canister estimated volume). This is used for automatic gas velocity estimate by LSDYNA: [method 1] INFLATVOL = blank must define LCVEL in card 6. [method 2] INFLATVOL > 0. 0 LCVEL = 0 in card 6 LSDYNA estimates inlet gas velocity automatically This sample model uses [method 1] [method 2] will be shown later … $ SIDTYP: 0=SGSID, 1=PSID; VNTTYP: 0=PSID, 1=PID, 2=SGSID; vol=60 lit=60 E 6 mm^3 *AIRBAG_ALE $1 SIDTYP RBID VSCA PSCA VINI MWD SPSF 1 1 0 0. 0 $2 ATMT ATMP=====| GC CC INFLATVOL=========| 298. 151. 01325 E-4 0. 0 8. 3144 1. 0 Livermore Software Technology Corporation 86

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 3: This card defines

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 3: This card defines some of the same coupling parameters defined in the *CONSTRAINED_LAGRANGE_IN_SOLID card. The defaults coupling settings for *AIRBAG_ALE may be a little different: NQUAD =4 = number of coupling points (start with 2). CTYPE =4 = penalty coupling method (also CTYPE=6). PFAC = 0. 1 = penalty force scaling factor. FRIC = 0. 0 = friction factor for coupling. FRCMIN = 0. 3 = minimum volume fraction to turn on coupling. NORMTYP =0 = use nodal-based normals for coupling direction. ILEAK = 2 = leakage control with energy compensation. PLEAK = 0. 1 = leakage penalty factor. (For CTYPE=4 DIREC=2 is implied. *AIRBAG_ALE $3 NQUAD CTYPE PFAC FRIC FRCMIN NORMTYP ILEAK PLEAK For CTYPE=6 DIREC is-11 controlled by the 4 4 0. 0 0. 3 code internally. ) 1 2 0. 1 Livermore Software Technology Corporation 87

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 4: This card defines

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 4: This card defines the venting option similar to that defined in the *AIRBAG_HYBRID card. Basically the Lagrangian segment set or part set making up the venting hole(s) are defined here. IVNTSET = a set id representing the vent hole(s). IVNTYP = set type: EQ. 0: PSID; EQ. 1: PID; EQ. 2: SGSID. IBLOCK = turn on|off venting blockage: EQ. 0: OFF; EQ. 1: ON. VENTCOEF = venting flow coefficient. *AIRBAG_ALE $4 IVNTSET 1 IVNTTYP 2 IBLOCK 1 VNTCOEF=========|=========| 0. 80 The porous flow will be modeled via the *MAT_FABRIC parameters (FLC, FAC, ELA). From which both the ALE and CV computations will obtain porous flow information. Livermore Software Technology Corporation 88

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 5: This card serves

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 5: This card serves two main purposes: [1] It specifies the options to define the background ALE mesh|part: (a) The users predefine the ALE mesh (with a mesh generator). (b) LS-DYNA automatically generates this mesh. The users must supply 2 more cards to define the “box” region over which a fluid mesh will be created. [2] It defines the mesh or system-group motion (translation, rotation *AIRBAG_ALE and expansion) via MOVERN and ZOOM parameters similar $5 ANEXPIDA NEYPIDG NEZ MOVERN ZOOM=========|=====| 45 44 32 0 0 to definitions by *ALE_REFERENCE_SYSTEM_GROUP card Livermore Software Technology Corporation (*ALE_REFERENCE_SYSTEM_NODE card may also be 89 needed).

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 5 – Mesh Generation:

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 5 – Mesh Generation: Method (b): LS-DYNA automatically generates ALE mesh. If NEX, NEY and NEZ are all defined they become NEX = number of elements to generate in the global X direction. NEY = number of elements to generate in the global Y direction. NEZ *AIRBAG_ALE = number of elements to generate in the global Z $5 ANEXPIDA NEYPIDG NEZ MOVERN ZOOM=========|=====| direction. 45 44 32 0 0 $5 B X 0 -446. 688 $5 C X 2 -446. 688 Y 0 -439. 19 Y 2 439. 19 Z 0 -73. 5 Z 2 -73. 5 X 1 446. 688 X 3 -446. 688 Y 1 -439. 19 Y 3 -439. 19 Z 1=========| -73. 5 Z 3=========| 559. 833 This sample model uses method (b). Method (a) will be shown later … z node 3 de 1 x node 2 y no Coordinates of the 4 nodes defining the “box” are required node 0 Livermore Software Technology Corporation 90

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 5 - Mesh Motion:

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 5 - Mesh Motion: Similar to *ALE_REFERENCE_SYSTEM_GROUP (ARSG) card: MOVERN = ALE mesh automatic motion option EQ. 0: ALE mesh is fixed in space. GT. 0: Reference node group id defining 3 nodes in the *ALE_REFERENCE_SYSTEM_NODE (ARSN) card for moving the ALE mesh similar to PRTYP=5 of the ARSG card. ZOOM = ALE mesh automatic expansion|contraction option EQ. 0: No ALE mesh expansion. *AIRBAG_ALE GT. 1: Expand/contract ALE mesh by keeping $5 ANEXPIDA NEYPIDG NEZ MOVERN ZOOM=========|=====| 45 44 32 0 0 all airbag $5 B X 0 Y 0 Z 0 X 1 Y 1 Z 1=========| -446. 688 -439. 19 -73. 5 parts (defined card 1) contained $5 C X 2 Y 2 Z 2 X 3 Y 3 Z 3=========| -446. 688 -73. 5 -446. 688 -439. 19 559. 833 within the ALE 439. 19 mesh (equivalent to PRTYP=9 of the Livermore Software Technology Corporation 91 ARSG card).

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 6: Some definitions required

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 6: Some definitions required by the traditional *AIRBAG_HYBRID card, hourglass control factor for the fluid elements, and the switch time: SWTIME = time to switch from ALE to CV computational method. HG = hourglass control factor for gases (typically 1. 0 E-5). NAIR = number of species making up the air. Each species will require 1 line under card 7 definition(s). NGAS = number of species making up the inflator gas. Each species will require 1 line under card 8 definition(s). NORIF = number of orifice(s) for inflator gas injection. Each orifice will require 1 line under card 9 definition(s). LCVEL = load curve for inlet gas velocity (if LCVEL=0 Then parameter INFLATVOL in card 2 must be defined). LCT = load curve for inlet gas stagnation temperature (absolute). *AIRBAG_ALE The associated input for this line follows $6 SWTIME=====| HG NAIR NGAS NORIF LCVEL LCT 40. 0000 0. 000 1. 0 E-5 1 1 8 3 2 Livermore Software Technology Corporation 92

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 7: Define the properties

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 7: Define the properties for each species of air … NAIR = number of species making up the air. Each species will require 1 line under card 7 definition(s) MWAIR = molecular weight of a species making up the air. INITM = initial mass fraction of a species making up the air. AIRA = 1 st coefficient of air molar heat capacity at constant pressure. AIRB = 2 nd coefficient of air molar heat capacity at constant pressure. NOTE: This “per-mole” heat capacity unit is molar heat capacity at constant AIRC = 3 rd coefficient of air similar to that from the pressure. *AIRBAG_HYBRID card. While the *MAT_GAS_MIXTURE card defines the heat capacities in “per-mass” unit! The users should be careful. *AIRBAG_ALE $7 AIR =========| MWAIR 0. 0288479 INITM 1. 0 AIRA 27. 055 AIRB 0. 0 AIRC 0. 0 Livermore Software Technology Corporation 93

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 8: Define the properties

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 8: Define the properties for species of the gas … NGAS = number of species making up the inflator gas. Each species will require 1 line under card 8 definition(s) MWGAS = molecular weight of a species making up the inflator gas. GASA = 1 st coefficient of gas molar heat capacity at constant pressure. NOTE: nd coefficient of gas molar heat capacity at constant GASB = 2 This “per-mole” heat capacity unit is pressure. similar to that from the GASC = 3 rd card. coefficient *AIRBAG_HYBRID While of thegas molar heat capacity at constant pressure. *MAT_GAS_MIXTURE card defines the heat capacities in “per-mass” unit! The users should be careful. *AIRBAG_ALE $8 GAS LCM=========| MWGAS=====| 1 0. 0225600 0. 0 GASA 35. 91369 GASB 0. 0 GASC 0. 0 Livermore Software Technology Corporation 94

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 8: Note: The Cp(T)

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 8: Note: The Cp(T) curve should be monotonically increasing for the whole operating T range. If Cp(T) curve has a maximum, then Cp(T) iteration will likely be stable for an operating T range about 200 -300 K below the maximum point. If not, numerical divergence may occur. Cp(T) to Mono d. T~300 K Cp(T) nica Max Cp(T) has maximum point sing T a e r c lly in OK Operating T range OK T Livermore Software Technology Corporation T Convergence problem possible 95

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 9: In card 6:

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL *AIRBAG_ALE card 9: In card 6: NORIF = number of orifice(s) for inflator gas injection. Each orifice will require 1 line under card 9 definition(s). Here: NODEID sources. VECTID point source. AREA *AIRBAG_ALE $9 = NODEID = node ID defining the location of each point = vector ID defining gas flow direction at each area of each point source. AREA=========|=========|=====| 100019 VECTID 1 13. 500 100018 8 13. 500 . . . Livermore Software Technology Corporation 96

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL Using [method 2] in cards 2&6

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL Using [method 2] in cards 2&6 LSDYNA estimates inlet gas velocity: Method 1= give velocity curve; Method 2=give inflator can volume Method 1 Method 2 $ SIDTYP: 0=SGSID, 1=PSID; VNTTYP: 0=PSID, 1=PID, 2=SGSID; vol=60 lit=60 E 6 mm^3 *AIRBAG_ALE $1 SIDTYP RBID VSCA PSCA VINI MWD SPSF 1 1 0 0. 0 $2 ATMT ATMP=====| GC CC INFLATVOL=========| $ 298. 151. 01325 E-4 0. 0 8. 3144 1. 0 60. 0 E 4 $3 NQUAD CTYPE PFAC FRIC FRCMIN NORMTYP ILEAK PLEAK 4 4 -11 0. 0 0. 3 1 2 0. 1 $4 IVNTSET IVNTTYP IBLOCK VENTCOEF=========|=========| 1 2 1 0. 80 $5 ANEXPIDA NEYPIDG NEZ MOVERN ZOOM=========|=====| 45 44 32 0 0 $5 B X 0 Y 0 Z 0 X 1 Y 1 Z 1=========| -446. 688 -439. 19 -73. 5 $5 C X 2 Y 2 Z 2 X 3 Y 3 Z 3=========| -446. 688 439. 19 -73. 5 -446. 688 -439. 19 559. 833 $6 SWTIME=====| HG NAIR NGAS NORIF LCVEL LCT $ 40. 0000 0. 000 1. 0 E-5 1 1 8 3 2 40. 0000 0. 000 1. 0 E-5 1 1 8 0 2 $7 AIR =========| MWAIR INITM AIRA AIRB AIRC 0. 0288479 1. 0 27. 055 0. 0 $8 GAS LCM=========| MWGAS=====| GASA GASB GASC 1 0. 0225600 0. 0 35. 91369 0. 0 $9 NODEID VECTID AREA=========|=========|=====| 100019 1 13. 500. . . 100018 8 13. 500 Livermore Software Technology Corporation 97

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL AIRBAG DEFINITION using predefined ALE meshes/parts:

Fluid-Structure Interaction Modeling with LS-DYNA *AIRBAG_ALE AIRBAG MODEL AIRBAG DEFINITION using predefined ALE meshes/parts: Note cards 5 b and 5 c must not be defined Predefined ALE mesh and PIDs $ SIDTYP: 0=SGSID, 1=PSID; VNTTYP: 0=PSID, 1=PID, 2=SGSID; vol=60 lit=60 E 6 mm^3 *AIRBAG_ALE $1 SIDTYP RBID VSCA PSCA VINI MWD SPSF 123 1 0 0. 0. 0. $2 ATMT ATMP=====| GC CC INFLATVOL=========| 298. 15 1. 0132 e-4 0. 0 8. 314 1. 60. e 6 $3 NQUAD CTYPE PFAC FRIC FRCMIN NORMTYP ILEAK PLEAK 4 4 -11 0. 0 0. 3 1 2 0. 1 $4 IVNTSET IVNTTYP IBLOCK VENTCOEF=========|=========| 1 2 0 0. 25 $5 ANEXPIDA NEYPIDG NEZ MOVERN ZOOM=========|=====| 50000 50003 0 $6 SWTIME=====| HG NAIR NGAS NORIF LCVEL LCT 10. 000 1. e-4 1 1 8 0 2001 $7 AIR =========| MWAIR INITM AIRA AIRB AIRC 0. 0288479 1. 0 27. 055 0. 0 $8 GAS LCM=========| MWGAS=====| GASA GASB GASC 1 0. 0225600 0. 0 35. 91369 0. 0 $9 NODEID VECTID AREA=========|=========|=====| 100019 1 13. 500. . . 100018 8 13. 500 *DEFINE_VECTOR $ vid xt yt zt xh yh zh 1 0. 0 -16. 250000 21. 213200 -16. 250000. . . 8 0. 0 -16. 2500001. 0000 e-06 30. 000000 -16. 250000 Livermore Software Technology Corporation 98

Fluid-Structure Interaction Modeling with LS-DYNA [IIId] ALE FSI FLAT AIRBAG MODEL Moving ALE Fluid

Fluid-Structure Interaction Modeling with LS-DYNA [IIId] ALE FSI FLAT AIRBAG MODEL Moving ALE Fluid Mesh Livermore Software Technology Corporation 99

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH MOVING

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH MOVING FLUID MESH: Ultimately, we want to attach an airbag to an inflator box, which itself is fixed to some moving platform (i. e. steering column). Consider a coarse mesh model: The input files are: abcoars 4_mvexp. k = Main ALE model input file abcoars 4_mvexp_ale. k = ALE part and coupling definitions mdot(t)=LCID 1; Tstag(t)=LCID 2; Vel(t)=LCID 3 The parts are: S 1 S 2 S 3 S 22 = = Top airbag shell Bottom airbag shell Inflator box shell structure rigid background panel = master => S 22 is tied to NSID 3 (S 3 + NID 9272) => move S 22 H 19 = inflator gas (no initial mesh): MGM, SPSM H 20 = background fluid mesh (with initial mesh): MGM, IGM PSID 1 = S 1 S 2 S 3 for coupling PSID 33 = S 1 S 22 for contact Livermore Software Technology Corporation 100

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH MODEL

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH MODEL GEOMETRY: S 1=top airbag H 20=background air mesh S 2=bot airbag S 3=inflator box S 22=rigid platform Livermore Software Technology Corporation 101

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH REAL

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH REAL PROCESS: In an automobile collision, a likely scenario is that the airbag will be deployed and at about the same time the steering which contains the airbag compartment may (move) be deformed, and passenger impacting the airbag. This requires that the airbag model system must be able to move in space. SIMULATION of REAL PROCESS: In this model, as the airbag is being inflated, the platform, to which the “airbag and inflator box system” is attached, is also moved. This is to illustrate that the point sources can be defined with respect to a moving reference system. Livermore Software Technology Corporation 102

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH MODELING

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH MODELING APPROACH: - Define an inflator box (S 3). - Define an airbag (S 1 & S 2) which is attached to the inflator box to form a closed volume system: S 1 -S 2 -S 3. - Define a rigid platform (S 22) to which the inflator-box-airbag system is attached. - Use the rigid platform (S 22) as a moving reference coordinate system and attach the inflator box (S 3) and Point-Source-NODE to it. Thus the airbag-inflator system and the point source will follow this motion. - Move the platform (S 22) during inflation process. Livermore Software Technology Corporation 103

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH LAGRANGIAN

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH LAGRANGIAN PART DEFINITIONS: Skipping the Lagrangian part definition details … S 1=top airbag S 2=bot airbag S 3=inflator box S 2=rigid platform *PART S 1 = TOP airbag shell part $ PID SECID MID EOSID HGID GRAV ADPOPT TMID 1 1 1 0 0 0 $======================================== *PART S 2 = BOT airbag shell part $ PID SECID MID EOSID HGID GRAV ADPOPT TMID 2 1 1 0 0 0 $======================================== *PART S 3 = inflator CANISTER shell part $ PID SECID MID EOSID HGID GRAV ADPOPT TMID 3 1 1 0 0 0 $======================================== *PART S 22 = DASHPANEL $ PID SECID MID EOSID HGID GRAV ADPOPT TMID 22 22 22 0 0 0 Livermore Software Technology Corporation 104

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH INITIAL

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH INITIAL BACKGROUND ALE AIR PART OR MESH: Using *MAT_GAS_MIXTURE material model for air requires an associated *INITIAL_GAS_MIXTURE card to indicate that this mesh can contain this AMMG (air = AMMG 2). PID 20 = mesh Heat capacities are in per-mass unit system PID 20 = mesh Reference density $ SSTYP= SLAVE set types: 0=SGSID ; 1=SHSID ; 2=PSID ; 3=PID ; 4=NSID ; 6=PSID *PART H 20 = initial background air mesh $ PID SECID MID EOSID HGID GRAV ADPOPT TMID 20 20 20 0 0 *SECTION_SOLID 20 11 0 *MAT_GAS_MIXTURE $ MID IECONSFLG GC 20 $ Cv 1_mas Cv 2_mas Cv 3_mas Cv 4_mas Cv 5_mas Cv 6_mas Cv 7_mas Cv 8_mas 719. 9000 $ Cp 1_mas Cp 2_mas Cp 3_mas Cp 4_mas Cp 5_mas Cp 6_mas Cp 7_mas Cp 8_mas 1009. 0 *HOURGLASS $ HGID IHQ QM IBQ Q 1 Q 2 QB QW 20 1 1. 00 e-04 *INITIAL_GAS_MIXTURE $ SID STYPE MMGID T 0 20 1 2 298. 00 $ RHO 1 RHO 2 RHO 3 RHO 4 RHO 5 RHO 6 RHO 7 RHO 8 1. 1823657 Allow AMMG=2 in thismesh Initial T Livermore Software Technology Corporation 105

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH ALE

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH ALE INFLATOR PART DEFINITION: ALE part definition details … $ SSTYP= SLAVE set types: 0=SGSID ; 1=SHSID ; 2=PSID ; 3=PID ; 4=NSID ; 6=PSID *PART H 19 = point sources for the inflator gas at inlet $ PID SECID MID EOSID HGID GRAV ADPOPT TMID 19 19 19 0 0 *MAT_GAS_MIXTURE 19 $ Cv 1_mas Cv 2_mas Cv 3_mas Cv 4_mas Cv 5_mas Cv 6_mas Cv 7_mas Cv 8_mas 1080. 000 $ Cp 1_mas Cp 2_mas Cp 3_mas Cp 4_mas Cp 5_mas Cp 6_mas Cp 7_mas Cp 8_mas 1414. 000 *HOURGLASS $ HGID IHQ QM IBQ Q 1 Q 2 QB QW 19 1 1. 00 e-04 *SECTION_POINT_SOURCE_MIXTURE $ SECID LCIDT NOTUSED LCIDVEL NIDLCOOR 1 NIDLCOOR 2 NIDLCOOR 3 19 2 0 3 272 254 233 $ LCMDOT 01 LCMDOT 02 LCMDOT 03 LCMDOT 04 LCMDOT 05 LCMDOT 06 LCMDOT 07 LCMDOT 08 1 $ NODEID VECTID AREA 9272 1 1. 00 e-04 *DEFINE_VECTOR 1 0. 0. 0. 1 Flow direction vectors are defined with respect to a local coordinate system defined by these 3 nodes inlet information Livermore Software Technology Corporation 106

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH FIXING

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH FIXING THE AIRBAG-INFLATOR-BOX TO THE PLATFORM: Here is one way to simulate moving airbag platform Attaching NSID 3 to the platform (S 22) NSID 3 = all nodes of S 3 + point source node (NID 9272) Moving the platform (S 22) $ tying a floating node to nodeset 3 then tying it to dummy S 4 *CONSTRAINED_EXTRA_NODES_SET $ PID NID/NSID 22 3 *SET_NODE_LIST_TITLE extra 3 0. 0 142 143 144 149 151 156 157 158 204 209 210 222 227 233 234 254 272 9272 $======================================== *NODE 9272 -1. 0937391 e-06 -2. 1874625 e-06 -0. 03000000 $======================================== $ new=move S 22 <=== old=move S 4 *BOUNDARY_PRESCRIBED_MOTION_RIGID $NID|NSID|PID DOF VAD LCID SF VID DEATH BIRTH 22 1 0 29 1. 0000000 0 Livermore Software Technology Corporation 107

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH AIRBAG-INFLATOR-BOX-RIGID-PLATFORM

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH AIRBAG-INFLATOR-BOX-RIGID-PLATFORM SELF-CONTACT: This simple model requires only very simple contact set-up. Efficient contact definitions for more complex system may be more sophisticated. Self-contact $ SSTYP= SLAVE set types: 0=SGSID ; 1=SHSID ; 2=PSID ; 3=PID ; 4=NSID ; 6=PSID *CONTACT_AIRBAG_SINGLE_SURFACE $ SSID MSID SSTYP MSTYP SBOXID MBOXID SPR MPR 33 0 2 0 0 0 $ FS FD DC VC VDC PENCHK BT DT 0. 1000000 0. 0000000 0 0. 000000 $ SFS SFM SST MST SFMT FSF VSF 0. 0000000 $ SOFT SOFSCL LCIDAB MAXPAR EDGE DEPTH BSORT FRCFRQ 2 0. 0000000 0 *SET_PART_LIST 33 1 2 22 PSID 33 = Airbag (S 1 & S 2) contacting the platform (S 22) Livermore Software Technology Corporation 108

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH ALE

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH ALE FSI COUPLING: Using the same coupling approach as shown in previous ALE examples $ SSTYP= SLAVE set types: 0=SGSID ; 1=SHSID ; 2=PSID ; 3=PID ; 4=NSID ; 6=PSID $ ALEMMID = 1 = PID 19 <== see *ALE_MULTI-MATERIAL_GROUP card, 1 st mat/PID *SET_MULTI-MATERIAL_GROUP_LIST 111 1 *CONSTRAINED_LAGRANGE_IN_SOLID $ SLAVE MASTER SSTYP MSTYP NQUAD CTYPE DIREC MCOUP 1 11 0 0 4 4 2 -111 $ START END PFAC FRIC FRCMIN NORM 0. 0 -321 0. 00 0. 3 1 $ CQ HMIN HMAX ILEAK PLEAK VLK_PLCID NVENT BLOKAGE 0 0 0 2 0. 10 *SET_PART_LIST 1 0. 0. 1 2 3 *SET_PART_LIST 11 20 *DEFINE_CURVE 321 $ A 1 O 1 0. 0 1. 0 e-3 8. 0 e 5 AMMG 1 Livermore Software Technology Corporation 109

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH MOTION:

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH MOTION: Mesh translation and expansion definitions Material interfaces Mesh translation Mesh expansion $ SSTYP= SLAVE set types: 0=SGSID ; 1=SHSID ; 2=PSID ; 3=PID ; 4=NSID ; 6=PSID *ALE_MULTI-MATERIAL_GROUP $ SID IDTYPE 19 1 20 1 $---------------------------------------$ This allows mesh 20 to move with 3 ref-nodes-sys => PRID=872 *ALE_REFERENCE_SYSTEM_GROUP $ SID STYPE PRTYP PRID BCTRAN BCEXP BCROT ICOORD 20 1 5 872 $ XC YC ZC EXPLIM DELAY VMAX PRTYP=0=EUL 0 $---------------------------------------*ALE_REFERENCE_SYSTEM_NODE $ NSID 872 $ N 1 N 2 N 3 N 4 N 5 N 6 N 7 N 8 272 254 233 $ N 9 N 10 N 11 N 12 0 $---------------------------------------$ This allows mesh 20 to expand/contract to envelop {PSID 1= S 1 & S 2 & S 3} *ALE_REFERENCE_SYSTEM_GROUP $ SID STYPE PRTYP PSID_PRID BCTRAN BCEXP BCROT ICOORD 20 1 9 1 $ XC YC ZC EXPLIM DELAY VMAX PRTYP=0=EUL 0. 0 0. 0 AMMG 1 AMMG 2 Mesh H 20 will move with NSID 872 Livermore Software Technology Corporation 110

Fluid-Structure Interaction Modeling with LS-DYNA [IV] SOME NEW ALE DEVELOPMENTS Livermore Software Technology Corporation

Fluid-Structure Interaction Modeling with LS-DYNA [IV] SOME NEW ALE DEVELOPMENTS Livermore Software Technology Corporation 111

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH NEW

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH NEW - *ALE_REFERENCE_SYSTEM_GROUP MESH MOTION 1 st concern: [1] Automatic mesh expansion is done uniformly over the whole ALE mesh Sometimes nonsymmetrical Lagrangian airbag expansion cause the predefined fine ALE mesh region around the inflator inlet to shift away NOT FINE ENOUGH ALE MESH TO RESOLVE THE INFLOW! Solution: Define a “center of expansion” via ICOORD parameter ICOORD = A Lagrangian NID defining a “center of ALE mesh expansion”. $ This allows mesh 20 to expand/contract to envelop {PSID 1= S 1 & S 2 & S 3} *ALE_REFERENCE_SYSTEM_GROUP $ SID STYPE PRTYP PSID_PRID BCTRAN BCEXP BCROT ICOORD 20 1 9 1 0 0 0 111 $ XC YC ZC EXPLIM DELAY VMAX FRACL 0 PAD 0. 0 0. 05 Livermore Software Technology Corporation 112

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH NEW

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH NEW - *ALE_REFERENCE_SYSTEM_GROUP MESH MOTION 2 nd concern: [2] Sometimes the Lagrangian airbag expand or move too close to the edge of the ALE mesh POSSIBLE FLOW LEAK OUT OF THE MESH OR UNSTABLE BOUDANRY SOLUTION! Solution: Make sure there is enough ALE mesh “padding” beyond the extreme reach of the Lagrangian airbag FRACL 0 PAD = A fraction of the characteristic length (L 0) of the ALE mesh (between 0. 01 to 0. 2) which guarantees that there will be a length of {FRACL 0 PAD*L 0} of ALE mesh beyond the Lagrangian bag’s extremity. $ This allows mesh 20 to expand/contract to envelop {PSID 1= S 1 & S 2 & S 3} *ALE_REFERENCE_SYSTEM_GROUP $ SID STYPE PRTYP PSID_PRID BCTRAN BCEXP BCROT ICOORD 20 1 9 1 0 0 0 111 $ XC YC ZC EXPLIM DELAY VMAX FRACL 0 PAD 0. 0 0. 05 Livermore Software Technology Corporation 113

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH NEW

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH NEW - *CONTROL_ALE COMPUTATIONAL EFFORT CONTROL [1] NUMBER OF LAGRANGIAN STEPS BEFORE 1 COUPLING COMPUTATION If the Lagrangian motion of the coupling materials is “small” Relative motion between Lagrangian and ALE materials are small We may not have to compute coupling every time step We can perform multiple Lagrangian steps before actually do 1 coupling calculation to save run time! NLAGSTEP = Number of Lagrangian cycles taken before 1 coupling step is to be done. $ This allows mesh 20 to expand/contract to envelop {PSID 1= S 1 & S 2 & S 3} *CONTROL_ALE $ DCT NADV METH AFAC BFAC CFAC DFAC EFAC 0 1 1 -1. 0 0. 0 $ START END AAFAC VFACT PEQUIL EBC PAMBIENT EBCXSID 0. 0 0 01. 01325 e-4 $ NLAGSTEP NLAGBUCKT 0. 0 Livermore Software Technology Corporation 114

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH NEW

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI FLAT AIRBAG MODEL MOVING ALE MESH NEW - *CONTROL_ALE COMPUTATIONAL EFFORT CONTROL [2] NUMBER OF LAGRANGIAN STEPS BEFORE 1 FULL BUCKET-SORT SEARCH FOR COUPLING POINT POSITIONS If the Lagrangian motion of the coupling materials is “small” Relative motions of Lagrangian VS. ALE coupling points are small We may increase the number of Lagrangian steps per bucket-sort We can perform multiple Lagrangian steps before actually do 1 fullbucket-sort search of the coupling point locations. NLAGBUCKT = Number of Lagrangian cycles taken before 1 full bucket-sort search for coupling is to be done. $ This allows mesh 20 to expand/contract S 1 &can S 2 & be S 3} lower. Typically between to 50 envelop and {PSID 200, 1=but *CONTROL_ALE $ DCT NADV 0 1 $ START END 0. 0 $ NLAGSTEP NLAGBUCKT 0 0 METH 1 AAFAC 0. 0 AFAC -1. 0 VFACT 0. 0 BFAC 0. 0 PEQUIL 0 CFAC DFAC 0. 0 EBC PAMBIENT 01. 01325 e-4 EFAC EBCXSID Livermore Software Technology Corporation 115

Fluid-Structure Interaction Modeling with LS-DYNA [V] SOME GENERAL REMARKS ON ALE MODELING & SUMMARY

Fluid-Structure Interaction Modeling with LS-DYNA [V] SOME GENERAL REMARKS ON ALE MODELING & SUMMARY Livermore Software Technology Corporation 116

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI - GENERAL REMARKS & SUMMARY VENTING FLOW

Fluid-Structure Interaction Modeling with LS-DYNA ALE FSI - GENERAL REMARKS & SUMMARY VENTING FLOW MODELING: - Vent hole area may be enlarged at high pressure. - The artificial fabric part located at the vent holes may alter the hole enlargement behavior due to its strength. - Livermore Software Technology Corporation 117

*KEYWORD Fluid-Structure Interaction *TITLE $======================================== *CONTROL_TERMINATION *CONTROL_TIMESTEP *CONTROL_ENERGY $======================================== *DATABASE_BINARY_D 3 PLOT *DATABASE_BINARY_D 3

*KEYWORD Fluid-Structure Interaction *TITLE $======================================== *CONTROL_TERMINATION *CONTROL_TIMESTEP *CONTROL_ENERGY $======================================== *DATABASE_BINARY_D 3 PLOT *DATABASE_BINARY_D 3 THDT $======================================== *CONTROL_ALE *ALE_MULTI-MATERIAL_GROUP *SET_MULTI-MATERIAL_GROUP_LIST *ALE_REFERENCE_SYSTEM_GROUP *ALE_REFERENCE_SYSTEM_NODE *CONSTRAINED_LAGRANGE_IN_SOLID $======================================== *PART *SECTION_SOLID *MAT_PIECEWISE_LINEAR_PLASTICITY *MAT_NULL *MAT_GAS_MIXTURE *MAT_FABRIC *EOS_IDEAL_GAS *EOS_GRUNEISEN *EOS_LINEAR_POLYNOMIAL *HOURGLASS $======================================== *INITIAL_VOLUME_FRACTION_GEOMETRY *BOUNDARY_AMBIENT_EOS *BOUNDARY_PRESCRIBED_MOTION_ *INITIAL_VELOCITY_ *RIGIDWALL_PLANAR *LOAD_SEGMENT_SET $======================================== *SET_PART_LIST *SET_SEGMENT $======================================== *DEFINE_CURVE $======================================== *NODE *ELEMENT_SOLID Modeling with LS-DYNA Livermore Software Technology Corporation 118