16th International LS-DYNA Conference

restrict search to current section

For a simple text search, enter your search term here. Multiple words may be found by combining them with AND and OR. The text in this field will be matched with items' contents, title and description.

• *ALE_STRUCTURED_FSI The New S-ALE FSI Solver

  Hao Chen (Ansys Livermore)

  The LS-DYNA® Structured ALE solver is developed in 2015. It is faster and more stable; uses less memory and storage; its input format is cleaner and much less confusing. It has been well received by users studying behavior of fluids, and especially their interaction with structures. During the past two years, the author worked on a new fluid-structure interaction (FSI) package dedicated to be used with S-ALE solver. The objective is to shorten the running time, stop leakage, and make the input deck user friendly. In this paper this new FSI package, together with its keyword -- *ALE_STRUCTURED_FSI is introduced.

• A Dedicated Forming Package LS-FORM for Stamping Simulation with LS-DYNA®

  Xinhai Zhu, Yuzhong Xiao, Jin Wu, Junyue Zhang, Yiquan Tang, Yi Xiao, Wei Ding (Livermore Software Technology, an ANSYS Company)

  One dedicated package providing the complete solution is always desired by LS-DYNA forming users. Being developed as a process-based package, LS-FORM achieves a seamless integration of pre-processing, LS-DYNA simulation and post-processing. An innovated tooling setup interface in pre-processing makes it easy to define complicated tool motions. The one-button submission provides the shortcut to the LS-DYNA solver to simulate a multi-stage stamping process. The post-processing module can perform a real-time analysis of the forming process by automatically chaining the multi-stage simulation results into a unified database. The user-friendly GUI, up-to-date graphic rendering and impressive stability will also make LS-FORM attractive to users.

• A Meso-Macro Scale Method for Jointed Structures and Their Failure Analysis

  Wei Hu, CT Wu, Xiaofei Pan, Youcai Wu (Livermore Software Technology, LLC)

  As automotive industry moves rapidly towards electrified and digitalized world, the use of lightweight materials and new joining technologies becomes crucial to counteract the weight of electronic and autonomous equipment for energy efficiency as well as to maintain safety and performance. Numerical modeling the joined structures including their failure behavior has been a big challenge in the modern lightweight vehicle safety design. In this study, a two-scale method developed in LS-DYNA® is introduced for modeling jointed structures and their connection failure. In the meso-scale, a new particle stabilization method via a velocity smoothing algorithm is developed for simulating the large deformation and material failure of joint models. The meso-scale joint model characterizing the baseline of joint structure is bridging with macro-scale shell structures using an immerse approach. As a result, a topological coupling between solid and shell formulations is achieved without the need of matching discretization. This two-scale method facilitates the modeling of most connection failures in different joint models and minimizes human interactions with software. A crushing tube example is utilized to demonstrate the effectiveness and applicability of the present method in modeling the joined structures and failure behavior for the modern lightweight vehicle safety design.

• A Methodology to Model the Statistical Fracture Behavior of Acrylic Glasses for Stochastic Simulation

  Marcel Berlinger, Stefan Kolling, Andreas Rühl (Technische Hochschule Mittelhessen), Jens Schneider (Technical University of Darmstadt)

  Acrylic glass made of PMMA bears great potential for automotive glazing. As a substitute for mineral glass, new freeform designs become possible with a simultaneous reduction of the structural weight. Under safety aspects of highly weight optimized components, a very precise knowledge of the material behavior is necessary since it is well known that PMMA is a material with high variability in its strength. In the present work, a methodology is proposed to determine the statistical probability distribution of fracture strains from experimental testing. Subsequently, a rate-dependent stochastic failure model is developed. By generation of uniformly distributed random numbers which represent the occurrence probability, *MAT_ADD_EROSION cards for LS-DYNA® are used, containing the tabulated fracture strains at different strain-rates. For stochastic simulation there are two possible procedures to apply the present model. The user provides a distinct probability quantile, e.g. 5 % occurrence probability, generates the corresponding *MAT_ADD_EROSION card and performs a worst-case simulation. Alternatively, the user generates a random set of probability quantiles, i.e. N values from zero to one, and performs N simulations. As an application, the last procedure is used in order to show the influence of a varying fracture strain on the head injury criterion (HIC) in validation tests on PMMA side windows. The example demonstrates the necessity of stochastic simulation for a reliable quantification of injury criteria in crashworthiness analysis.

• A Model for the Stochastic Fracture Behaviour of Glass

  Christopher Brokmann, Stefan Kolling (Technische Hochschule Mittelhessen)

  The failure of glass is caused by initial flaws that are induced during the manufacturing process. These micro-cracks are randomly distributed on the surface of glass, which is why failure is a random process and stress at failure is a non-deterministic parameter. In the present work, a model for the stochastic fracture behaviour of glass is proposed and implemented as a user subroutine in LS DYNA® for shell elements. Due to the stochastic fracture behaviour of glass, a large scattering can be expected when determining the head injury criteria (HIC) in the case of a pedestrian head impact on an automotive windscreen. In this case a high experimental effort would be necessary to evaluate the stochastic scattering. This can be reduced by numerical simulation using a stochastic failure model. The present model generates failure strengths out of a Weibull distribution obtained by coaxial ring-on-ring tests. The generated stresses are used to calculate initial crack lengths by recalculate the subcritical crack growth during experiments. These initial cracks slowly grow, i.e. subcritical, in dependence on the applied stress rate until the critical stress intensity is reached and failure occurs. In order to validate the model, coaxial ring-on-ring tests with different test setups are simulated and compared to experimental values and analytical solutions using the Weibull surface shift.

• A New Keyword to Apply Moving Temperature Boundary Conditions and its Application in Fused Filament Fabrication

  Jinglin Zheng, Xinhai Zhu (Livermore Software Technology, an ANSYS company), Danielle Zeng (Ford Motor Company)

  This paper introduces a new keyword, *BOUNDARY_TEMPERATURE_TRAJECTORY, which is capable of applying a prescribed temperature boundary condition within a specified volume which moves along a designated path at a given speed. Combined with LS DYNA®’s computational welding material model, it is able to simulate the process of extruding and depositing melted material onto the build platform, such as the process of fused filament fabrication. An example of printing a dog-bone-shaped plastic part is given to demonstrate the usage of this keyword. The simulation result agrees very well with experimentally measured local temperature histories, which demonstrates the validity and accuracy of this simulation approach.

• A Path Towards Including Batteries in Electric or Hybrid Car Crash Simulations with LS-DYNA®

  Pierre L’Eplattenier, Iñaki Çaldichoury (Livermore Software Technology, LLC)

  Safety is an important functional requirement in the development of large-format, energy-dense, lithium-ion (Li-ion) batteries used in electrified vehicles. Many automakers have dealt with this issue by enclosing the batteries into protective cases to prevent any penetration and deformation during the car crash. But with the range of electric vehicle increasing and thus the size of the batteries, a more detailed understanding of a battery behavior under abuse becomes necessary. Computer aided engineering (CAE) tools that predict the response of a Li-ion battery pack to various abusive conditions can support analysis during the design phase and reduce the need for physical testing. In particular, simulations of the multi-physics response of external or internal short circuits can lead to optimized system designs for automotive crash scenarios. The physics under such simulations is quite complex, though, coupling structural, thermal, electrical and electrochemical. Moreover, it spans length scales with orders of magnitude differences between critical events such as internal shorts happening at the millimeter level, triggering catastrophic events like the thermal runaway of the full battery. The time scales also are quite different between the car crash happening in milliseconds and the discharge of the battery and temperature surge taking minutes to hours. A so called “distributed Randles circuit” model was introduced in LS-DYNA in order to mimic the complex electrochemistry happening in the electrodes and separator of lithium ion batteries [1][2][3]. This model is based on electrical circuits linking the positive and negative current collectors reproducing the voltage jump, internal resistance and dumping effects occurring in the active materials. These circuits are coupled with the Electromagnetics (EM) resistive solver to solve for the potentials and current flow in the current collectors and the rest of the conductors (connectors, busses, and so forth). The EM is coupled with the thermal solver to which the joule heating is sent as an extra heat source, and from which the EM gets back the temperature to adapt the electrical conductivity of the conductors as well as the parameters of the Randles circuits [1]. One of the advantages of the Randles circuit model is the relative easiness to introduce internal short circuits by just replacing the Randles circuits in the affected area by a short resistance [1][3]. The Randles circuit model also is coupled with the mechanical solver of LS-DYNA where the deformations due to a battery crush allow the definition of criteria to initiate internal shorts [1]. The Randles circuit model can be used either on a solid element mesh that include all the layers of a cell [1][2][3], or using composite Tshells [4][5]. In the second case, the mechanics is solved on the composite Tshell, but an underlying solid mesh with all the layers still has to be built to solve the EM and the thermal. This implies very large meshes and hence simulation times when dealing with many cells, let alone modules, packs or a full battery. This new Battery Macro (BatMac) model allows simulating a cell with very few layers of elements (down to one). Two fields exist at each node of the mesh, representing the potential at the positive and negative current collectors. These two fields are connected by a Randles circuit at each node. It still is possible to include external and internal shorts. The internal shorts can be locally created depending on local values of different mechanical, thermal or EM parameters. The Joule Heating generated by the current leaking through the short resistance is sent to the thermal solver.

• A Simple Ejection Mitigation Device to Increase Survival of Standing Gunner

  M.S. Hamid (Advanced Computational Systems, LLC)

  It is a challenging task to provide warfighters protection due to impact related injuries such as skull fracture, lower leg and ankle fractures and neck and spinal injuries, which might occur during underbody blast. Tremendous improvements in design of ground vehicles have minimized these injuries. Further improvement can be achieved by new technology solutions by adding a simple passive device to existing design with minimal changes in order to mitigate occupant ejection for a vehicle occupant experiencing underbody blast while standing at nametag, defilade through round hatch opening as part of their operational duty. The patented device is an energy absorbing box with corrugated sheets in all four sides and welded to the top and bottom plates. The top and bottom plates are connected with brackets. The top plate and the brackets are connected with tension failure plates. The corrugated sheets act like energy absorbing (EA) bellows. The EA capacities can be increased by adding collapsible stiffeners inside the box. The device is virtually evaluated in an occupant standing position in a vehicle using LS-DYNA® and LS-PrePost®. The standing occupant is modeled using ATD from LSTC. Modeling of the Gunner Restraint System (GRS) used to keep the occupant in position is developed by using Seatbelt Fitting application module in LS-PrePost. An acceleration pulse is applied to the vehicle in order to represent the effect of blast load. The EA mechanism is evaluated for various side plate thicknesses and heights of the device. The reductions in tibia force and head excursion relative to hatch are presented.

• A Study of LS-DYNA® Implicit Running the Rolls-Royce® Large Representative Engine Model with Intel® Optane™ DC Persistent Memory Technology

  Nick Meng (Intel Corporation), Roger Grimes, Robert Lucas, Francois-Henry Rouet (Livermore Software Technology (LST), an ANSYS company), James Ong (Rolls-Royce Holdings Corp.)

  In this paper we discuss Intel’s continued efforts to optimize the performance of LS-DYNA Implicit. We focus on the Rolls-Royce® Large Representative Engine Model (LREM), the largest implicit model known to Livermore Software Technology. Performance analysis indicated three opportunities for improvement: shared memory parallelization of the LS-GPart reordering code, optimization of the multifrontal linear solver, and usage of Intel® Optane™ DC Persistent Memory. We present results taken while running the LREM model with a tuned hybrid version of LS-DYNA R12.r144413 HYBRID_DP on 2nd Generation Intel® Xeon™ Platinum 8260L scalable processor (formerly Cascade Lake) cluster with Intel’s Optane persistent Memory. We depict the benefits of Intel Optane persistent Memory technology and discuss the techniques needed to optimize LS-DYNA.

• A Study on the Transfer of GISSMO Material Card Parameters from 2D- to 3D-Discretization

  Daniel Sommer, Peter Middendorf (University of Stuttgart), Florian Schauwecker (University of Stuttgart / Daimler AG)

  This study presents basic strategies for transfer of LS-DYNA® material card parameters from 2D- to 3D-discretizaion without extensive recalibration. The responses of a material card, calibrated on two-dimensional shell elements, are shown on single-element tests with different element formulations, on multi-element patches and on coupons. From this, magnitudes of error are ascertained and quick recalibration measures on material model parameters are derived. An evaluation of the stress-state of typical GISSMO-type specimen in different thicknesses and discretization-lengths is given and typical stress-states within the Lode-triaxiality stress-space of these specimens are highlighted. Therefrom, a measure of deviation from the calibrated state can be derived and used as a measure for allowable deviation from the 2D stress-state. Furthermore, the information on the three-dimensional stress-states, even in thin specimen, can be used for quick recalibration of parameters on the failure surface.

• A Unified SPH-DEM-FEM Approach for Modeling of Debris Flow Impacts on Protective Structures

  Chun Liu, Zhixiang Yu (Southwest Jiaotong University), Qiang Wang (Shanghai Fangkun Software Technology Ltd)

  Debris flows are rapid gravity-driven unsteady flows of highly concentrated mixtures of water and solid particle material, destroying numerous mountain building structures and traffic facilities. Investigation of debris flows is thus of significance to hazard prevention and mitigation. This paper aims to provide a numerical model capable of reproducing the debris flow impact estimation by accounting for complicated fluid-particle-structure interaction (FPSI) with a unified Smoothed Particles Hydrodynamics (SPH), Discrete Element Method (DEM) and Finite Element Method (FEM) approach. The fluid phase is represented by SPH. The solid phase consists of physical particle(s) is represented by DEM, and deformable solid structure is represented by FEM. The Wenjia gully debris flow is carried out to demonstrate the capability of the coupled model for simulating FPSI as the application of the debris flow impact simulations. Compared the actual situation, the propagation of the debris flow and destruction of structures were predicted. Then, the effectiveness of the treatment measures of the Wenjia gully debris flow was clarified from the impact evolution, impact force. The developed method will contribute to a better understanding of FPSI and is a promising tool for hazard analysis and mitigation.

• Acoustic Radiated Power and Radiation Efficiency Calculation with LS-DYNA®

  Yun Huang, Zhe Cui (Livermore Software Technology, an ANSYS Company)

  The keyword *FREQUENCY_DOMAIN_SSD in LS-DYNA not only provides convenient solution for steady state vibration analysis for structures, but also raises the possibility for acoustic simulation. For example, it can be combined with acoustic boundary element method (keyword *FREQUENCY_DOMAIN_ACOUSTIC_BEM) or acoustic finite element method (keyword *FREQUENCY_DOMAIN_ACOUSTIC_FEM), to compute the acoustic pressure and sound pressure level for vibro-acoustic problems. In addition, with the option ERP for this keyword, one can perform ERP (Equivalent Radiated Power) analysis to get a quick solution for radiated noise, based on the plane wave assumption for the acoustic waves. A new parameter RADEFF has been added to the keyword *FREQUENCY_DOMAIN_SSD_ERP to run acoustic radiated power computation for baffled plates, and also computes the radiation efficiency. With some examples, this paper explains the difference between the ERP (equivalent radiated power) and ARP (acoustic radiated power) and shows how to use this new parameter to compute the acoustic radiated power and radiation efficiency for vibrating structures.

• Adaptive Smoothed Particle Hydrodynamics and Higher Order Kernel Function in LS-DYNA®

  Jingxiao Xu, Wei Hu, Bo Ren, Youcai Wu, Xiaofei Pan, C. T. Wu (LST)

  This paper presents the implementation of an adaptive smoothed particle hydrodynamics (ASPH) method for high strain Lagrangian hydrodynamics with material strength in LS-DYNA. In standard SPH, the smoothing length for each particle represents the spatial resolution scale in the vicinity of that particle and is typically allowed to vary in space and time so as to reflect the local value of the mean interparticle spacing. However, in the presence of strongly anisotropic volume changes which occur naturally in most of the applications the local mean interparticle spacing varies not only in time and space, but in direction as well. In ASPH, the isotropic kernel in the standard SPH is replaced with an anisotropic kernel whose axes evolve automatically to follow the mean particle spacing as it varies in time, space, and direction around each particle. By deforming and rotating these ellipsoidal kernels so as to follow the anisotropy of volume changes local to each particle, ASPH can capture dimension-dependent features such as anisotropic deformations with a more generalized elliptical or ellipsoidal influence domain. Some numerical examples are investigated using both SPH and ASPH, also higher order kernel function is studied for both SPH and ASPH formulation. The comparative studies show that ASPH has better accuracy than the standard SPH when being used for high strain hydrodynamic problems with inherent anisotropic deformations, also higher order kernel function has better accuracy than the standard cubic kernel function.

• Adiabatic Shear Band Modeling in Inconel-718 Alloy

  Stefano Dolci, Kelly Carney, Paul Du Bois, Cing-Dao Kan (George Mason University)

  The failure mechanism for thick metal plates impacted by blunt projectiles is known as an adiabatic shear band (ASB), which results in a catastrophic failure due to concentrated shear deformation. ASB is generally considered to be a material or structural instability and as such is not controllable. ASBs are a thermodynamic phenomenon occurring at high strain rates and are characterized by large deformations, localized in a narrow band consisting of highly sheared material. Due to the extreme localization of the shear band it is difficult to model it using FEM, because the mesh size needed to capture it is often not practical for real applications. ASBs can control failures in ballistic impacts. For example, the strength characteristics of Inconel 718 are superior to those of Titanium 6Al4V, in quasi-static conditions. Hence, the ballistic limit for the same geometry plates of these materials should be very different. Unexpectedly, tests show that the materials have a similar ballistic limit for a 0.5” plate. The explanation for this similarity is that, inside the ASB the temperature rises to above 700°C, where Inconel undergoes a phase transformation that makes it brittle, causing a sudden failure. Ti6Al4V HPC lattice has a different microstructure than Inconel’s, and so its ASB behavior also differs. 2D analyses are presented, with progressively reduced element sizes until ASB appearance. ASBs of Inconel are shown to have a width of approximately 5 μm. This is far smaller than an element from an “industrial size” mesh. The elements of the mesh need to be smaller than the ASB width in order to capture the localization of shear and the consequent temperature rise in that region. As adiabatic shear band modeling exhibits a strong dependency on mesh size, we propose to neutralize the effects of different mesh dimensions by implementing a new regularization algorithm. The new algorithm will increase the amount of plastic work converted into heat as function of mesh size, where ASBs will occur. The updated model will replicate the ASB characteristics obtained with the ultra-fine 2D mesh using an “industrial size” mesh.

• Adiabatic Shear Band Modeling in Inconel-718 Alloy

  Stefano Dolci, Kelly Carney, Paul Du Bois and Cing-Dao Kan (George Mason University)

  The failure mechanism for thick metal plates impacted by blunt projectiles is known as an adiabatic shear band (ASB), which results in a catastrophic failure due to concentrated shear deformation. ASB is generally considered to be a material or structural instability and as such is not controllable. ASBs are a thermodynamic phenomenon occurring at high strain rates and are characterized by large deformations, localized in a narrow band consisting of highly sheared material. Due to the extreme localization of the shear band it is difficult to model it using FEM, because the mesh size needed to capture it is often not practical for real applications. ASBs can control failures in ballistic impacts. For example, the strength characteristics of Inconel 718 are superior to those of Titanium 6Al4V, in quasi-static conditions. Hence, the ballistic limit for the same geometry plates of these materials should be very different. Unexpectedly, tests show that the materials have a similar ballistic limit for a 0.5” plate. The explanation for this similarity is that, inside the ASB the temperature rises to above 700°C, where Inconel undergoes a phase transformation that makes it brittle, causing a sudden failure. Ti6Al4V HPC lattice has a different microstructure than Inconel’s, and so its ASB behavior also differs. 2D analyses are presented, with progressively reduced element sizes until ASB appearance. ASBs of Inconel are shown to have a width of approximately 5 μm. This is far smaller than an element from an “industrial size” mesh. The elements of the mesh need to be smaller than the ASB width in order to capture the localization of shear and the consequent temperature rise in that region. As adiabatic shear band modeling exhibits a strong dependency on mesh size, we propose to neutralize the effects of different mesh dimensions by implementing a new regularization algorithm. The new algorithm will increase the amount of plastic work converted into heat as function of mesh size, where ASBs will occur. The updated model will replicate the ASB characteristics obtained with the ultra-fine 2D mesh using an “industrial size” mesh.

• Advanced Pedestrian Legform Impactor (aPLI)

  Katharina Stielau (CDH AG), David Blauth (ATD-MODELS GmbH)

  Pedestrian protection aims to reduce injuries in car-to-pedestrian impacts. It is the subject of increasingly stringent worldwide statutory and consumer rating requirements and is thus increasingly important in current vehicle development. With electric vehicles, which are not always readily noticed by pedestrians, there is an even higher risk of pedestrian injury. Passive pedestrian protection is, active safety systems notwithstanding, the final countermeasure to reduce serious and fatal pedestrian injury. The current test assessment for pedestrian protection uses three “impactors” in different loading scenarios designed to represent injuries to specific parts of the human body, such as the head, the upper leg and the lower leg. Recent development of the Advanced Pedestrian Legform Impactor (aPLI) focuses on enhanced biofidelity of the lower extremities to address long-bone fractures, knee ligament injuries and pelvis fractures in a single impactor instead of the separate, independent, lower leg and upper leg impactors. To date, the requirements for the aPLI are applied to the EuroNCAP 2022 regulations. But as seen with the previous development of the FlexPLI, it is to be expected that the global statutory requirements will be influenced by global consumer rating requirements. The CAE development process focuses on high accuracy simulations within the vehicle development process, using reliable and robust FEM vehicle and impactor models. The aPLI-FE models represent the Cellbond aPLI SBL-A hardware and have been developed in close cooperation with the hardware manufacturer. Utilizing many years of experience in the field of occupant protection with HIII models, the FE model ATD-aPLI was also developed with extensive feedback from partners in the German automotive industry. Special modelling techniques were developed and applied in LS-DYNA®. The ATD-aPLI model was validated by material and component investigations and by experiments in generic test rigs. Models for future hardware versions will be released regularly in preproduction and final versions. Automotive manufacturers face new challenges in the CAE process and hardware tests using the new aPLI. Despite similarities with the previous FlexPLI, the attached upper body mass of the aPLI has a strong influence on the impactor kinematics and particularly the femur load. This paper highlights some characteristics of the aPLI and describes a sensitivity analysis using LS-DYNA within the framework of a typical CAE development process.

• Aircraft NPP Impact Simulation Methodology

  Yu.V. Novozhilov (CADFEM CIS JSC), A.N. Dmitriev, D.S. Mikhaluk (CEPSA JSC), N.A Chernukha, L.Yu. Feoktistova ((ATOMPROECT JSC), I.A. Volkodav (SEC DDC JSC)

  The IAEA safety requirements imply that the design of nuclear power plants (NPP) considers both the potential technological disaster and acts of nature. One of the modern mandatory requirements of IAEA standard SRS 87 is to consider a possible massive commercial airplane crashing (APC) into or attacking NPP reinforced concrete (RC) structures. The basis for the methodology is well-validated RC structures modeling. LS-DYNA® constitutive models have been tested by solving a set of verification problems. Selected problems set describes different loading conditions and scales: single finite element study, quasi-static loading, low-speed impact, deformable missile impact, RC wall APC load, shock wave load, perforation by a kinetic missile. The following model parameters are examined: mesh convergence, contact algorithms work, immersed reinforcement coupling, nonlinear stability. The second part of the paper presents a universal method of APC events direct modeling based on the finite element method in the Euler formulation. The relations allowing to identify geometrical, strength, and mass parameters of the airplane finite element model by the given load curve and impact spot are developed. The model of the aircraft obtained in this way allows the transfer of loads on complex-shaped civil structures or when considering the impact at an angle to the surface, with great fidelity. The last part describes the simulation results processing and analysis procedure. Criteria for determining the strength based on the analysis of displacement, strain, and damage of both concrete and reinforcements are proposed. Approach to the evaluation of penetration, perforation, and fracture speed behind the barrier estimation is developed.

• An Adaptive Thick Shell Element for Crashworthiness Assessment of Laminated Composites

  Johannes Främby, Martin Fagerström (Chalmers University of Technology), Jesper Karlsson (DYNAmore Nordic AB)

  The automotive industry is strongly dependent on efficient numerical tools in order to assess the crashworthiness of laminated composites. Unfortunately, to achieve predictive assessments of fracture in laminated composites one must resort to computationally costly, high-fidelity layered models, which in practice makes full vehicle crash simulations very difficult (or even impossible). One solution to this is to use an adaptive modelling technique where an initially coarse model is automatically refined, when and where needed, during the analysis. In this context, we have developed an LS-DYNA® user element which can be adaptively refined through the thickness to allow for both so-called weak discontinuities (discontinuities in strain at material interfaces) and strong discontinuities (discontinuities in displacements, i.e. delamination cracks). Furthermore, we have proposed a remedy to the numerical instabilities which arise from using adaptive refinement in a dynamic explicit solver. This adaptive element proves capable of reproducing the result of high-fidelity models, although at a lower computational cost.

• An Approach for Modeling Shock Propagation Through a Bolted Joint Structure

  Pouya Shojaei, Mohamed Trabia, Brendan O'Toole, Jed Higdon (University of Nevada, Las Vegas)

  Impact loading is typically characterized by a relatively large load happening over an extremely short duration and inducing broad range of vibration frequencies. Standard design approaches of bolted joints based on static or quasi-static criteria may not be effective under these conditions. This study focused on simulating a drop-weight tower experiment where a free-falling mass impacted a target plate, which was bolted to a cylindrical structure. An accelerometer was used to record transmitted acceleration to the cylindrical structure. An approach for simulating the shock propagation was proposed using LS-DYNA® Explicit finite element code. To reduce computational time, thread was not included. Instead, bolts were represented as cylinders with cross-sectional areas equal to the tensile stress area of the bolts. The results showed good agreement between the finite element and experimental results.

• An Engineering Approach of an X-Ray Car Crash Under Reverse Small Overlap Configuration

  Y. Leost, P. Bösl, I. Butz, T. Soot, M. Kurfiß, S. Moser (Fraunhofer Institute for High-Speed Dynamics, Ernst-Mach-Institut, EMI), A. Nakata, F. Kase, T. Hashimoto, S. Shibata (Honda R&D Co., Ltd.)

  During a crash event, conventional optical measuring systems provide information about the deformation of parts that are directly visible. The new measuring method called X-Ray Car Crash (XCC) developed at Fraunhofer EMI allows accessing the crash kinematic of specific parts inside the vehicle. This method provides precious information that is currently not accessible in a crash test and allows for better comparison with FEM simulations. The present paper describes a preliminary study performed in collaboration with Honda R&D Co., Ltd. The load case under consideration is a reverse variant of the IIHS Small Overlap with integrated X-Ray technology. Fraunhofer EMI Research Crash Center aims at developing new measurement methods to investigate non-standard high-speed dynamics safety issues. Most of these specific requests are coming from car manufacturers. In order to achieve maximum test reproducibility and simplify boundary conditions, the facility is equipped with a propelled sled system on rails. Thus, it enables to perform impactor to vehicle scenarios with moving barriers up to 3000 kg by 22 m/s. The standard Small Overlap at 64 km/h belongs to the vehicle to barrier scenario and requires some preliminary computations to adapt it for the EMI Crash test facility. Special consideration was given to energy balance in order to determine the right barrier velocity and mass to achieve a similar intrusion in the car to in the standard configuration. Numerical simulations were required at each step to meet the different challenges of this study. This paper describes first the numerical assessment of the validity of the reverse scenario. FEM simulations were then used extensively for developing a special moving barrier presenting maximal structural robustness, well-balanced dynamic behavior and allowing on-sled measurement technics and braking system. Then, LS-DYNA® simulations provided necessary data to perform ray tracing simulations and thus finding the right placement for X-ray source and X-ray detector. Finally, numerical simulations played an important role for an enhanced test setup, by finding the best balance between appropriate mechanical robustness of supporting structures (so called Pit-cover) and low X-ray attenuation.

• An Investigation of Maple Wood Baseball Bat Durability as a Function of Bat Profile using LS-DYNA®

  Blake Campshure, Patrick Drane, James Sherwood (University of Massachusetts Lowell)

  During the 2008 Major League Baseball (MLB) season, there was a perception that the rate at which wood bats were breaking was increasing. MLB responded by implementing changes to the wood bat regulations that were essentially transparent to the players, e.g. changing the orientation for the hitting surface on maple bats, setting a lower bound on wood density, and reducing the allowable range for the slope of grain (SoG) of the wood used to make bats. These new regulations resulted in a 65% reduction in the wood-bat breakage rate. It is proposed that a further reduction can be realized by accounting for the role that bat profile plays with respect to bat durability. To begin to develop an understanding of this relationship, a parametric study was conducted using a finite element model on three baseball bat profiles made from maple wood. These bat profiles that span a range of volumes were examined using LS-DYNA to observe their response to bat/ball impacts over a range of game-like speeds. The Altair HyperMesh and LS-PrePost® pre-processors were used in the making of the geometric representation and finite element meshing of the models. The mechanical behavior of the maple wood and its fracturing were modeled using the *MAT_WOOD material model in combination with the *MAT_ADD_EROSION option, respectively. The effective wood material properties were varied as a function of wood density. Results include how bat profile and SoG influence bat durability, where durability is defined as the relative bat/ball speed that results in crack initiation, i.e. the higher the breaking speed, the better the durability. The effective durability of the bat as a function of profile was found to be well predicted by the LS DYNA modeling.

• Analysis and Optimization of Aluminum Automobile Side Door Design Using LS-DYNA® Implicit and LS-OPT®

  Akshay Kulkarni, Richard Newton (Novelis Inc.)

  Vehicle design and development involves various types of linear and non-linear finite element analyses. Using different solvers for different types of analyses is not always a cost-effective solution. Using one solver for multiple types of analysis saves cost as well as it allows sharing CAE models between various disciplines. Although LS-DYNA historically has been used for explicit analysis, recent enhancements in LS-DYNA Implicit enable it to be used for various implicit analyses. This work focuses on the analysis and optimization of an automobile side door assembly made of aluminum, using LS-DYNA Implicit Solver for multiple load cases. A combination of Novelis’s high formable and high strength aluminum alloys were used in the door design. Implicit load cases used for this analysis were – modal analysis, door sagging, door frame stiffness, and beltline stiffness. LS-DYNA Implicit models were further used for setting up DOE and design optimization. LS-OPT tool was used to conduct multi-response DOE studies and optimization to minimize the door weight while meeting all the performance requirements. Additionally, the DOE runs results were combined with Excel cost model results to choose an optimal design that balanced the total mass of the door versus the cost to manufacture. A final validation simulation was run to demonstrate the optimized design met all expected performance requirements.

• Application of *MAT_258 for Bending and Crushing of Extruded Aluminum Profiles Using Shell Elements

  Jens Kristian Holmen, Joakim Johnsen (Enodo AS), David Morin, Tore Børvik, Magnus Langseth (Norwegian University of Science and Technology (NTNU))

  *MAT_258 (*MAT_NON_QUADRATIC_FAILURE) is a through-thickness failure regularization model for shells in LS-DYNA®. In this model the failure parameter is computed as a function of the size of the element as well as its bending-to-membrane loading ratio. The constitutive behavior and fracture surface in *MAT_258 are represented by well-known analytical expressions which simplify the calibration process. This means that ductile failure initiation can be predicted in thin-walled metallic structures with minimal calibration effort and cost. In this study, we go through the calibration process of *MAT_258 for two aluminum alloys before the calibrated material cards are applied in shell element simulations of double-chamber aluminum extrusions in both three-point bending and crushing.

• Application of a Composite Material Shell-Element Model in Ballistic Impact and Crush Simulations

  Tobias Achstetter, Kelly Carney, Paul Du Bois, Cing-Dao Kan (George Mason University), Sheng Dong (The Ohio State University), Allen Sheldon (Honda R&D Americas)

  A new orthotropic material model with tabulated hardening curves for different loading directions, strain-rate and temperature dependency, damage, and a new strain-based generalized tabulated failure criterion was utilized to simulate ballistic impacts and a C-channel under crush loading. These validation simulations of the material model were performed to test the physical usefulness and robustness of the developed material model. Ballistic impact tests were chosen to highlight the capabilities of the material model in high speed impact applications. For the tested unidirectional composite material in the ballistic impact, extensive material data was available. In a recent study, Dong et al. calibrated an existing material model in crush simulations to match force-displacement characteristics of several crush experiments and a match between tests and simulations was achieved after several rounds of optimization. To highlight the capabilities of the new material model in crush load cases, its results were compared to the force-time history obtained in tests and simulations using MAT58. For both the ballistic impact and crush simulations, the same modeling approach was used.

• Application of FSI/ALE on Mower Grass Cutting Simulation

  Vincent Zou, John Cox (Stanley Black & Decker)

  A mower’s grass mowing quality and energy consumption are two very important factors for battery powered mower development. The challenges include developing a highly efficient cutting blade that matches with the deck for creating the ideal air flow for bagging, mulching and side discharging and consumes as little energy as possible, which, improves the battery running time. It is critical for the development engineers to understand the mower’s air flow inside the deck, the blade’s energy consumption for air flow and clippings transportation during mowing. In this paper, the LS-DYNA® FSI/ALE was used to simulate the mower mowing process. The simulation model was validated, and the result was used for improving the mower’s deck and blade design.

• Automatic Analysis of Crash Simulations with Dimensionality Reduction Algorithms such as PCA and t-SNE

  David Kracker (Dr. Ing. h.c. F. Porsche AG), Jochen Garcke (Fraunhofer SCAI), Axel Schumacher University of Wuppertal, Pit Schwanitz (University of Wuppertal)

  The increasing number of crash simulations and the growing complexity of the models require an efficiently designed evaluation of the simulation results. Nowadays a full vehicle model consists of approximately 10 million shell elements. Each of them contains various evaluation variables that describe the physical behavior of the element. Therefore, the simulation models are very high dimensional. During vehicle development, a large number of models is created that differ in geometry, wall thicknesses and other properties. These model changes lead to different physical behavior during a vehicle crash. This behavior is to be analyzed and evaluated automatically. In this article, potentials of several algorithms for dimensionality reduction are investigated. The linear Principal Component Analysis (PCA) is compared to the non-linear t-distributed stochastic neighbor embedding (t-SNE) algorithm. For those algorithms, it is necessary that the input data always has an identical feature space. Geometrical modifications of the model lead to changes of finite element meshes and therefore to different data representations. Therefore, several 2D and 3D discretization approaches are considered and evaluated (sphere, voxel). In order to assess the quality of the results, a scale-independent quality criterion is used for the discretization and the subsequent dimensionality reduction. The simulations used in this paper are carried out with LS-DYNA®. The aim of the presented study is to develop an efficient process for the investigation of different data transformation approaches, dimensionality reduction algorithms, and physical evaluation quantities. The resulting evaluation method should represent physically relevant effects in the existing simulations in a low-dimensional space without human interaction and thus support the engineer in the evaluation of the results.

• Automatic Evaluation of LS-DYNA® Simulation Results Using Statistical Database and Python

  Daxin Wu, Olaf Hartmann (ARRK Engineering)

  Thanks to the significant increase of computational power, full-vehicle crash simulation has become a standard procedure in vehicle crashworthiness design. It has enabled engineers to accommodate constantly shortened development period in automotive industry. On the other hand, with eased access to simulation tools, simulation results flood during the project development. Most of the time, only a small specific portion of the results is analyzed by engineers. Based on internal statistics gathered from various projects, the number of curves in history outputs from a full-vehicle crash simulation varies from 14,000 to 500,000, depending on the model complexity and conventions of the project. Nevertheless, only 0.1 to 1 percent of the entire outputs are used. The rest of the curves are not analyzed mainly due to two reasons, namely little relevance to the focus of the analysis and absence of post-processing method for systematic analysis of a large amount of outputs. However, considering the size of a full vehicle crash model and the complexity of the crash event, a lot of information can be overlooked. Using data mining, history outputs from past simulations can be systematically gathered, processed in a statistical manner and then stored in a database to serve as a reference for future simulations and even as the basis for more advanced evaluations methods. In the scope of this work, we developed a method for automatic evaluation of history outputs of LS-DYNA simulations by comparing them with a reference database created from previous simulation results, which are predecessors or comparable with the new simulation. The comparison identifies the history outputs with the most significant deviations and records the time points, when such discrepancies occur. Furthermore, the spatial information of these history outputs are extracted, namely the positions on the vehicle. The comparison result therefore shows both the time when deviations occur and the structural regions which are most likely responsible for the deviations. This helps determine the sequence of different structural behaviors and their interdependencies. Significant deviations can come from initial differences in the model, which may indicate modelling errors, or arise over time. This tool was adopted in the crash development of a sports car in order to ensure model quality and identify the sequence of different structural behaviors and their causes. In this paper, we present the possibility of automatic evaluation of all LS-DYNA history outputs using python. This serves as a foundation for further evaluation techniques based on big data analysis.

• Ballistic Impact Simulations of an Aluminum 2024 Panel Using *MAT_224 in LS-DYNA® Considering Oblique Incidence and Attitude Angles of a Rectangular Projectile

  C. K. Park, K. Carney, P. Du Bois, C. D. Kan (George Mason University), G. Queitzsch (retired, Federal Aviation Administration), D. Cordasco, W. Emmerling (Federal Aviation Administration)

  The objective of this study is (1) to validate the *MAT_224 model for Aluminum 2024 with complex impact conditions, (2) to evaluate its predictability of ballistic limit and residual velocities of a projectile under various impact conditions, and (3) to investigate the effects of oblique and attitude angle variations of a projectile on penetration to a target plate. The newly developed *MAT_224 model (version 2.0) for Aluminum 2024 was utilized to simulate a series of ballistic impact tests conducted by NASA using a rectangular block projectile of Inconel 718 with sharp edges and corners, impacting Aluminum 2024 flat panels at oblique angles of incidence. A full ballistic impact simulation model was created with over twenty million solid elements and used to conduct approximately one hundred ballistic impact simulations. Overall, the ballistic impact simulations showed highly comparable results with the NASA tests in terms of projectile residual velocities, failure shapes of the target plates, and projectile penetration behavior. Based on a series of ballistic impact simulations, the ballistic limit velocities of the projectile were predicted.

• Belt Modelling in LS-DYNA®

  Mikael Dahlgren, Abhiroop Vishwanatha, Anurag Soni (Autoliv Sweden, India, North Germany), Klas Engstrand, Jimmy Forsberg (DYNAmore Nordic AB), Isheng Yeh (LST LCC)

  Belt modelling in LS-DYNA has gone from 1D-belt elements, through hybrid belt modelling into 2D seatbelt elements. The current paper investigates recently developed features in LS-DYNA such as bending, strain-rate and orthotropic material behavior. A short description of the evolution of belt modelling is also given. Belt modelling in LS-DYNA has in the past and in the closest future included modelling of sliprings which are points in space where the belt elements pass through. Typical, these are located at sharp directional changes in the belt routing, e.g. B-pillar, D-ring and buckle tongue. Currently, it is the slipring functionality that inhibits users from using an ordinary element and material, of their own selection, to model the belt. Both 1D and 2D belt elements are assigned with *MAT_SEATBELT as it is the only material compatible with sliprings. 2D belt elements in LS-DYNA is a combination of 1D-belt elements along the length direction of the belt and 2D membrane elements made of a MAT_FABRIC material created internal in LS-DYNA. The MAT_FABRIC is created based on *MAT_SEATBELT values. This means current *MAT_SEATBELT in LS-DYNA do not carry any bending loads. The new feature development makes the coating functionality found for MAT_FABRIC available in *MAT_SEATBELT_2D. This enables the bending load carrying possibilities for the 2D belt elements in LS-DYNA. Apart from this feature strain-rate dependency and orthotropic material behavior have been added to the 2D belt elements. It will be shown that the interaction between occupant and belt is improved. It will also be shown that the behavior of an unloaded belt is improved. Finally, there is a short outline of foreseen future needs regarding belt modelling features.

• Calibration and Application of GISSMO and *MAT_258 for Simulations Using Large Shell Elements

  Joakim Johnsen, Jens Kristian Holmen (Enodo AS), Gaute Gruben (SINTEF Industry), David Morin, Magnus Langseth (Norwegian University of Science and Technology (NTNU)

  For many industrial applications finite element (FE) models are becoming increasingly large, making shell elements a necessary tool to maintain a reasonable computational time. Shell elements describe a plane stress state and phenomena like local necking and failure under bending must be appropriately dealt with. Thickness-to-length ratios larger than two are not uncommon for shell elements. This is often larger than the elements used for material model calibration and can sometimes lead to challenges in describing the geometry and the stress state. In this study, we evaluate the accuracy of *MAT_258 and a standard GISSMO calibration. The material and component tests are made of Docol 1400M. Results from *MAT_258 and GISSMO are compared to several component tests spanning a wide range of stress states.

• Cardiac Electrophysiology Using LS-DYNA®

  Pierre L’Eplattenier, Iñaki Çaldichoury, Facundo Del Pin, Rodrigo Paz, Attila Nagy, Dave Benson (Livermore Software LLC)

  Heart disease is among the leading causes of death in the Western world; hence, a deeper understanding of cardiac functioning will provide important insights for engineers and clinicians in treating cardiac pathologies. However, the heart also offers a significant set of unique challenges due to its extraordinary complexity. In this respect, some recent efforts have been made to be able to model the multiphysics of the heart using LS-DYNA. The model starts with electrophysiology (EP) which simulates the propagation of the cell transmembrane potential in the heart. This electrical potential triggers the onset of cardiac muscle contraction, which then results in the pumping of the blood to the various organs in the body. The EP/mechanical model can be coupled with a Fluid Structure Interaction (FSI) model to study the clinically relevant blood flow parameters as well as valves or cardiac devices. This paper concentrates on the EP part of the model. Other papers in this conference will present the mechanical and FSI parts. Different propagation models, called “mono-domain” or “bi-domain”, which couple the diffusion of the potential along the walls of the heart with ionic equations describing the exchanges between the inner and the outer parts of the cells have been implemented. These models were first benchmarked against published results obtained from other EP research codes on a simple cuboid heart tissue model. More recently, we also performed benchmarks proposed by the FDA against analytical solutions. Other features of the EP solver will also be presented such as coupling between the tissue and a surrounding bath, coupling between mono and bi-domain in the same model, and coupling of the mono/bi domain models with a Purkinje Network. Finally, multi-physics simulations with the EP coupled with mechanical deformation and FSI for the blood flow will be presented. These models include arteries and valves for a realistic model of a ventricular pump.

• Challenges of Simulative Consideration of Aluminium Hardening Caused by Paint-Bake

  David Koch, André Haufe, Christoph Wilking, Christian Liebold (DYNAmore GmbH), Markus Feucht, Sebastian Roller (Mercedes-Benz AG)

  Local pre-straining in the forming process and thermal loading in the paint bake process have a significant effect on the further development of local material properties. In many cases it is therefore unavoidable to take such effects into account in the further course of the simulation along the part processing chain. This is particularly evident when the hardening of aluminium components during artificial ageing in the cathodic dip painting furnace is to be considered. During the precipitation processes, dislocations form between the metal lattices depending on the pre-straining and the applied temperature load, which ultimately lead to a further increase in strength. In the following it is shown how this strengthening can be estimated for use in subsequent CAE simulation processes and thus made accessible to crashworthiness simulations. The basis for this approach is the JMAK equation that describes the degree of local hardening phenomenologically. Tensile tests are carried out on differently pre-strained and heat-treated samples. Based on the results of the tests the parameters of the JMAK equation are derived. Using *MAT_TAILORED_PROPERTIES in LS-DYNA® allows the consideration of several locally varying state variables. Here the yield stresses are to be defined as a function of i.e. the degree of hardening and the pre-strain applied during the forming process. The corresponding quantities are mapped from the discretization of the forming simulation (i.e. equivalent plastic strains) and the thermal simulation of the paint-bake-oven (i.e. the computed degrees of hardening) onto the spatial discretization used in crashworthiness. Thus, it is possible to represent the dependence of the yield curves on the decisive influencing variables of the production and hence consider locally individual material behaviour.

• Characterization and Material Card Generation for Thermoplastics

  M. Helbig, A. Erhart, A. Haufe (DYNAmore GmbH)

  Modelling polymer materials for crashworthiness applications is still an ongoing and challenging topic. Besides the constitutive model the spatial discretization plays a significant role in setting up predictive models in impact scenarios where damage and fracture are dominating the part deformation. Therefore, the limits of the chosen spatial discretization shall always be kept in mind. However, the present contribution is focused on recent enhancements of constitutive models for polymeric materials. This topic is as well ongoing for many years and has been tackled almost 2 decades ago by the development of MAT_SAMP-1 (#187) in LS-DYNA®. Many years of continuous improvement lead to a versatile and usable as well as predictive model. Unfortunately for the cost of slow execution speed if the parameter set or the various curve definitions were not chosen wisely. Therefore, a simplified model with the aim to have a more competitive model available when it comes to computing speed was developed. The so called SAMP_LIGHT (#187L) model comes with a complete redesign for speed but also with a number of limitations (due to the speed argument) and still seems to be versatile enough for everyday simulations. The present contribution recalls the features of SAMP-1 and discusses some of the issues that may lead to exaggerated execution time. Then the reduced model is described and a viable approach to convert available SAMP-1 constitutive data towards SAMP_LIGHT is presented. Clearly the limited model may not be as predictive as the fully flavored one – but the drawbacks may not be severe enough to not give it a try.

• Cohesive Zone Modeling of Adhesively Bonded Interfaces: The Effect of Adherend Geometry, Element Selection, and Loading Condition

  Devon C. Hartlen, John Montesano, Duane S. Cronin (University of Waterloo)

  Cohesive zone modeling (CZM) is an efficient technique for modeling adhesively bonded interfaces in structure-level simulations. CZM provides a reduction in computational expense compared to using solid elements and provides a more accurate representation of adhesive response, damage, and failure compared to tiebreak contact methods. Despite these benefits, the adoption of CZM is slowed, in part, due to a lack of information and guidelines as to their appropriate use in industrially relevant modeling situations. While many research studies have applied cohesive elements in conjunction with solid elements to model the adherends, most industrial applications such as automotive structures, are commonly modeled with shell elements. Using CZM to join parts meshed with shell elements requires understanding of how forces and moments are transferred from adherend to adhesive, how the bond line should be modeled geometrically, and how adhesive and adherend parts should be connected numerically. While LS-DYNA® contains several cohesive element formulations, including one specifically developed to be compatible with adherends meshed with shells, and a host of connection techniques, there remains a general lack of understanding as to how CZM should be implemented. To guide the use of CZM, a parametric study was undertaken to examine the effect of cohesive element formulation, adherend element selection, bond line geometry, adherend geometry, and the effect of various connection methods on model response. Two exemplar geometries were explored to examine the effect of different loading condition: a double cantilever beam test and a single lap shear test. It was found that changes in element selection and geometry could dramatically affect the stability and accuracy of the responses of a simulation. Based on the results of this study, this paper presents guidelines and recommended practices for applying CZM to adhesive bond lines for a range of possible modeling scenarios.

• Conjugate Heat Transfer in LS-DYNA®: An Update of the ICFD-Structure Coupling Capabilities for Hot Stamping

  Iñaki Çaldichoury Rodrigo Paz, Facundo Del Pin, Chienjung Huang (Livermore Software Technology, an Ansys company)

  Hot stamping is an essential process in the long and complex assembly chain that will lead to the manufacturing of a complete vehicle. The three main steps consist in a metal sheet being heated, formed and rapidly cooled, yielding parts with high strength and light weight. The main focus of research in that area resides in reducing cycle times, i.e. the time it takes for the tools to cool down, while simultaneously maintaining a high integrity of the formed workpieces. To that effect, the design of cooling systems has drastically increased in complexity and new guiding models are needed for the engineer to be able to identify and then correct potential ‘dead flow zones’, ‘hot spots’ and other problematic areas. Simulation is increasingly viewed as the most general and versatile tool to tackle those challenges. The physics require a coupled thermal, fluid and often mechanical simulation for which a Multiphysics code is needed. Within LS-DYNA, the ICFD solver offers such capabilities and efforts have been continuous over the years to improve existing capabilities and add new ones in conjunction with user feedback [1] [2]. This paper will offer an overview of the existing capabilities, reveal some best practice approaches as well as introduce the latest developments, with a special focus on the fluid-structure interface and how turbulent effects may affect the heat transfer.

• Constrained Multidisciplinary Topology Optimization

  Willem Roux, Imtiaz Gandikota, Guilian Yi (Livermore Software Technology – an ANSYS Corporation)

  For multidisciplinary topology optimization it can be difficult to select the weights for each load case. This becomes even harder if there are multiple design considerations per case. But if constraint values can be defined, then the problem is solvable, because the problem is transformed into one of satisfying the constraints. The most difficult constraint to control is that of the crash pulse, because the existing linear methodologies cannot be used – solutions such as multipoint strategies and spatial kernels must be introduced instead. The NVH constraints are however linear and solving the NVH constraints in combination with the crash pulse becomes a two-level problem. In this paper we show multidisciplinary design optimization considering constraints from impact, linear statics, and frequency load cases.

• Coupled Crash Live Deployment Simulation Using LS-DYNA® Functional Mock-up Interface

  Ke Dong (General Motors Company), Xiaomeng Tong, Isheng Yeh (ANSYS)

  Advanced driver-assistance systems (ADAS) consist of multiple advanced sensors. Information from these sensors has great potential to improve the performance of existing crash sensing systems. Integrated simulation of the active and passive sensing system is a key enabler to assess the potential benefit. Since ADAS sensing and crash sensing simulation are usually conducted in different environments, co-simulation capability is necessary to bring multiple different environments together to achieve the same simulation goal. In this paper, new co-simulation features have been developed in LS-DYNA using Functional Mock-up Unit (FMU). LS-DYNA is able to generate FMU and co-simulate with other software though Functional Mock-up Interface (FMI). A plugin package “FMU manager” built upon FMI 2.0 standard is provided to LS-DYNA users to implement the FMU import and export. The newly developed features were applied to a coupled live development crash simulation between Matlab/Simulink and LS-DYNA. The crash sensing algorithm was built in Matlab/Simulink and finite element vehicle and airbag models were built in LS-DYNA. Co-simulation between them demonstrated the capability of live deployment simulation under different crash scenarios without modifying the models. In future, the coupled live deployment model will be further integrated with ADAS sensing simulations to explore the benefits of the integrated sensing system.

• Coupled Fluid-Structure Interaction Simulation of Prosthetic Heart Valves

  Facundo Del Pin, Iñaki Çaldichoury, Rodrigo R. Paz, Chien-Jung Huang (Livermore Software Technology LLC)

  Artificial heart valves are medical devices that are implanted in patients to replace a diseased native heart valve. They could be classified according to their shape and materials used to manufacture them into mechanical, biological, tissue-engineered and polymeric valves. Approximately 2% of the US population suffer from valvular heart disease (VHD) with the most common causes being aortic stenosis (AS) mostly due to calcification of the aortic valve and aortic valve insufficiency. This paper deals with the numerical simulation of a biological prosthetic aortic valve (AV). This type of valves is composed of three leaflets configured in a complex hemispherical geometry. The leaflets have a variable thickness distribution being thicker at the attachments and free edges and thinner at the belly of the leaflet. Important design parameters for PHVs include effective orifice area, jet velocity, pressure gradient, regurgitation and thrombogenic potential. The objective is to showcase a framework within LS-DYNA® to perform a coupled Fluid Structure Interaction simulation (FSI) of a prosthetic valve and the possible different procedures used to evaluate the design parameters which can be used for a later optimization procedure.

• Cross-Platform Co-Simulation for Vehicle Safety Analysis

  Xiaomeng Tong, Isheng Yeh (Livermore Software Technology, an ANSYS Company)

  Cross-platform co-simulation is gaining more popularity nowadays for vehicle safety analysis. Essential elements, such as ADAS (advanced driver-assistance system) sensors, vehicle dynamics, occupant posture, and controller, can be individually solved in each software and effectively connected to the toolchain. The concept of co-simulation well suits the vehicle integrated safety analysis, which consists of both (1) the active safety features, such as autonomous emergency braking, lane keeping, etc., and (2) the passive safety features, such as the airbag, seatbelt pretensioner, etc. The co-simulation also extends the vehicle safety analysis from the traditional in-crash to a more comprehensive inclusion of pre-crash so as to evaluate the dummy posture and injury more precisely. To achieve this purpose, LS-DYNA® develops a co-simulation feature based on the Functional-Mockup-Interface (FMI), which allows LS-DYNA to remotely exchange data with any 3rd party software supporting this standard. Two cases are demonstrated hereby: the first is a passive safety co-simulation between LS-DYNA and MATLAB, where MATLAB controls the seatbelt pretension force, timing and the airbag deployment in LS-DYNA; the second case is the integrated safety case focusing on the active seatbelt control, where ANSYS VRX Driving Simulator solves the vehicle dynamics, and MATLAB provides the controller of braking/acceleration in VRX as well as the seatbelt/airbag in LS-DYNA. Both cases reveal that a more accurate occupant posture and significant improvement of occupant injury can be achieved by optimizing the active/passive safety features through the co-simulation.

• Demonstrating LS-DYNA®’s Capabilities in Welding Simulations by Experiments

  Maarten Rikken, Gianmarco Montalbini, Bruna Frydman, David Gration (ARUP)

  This study examines the welding residual stress formation in a bead-on-plate welded specimen using LS-DYNA and a series of physical experiments to validate the simulation approach. Calculating residual stresses numerically could be used to improve welding procedures or to assess the structural integrity of welded joints in service conditions. Dilatometer and tensile tests at elevated temperature were conducted to obtain thermal expansion behavior and stress-strain curves of the S355G10+M base material. The software package JMatPro provided the other material properties required for the welding simulations in LS-DYNA. A thermal analysis was set up in LS DYNA to simulate a bead-on-plate welded specimen for which the weld heat input was modelled with the Goldak heat flux distribution. Welding experiments were carried out and the transient temperature distribution during welding was measured with thermocouples. This was used to calibrate the thermal analysis in LS-DYNA. Macrographs of the welded specimen helped to validate the fusion zone shape and cooling rate in the heat affected zone. The thermal analysis results were subsequently coupled with a mechanical analysis to calculate the thermal strains and residual stress formation. *MAT_270/CWM was used for this analysis as it is able to reproduce the transient weld material deposition. The residual stress over the full depth of the specimen was compared to the experimentally obtained residual stress state. The crack compliance method was used to experimentally measure the residual stress over the full depth of the specimen. The numerically and experimentally determined residual stress distributions are in good agreement. This study demonstrates the capabilities of LS-DYNA to simulate welding procedures and validates the corresponding results using physical experiments.

• Development and Implementation of a Composite Material Shell-Element Model

  Tobias Achstetter, Paul Du Bois, Kelly Carney, Cing-Dao Kan (George Mason University), Allen Sheldon (Honda R&D Americas), Sheng Dong (The Ohio State University), Gunther Blankenhorn (Livermore Software Technology Corporation)

  While the response to loading of traditional engineering materials, such as plastics and steel, is well understood and can be simulated accurately, designers of composite structures still rely heavily on physical testing of components to ensure the requirements of load bearing capabilities are met. The majority of composite material models that have been developed rely on non-physical material parameters that have to be calibrated in extensive simulations. A predictive model, based on physically meaningful input, is currently not available. The developed orthotropic material model includes the ability to define tabulated hardening curves for different loading directions with strain-rate and temperature dependency. Strain-rate dependency was achieved by coupling the theories of viscoelasticity and viscoplasticity to allow for rate dependency in both the elastic and plastic regions of the material deformation. A damage model was implemented, where a reduction of stiffness and stress degradation in the individual material directions can be tracked precisely. Modeling of failure and Finite Element erosion was achieved by implementing a new strain-based generalized tabulated failure criterion, where failure strains can be precisely defined for specific states of stresses. Composite materials are generally used in a layup of plies with different fiber directions. These individual plies are very thin, which leads to impractically small mesh sizes when modeled with three dimensional solid elements. The developed material model is, therefore, made available for shell elements. The presented material model is a step towards the goal of a truly predictive material model for composite materials.

• Development of a Regression Model for Blast Pressure Prediction in Urban Street Configurations

  Sungmin Lee (Michigan Engineering Services, LLC), Nickolas Vlahopoulos (University of Michigan), Syed Mohammad (Department of Homeland Security)

  Structural damage assessment due to explosive detonations in an urban setting requires a prediction tool for air-blast loads on buildings within the region. In this study, dead-end and cross-roads configurations are considered for blast wave simulations using Multi-Material Arbitrary Lagrange-Eulerian (MM-ALE) with mapping, with aim at developing a prediction model using regression analysis. Variations in street width and charge size and location are considered in constructing these street configurations. For all simulation models we use the uniform building height of 50 m and the identical street length of 50 m, and assume a vehicle bomb, meaning that a charge is carried by a vehicle such as pickup truck and detonates 1.25 m above ground. MM-ALE simulations with mapping, which is available in LS-DYNA®, will be used to achieve accuracy with reasonable amount of computational efforts. Mapping of solutions from 1D to 2D and then from 2D to 3D constitutes our three-step multi-material ALE simulation approach. 1D ALE analysis is performed for the spherically symmetric region between the explosive charge and the ground; 2D ALE analysis for the axi-symmetrical region from explosive location to closest wall; and 3D ALE analysis for the rest of the analysis domain. The ALE mapping approach is validated by comparing its simulation results to experimental data from literature. For the development of a fast running blast model we use regression analysis to estimate the relationship between an important blast simulation output variable (peak pressure) and input variables including street width, explosive size, explosive location, and type of street configuration. Regression analysis results are compared with actual simulation results.

• Drag Coefficient Optimization for a Sports Car Using the Coupling Between LS-DYNA® ICFD Solver, LS-OPT® and DEP MeshWorks Software

  M. Seulin, M. Le Garrec, A. Poncet (DynaS+), I. Çaldichoury (ANSYS LST), K. Gudlanarva (Detroit Engineered Product)

  Vehicle aerodynamics are one of the key points allowing to improve the vehicle dynamic behavior, to improve performance and to reduce fuel consumption. The vehicle aerodynamics have been studied in wind tunnels for several decades. Numerical simulations are increasingly used in addition of physical testing and permit to increase the number of design experimentations with cost and time savings. When CFD engineers are looking into optimizing the global aerodynamics of a car, numerous factors are taking into considerations. A car is a very complex assembly that must fit with multi-physical requirements updated along the vehicle project (design aesthetic, crash safety, weight, vibrations, noise, performances, design manufacturing, etc.) to find the best compromise according to initial specifications. DynaS+, ANSYS-LST and DEP are working closely with automakers across the world on various applications. Often, the automakers are sharing their work in conferences only several years after for obvious innovative competitive reasons. The aim of this work is to demonstrate what the current innovative technologies are, and methodologies used on aerodynamic applications using open source sports car data. Like in the majority of aerodynamic studies, in the present work, the objective was to reduce the aerodynamic drag coefficient of our model. A design optimization was performed on the initial design with the help of the advanced morphing capabilities of the DEP MeshWorks© solution coupled with the optimization software LS-OPT and the Incompressible Computational Fluid Dynamics (ICFD) solver LS-DYNA.

• Driving Through Flooded Road

  Bijoy Paul, Rachel Hysong, Babak Tehrani, Elizabeth Welch, John Davis, Amit Wavde (General Motors)

  Driving through flooded roads is always a challenge. Hydrostatic as well as hydrodynamic pressure can cause serious damage to the vehicle. Damage can adversely affect the performance of the vehicle in many ways. For example, high stress and strain can cause part failure, water ingestion into electrical components can lead to instant shutdown of the electrical system, corrosion, due to interaction with water can affect the performance and cosmetics of the vehicle. All of these can be costly fixes that are extremely dissatisfactory to our customers. General Motors has been designing and building state-of-the-art vehicles for more than a century. Safety, structural durability, part integrity, and performance are key features of every vehicle that General Motors produces. General Motors constantly invests in new technology and methods to improve quality, performance, and customer satisfaction. Smooth Particle Hydrodynamics (SPH) was developed in the late 70s. This mathematical advancement was transformed in the form of application in the recent past. The application has now been widely accepted by the CAE analysis community to study Fluid-structure-interface, water path analysis, and other hydrodynamic behavior, related to water and oil. General Motors has worked with LS DYNA® to improve performance issues of its vehicles in many areas of interests such as occupant safety, crash worthiness, structural durability and most recently, on water intrusion issues using SPH. This study involved structural durability analysis of a vehicle when driven on a flooded road. SPH particles were created to mimic the flooded road. A non-linear transient (crash) model was selected for the analysis. A node-to-surface contact was established between the SPH particles and vehicle. The vehicle was given an initial velocity of 30 km/hour, and the wheels were let to spin with the calculated rotational velocity. The LS-DYNA simulation was run for 400 milliseconds and plastic strain outputs were measured. A physical test was scheduled. Strain gauges and strain rosettes were affixed at the areas where computer simulation results were measured and recorded. The physical test was then performed at the General Motors Milford Proving grounds. Analysis results were then compared with the physical test results. In conclusion, a good correlation was observed between CAE (SPH analysis) and test results. SPH analysis is computationally very intense. Therefore, steps are being discussed to shorten the total computational time.

• Effect of Thickness Changes and Friction in Thermoforming Process Simulations in LS-DYNA® for UHMWPE Unidirectional Cross-Plies

  Kari D. White, James A. Sherwood (University of Massachusetts Lowell)

  This paper discusses the use of LS-DYNA for the modeling of thickness changes and frictions of DSM Dyneema® HB210, an Ultra-High Molecular Weigh Polyethylene (UHMWPE) unidirectional cross-ply thermoplastic laminate, during a thermoforming process. The thermoforming process being investigated consists of the preform phase that transforms the laminates to near net shape ply stacks and the subsequent consolidation phase that employs pressure and heat to join the preforms into a final part. During the preform phase, interply (ply/ply) and ply/tool frictions induce in-plane tension in the sheet of thermoplastic lamina. Knowing the effective tool/ply and ply/ply frictions such that the binder force can be prescribed is critical to preventing defects such as wrinkling, waviness and tears during the preforming process. The main mode of deformation of the laminate during the preform phase of the manufacturing process is in-plane shearing of the laminate, which can lead to variations in thickness. When multiple preform layers are compressed in the consolidation phase, the compounding of the thickness variations can adversely affect the uniformity of pressure distribution between matched die tooling, resulting in inconsistent consolidation. The modeling of the preform and consolidation steps can guide design changes in the processing conditions and ply blank geometries to achieve a well consolidated part. The temperature-dependent material properties derived from shear, bending, tensile and friction tests are implemented in a LS-DYNA simulation with a discrete-mesoscopic user subroutine for the material behavior of the cross-ply laminates. Beam elements capture the fiber orientations and carry the tensile and bending loads, while shell element exhibit the shear stiffness as a function of shear angle. The effect of thickness change of the laminate is investigated through the comparison of general shell elements without thickness change to thickness stretch shell elements (Elform=25) that change in thickness due to shearing, stretching and compression. The sensitivity to tool/fabric, as well as fabric/fabric, friction is also investigated in combination with thickness changes. Single-layer and triple-layer preforms are simulated and results produced support the need for both accurate friction inputs as well as including changes in thickness in the simulation.

• Electrochemical-Thermal-Mechanical Coupling of Lithium-Ion Battery Model in LS-DYNA®

  Kyoungsu Im, Z.-C. Zhang, and Grant Cook, Jr. (Livermore Software Technology, an ANSYS Company), Jaeyoung Lim (Hyundai Motor Group R&D Division), Kyu-Jin Lee (Myongji University)

  In this paper, we report a new development of battery-thermal-structure-interaction (BTSI) based on previously developed electrochemical Lithium-Ion models: i) a single insertion lithium metal model, and ii) a dual insertion composite model. In 10 cells of a lithium ion battery stack, each cell consists of Graphite(LiC6) anode/Separator/high performance layered LMO(LiMn2O4) or NCM(LiNi1/3Co1/3Mn1/3O2) cathode, which has been strongly proposed as a candidate for automotive batteries because of its high capacity, thermal stability, and low volume change rate (cycle performance). For the thermal-mechanical analysis, each layer in a cell and outside case are modeled corresponding to their material properties. Then, a rigid ball impacts center top position of the cell stack in order to investigate the thermal and mechanical responses of a lithium ion battery stack. To see the cell responses in different state of charge (SOC), we selected the first 20 second of the discharging processes. The results show that after the ball impact the cell stack, then the mechanical deformation started and 6 seconds after the ball compressed, a strong hot spot developed inside cell stack and the temperature increased exponentially over the melting point of the lithium, 453K. Although we demonstrated a simple impact problem to show how to simulate the electrochemical-thermal-mechanical problem, the current solver can be used to solve more practical problems such as a cellular phone drop test, notebook battery impacting test, and even deformation test of the scaled-up electric vehicle(EV) battery pack.

• Electrostatics and EM-ICFD Coupling in LS-DYNA®, a Glimpse of Things to Come

  Iñaki Çaldichoury Pierre L’Eplattenier (Livermore Software Technology, an Ansys company)

  The EM LS-DYNA solver’s primarily focus is on Electromagnetic metal forming, Inducting heating, Resistive heating. Recently, its capabilities have been extended in the domains of battery charge/discharges, electrophysiology and spot welding. However, there is a domain of applications that gets periodically inquired about and users sometimes wonder whether LS-DYNA possesses any capabilities in that area. This area would be electrostatics. As it happens, the existing resistive heating solver can be used for certain applications by proceeding with an analogy between the Poisson equation for electrostatics and the Poisson equation for resistive heating, itself a derivative of Ohm’s law. Still, electrostatics often involves the calculation of the Coulomb force which is a surface force typically acting on the parts of a capacitor and for which no option was available for the user to do a coupled EM-structure analysis. From this, the idea sprung to solve the EM fields on the same mesh as the one provided by the ICFD solver. Indeed, the ICFD solver specializes in fluid structure interaction problems and has got extensive capabilities in transferring forces from the fluid surface to the solid as well as advanced dynamic mesh movement tracking and adaptive remeshing. In this paper, the current capabilities allowing the merging of the EM and ICFD solvers will therefore be described. Going beyond the domain of electrostatics, further applications which combine the domains of electromagnetic and fluids such as Magnetic Hydrodynamics (MHD) and Magnetorheological fluid (MRF) will be discussed.

• Enhancement of Deformation Sub-Model in an Orthotropic Material Model

  Loukham Shyamsunder, Bilal Khaled, Subramaniam D. Rajan (Arizona State University), Kelly S. Carney, Paul DuBois (George Mason University), Gunther Blankenhorn (LSTC)

  A generalized tabulated three-dimensional orthotropic material model currently available in the dev version of LS-DYNA® as MAT_213 is enhanced with new features. MAT_213 has a modular constitutive model architecture consisting of deformation, damage and failure sub-models. The deformation sub-model has been enhanced with visco-elastic-plastic formulation with rate and temperature dependencies as well as strain-smoothing techniques to improve the stability of the analysis. Verification tests are carried out with experimentally obtained stress-strain curves at quasi-static and at higher rates of loading for the T800-F3900 unidirectional composite. Validation tests are carried out using data from high-speed projectile impacts on stacked-ply composite panels. Results show that the developed framework provides reasonable predictive capabilities.

• ES-2/re Model Validation on FAA Requirements for Aircraft Side Impact with LS-DYNA®

  Alexander Schif, Yupeng Huang, Sebastian Stahlschmidt (DYNAmore GmbH)

  For many years ES-2 and ES-2re dummy models are used in car side crash simulations. The use of the ES-2 and ES-2re dummy models in these simulations is precisely defined. Until recent work in the aerospace industry within Aerospace Recommended Practice (ARP) 5765 Revision B by SAE International (SAE) there were no instructions available for the exact use of the ES-2 and ES-2re dummy. SAE ARP 5765 Revision B aims for an easier seat certification process to fulfill Federal Aviation Administration (FAA) requirements giving best practice advice of how to work with ES-2re in side facing impact aircraft simulations. In connection with SAE ARP 5765 Revision B new side facing sled tests were performed by the FAA with special pulses. Based on these new side facing sled tests the DYNAmore ES-2 and ES-2re model was further validated to meet these new demands. With the end of the validation process the ES-2 and ES-2re V8 model, suited for car and aircraft side crash simulations, was released. Besides much better performance in the ARP side facing sled tests, also the overall performance of the already existing Partnership of Dummy Technology and Biomechanics (PDB) car side facing sled tests was increased. To support the visualization of the increasing performance of the ES-2 and ES-2re dummy CORA ratings were created for the last three release versions of ES-2 and ES-2re. The ratings are available for the new FAA sled tests, the PDB sled tests and all the certification tests of the dummy.

• Experimental Design for Negative Triaxialities: Ductile Fracture Under Combined Uniaxial Tension and Hydrostatic Pressure

  Robert L. Lowe, Luke D. Hoover, Christopher A. Negri (University of Dayton)

  Many modern continuum-scale approaches for modeling the ductile fracture of metals regard the equivalent plastic strain at fracture as a function of the stress triaxiality and Lode parameter, a pair of invariant-based quantities that together characterize the three-dimensional state of stress at a point. Generally, these ductile fracture models (whether parameterized or tabulated) are calibrated using standard mechanical tests, e.g., notched axisymmetric (round), plane stress (thin), and plane strain (thick) specimens subjected to tensile loading. However, these standard tests are only able to capture a limited window of stress states, leaving potentially important “regions” of the ductile fracture model unpopulated with experimental data. For instance, although previous research has suggested that fracture will not occur below a triaxiality of 0.33 (the “cut-off” value), recent ballistic impact simulations involving 0.5-inch-thick titanium Ti-6Al-4V target plates predicted large negative (compressive) triaxialities in the vicinity of the adiabatic shear band. These results not only suggest the potentially unanticipated importance of the negative triaxiality (compressive) region of Lode-triaxiality stress space, but also the need to experimentally revisit previous interpretations of the “cut-off” value of the triaxiality. As a first step, this paper presents a novel physical interpretation of the Lode parameter = 1 (constant) meridian over a range of triaxialities, spanning positive (tensile) to negative (compressive). Guided by this physical insight, ductile fracture experiments that employ hydrostatic pressure superposed on uniaxial tension are proposed and numerically simulated in LS-DYNA®, with initial efforts focusing on 2024-T351 aluminum. Our numerical simulations provide a promising “virtual” proof-of-concept demonstrating that stress triaxiality can be tuned (at constant Lode parameter) by adjusting the magnitude of the applied pressure, allowing a wide range of stress states to be accessed through a single experimental test setup.

• Finite Element Modeling of Reconstructed Vehicle Rear Seats with Adult Male ATDs

  Keegan Yates, Costin Untaroiu (Virginia Tech)

  Most car crash fatalities occur in the front seats, so experimentation and regulations involving car crash occupant protection typically focus on the front seats. Because of this, the safety of the front seats has increased greatly over the years, and in some circumstances, the front seats now perform better than rear seats. This represents a problem because the rise of ridesharing transportation and automated driving systems has the potential to increase rear seat occupancy by adults, which could result in an increase in injury and death. To help inform the design of new vehicle rear seat safety systems, it is important to understand the performance of current vehicle rear seats with adult occupants. The rear seats of eight vehicles were reconstructed from scans of the seat surfaces as well as the seat pan and seatbelt components. Seat foam material properties were taken from quasistatic tests of each seat. The THOR and Hybrid III male 50th percentile ATD FE models were positioned and settled in each seat. The vehicles frontal NCAP crash pulse as well as a less severe pulse were applied to each vehicle in LS-DYNA®. Injury likelihood was assessed by a summary of the AIS3+ risk curves for the head, neck, chest, and femurs. Overall, the results with a frontal NCAP pulse ranged from a near certainty of AIS3+ injury to around a 35% chance. Additionally, the best performance was seen with vehicles that contain pretensioners and load limiters in the rear seats. These results indicate that such technologies may be necessary in the rear seat to improve crash performance. Additionally, these results have helped select a range of vehicles for further experimentation and identified variables of interest for further simulation.

• FSI Based on CESE Compressible Flow Solver with Detailed Finite Rate Chemistry

  Kyoungsu Im, Grant Cook, Jr., Zeng-Chan Zhang (Livermore Software Technology, an ANSYS Company)

  We have developed a new module of the modeling fluid structure interaction with the finite-rate chemistry in compressible CESE solver, which is based on the immersed boundary FSI method, and fully coupled with the LS-DYNA® structural FEM solver. In the CESE fluid structure interaction solver, we have two principal treatment methods, i.e., the immersed boundary method with a direct-forcing strategy and the moving mesh method. Although the moving mesh method is more accurate than the immersed boundary method, the latter is most efficient and robust when the problem involves large deformation such as a structure demolition by explosion. In the present report, we have demonstrated most practical fluid structure interaction problems by using the immersed boundary method with chemistry: i) shock-induced combustion in front of a spherical projectile moving at supersonic speed, ii) the blast relief wall simulation in methane and air mixture (CH4/Air), and iii) the fracture of the shell and solid structures by high explosive spots in an H2/O2 premixed environment. The results are validated with existing experimental data and descriptions of the keyword setup are provided in detail for users.

• Further Validation of the Global Human Body Model Consortium 50th Percentile Male Pelvis Finite Element Model

  Daniel Grindle, Yunzhu Meng, Costin Untaroiu (Virginia Tech)

  Road traffic accidents are the eighth leading cause of death worldwide, killing 1.35 million annually. The Global Human Body Model Consortium (GHBMC) has previously created and validated a finite element model of a 50th percentile male pedestrian in LS-DYNA® to investigate vehicle-pedestrian impacts. To assure and improve the model’s biofidelity, additional model improvements were made to the GHBMC pelvis model. These pelvis developments included the addition of acetabular cartilage and the optimization of material properties. The updated pelvis model was calibrated against Post-Mortem Human Surrogate (PMHS) component tests: dynamic lateral acetabulum loading, dynamic lateral iliac wing loading, and quasi-static sacroiliac joint loading. After new material properties were established for the pelvis model, the updated properties were applied to the whole-body GHBMC model. The updated model pelvis injury response was validated against whole-body PMHS lateral vehicle impact tests. More biofidelic biomechanical responses were observed in the updated pelvis model in the majority of component level validations. In addition, the fracture patterns of the updated pelvis matched the PMHS fracture patterns in whole-body impacts. This updated pelvis model will be incorporated into the next generation of GHBMC models. In future, it can be used to properly investigate pelvis injury mechanisms in impact scenarios to reduce pedestrian injuries in traffic accidents.

• GISSMO Based Material Characterization Using Workflows in d3VIEW

  Shashank Dhanakshirur, Apurva Walke (d3VIEW), Suri Bala (d3VIEW and Ansys/LST), Paul DuBois (Consultant)

  In times when the automotive industry is trying ways to reduce the R&D time for a new vehicle, material characterization using Workflows provide a structured way to develop a material card quickly and efficiently. GISSMO Material Development workflow greatly reduces the time required to generate and edit LS-DYNA® input files, visualize the results and automate decision-making aimed for optimization. The workflow is divided into sections as hardening curve optimization, Triaxiality curve optimization, mesh regularization, evaluation runs and result generation to make the challenging and complicate process of developing material model seem easy and user-friendly. Advantages of the workflow are timesaving, ease of use, efficient and flexible to adapt to the requirements of different materials.

• Hybrid III 95th Percentile Large Male Finite Element Model Neck Alteration

  Eric Day, Christoph Maurath Sommer (Livermore Software Technology), Jacob Putnam, Chuck Lawrence, Preston Greenhalgh (NASA)

  The motivation behind the project was to update the Livermore Software Technology Corporation Hybrid III 95th percentile finite element model, such that the neck assembly response under varying simulated loading conditions equals that of the federally regulated Hybrid III 95th percentile anthropomorphic testing device (ATD). The former neck model was poorly correlated to that of the physical Hybrid III neck in corresponding tests. Adjustments were made to mass and geometry, element formulation, and element discretization to improve model durability and accuracy. Test data from a physical compression test and NASA-performed Neck Sled Tests were collated with data from simulation to adjust material properties. The neck rubber material was further calibrated according to Code of Federal Regulations (CFR) neck calibration test response requirements. The resulting neck model developed in LS-DYNA® exhibited improved dynamic characteristics and reliability under both low and high severity loading. Computational efficiency was enhanced along with model stability under excessive loading. The revised neck model will be adopted by NASA for use in predicting potential occupant injury during spacecraft landing.

• Immersed Interface Development in Incompressible CFD

  Facundo Del Pin, Iñaki Çaldichoury, Rodrigo R. Paz, Chien-Jung Huang (Livermore Software Technology LLC)

  The pre-processing of complex geometries exported from CAD programs is a big challenge in the Finite Element analysis of fluid problems. There are many situations where a detailed high quality mesh is preferred and possible mandatory. Such is the case for problems where shear stresses are an important component of the total force, i.e. ground vehicle aerodynamics, aircraft drag prediction and some bio-mechanics applications where the stresses in the endothelium are need to predict the development of some diseases. There are many other applications though where pressure forces are enough in terms of accuracy or where rapid prototyping of engineering parts do not need the accuracy required in the final stages of engineering design. In these cases, the geometry could be simplified by approximating the domain walls immersing them inside a much simpler domain. This simplification becomes even more appealing in the presence of an internal structure that interacts with the fluid which in many engineering applications are modeled as thin shell structures. In the current work the sub-element interfaces of the geometry will be approximated by level set distance functions. The walls could be part of a flexible structure or they could be rigid and they may be "wet" on both sides. The pressure discontinuity across the wall (in the case of shell structural elements) will be approximated by discontinuous shape functions as described in [1]. One of the main advantages of this approach is that it is easily adapted to an existing solver since no additional degrees of freedom need to be added. The presentation will include details of the additions that the existing solver needed such as: 1) boundary recognition; 2) level set representation; 3) sub-element splitting; 4) computation of the new interpolation functions and integration; 5) assembly and solution.

• Incompressible Smoothed Particle Galerkin (ISPG) Method for an Efficient Simulation of Surface Tension and Wall Adhesion Effects in the 3D Reflow Soldering Process

  Xiaofei Pan, C. T. Wu, Wei Hu (Livermore Software Technology (LST), an Ansys Company)

  A new numerical method in LS-DYNA®, the incompressible smoothed particle Galerkin (ISPG) method, is developed for the simulation of shape evaluation of solder joints in electronic equipment during the reflow process. The ISPG method is aiming to suppress key numerical instabilities observed in the simulation of incompressible free surface fluid flow using strong form Lagrangian particle methods such as SPH. In ISPG method, a momentum-consistent smoothing algorithm is utilized to offer the desired numerical stability associated with the velocity field in the fluid particle integration scheme. To stabilize the pressure field in Navier-Stokes equations, a second-order generalized rotational incremental pressure-correction scheme is developed for the incompressible fluid flows. To simulate the shape evolution of solder joints during the reflow process, a numerical procedure considering the surface tension and wall adhesion effects is introduced. Several numerical examples are studied to demonstrate the accuracy and the efficiency of the new method.

• Incremental Damage Model for Fatigue Life Assessment in Complete Machinery Simulation

  Marcus Lilja, Jesper Karlsson, Anders Jonsson, Daniel Hilding (DYNAmore Nordic AB), Stefan B. Lindström, Daniel Leidermark, Peter Schmidt (Linköping University)

  In CAE today a transition towards “complete machinery simulation”, away from the traditional component or sub-assembly simulation, is seen. The complete, assembled and pre-loaded machine is simulated with real loads and boundary conditions which minimizes the risk of errors in the boundary conditions and loading. The longer simulation time is mitigated by the reduction in the number of load cases needed and that a single simulation yields the results for all components. This “complete machinery simulation”-approach is not new, e.g. in the automotive industry LS-DYNA® has been used for realistic simulations for many years and this approach has now reached other industry sectors as well. When developing e.g. heavy industrial equipment, static strength is not a common failure mode, but fatigue is. Fatigue life estimation of a product is crucial and since fatigue tests are both expensive and time-consuming there is a need for accurate fatigue simulation methods. Fatigue analysis within the CAE-process is commonly based on the rainflow count method for cycle counting and the Palmgren-Miner’s linear damage accumulation model. The fatigue life prediction is performed on the result history from a previous analysis and is dependent on the output frequency so that all peaks and valleys of the result variation are identified. This method is widely used and is well-suited for most of the common fatigue scenarios today. However, when using complete machinery simulation, shortcomings in the above method have been identified to be caused by the combination of very large models, high frequency output, and non-proportional loading. This tends to result in a great amount of data for the subsequent fatigue analysis. The amount of data makes post processing and fatigue analysis cumbersome and since development is an iterative process, disk space may become a critical factor. This paper presents an implementation of the incremental fatigue model of Ottosen and co-workers [Int. J. Fatigue, 30:996-1006 (2008)] as a user-material for LS-DYNA. The model offers a uniform framework for multiaxial, non-proportional and non-cyclic loading. With this model, the fatigue assessment is made on the element level during the simulation. The model enhances performance in terms of faster integration, less data storage, and easier usage. A comparison of the fatigue life predicted using the new method to the standard rainflow count method for selected grades of steel and aluminum is presented.

• Intelligent Multiscale Simulation Based on Process-Guided Composite Database

  Zeliang Liu, Haoyan Wei, C.T. Wu (Livermore Software Technology LLC), Tianyu Huang (Livermore Software Technology LLC / Northwestern University)

  In the paper, we present an integrated data-driven modeling framework based on process modeling, material homogenization, mechanistic machine learning, and concurrent multiscale simulation. We are interested in the injection-molded short fiber reinforced composites, which have been identified as key material systems in automotive, aerospace, and electronics industries. The molding process induces spatially varying microstructures across various length scales, while the resulting strongly anisotropic and nonlinear material properties are still challenging to be captured by conventional modeling approaches. To prepare the linear elastic training data for our machine learning tasks, Representative Volume Elements (RVE) with different fiber orientations and volume fractions are generated through stochastic reconstruction and analyzed using the LS-DYNA® RVE package. More importantly, we utilize the recently proposed Deep Material Network (DMN) to learn the hidden microscale morphologies from data. With essential physics embedded in its building blocks, this data-driven material model can be extrapolated to predict nonlinear material behaviors efficiently and accurately. Through the transfer learning of DMN, we create a unified process-guided material database that covers a full range of geometric descriptors for short fiber reinforced composites. Finally, this unified DMN database is implemented and coupled with macroscale finite element model in LS-DYNA to enable concurrent multiscale simulations. From our perspective, the proposed framework is also promising in many other emergent multiscale engineering systems, such as additive manufacturing and compressive molding.

• Introducing Arup-Cellbond MDPB Shell Model

  Laura Rovira Crespo, Mattia Bernardi, Dong-Ling Li, Zhao Bao, Zhi Zou, Yi-Ning Ding, Maya Shinozaki, Francois Lancelot (ARUP)

  This Mobile Offset Progressive Deformable Barrier (MPDB) for frontal impact model has been developed to take advantage of the latest developments in the LS-DYNA® code and is designed to provide robust and efficient analysis. In this paper, some details of the calibration and validation process will firstly be presented, which not only satisfies performance requirements set by regulations (Euro NCAP 2020 Dynamic Tubular Impactor test) but also goes beyond through rigorous calibration against other physical tests (Vertical/Rounded Impactor Test, Quarter Wall test etc). Arup and Cellbond worked closely with Jaguar Land Rover in the UK for the development of this barrier model through ensuring correlation to full speed, real vehicle tests. Finally, methods of how to automate post-processing of results (including the calculation of Euro NCAP compatibility modifier) using the Oasys Software will be demonstrated.

• Introduction of Sliding Capabilities in the ICFD LS-DYNA® Solver

  Iñaki Çaldichoury, Chienjung Huang, Facundo Del Pin, Rodrigo Paz (Livermore Software Technology, an Ansys company)

  Sliding mesh is a technique that prevents excessive re-meshing in problems that involve rotating parts. It is ideal for solving transient problems in turbo-machinery. Overset mesh techniques on the other hand typically contain the body of interest for the study around which a fine fluid mesh in constructed. That initial domain is then superimposed on a background mesh containing the surrounding geometry with data being interpolated between the two. Other techniques include using a non-inertial rotating frame or using immersed FSI techniques. Within LS-DYNA, the ICFD solver has seen a continuous growth of users that wish to simulate increasingly complex multiphysics problems involving moving structures, thermal heating, particle displacement and sometimes magnetic fields. As such, it is imperative to offer as many advanced CFD capabilities and solving tools to the users as possible. Among those, sliding mesh has been amongst the most prominent requests. In this paper, the current state of development will be presented, along with a description of the algorithm used as well as some examples and some benchmarking results.

• Investigation of Mesh Regularization in MAT_224 for Subsequent Use in Impact Simulations

  Troy Lyons, Kiran D’Souza (The Ohio State University)

  This work is focused on the use of mesh regularization, which is an attempt to remove the mesh dependence from finite element simulations. Mesh regularization techniques are often used to help match experimental data and allow for reduced computational cost. The MAT_224 material model within LS-DYNA® allows users to define a failure criterion that is dependent on temperature, strain rate, stress state, and element size. The element size dependence in the failure criterion can help reduce the influence of mesh size on simulated results under certain circumstances. However, some issues may arise when the MAT_224 material model is applied to different geometries, stress states, and element sizes than the regularization curve was originally created from. In this work, the conditions to best use mesh regularization are investigated, which is done with various comparisons using experimental and simulated data, both with and without mesh regularization.

• Investigation on Transversal Anisotropy of an Aluminum Sheet for Crash Applications

  F. Andrade, C. Wilking, D. Koch (DYNAmore GmbH), M. Feucht (Mercedes-Benz AG)

  In this paper, we concentrate our efforts on the simulation of an aluminum sheet material used in the automotive industry. A series of experiments using samples with different geometries is performed in order to characterize the material under different stress states. Plasticity is considered using a von Mises based and a Barlat based material model (respectively, *MAT_024 and *MAT_036 in LS DYNA®). For the Barlat-based model, it is assumed that the R-values are the same for all material directions, a suitable assumption for 6000 aluminum sheets. This means that anisotropy is only present through the thickness and not in the plane of the material. In turn, this allows a more straightforward usage of *MAT_036 in complex parts for which no mapping of material directions have to be undertaken because the thickness direction for shell elements is known a priori. Comparison with experimental data (including strain fields measured with DIC) shows that this strategy leads to a somewhat better description of the material deformation observed in physical tests when compared to the predictions of the isotropic model *MAT_024. Finally, the GISSMO failure/damage model is adopted for the failure description in LS-DYNA. It is shown that the numerical results agree very well with the experiments and not only the global force-displacement curve but also the local strain fields.

• J-Composites/Compression Molding - Introducing New Simulation System for FRP Composites

  Shinya Hayashi, Shaun Dougherty, Shinya Hiroi, Yoshida Atsushi (JSOL Corporation)

  Composite materials like fiber reinforced plastics (FRP) are becoming more widely used in the automotive industry and have been found very effective in reducing vehicle weight. Recently, discontinuous long carbon fiber reinforced plastics are increasingly used for lightweight structural parts with high stiffness, strength and energy absorption performance. Compression molding is considered one of the most efficient manufacturing processes to mass produce FRP parts for automotive applications. Compression molding can form discontinuous long fiber reinforced plastics into complex shapes with relatively low manufacturing cost and short process time. LST and JSOL developed new compression molding simulation techniques for discontinuous long fiber reinforced plastics using a beam-in-solid coupling function in LS-DYNA®. Then JSOL developed a modelling tool called J-CompositesⓇ/Compression Molding to generate an input deck for this new compression molding simulation. In this paper, main features of J-Composites/Compression Molding are introduced and the latest compression molding simulation result of a large scale component model created by J-Composites/Compression Molding is presented.

• Lithium-Ion Battery Multi-Physics Simulations Using LS-DYNA®

  Jie Deng, Chulheung Bae, Theodore Miller, Min Zhu (Ford Motor Company), Pierre L’Eplattenier, Inaki Caldichoury (Livermore Software Technology LLC)

  The market share of electrified vehicles grows rapidly in recent years. One of the top priorities in the electrified vehicle design is to improve the robustness of lithium-ion battery system during a crash. Various abuse tolerance tests have been developed to evaluate the performance and robustness of lithium-ion batteries. These tests can be resources intensive, and in some cases, provide limited information on the failure mechanisms of batteries. As such, computational modeling becomes an important tool to evaluate the battery under different abuse scenarios. Here we present a multi-physics battery model that can predict coupled mechanical, thermal, electrical and electrochemical responses of automobile lithium-ion batteries under abusive conditions. In this model, the electrochemical behavior of batteries is described by a spatially distributed equivalent circuit model, where polarization and damping effects are captured by a resistance-capacitance network. During simulations, the mechanical solver predicts the onset of short circuit, and then the coupled thermal, electrical and electrochemical solver captures the evolution of temperature, voltage and current distribution after short circuit initiation. In order to make the proposed model applicable to module or pack level simulations, various element formulations and strategies have been developed to improve computational efficiency without scarifying much accuracy. Details of model set up, parameters evaluation, and case studies that demonstrate the model capabilities will be presented. Experimental validation of model prediction and the future development of this framework will also be discussed.

• Load Balancing Update

  Brian Wainscott (LSTC)

  One of the keys to efficient parallel processing is having an evenly distributed workload so that every processor has the same amount of work to do. Because there are unavoidable synchronization points in the simulation process, when one processor has more work then the other processors will have to wait, which wastes CPU time. LS-DYNA® has many options available for controlling the initial distribution of elements to processors to help achieve a good load balance. Unfortunately, for many simulations the distribution of work changes during the run. This can make it nearly impossible to have a good load balance over the whole simulation with a static decomposition. The capability to move nodes and elements between processors during the simulation to maintain good load balance has been under development for some time. The current state of this ongoing work is presented.

• LS-DYNA® Material Model 263 and Its Application to Earing Predictions in Cup-Drawing

  Jinglin Zheng, Xinhai Zhu (Livermore Software Technology, an ANSYS company), Yanshan Lou (Xi’an Jiao Tong University), Saijun Zhang (South China University of Technology), Jeong Whan Yoon (Korea Advanced Institute of Science and Technology (KAIST)

  This paper introduces a newly implemented metal forming material model, material type 263, in LS-DYNA material library. The yield function of this model is based on a recent theoretical development of extending the original Drucker function into an anisotropic form. The flexibility of the yield function is further improved by adopting the non-associated flow rule. The paper also outlines how to use LS OPT® to calibrate the material parameters used in the model, followed by a cup-drawing analysis which demonstrates the model’s capability of capturing the cup earing profile, especially when paired with LS-OPT for material parameter identification.

• Mesomechanical Modeling of the Mechanical Behavior of Parachute Suspension Lines using LS-DYNA®

  Catherine P. Barry, Scott E. Stapleton, David J. Willis, James A. Sherwood (University of Massachusetts Lowell), Francesco Panerai (University of Illinois at Urbana Champaign), Keith Bergeron (U.S. Army Corps of Engineers), Christine Charette, Gregory Noetscher (United States Army Combat Capabilities Development Command-Soldier Center)

  Parachute suspension lines can develop vortex-induced vibrations while in flight, degrading the flight performance and creating noise. To understand the braid design factors associated with this vibration, the mechanical behavior of the braided parachute suspension line can be investigated using a fluid-structure interaction (FSI) analysis. Such an FSI analysis requires a well characterized macroscopic model of the axial, bending and torsional stiffnesses of the line. In the current study, a novel model using a combination of truss elements embedded in solid elements for capturing the asymmetric axial tension-compression stiffness of a tow as well as the lateral compression is presented. Tensile and transverse compression experiments were performed on the individual tows to characterize the axial and transverse stiffnesses to be used with the finite element model. Finite element models of the tow material characterization tests were compared with experimental data to calibrate the material model parameters and to validate the modeling approach. The mesoscale of a suspension line was resolved under zero load using X-ray computed tomography and data were used as the baseline image to generate a finite element model of a representative unit cell of the line. The Virtual Textile Morphology Suite (VTMS) and LS PrePost® were used to obtain the geometry and mesh, respectively. The feasibility to use the novel modeling approach for capturing the mechanical behavior of the braid is demonstrated.

• Modeling and Validation of Failure Behaviors of Composite Laminate Components using MAT_262 and User Defined Cohesive Model

  Masato Nishi1, Masato Iimori, Kei Saito (JSOL Corporation), Tsuyoshi Nishihara, Chikara Kawamura, Shunsuke Kanemoto, (Mazda Motor Corporation), Stefan Hartmann (DYNAmore Corporation)

  The objective of the present study is to develop a finite element (FE) approach to predict the changes in failure behavior of a unidirectional carbon fiber reinforced plastic (CFRP) material for different laminate configurations in LS-DYNA®. Damage related parameters for an intra-lamina material model are often adjusted by reverse engineering. However, in our study, we identified these parameters in material type 262 based on a crack resistance curve, which shows the relationship between fracture toughness and crack length. A user defined cohesive zone model was also developed to take into account anisotropic inter-laminar fracture toughness depending on the fiber orientation. The changes in fracture behavior observed in the different laminate configurations in experiments can be represented in four-point bending simulations of a CFRP laminated component using the developed FE model.

• Modeling of Vehicle Fuel via Smoothed Particle Hydrodynamics (SPH) Method in LS-DYNA® for Vehicle Crash Virtual Simulation

  Tushar Phule, Scott Zilincik (FCA US LLC)

  This paper describes advantages of modeling a fluid in a vehicle fuel tank using the Smoothed Particle Hydrodynamics (SPH) method in vehicle crash computer aided engineering (CAE) simulations in order to achieve appropriate fuel/fluid behavior during a crash event. SPH is a mesh free Lagrangian particle fluid modeling technique used for simulating fluid flows, whereas, the legacy CAE modeling method uses a solid tetra element mesh with MAT1F MAT_ELASTIC_FLUID to model the fluid in the fuel tank. The SPH method has many advantages over the legacy modeling method in terms of capturing important fuel tank responses, such as: correct tank internal fluid pressure, proper tank shell deformation, and tank clearances to the surrounding environment. Accurate simulation of these are important to meet NHTSA FMVSS 301 and fuel system integrity requirements. To compare the responses from the fuel tank, a study has been carried out by comparing SPH and legacy CAE methods with a physical test. The internal fluid pressure at the fuel tank control valve from a 35 mph flat frontal rigid barrier impact CAE model was plotted against a physical sled test that corresponds to the 35 mph flat frontal rigid barrier impact event. It was found that the SPH method provides better correlation over the legacy modeling method. In addition, tank shell deformation and tank clearances for both methods were compared with the physical sled test; It has been observed that the SPH method provides a more accurate tank shell deformation when compared to the legacy modeling method. There is an increase in computation time for the SPH method, however, this method ensures the result accuracy during fluid structure interaction.

• Modeling Plastic Clips in LS-DYNA® for Low-Energy Impact Analyses

  Kenneth E. Freeman, Alexander Gromer (DYNAmore Corporation), Brian O’Hara, Cameron O’Keeffe (Honda R&D Americas, Inc.)

  Through several different low-energy automotive impact simulations, it was discovered that capturing plastic clip behavior played a substantial role in predicting the system response. Therefore, a methodology for modeling plastic push-in rivets and snap-fit clip connections was developed in LS-DYNA for use in these low-energy automotive impact analyses. The required geometric discretization, contact definitions, material models and constraints that make up the models are discussed in detail. Pull-out force data was utilized to correlate the response and failure modes of the clip models. In addition, three different levels of clip model complexity were compared with respect to their suitability for different load cases. Simple clip model approaches were easy to pre-process and sufficiently captured most of pull-out failure modes. However, these did not capture shear or off-angle failure. More complex clip models sufficiently captured shear and off-angle failure, but come at a greater pre-processing and development effort. Lastly, some pre-processing methods are discussed to demonstrate how hundreds of clips can be incorporated in a model in very little time.

• Modified Dynamic Time Warping for Utilizing Partial Curve Data to Calibrate Material Models

  Katharina Witowski (DYNAmore GmbH), Nielen Stander (Livermore Software Technology, Livermore)

  Material calibration can be solved as a non-linear regression problem in which a parametric model of the material test is calibrated to its experimental result. In typical material testing, temporal and/or spatial data is produced experimentally and compared to its computational equivalent. At its basic level a single response comparison consists of two curves which are matched to produce a distance between them. The calibration requires minimizing the distance measure. The difficulty of the comparison is determined by phenomena such as noise, hysteresis and differences in geometric curve length (length compatibility). While noise and hysteresis problems have been solved in this context using LS-OPT® in the distant past, the question of curves having substantially different lengths has remained a challenge until recently. In one example, the computational output extracted from LS-DYNA® causes parts of the output to not be relevant to the test data. In this case most distance measures produce spurious distance calculations. This paper introduces a method to address this question. The approach is based on a modification of the Dynamic Time Warping distance measure and referred to as Modified Dynamic Time Warping or DTW-p (for partial). It consists of the trimming of the DTW path as well as iterative mapping to produce a uniform map. An example based on output of the GISSMO model in LS-DYNA is used to demonstrate the effectiveness of the method.

• MPDB Pre- and Postprocessing in Generator4 and Animator4

  Leyre Benito Cia, Christian Mospak, Christoph Kaulich (GNS mbH)

  While the development of modern cars reduces the risk of injuries for the occupants in a lot of different load cases which are well-defined for different scenarios, the risk to get injured when hitting another car with a partial overlap between the two vehicles is still high. The structure of the two involved cars is not able to completely absorb the energy from the occupants. Around 2010, a new type of barrier and test procedures were developed in order to simulate this type of accident. A moving honeycomb barrier hitting a driving car with an offset, should help improving modern cars. This test is used as procedure for different NCAP organizations around the world from 2020. Generator4 and Animator4, the FEA pre- and postprocessors from GNS mbH, can help the engineer to set up, start and evaluate simulations of MPDB barriers, calculating the loads on the occupants and the deformations in the barrier.

• Multi-Material Topology Optimization in LS-TaSC™ Using Ordered SIMP Interpolation

  Satchit Ramnath (The Ohio State University), Mariusz Bujny, Stefan Menzel (Honda Research Institute Europe GmbH), Nathan Zurbrugg, Duane Detwiler (Honda Research & Development Americas)

  Topology optimization allows for the design of structures with an optimum distribution of material for a given set of load cases that may have conflicting requirements. Though the methods used in topology optimization help to generate better and novel designs, they are generally limited to a single material only. In contrast, modern vehicle structures are composed of parts made of multiple materials, exhibiting usually superior performance compared to single-material designs. In order to support the design process of such structures, this problem requires the ability to use multi-material optimization methods within commercially available software like LS-TaSC, to optimally distribute multiple materials within a single design domain. In this paper, a method for integration of the ordered SIMP in LS-TaSC to realize multi-material TO is proposed and evaluated using a solid beam that is subject to static and crash load cases. The optimization uses the updated material distribution, based on ordered SIMP, to assign material/density values to elements in the design domain. To demonstrate the potential of the multi-material TO and validate the results, the obtained topologies are compared to single material designs as well as to the structures optimized using the state-of-the-art gradient-based approach based on ordered SIMP. The results show that LS-TaSC can be successfully used for deriving multi-material structures superior to the single-material designs. Finally, due to the low computational costs, the method seems to be suitable for the optimization of large-scale industrial models.

• Multiaxial Fatigue Analysis with LS-DYNA®

  Yun Huang (Livermore Software Technology, an ANSYS company), Anders Jonsson, Marcus Lilja (DYNAmore Nordic AB)

  Fatigue life is an important dimensioning criterion within product development. Several tools and software are today available and are widely used for fatigue assessment within the CAE process. To further improve the capabilities for integrated fatigue analysis in LS-DYNA, a time domain fatigue solver has been developed and implemented by LST (an ANSYS company), as a compliment to the already existing frequency domain fatigue solvers. As of coming releases of LS-DYNA, different options for fatigue analyses will be available, based on the results from general load cases and structures including e.g. non-linearities, non-proportional and multiaxial loading conditions. The time domain fatigue analysis can be based on stress or strain results from all time domain solvers (implicit, explicit, thermal, FSI, etc.) in LS-DYNA. The stress or strain state of the elements is usually three dimensional, especially for the parts under multiaxial loading cases like bending or twisting. However, the standard procedure to obtain the SN curve or EN curve is based on nominal stress or strain of the specimen, which is a scalar not a tensor. Several options to deal with the multiaxial stress state for fatigue analysis have been implemented in LS-DYNA (keyword *FATIGUE_MULTIAXIAL). They include 1. Running fatigue analysis based on an equivalent stress index (e.g. von Mises stress); 2. Running fatigue analysis on multiple planes and picking the highest damage ratio across the planes as the fatigue damage ratio of the element; 3. Locating a critical plane first and projecting the whole stress history to the critical plane and then running fatigue analysis on the critical plane. Several examples are given in this paper, to discuss the different options for multiaxial fatigue analysis, including a crankshaft model and a cylinder bar model with a groove. Validation has been performed by comparing the simulation results from simple test cases to analytical results from the same load cases. Also, a comparison of the fatigue analysis results from LS-DYNA to the results from the fatigue postprocessing module mFAT (a plug-in to the post-processor META) is presented in this paper.

• Multiphase Flow CESE Solver in LS-DYNA®

  Zeng-chan Zhang, Grant Cook, Jr., Kyoung-su Im (Livermore Software Technology, an ANSYS Company)

  In this paper, we will introduce a new capability of multiphase flow simulations in the LS-DYNA CESE compressible solvers. It is a hybrid multiphase flow model proposed by L. Michael [1]. This model is targeted for high-speed explosions, especially shock-to-detonation transition in liquid nitromethane. While the space-time conservation element and solution element (CESE) method, originally proposed by Chang [2], is designed for solving compressible flows, it is especially good for high-speed flows with complicated flow patterns. So we will use the CESE method to solve this hybrid multiphase flow model, and this approach will avoid a lot of complicated and time consuming treatments such as Riemann solvers and the Strang-splitting that are used in Ref.[1]. Our numerical examples show that we can get similar results using the CESE method. In the next sections, we will first give a brief introduction to the hybrid multiphase model, then the CESE method. Finally, we will give some numerical examples.

• New Design Considerations for the Calibration of Rubber-Like Materials

  Y. Lev, K.Y. Volokh (Technion Institute of Technology), A. Faye (Indian Institute of Technology)

  Rubber-like materials have unique characteristics that make them industrially very attractive. These materials can reach up to stretches of about 6-7 while staying hyper-elastic. Only a few studies that deal with these high level up to failure stretches are available. In addition, information regarding the temperature effect on fracture is absent in the literature. This work summarizes some recent aspects regarding the calibration of rubber-like materials reaching ultimate deformation (strength) under high operating temperatures. Our approach to modelling rubber fracture is based on the elasticity with energy limiters theory. Constitutive relations have been developed to generalize this description in order to include the thermo-elastic behavior. A relation for the temperature dependent energy limiters, and a new form for the thermal energy contribution are offered. The presented theory is used for calibration of rubber-like materials using LS-DYNA®. Our work also includes the design and set-up of a homemade test chamber for the uniaxial and bulge test (inflation of balloon test) cases. These tests are subjected to temperatures in the range of 25℃ to 90℃. The equi-biaxial conditions are extracted indirectly from the bulge test data by performing iterative finite element simulations that are done up to a sufficient fit to the bulge experiment results. The importance of a simultaneous calibration using both uniaxial and biaxial load cases together is highlighted. Material parameters found are significantly better than the parameters extracted by rubber manufacturers and labs that usually use uniaxial tests in room temperature only. The methodology used allows the correct modeling of ultimate properties as a function of high common operating temperatures for rubber-like materials. The findings can serve as new design considerations for engineers using these materials.

• New Generation Iterative Solvers in LS-DYNA®

  Cleve Ashcraft, Roger Grimes, Robert Lucas, François-Henry Rouet (Livermore Software Technology)

  Iterative solvers for sparse linear systems are used as the default option in a few places in LS-DYNA, e.g., thermal analysis and incompressible fluid flow. They are also available as a non-default option for implicit mechanics, electromagnetics, and acoustics. Until now, our suite of iterative solvers was limited to a few simple solvers (Conjugate Gradients and GMRES), and a few simple preconditioners. We recall that a preconditioner is an approximate inverse of the matrix (e.g., the stiffness matrix) that aims at improving convergence. Our preconditioners were limited to simple techniques like diagonal scaling and basic domain decomposition techniques that discard the coupling terms between processors, in MPP. Problems from customers are growing faster than memory size, making it difficult to use direct solvers. They are also often too numerically challenging to use simple iterative solvers, in particular in implicit mechanics. This has pushed us to revisit our suite of iterative solvers and preconditioners. In particular, we have been investigating the use of Block Low-Rank factorizations (BLR) and the use of Algebraic Multigrid (AMG). In the talk, we will compare these new options across all the different applications that make use of linear solvers. We will discuss convergence, memory usage, and scalability. For end users, the takeaway will be a better understanding of which solver options to use for different kinds of problems, and what to expect from them.

• New Implementation of a Weakly Thermal-Mechanical Coupling Scheme in LS-DYNA®

  Thomas Kloeppel (DYNAmore GmbH)

  In this paper a novel implementation of a weakly one-way coupled approach for thermal-mechanical coupled problems is intro-duced. The core of the implementation is the new keyword *LOAD_THERMAL_BINOUT. The main advantage is a very flexible input structure, which allows defining the results of several thermal simulations, for example the temperature evolution in different weld stages, as time-dependent boundary conditions for a structural simulation. The temporal order, in which the temperature boundary conditions are processed, can easily be modified. Hence, the effect of, for example, a modified weld order can be considered without recalculation of the thermal results. Several examples will demonstrate the application and advantages of the new approach in the context of the manufacturing process chain. Limitations of the approach, further possible simplifications for coupled simulations and future implementations are also discussed.

• New Metal Forming Keywords in LS-DYNA®

  Xinhai Zhu, Yuzhong Xiao, Jin Wu (Livermore Software Technology, an ANSYS Company)

  Newly developed forming keywords have been optimized to be easy for users. Besides the unification of all the control cards, the tool motion definition has been simplified dramatically and time-related tooling motion curve definition is no longer needed. Various contact algorithm parameters are also treated internally by the LS-DYNA solver. More realistic contact features are available to simulate draw beads and pins. New metal forming keywords therefore achieve an input deck with well-organized input formats which directly describe actual forming processes.

• Occupant Injury Risk Assessment during a Car–to-End Terminal Crash under Crash Test Conditions and Extended Scenarios

  Yunzhu Meng, Costin Untaroiu (Virginia Tech)

  The safety performance of ET-Plus, the most common energy-absorbing guardrail end terminal used in the U.S., was evaluated based on the crash tests recommended by the National Cooperative Highway Research Program (NCHRP). However, while the NCHRP standards were updated to Manual for Assessing Safety Hardware (MASH) in 2009, the safety performance of ET-Plus was not re-evaluated by the updated tests. Also, the occupant injury was evaluated on the full-body level and no seatbelt or airbag usage was considered. Therefore, the main objectives of this study were to evaluate the safety performance of ET-Plus under the MASH test conditions and to assess occupant body-region injuries under varied impact conditions for the first time. Yaris car to ET-Plus crashes simulations were developed based on two sets of MASH test conditions to evaluate the safety performance of ET-Plus. Furthermore, extended scenarios were developed in this study to evaluate the occupant injury risk in varied impact conditions. Three impact velocities (80, 100, and 120 km/h), two impact angles (0, and 15 degrees), and two impact overlap (none, and 25% passenger-side) were used as the pre-impact condition parameters. In each simulation, Occupant Impact Velocity (OIV) and Occupant Ridedown Acceleration (ORA) were calculated and compared to the body-region injury probabilities. The body-region injury probabilities were calculated based on the kinematic responses of the dummy head, neck, and chest. The injury potential was evaluated (HIC, Nij, maximum chest deflection, and chest acceleration) and the severe injury probabilities were then assessed for head and neck injury, and chest injury. For the two simulations developed based on the MASH test conditions, the one with a small overlap (test 30) passes all the requirements while the other one (test 32) failed because the OIV longitudinal exceeds the threshold. Considering the extended scenarios, the average OIV longitudinal was observed to increase with pre-impact velocity: they were recorded as12.3, 12.6, and 14.0 m/s while the pre-impact velocity was 80, 100, and 120 km/h, respectively. Meanwhile, OIV longitudinal was observed to be a good predictor for chest injury while it cannot be used to predict head and neck injury. The OIV lateral was found to be correlated to the head and neck injury. However, it is not recommended to be used to do accurate predictions since the p-value is close to 0.05. On the other hand, the ORA, both longitudinal and lateral, were observed to have no predictability for either head and neck injury or chest injury. This study indicated that the ET-Plus may have the weak capability to protect the occupant during a vehicle to end terminal collision because it would fail MASH test 32 based on the simulation results. Meanwhile, OIV and ORA were observed to have a low capability to predict occupant body-region injuries. Only the OIV longitudinal has predictability for chest injury probability. Head and neck injury, which is a common occupant injury, cannot be assessed by any vehicle-based metrics. Therefore, the usage of dummies should be recommended to the current test requirements. The numerical simulation methods could also supplement the development of new crash tests with varied impact conditions and the optimization of the design of guardrail end terminals.

• On Accuracy and Stability of Implicit Time Integration Schemes for Rotating Structures

  Thomas Borrvall (DYNAmore Nordic AB), Mike R. Lawson (Imperial College)

  In the context of finite element analysis, the Newmark time integration scheme is the most commonly used for nonlinear implicit dynamic applications. While it is characterized by unconditional stability and energy conservation, it is also prone to numerical instability when the models are subjected to rotational motion. To this end, LS-DYNA® offers a selection of alternate integration schemes to remedy this deficiency; Bathe, HHT (Hughes-Hilber-Taylor) and FRD (Finite Rotational Dynamics). The intention with this paper is to tentatively discuss these instabilities and investigate to what extent they can be resolved by incorporating more sophisticated schemes.

• On Composite Model Calibration for Extreme Impact Loading Exemplified on Aerospace Structures

  A. Haufe, S. Cavariani, Chr. Liebold, T. Usta (DYNAmore GmbH), Th. Kotzakolios, E. Giannaros, V. Kostopoulos (University of Patras), A. Hornig, M. Gude (Technical University of Dresden), N. Djordjevic, R. Vignjevic (Brunel University London), M. Meo (University of Bath)

  This contribution will present some simulation related work carried out within a public funded Horizon2020 project of the European Community. Focus is set on composite damage and fracture modelling available in the finite element solver LS-DYNA® and the constitutive models developed within the project. Based on use-cases identified within the project EXTREME, experimental testing and numerical modeling techniques for continuous fiber reinforced aircraft structures such as turbine blades and wing sections are shown. The contribution will showcase results of work packages of the project, such as physical tests conducted to determine the various model parameters which are needed to accurately describe the anisotropic material behavior on a macroscopic length scale that is considered being state-of-the-art in numerical simulations.

• On the Performance and Accuracy of Enhanced Particle Finite Element Method (PFEM-2) in the Solution of Biomedical Benchmarks

  Chien-Jung Huang, Facundo Del Pin, Iñaki Çaldichoury, Rodrigo R. Paz (LST LLC, ANSYS)

  The numerical methods can be helpful on the R&D of medical devices to reduce the costly and lengthy process that clinical trials take for the US Food and Drug Administration (FDA) to approve a medical device. The FDA and academia are working together to create laboratory experiments that will help the industry gain confidence in numerical techniques as well as provide software developers with insights on the strengths and weaknesses of numerical software. In this study, three benchmarks proposed by the FDA are used to compare experimental results with LS-DYNA® ICFD solver with a Finite Element Method (FEM) formulation and Enhanced Particle Finite Element Method (PFEM-2) formulation. In PFEM-2, on top of the finite element mesh, the advection effects are approximated in a Lagrangian way using flow property carrying particles. This means no stabilization is needed for the Galerkin approximation of the advection term in the transport equations. The PFEM-2 enables analyzing a problem with a large time step and is a big advantage in problems with flows at high Reynolds number. The first and second benchmark problems are the flows in a nozzle containing a gradual and a sudden change of diameter and flow in a simplified centrifugal blood pump with the goal of predicting hemolysis. The third benchmark studies the steady flow in a patient-averaged inferior Vena Cava.

• One-step Method for Tri-Axial Carbon Fiber Reinforced Composites in LS-DYNA®

  In this work, an algorithm for one-step analysis approach of tri-axial carbon fiber reinforced composites (CFRC) modeling is introduced and successfully implemented in LS-DYNA. Local fiber rotations during the forming process of fiber reinforced composites are almost inevitable. These rotations have significant effect on the material behaviors of the composite, especially for composites with tri-axial carbon fibers embedded in. In the current work, rotation effects of the embedded fibers are considered and new implementation is capable of handling composites with tri-axial carbon fibers. The prediction ability of the algorithm is demonstrated through modeling of a double dome part with tri-axial carbon fiber composites. Good agreement is obtained in the initial composite shape prediction as compared to experimental data.

• Optimising Run Times for Sheet Metal Forming Simulation

  Trevor Dutton (Dutton Simulation Ltd), Annika Weinschenk (Hexagon)

  Many practitioners of sheet metal forming simulation work in the lower tiers of the manufacturing sector where they tend to rely on PC technology (such as desktop workstations or laptops) to solve their models. Fast turnaround of analyses is critical to success during the process engineering and tool design stage so optimising the model run time is very important. Most PC hardware now provides multi-core chip technology as standard so this paper examines the differences in run times with different numbers of solver cores across a range of model sizes in order to establish best practice, and indirectly best value for money when investing in both simulation software and hardware. The paper considers a selection of LS-DYNA® SMP and MPP solvers, examining the scalability across different numbers of cores (up to 16 maximum) and the interaction between solver scalability and the use of adaptivity – the differences from running with a fixed, uniform, small element size vs. an initial large element size mesh with varying levels of adaptivity are described. The method of mass scaling (standard vs selective) is also examined, to determine the influence of time step on both run time and results. A number of other factors that can influence run time are also reviewed, including adaptivity parameters such as fusion settings, and the type of storage disk hardware used (HDD vs SSD). A number of recommendations are offered, based on model size and available hardware, in the hope that workers in the field of sheet metal forming will be able to efficiently apply the LS-DYNA solver for this important and widespread application.

• Overview of the CESE Compressible Fluid and FSI Solvers

  Zeng-chan Zhang, Grant Cook, Jr., Kyoung-su Im (Livermore Software Technology, an ANSYS Company)

  The Original CESE solver in LS-DYNA® is a compressible fluid solver. Over time, more and more capabilities and applications have been added, especially coupled with the LS-DYNA structural FEM solver to solve different fluid/structure interaction (FSI) problems. Among the many problem types suited to using the CESE solver are the following applications: flimsy vacuum sucking in tissue processing, airbag deployment, blast wave FSI, cavitation in fuel injection, etc. In this paper, there are three parts: (i) a brief review of the current CESE solver; (ii) introduction of our new dual CESE solver; (iii) our two different FSI solvers and their applications.

• Parameters Identification for Wood Material (*MAT_143) and its Application on the Modeling of a Typical Timber Nuki Joint

  Benshun Shao, Aliz Fischer, Yuli Huang, Francois Lancelot (ARUP), Nicolette Lewis (University of Washington)

  Understanding the mechanical properties of the timber joint is a crucial aspect of modern wood construction. Numerical simulation of timber joints can provide valuable insights. However, due to the anisotropic nature of the wood, the sensitivity to local sliding and contact effects, the stiffness and strength modelling of timber joint connections is complex. This study explores the capability of the wood material formulation (*MAT_143/*MAT_WOOD) in LS-DYNA® for simulating the bending behavior of a typical Japanese carpentry connection, the Nuki joint. The material parameters identification was first conducted based on various experimental data. To provide robust results in an efficient manner, the usage of LS-OPT® was explored in this process. A 3-D Nuki joint model was then constructed and its bending behavior was compared with the experimental measurements. Sensitivity studies were conducted on the key contact modeling parameters in LS-DYNA. The study describes an efficient workflow for calibrating the wood material parameters. It also discusses the challenges involved in the modeling of timber joinery mechanics in LS-DYNA and offers suggestions for future research on this topic.

• Pedestrian Head Impact, Automated Post Simulation Results Aggregation, Visualization and Analysis Using d3view

  Milind Shivaji Parab (FCA US LLC), Sreenath Mallela (FCA Engineering India Pvt. Ltd.), Suri Bala (ANSYS/d3VIEW)

  Euro NCAP Pedestrian head impact protocol mandates the reduction of head injuries, measured using head injury criteria (HIC). Virtual tools driven design comprises of simulating the impact on the hood and post processing the results. Due to the high number of impact points, engineers spend a significant portion of their time in manual data management, processing, visualization and score calculation. Moreover, due to large volume of data transfer from these simulations, engineers face data bandwidth issues particularly when the data is in different geographical locations. This deters the focus of the engineer from engineering and also delays the product development process. This paper describes the development of an automated method using d3VIEW that significantly improves the efficiency and eliminates the data volume difficulties there by reducing the product development time while providing a higher level of simulation results visualization. This method reduces post-simulation analysis time through automation thereby eliminating the effort and time of manual data management and visualization. d3VIEW is tightly integrated with LS-DYNA® and as the raw LS-DYNA data is processed on HPC, the resulting output data stored in d3VIEW server is considerably small and eliminates the issue of data bandwidth throttling when the output is to be made available across different geographical regions. Besides score calculation, the capability of d3VIEW has been challenged to include the generation of automatic opportunity chart that can highlight potential locations which require a minimal design change to improve overall score. In summary, d3VIEW platform provides significant benefit in reducing product development process and provides an efficient guiding tool in pedestrian head Impact analysis that can also be extended to other regulatory requirements.

• Performance Study of In Core Adaptivity in LS-DYNA®

  Houfu Fan, Brian Wainscott, Li Zhang and Xinhai Zhu (LST, an ANSYS company)

  Adaptivity is a key technique in metal forming applications, to save computation time while maintaining the accuracy of the result at locations of interest. Although very useful, the traditional implementation of adaptivity in LS-DYNA has certain inefficiencies in its execution. A new approach for adaptivity has been being developed for a couple of years in MPPDYNA. Recently load rebalance was also added into the code right after each adaptive step, which dramatically improves the computation efficiency. A performance study on the current status of this work is presented.

• Performing DOE Studies in Occupant Protection Using BETA CAE Tools

  Thanassis Fokylidis, Nikolaos Tzolas (BETA CAE Systems S.A.)

  One of the most important studies during the design process of a vehicle is its strength in different Crash scenarios. In particular the Safety of its occupants is one of the uttermost goals. Engineers try to cover as many possible cases from the reality by producing different simulations. DOE studies are inevitable to achieve that. ANSA and META, the pre and post processors of BETA CAE systems, offer a wide range of tools for the handling of ATDs, seats and seat-belts as well as tools for the automatic setup of different load-cases, the setup of DOE studies and the evaluation of the results. In this paper LS-DYNA® is used for a DOE study which is setup from the Optimization Tool of ANSA and examines how different parameters like the friction between the ATD and the seat-belt and between the ATD and the seat and the actual position of the dummy can affect the occupant's injury results.

• Preliminary Assessment of Precast Reinforced Concrete Columns against Close-in Air Blast

  Swee Hong TAN, Hui Qi LOH, Jiing Koon POON (Ministry of Home Affairs, Singapore)

  In this contribution, a series of initial numerical results with respect to structural response of precast reinforced concrete (RC) columns subjected to close-in air blast are discussed. Although it is widely established that precast components generally possess limited blast resilience due to their non-monolithic connections, the underlying mechanisms are not well understood. To this end, the present study seeks to gain further insights via explicit modelling of a typical grouted sleeve connection, involving the bond behavior along reinforcement laps and the contacts between interacting concrete surfaces at the column base. Eurocodes have replaced British Standards as Singapore's prescribed building codes for structural design since 2015. CEB-FIP Model Code 1990 has served as an important basis for Eurocode 2: Design of Concrete Structures. In absence of experimental data, this study adopts the relevant guidance from the revised fib Model Code 2010, in attempt to incorporate the latest recommendations numerically. Two key departures are observed vis-à-vis the 1990 version. First, the fracture energy, which characterizes the tensile softening phenomenon, is now solely a function of mean unconfined compressive strength, i.e. independent of maximum aggregate size, while second, the local bond stress-slip analytical model that predicts the interaction between reinforcing bar and concrete, has largely remained the same, albeit with different input parameters.

• Random Vibration Fatigue Analysis Model Development from Explicit to Implicit in LS-DYNA®

  Hwawon Lee, Parvath Police (General Motors), Amit Nair (LSTC)

  Fatigue damage evaluation on vehicle body and hang-on components is one of the most critical paths for the vehicle development stage. Conventionally, fatigue analysis model has been characterized by linear static or dynamic model in which non-linearity of material and contact among vehicle parts are not properly considered, whereas the crash model using Explicit code takes both factors into account. Recent trend of crash event modeling is to increase number of elements up to several millions of finite elements, which is aided by rapidly improving computing power, enabling full vehicle simulations in a very short period of turnaround time. Currently, more focus is on the automotive industry to create larger FE models in a shorter period of time. This is to minimize or reduce vehicle development time caused by the size of the fatigue evaluation model. There were certain efforts to reduce modeling time by converting Explicit crash model to Implicit fatigue evaluation model without losing model contents in a very short period time. This improvement can be achieved in LS-DYNA. This paper demonstrates how to build random vibration fatigue analysis models on MAST (Multi Axis Shaking Table) from Explicit crash model and how to predict fatigue life under random vibration cyclic loading. The first model is a full pick-up truck box, and the other one is a simplified end-gate hinge. A series of parameter study has been attempted to achieve a good correlation between the simplified fatigue testing and such parameters including mesh size, shell element formulation, number of thru thickness integration point & forming effect. The most critical parameter affecting damage ratio in the pick-up truck box is identified by comparing the corresponding test and the proposed model to achieve reasonable fatigue life predictions.

• Recent Development of LS-DYNA® XFEM Shells for Dynamic Ductile Failure Analysis: XFEM with GISSMO Damage Model

  Y. Guo, C. T. Wu (Livermore Software Technology (LST), an ANSYS company)

  This paper presents a coupling of LS-DYNA XFEM shell method [30] and GISSMO damage model for dynamic ductile failure in shell structures. The XFEM shell formulation adopts the finite element continuous-discontinuous approach with the phantom-node technique [17] employed to incorporate velocity discontinuities in standard shell finite element formulations. The generalized incremental stress-state dependent damage model (GISSMO) adds damage and failure to most material models in LS-DYNA that do not allow failure. With the stress-triaxiality dependent failure criterion provided in GISSMO model, XFEM can better simulate material failure and crack propagation in mixed modes and under more complicated loading conditions. The option of element-size dependent regularization factor in GISSMO model removes the strain localization existing in the standard continuum damage model and suppresses the element-size sensitivity of ductile fracture, which is similar to the regularization zone approach in our original XFEM shell method for ductile fracture [30-32]. Unlike element erosion, when an element fails after certain number of integration points reach failure criterion, a crack (discontinuity)is inserted into the element with its direction depending on the stress state or other propagation option, and the element becomes an XFEM element comprised of two phantom elements. XFEM formulation allows crack propagation across elements without the sensitivity on mesh discretization and maintains the conservations of mass and momentum. Several numerical benchmarks and examples are tested using the explicit dynamics analysis to demonstrate the effectiveness and accuracy of the method described in this paper.

• Recent Developments in LS-DYNA® S-ALE

  Hao Chen (Ansys Livermore)

  The LS-DYNA ALE/FSI package is widely used in studying structures under blast loading. Generally, the ALE mesh is necessarily unstructured to accommodate complex geometries; however, for simple rectilinear geometries, a structured, logically regular, mesh can be utilized. Recognition of this latter case leads to algorithmic simplifications, memory reductions, and performance enhancements, which are impossible in unstructured mesh geometries. In 2015, LS-DYNA introduced a new structured ALE (S-ALE) solver option dedicated to solve the subset of ALE problems where a structured mesh is appropriate. As expected, recognizing the logical regularity of the mesh brought a reduced simulation time for the case of identical structured and unstructured mesh definitions. In this paper we will introduce the new developments and enhancements in LS-DYNA S-ALE for the past two years.

• Recent Developments in Time Domain Fatigue Analysis with LS-DYNA®

  Zhe Cui, Yun Huang (Livermore Software Technology, an ANSYS Company)

  A series of new options were implemented to the time domain fatigue analysis features since the last international LS-DYNA User’s Conference 2018. They include:  Fatigue mean stress correction methods  Load steps definition  Fatigue damage evolution  Fatigue failure simulation  Multiaxial fatigue analysis  Fatigue summation This paper gives a brief review of these new options for time domain fatigue analysis with LS-DYNA. Some examples are provided to demonstrate the new feature of LS-DYNA and show how to use this feature towards different loading cases.

• Recent Developments of Smoothed Particle Galerkin (SPG) Method for Joint Modeling

  Youcai Wu, Wei Hu, Xiaofei Pan, C.T. Wu (Livermore Software Technology, an ANSYS company)

  This paper presents the most up-to-date status of SPG (Smoothed Particle Galerkin) development with a focus on the establishment of a failure criterion library, a default keyword parameter setting, and its application in joint modeling. In the recent years, SPG bond failure criterion has been extended from effective plastic strain to 1st/3rd principal strain, maximum shear strain, 1st principal stress and several other quantities defined through *MAT_ADD_EROSION (e.g. effective stress/strain and GISSMO damage). Meanwhile, to minimize users’ work in setting up an SPG simulation, default parameters have been provided so that user can set up the SPG material failure analysis easily with as few as one prescribed parameter for the failure criterion. To demonstrate the effectiveness of SPG method with the new features, the failure modeling of FDS (flow drill screwing) and spot welding is studied.

• Resistance Spot Welding in LS-DYNA®: An Overview of Current Capabilities

  Iñaki Çaldichoury Pierre L’Eplattenier (Livermore Software Technology, an Ansys company)

  Resistance spot welding is perhaps the most frequently encountered joining method for steel sheet in the automotive industry. It is accomplished by passing an electrical current through metal sheets via electrodes. The sheets are held together under the pressure exerted by the electrodes and heat is induced by the electrical current which generates a molten nugget between the sheets. The molten nugget then solidifies to form a bond. During the spot welding process, important changes occur in mechanical and metallurgical properties of the spot welded areas and heat affected zones appear. Although routinely used by the industry, the physics involved in the process are far from trivial, and generally involve a combination of electrical, mechanical, thermal, and metallurgical fields. In particular, the contact area between electrode and workpiece generates an additional electric contact resistance dependent on the models parameters. This contact resistance will have a decisive impact on the shape and size of the nugget and therefore the weld’s quality. Furthermore, the development of new materials such as advanced high strength steels or the replacement of steel by aluminum in certain automotive parts further increases the complexity of the process. Numerical tools and finite element analysis (FEA) can on the other hand offer a crucial assistance in the comprehension of the phenomena involved. Numerically, setting up the RSW model consists in a challenging and highly non-linear problem where solid mechanics, thermal and EM quantities interact with each other. The interface area is especially critical, a robust electric contact algorithm is needed to accurately distribute the local extra resistance to the faces that are in contact with one another so that the correct heating can be calculated and passed on to the thermal solver, even in complex cases with different density meshes and shapes. Over the years, several developments have been introduced in LS-DYNA in order to tackle this problem and in this paper, an overview of the current capabilities will be given along with some example description so that potential users can gain a better understanding of what to expect.

• Response Spectrum Analysis and DDAM Analysis in LS-DYNA®

  Yun Huang, Zhe Cui (Livermore Software Technology, an ANSYS Company)

  Response spectrum analysis (keyword *FREQUENCY_DOMAIN_RESPONSE_SPECTRUM) evaluates the peak response of structures subjected to various loads like ground motions in an earthquake. It combines contribution from each vibration mode of the structures. This feature has important application in Civil and hydraulic engineering, where seismic analysis is critical to the design and safety evaluation of the large scale buildings. DDAM (Dynamic Design Analysis Method) is a U.S. Navy-developed analytical procedure for shock design. It helps validate the design of onboard equipment and structures subject to dynamic loading caused by underwater explosions (UNDEX). It is a widely accepted procedure for safety evaluation for civil and military ship building. The keyword for response spectrum analysis (*FREQUENCY_DOMAIN_RESPONSE_SPECTRUM) in LS-DYNA has been extended to run DDAM analysis for shipboard components, with the option _DDAM. This paper first gives a brief review of the theory for response spectrum analysis and DDAM analysis. Then, with several examples, this paper shows how to run response spectrum analysis and DDAM analysis with LS-DYNA and how to perform post-processing of the results. For purpose of cross-validation, the results of DDAM analysis with LS-DYNA are compared with that given by other commercial code.

• Sensitivity Study of Self-Piercing Rivet Insertion Process Using Smoothed Particle Galerkin Method

  Li Huang, Garret Huff, Andrey Ilinich, Amanda Freis, Shiyao Huang, George Luckey (Ford Motor Company), Youcai Wu (Livermore Software Technology Corporation)

  Self-piercing rivets (SPR) are efficient and economical joining elements for automotive body structures manufacturing. The Smoothed Particle Galerkin Method has been initially proven as a potentially effective way to assess the SPR joining process. However, uncertain CAE parameters could result in significant mismatches between the CAE predictions and physical tests, and therefore the sensitivity study on critical model parameters is important to guide the modeling of the SPR insertion process. In this paper, a meshfree, i.e., Smoothed Particle Galerkin (SPG), method was applied to the simulation of the SPR insertion process with LS DYNA®/explicit. The severely deformed upper sheet was modeled using the SPG method with activated bond failure, while the rest of the model was modeled using the traditional finite element approach. An extensive sensitivity study is conducted to understand the effect of a set of model parameters. This work provides a foundation for CAE model calibration for the SPR insertion process using SPG.

• Sequential Optimization & Probabilistic Analysis Using Adaptively Refined Constraints in LS-OPT®

  Anirban Basudhar, Imtiaz Gandikota (Livermore Software Technology LLC, Ansys Group), Katharina Witowski (Dynamore GmbH)

  This paper presents some of the sequential optimization and probabilistic analysis methods in LS-OPT with particular emphasis on the use of classifiers for accuracy and efficiency improvement. Classifiers were first introduced in LS-OPT 6.0 for the handling of constraints. This paper provides a review of the basic classification-based constraint handling method and its applications and advantages for specific types of problems. Additionally, the application of classifiers is extended to adaptive sampling using EDSD (explicit design space decomposition) sampling constraints in LS-OPT 6.1. The different adaptive sampling options and approaches are presented through the examples. Another aspect of this paper is the extension of the probabilistic analysis method in LS-OPT from single iteration to sequential. The sequential analysis can be performed with or without EDSD sampling constraints, but sampling constraints, if used, are can guide the samples adaptively to important regions. Although the EDSD sampling constraints are defined using support vector machine (SVM) classifiers, the adaptive samples are useful in enhancing the constraint boundary accuracy even if it is defined using metamodels.

• Side Curtain Airbag Folding Methodology

  Pablo Alberto Rodríguez Calzada, Hector Hernández Hernández, Alejandro García Pérez, Carlos Gómez González (Ford Motor Company)

  In recent years, CAE simulations have been substantially improved as a result of the growing need to achieve full vehicle developments in a shorter time span while also attending the demand of cost reduction in such developments. One of the most critical components regarding the passive safety systems of a vehicle is the Side Curtain Airbag, therefore the necessity to involve this critical component in an agile product development process becomes compulsory. Consequently, when the validation using numerical methods of such component is performed, a full deployment of the airbag is needed to be evaluated and analyzed, having as a key objective the monitoring of its dynamic behavior caused by the effect of interacting with nearby components. In view of the foregoing, the folding process of the airbag plays a key factor in its whole operation. This study describes a hybrid methodology to fold a Side Curtain Airbag by means of a geometrical and simulation-based routine, which can be defined entirely on LS-PrePost®, using the embedded tools in the occupant safety applications. This work aims to englobe the tools and steps followed in order to obtain, within a short period of time, a LS-DYNA® CAE model of the airbag, capable of representing efficiently and accurately a deployment, which might be used in early stages of numerical analysis for areas such as Interior Trim integrity and safe interaction. Using this CAE methodology, a new scope of problem-solving techniques originates. Applying the novel approach described in the preceding paragraph, a folding scenario could be useful to control the dynamics of the airbag in order to achieve a faster deployment in a certain zone, to avoid an undesired interaction with the interior trim of the vehicle, or to simply evaluate the aperture time of the system overall. All this adds up to a feasible cost reduction alternative to the most common techniques that involve modifying and adapting geometries including supplementary components, that impact directly in the prime cost of a vehicle.

• Simplifying the Pre-Simulation Set Up of Airbag Folding in LS-DYNA® Using ANSA

  Thanassis Fokylidis, Stavros Porikis (BETA CAE Systems S.A.)

  Crash and Safety simulations of virtual models hold a key role during the design process of a vehicle. Occupant protection analyses combined with laboratory tests ensure that models will offer the necessary safety to passengers. Airbags are one of the many protection systems that ensure this safety. During model set up for occupant safety simulations, significant time is consumed to fold the airbags properly and fit them in the respective cases. Airbag folding in virtual models follows the exact specifications given for the real airbags. LS-DYNA, as a standard solution in Crash and Safety analysis supports the folding of an airbag. The loadcase set up, for preparing the respective pre-crash simulation, may include the definition of several LS-DYNA keywords. BETA CAE Systems has developed an assistant functionality, in the ANSA pre-processor, which streamlines all the necessary steps to set up properly a LS-DYNA pre-simulation for airbag models. The current paper introduces this assistant and demonstrates the advantages of the procedure by ensuring a convenient LS-DYNA loadcase set up with the minimum effort.

• Simulating Prepreg Platelet Molding Compound Flexure Coupons in LS-DYNA® Using MAT54

  Rebecca A. Cutting, Anthony J. Favaloro, Johnathan E. Goodsell (Purdue University)

  Prepreg platelet molding compound (PPMC) is a composite material system primarily for compression molding comprised of rectangular platelets made from slit and chopped unidirectional prepreg. Prior to processing, the platelets are assumed to be in a planar random orientation state; however, this changes during molding as the platelet orientation state evolves in response to flow. The final orientations of the platelets affect the mechanical performance of manufactured components. As such, it is necessary to incorporate platelet orientations in simulations of PPMC parts to capture the material system behavior. This work introduces a modeling method for simplified PPMC geometries using MAT54 in LS-DYNA. A platelet generation code creates virtual PPMC samples with varying global orientation distributions. These orientations are input into LS-DYNA models of flexure coupons. A study is completed to understand how global platelet orientation affects the flexural stiffness and strength of the virtual samples. In addition, the results of the flexure simulations are compared to experimental results of flow-aligned PPMC coupons to validate the modeling method. The simulation study reveals that an increase in platelet alignment along the longitudinal axis of the sample results in an increase in flexural stiffness and strength in the models. This trend is confirmed experimentally, and the accuracy of the modeling method is discussed.

• Simulation Data Management from CAD to Results with LoCo and CAViT for Large Scale LS-DYNA® LEGO® Crash Models

  Thorsten Gerlinger, David Koch, Andre Haufe (DYNAmore GmbH), Nils Karajan, Thomas Weckesser (DYNAmore Corporation), Pierre Glay (DYNAmore France SAS), Alexandru Saharnean, Marko Thiele (SCALE GmbH)

  Given that in our professional lives we are dealing with highly sophisticated crash models on a daily basis, it seems obvious that we instantly thought we should be able to simulate a crash of a LEGO® Porsche Technic Model using the LS-DYNA FEM solver after seeing a video of a physical crash of this model on YouTube. Setting up a process, which involves every aspect of working with CAD data, meshing, dealing with solver files, submitting and monitoring the simulations, and finally handling the result files of simulations, is an important step when developing a Simulation Data Management (SDM) system such as LoCo and CAViT. Therefore we decided to use this LEGO® crash as a challenge and benchmark for our software. The real LEGO® models are often assembled with thousands of bricks. Handling so many parts in a SDM system on one hand and maintaining the ability to work on such models in a collaborative way with multiple users on the other is quite challenging. Initially, we set up the whole simulation process for the Porsche which is composed of 2704 individual bricks. But when we showed the results of the LEGO® Porsche crash simulation to the c’t magazine (a widely read German computer magazine) and ADAC (General German Automobile Club) who had performed the initial physical crash test in 2017, they suggested doing another LEGO® crash scenario. This time a LEGO® Porsche was supposed to crash into a LEGO® Bugatti model at 60km/h and a LS-DYNA simulation should predict the outcome of the crash before the physical test was going to be conducted. The LEGO® set number 42083 of the Bugatti Chiron is even bigger than the Porsche model and consists of 3599 bricks. The results of the simulation were then evaluated and presented to ADAC and c’t magazine to provide our prediction of the upcoming real physical crash. Later, the comparison of our prediction with the real crash results revealed that many details have been predicted correctly by the simulation. The final LS-DYNA model of both car models consisted of more than 45 million elements. Preprocessing as well as getting the model to run on an HPC system and handling the few but large result files has proven to be challenging in many ways.

• Simulation of Compression Behavior of Paper Product Using *MAT_PAPER

  Sunao Tokura (Tokura Simulation Research Corporation), Kunio Takekoshi (Terrabyte Co., Ltd.)

  Environmental pollution caused by plastic products is now a global serious problem. Plastics dumped in the ocean become into micro plastics and threaten the living environment of various organisms. In order to reduce such environmental pollution, products are being developed using paper materials with less environmental impact as an alternative to plastic products. In developing a paper product, it is necessary to design a strength suitable for the application. Therefore, accurate prediction of product strength by simulation is considered to be very important for product development. In this study, we attempt to predict the strength of paper products using *MAT_PAPER, which is a paper material model implemented in LS-DYNA®. Since *MAT_PAPER is a complex anisotropic elasto-plastic composite constitutive equation, properties of paper materials were measured, and the reliable input parameters were determined. A paper cup compression test and simulation were performed using the paper material model which was constructed from the test. As a result, a good agreement and some differences between the experimental results and the simulation results is shown.

• Simultaneous Exploration of Geometric Features and Performance in Design Optimization

  Nivesh Dommaraju (Technical University of Munich), Mariusz Bujny, Stefan Menzel, Markus Olhofer (Honda Research Institute Europe GmbH), Fabian Duddeck(Technical University of Munich / Queen Mary University of London)

  Topology optimization (TO) algorithms generate novel concepts to inspire and propel the design iteration process. LS-TaSC® is an industrial tool that implements TO algorithms and generates designs optimized for performance objectives such as maximum stiffness or energy absorption, under specified constraints e.g. allowed mass fraction of material in the design space. A multi-objective design exploration framework based on LS-OPT® and LS-TaSC to generate designs is already available. The framework yields a Pareto set of designs by varying a parameter representing the relative preference of the user among the different objectives. A challenge persists as to how potentially large datasets of designs, generated using such an approach, can be reviewed efficiently by a designer. In this paper, we propose a method to identify a few representative design prototypes, which can be more easily reviewed by a designer. More concretely, the approach identifies classes of designs that look significantly different from a geometric point of view. For this purpose, we encode the information about the geometry using a voxel representation of the design. Subsequently, we use Principal Component Analysis (PCA), to reduce the high dimensionality of the representation, and extract features that encapsulate the geometric variation in the set of designs. Design prototypes are derived based on clustering algorithms using weights of principal components as features. To evaluate the proposed approach, we consider a solid beam model that is optimized for high stiffness under a static load case and high-energy absorption in a crash load case. Similar design problems are especially common in the car body design. We generate a Pareto set of designs for this test case and identify design prototypes. An interesting application of this method is to find designs with similar geometric appearance but very different performances. This can help us to estimate the robustness of a design. By helping in design exploration and selection, the proposed approach shows promise in large-scale industrial applications.

• Stacked Shell Modeling for Evaluation of Composite Delamination in Full Vehicle Simulations

  Olaf Hartmann (ARRK Engineering)

  Accurate prediction of delamination in composite materials is a challenge and often limits the application of lightweight materials in safety relevant components, as it may reduce the available strength significantly. Very detailed modeling (e.g. with solid elements) can be employed to correctly recreate this phenomenon, but this can normally not be simulated in a full vehicle simulation in an acceptable amount of time. Single shell modeling is widely used in full vehicle simulations because of its high runtime performance but cannot support the physical separation of layers. In order to correctly evaluate delamination of composites while retaining a good runtime performance, a new modeling approach in LS-DYNA® was studied in this paper. A stacked shell modeling technique was developed. The new modeling approach was firstly investigated at coupon level with comparison with experimental results for assessing its accuracy and capability of delamination prediction. Furthermore, stacked shell modeling was adopted into components under more complex loading and its performance was evaluated in terms of accuracy and run time compared with conventional modeling. At the end, this modeling technique was studied in full-vehicle simulation. Our stacked shell modeling approach has shown promising results at coupon and component level. At the full-vehicle simulation scale, the new modeling approach has presented robust delamination prediction capability while still retaining high run time performance. The approach presented in this paper can be adopted in full-vehicle crash and also aerospace simulations in order to evaluate composite delamination.

• Strength Assessment of a Plastic Component considering local Fiber Orientation and Weld Lines

  Natalja Schafet, Marta Kuczynska (Robert Bosch GmbH), Sascha Pazour, Wolfgang Korte, Marcus Stojek (PART Engineering GmbH)

  The aim of the study was to provide static strength assessments for a short fiber reinforced plastic part considering anisotropic material properties and strength drop caused by weld lines. The sequential coupling of process and structure simulation opens up additional potential for the development process. It is shown how significantly this additional information influences the quality of the assessment. In focus are methods used to assess the strength distribution in the plastic part as well as the comparison between the results of the measurement and final structural-mechanical simulation. More over various failure assessment approaches, ranging from simple estimating procedures, like reduction factors, to the full consideration of fiber orientations, are presented to compare their performance.

• Structured ALE Solver with Large Models

  Hao Chen (Ansys Livermore)

  In 2015, LS-DYNA® introduced a new structured ALE (S-ALE) solver option dedicated to solve the subset of ALE problems where a structured mesh is appropriate. As expected, recognizing the logical regularity of the mesh brought a reduced simulation time for the case of identical structured and unstructured mesh definitions. In this paper we will introduce the recent enhancements to facilitate running large models using S-ALE solver.

• The Effect of InfiniBand and In-Network Computing on LS-DYNA® Simulations

  Ophir Maor, Gilad Shainer, David Cho, Gerardo Cisneros-Stoianowski, Yong Qin (HPC-AI Advisory Council)

  High-performance computing (HPC) technologies are used in the automotive design and manufacturing industry. One of the applications is computer-aided engineering (CAE), from component-level design to full analyses for a variety of use cases, including: crash simulations, structure integrity, thermal management, climate control, modeling, acoustics, and more. HPC helps drive faster time-to-market, significantly reducing the cost of laboratory testing and enabling tremendous flexibility. HPC’s strength and efficiency depend on the ability to achieve sustained top performance by driving the CPU performance toward its limits. The motivation for high-performance computing has long been associated with its tremendous cost savings and product improvements; the cost of a high-performance compute cluster can be just a fraction of the price of a single crash test, for example, and the same cluster can serve as the platform for every test simulation going forward. Recent trends in cluster environments, such as multi-core CPUs, GPUs, and advanced high-speed and low latency interconnect with In-Network Computing capabilities, are changing the dynamics of cluster-based simulations. Software applications are being reshaped for higher degrees of parallelism and multithreading, and hardware is being reconfigured to solve new emerging bottlenecks to maintain high scalability and efficiency. Applications like LS-DYNA and others are widely used and provide better flexibility, scalability, and efficiency for such simulations, allowing for larger problem sizes and speeding up time-to-results. CAE applications rely on Message Passing Interface (MPI), the de-facto messaging library for high performance clusters that is used for node-to-node inter-process communication (IPC). MPI relies on a fast, unified server and storage interconnect to provide low latency and a high messaging rate. Performance demands from the cluster interconnect increase exponentially with scale, due in part to all-to-all communication patterns. This demand is even more dramatic as simulations involve greater complexity to properly simulate physical model behaviors. In this paper, we will focus on the value of HDR InfiniBand interconnect technology for LS-DYNA applications, by comparing different InfiniBand network transport options and MPI libraries.

• The Latest Developments of the ANSA Preprocessor for IGA Applications of LS-DYNA®

  Lambros Rorris (BETA CAE Systems International AG), Ioannis Chalkidis (BETA CAE Systems SA)

  Iso Geometric Analysis (IGA), is maturing and becoming capable to be incorporated in industrial applications. Widely used in the automotive industry for crash analysis, LS-DYNA is the first commercial solver to provide IGA features. Better accuracy, potential shorter run times, accurate geometry representation make IGA effective for crash analysis. Nevertheless, the complexity of the current automotive models and the maturity of the already established methods and processes require the development of the respective IGA tools and processes to reach and exceed the current levels of effectiveness. The new technical challenges give the opportunity for new solutions and improvements in engineering simulation technology. During the last year ANSA has developed the needed tools and algorithms to successfully convert CAD geometry to IGA ready part descriptions, thus making the first successful complex hybrid IGA - FEA models possible. We continue our work towards full integration of IGA in current complex LS-DYNA FEA crash models and processes. All latest development will be presented.

• The Shotgun Pellets Interior Ballistics Analysis by Discrete Element Method (DEM) of LS-DYNA®

  Shigan Deng, Ta-Chiang Wu (National Defense University, Taiwan), Jason Wang (Livermore Software Technology Corporation)

  This research used the Discrete Element Method (DEM) of Finite Element Software (LS-DYNA) successfully simulated the interior ballistics performance of shotgun pellets and the deformation of wad and shot cup. This research used 12-gauge shotgun with #9 bird shot which has 433 pellets inside the shot, as an example. There are four components are modeled by finite element: barrel, wad with shot cup, case, and pellets of the shot. The input forces are the chamber pressure on the bottom of wad, which is calculated by Vallier-Heydenreich empirical formula, and the air resistance in the front faces of shot cup. The outputs are velocity/acceleration history of the wad with shot cup and pellets, the distribution of pellets after exits muzzle, and the deformation of the wad with shot cup. The calculated muzzle speed of pellets is 423m/s, compare to 412m/s of Sporting Arms and Ammunition Manufactures Institute (SAAMI) tested, there is only 2.67% error. The results of this research could lead the traditional test-oriented shotgun industry into contemporary Computer Aided Engineering (CAE) by introducing the DEM of FEM to the industry.

• The Use of User Defined Elements and Extra Degrees of Freedom

  Kelly S. Carney, Paul DuBois (Forming Simulation Technology LLC), Thomas Borrvall (DYNAmore GmbH)

  The capabilities of the LS-DYNA® User Defined Features, allowing for modifications to LS-DYNA analyses, are powerful. It would be impractical to fully exercise their extensive capabilities, and so many of the features have also not been fully documented. Specifically, several undocumented actions are required to use User Defined Elements, when using extra degrees of freedom. This brief paper will explain the additional steps required for the LS-DYNA integrator to update the extra degrees of freedom, when using User Defined Elements. Additional details of extra degrees of freedom usage will also be given.

• Transient Fluid Structure Simulation of Ground Vehicles

  Facundo Del Pin, Iñaki Çaldichoury, Rodrigo R. Paz, Chien-Jung Huang (Livermore Software Technology LLC)

  Ground vehicle aerodynamics is an important stage in the design of a car. The field is currently well established, and the final goal is to decrease the lift to drag ratio functional by maintaining certain design constraints. Traditionally the studies are performed in wind tunnels but with the advance of hardware and software in the past two decades most of the design is now performed by simulation using Computational Fluid Dynamics (CFD). The automotive industry has embraced these techniques due to their low cost and high accuracy. But in recent year the tougher environmental regulations together with the natural evolution of technology have pushed the industry into new lighter materials, thinner panels and more compact parts distribution. These changes bring new challenges to the design process. The clay models traditionally used in wind tunnels cannot predict the response of the real structures subject to aerodynamic or thermal loads. Traditional CFD simulations are faced with similar limitations. Furthermore the internal organization of the CAE departments are not catching up fast enough to adapt to this new reality and the result is that last minute unexpected behaviors happen during drive test conditions forcing late modifications and the retooling of parts, greatly increasing the design cost. It is our goal to introduce in the design cycle intermediate steps where coupled Multiphysics simulations will be used to anticipate unexpected behaviors and correct the design before the tooling stage. In this work a real world model of a mid-size sedan is used as a showcase of the different possibilities that are available in LS-DYNA® to perform CFD together with Fluid Structure Interaction (FSI) to study the response of different structural parts of the vehicle subject to aerodynamic loads. One of the main advantages is that complex structural parts are "borrowed" from a LS-DYNA crash model and easily introduced into the CFD model greatly simplifying the process. All the material settings and geometry will be automatically ready for the FSI simulation.

• Use of Prepreg Carbon and Aluminum in Satellite Shielding Submitted to High Velocity Impacts

  Tess LEGAUD, Morgan LE GARREC, Vincent LAPOUJADE (Dynas +), Hakim ABDULHAMID, Paul DECONINCK (Thiot Ingénierie)

  A substantial number of debris coming from human production gravitates around the Earth. Their size, nature, orbit and velocity can extremely vary, but all these debris represent an increasing collision risk and a threat for the current and future spatial activity. The spatial researchers are looking for solutions to limit this risk, by better controlling the launched objects number and by improving the protection of their structures. All those debris are classified depending on their size. The ReVuS European project showed that the most dangerous debris, according to the satellite mission failure probability, have a diameter included in the range 1mm to 5mm. Following this reference, the aim of the ATIHS project, funded by French region Occitanie, is to improve the satellite protection from millimetric debris impacts. Multiple solutions exist in order to do so, ATIHS focuses on the shielding one. The project global aim consists in: - Improving the satellites resistance on strategic locations to prevent it from the mission failure, - Working on the secondary debris generation limitation during a non-lethal impact in order to minimize the satellite contribution in the debris increasement. The project is composed of three main tasks: - Evaluating new material solutions showing an optimised mass/resistance combination, - Evaluating new hypervelocity testing devices which should permit to go further than the currently available devices (goal: 8 to 12 km/s for millimetric to centimetric projectiles), - Setting up numerical methodologies that should permit to increase the capacities and the hypervelocity computations reliability, by accurately modelling the materials behaviour during this kind of extreme solicitations. This article focuses on the hypervelocity impact response. It especially deals with the evaluation of new structures composed of carbon prepreg or/and zylon laminates to better protect the satellite equipment and there SPH modelling methodologies. As a first step, some tests have been performed on a unique sheet made of composite (carbon prepreg or zylon). Then, it is used to compose a mixed Whipple shield such as the assembly of two skins using various materials.

• Using *MAT_213 and *MAT_187 to Predict Failure in Unidirectional Composites

  Bilal Khaled, Loukham Shyamsunder, Josh Robbins, Yatin Parakhiya, Subramaniam D. Rajan (Arizona State University)

  Failure in composite materials is due to various complex mechanisms often occurring simultaneously. The heterogeneous, anisotropic nature of composites provides challenges in deriving analytical models for failure similar to what has historically been done with homogeneous, isotropic metals. However, as composites continue to be used in the design of large structures, finite element material models which homogenize the composite response become the only logical choice as modeling the entire microstructure is currently impractical. Thus, relating the microscale behavior caused by the macroscopic excitations is required. A modeling methodology where plasticity, damage, and failure related experimental data are obtained for each constituent and subsequently used to generate high fidelity computational micromechanical models. The ultimate goal is to utilize information from the micromechanical computational models to drive the failure sub-model of *MAT_213 in LS-DYNA®. The first step is to obtain high fidelity experimental data and refine the respective material models for each constituent. This research presents the experimental results from tests performed on the F3900 epoxy resin. The data is then used to populate the input deck for *MAT_187 in LS-DYNA. Verifications tests are presented showing how the derived experimental data performed in virtual finite element tests.

• Using the CESE Immersed Boundary FSI Solver to Simulate the FSI of the Front Portion of a Turbofan, including Damage

  Grant Cook, Jr, Zeng-Chan Zhang, Gunther Blankenhorn (Livermore Software Technology, an ANSYS Company)

  In this paper, we demonstrate use of the LS-DYNA®. Conservation Element/Solution Element (CESE)[1] solver doing fluid-structure interaction (FSI) calculations employing its immersed boundary FSI method[2,3]. The fan rig used is a portion of the fan blade-off rig test for a generic fan rig model. The model is available through the LS-DYNA. Aerospace Working Group (AWG) at http://awg.lstc.com. In this model, programmed failure of some of the structural elements is set up in one blade near the fan hub. Two cycles of operation are analyzed with and without FSI, and then the failure situations will be demonstrated with and without FSI active.

• Using the Latest Cloud Technology to Accelerate LS-DYNA®

  Rodney Mach (TotalCAE)

  This paper will focus on the latest technologies available on cloud to accelerate LS-DYNA using the major IaaS vendors including Amazon AWS and Microsoft Azure, and how they impact the total simulation job cost of a LS-DYNA benchmark model.

• Visualising Vehicle Platoon Aerodynamics Using ICFD in LS-DYNA

  Edward Pettitt, Max Resnick (ARUP)

  Since the fuel crisis in the 1970s, platooning of vehicles has been considered as a method of reducing fuel consumption and improving traffic congestion. Up till now, research has primarily considered homogeneous platoons with varying separation distances. Furthermore, the basis of much of the work so far has been experimental instead of computational and so fluid flow interactions between the platooning vehicles are not well understood. The Incompressible Computational Fluid Dynamics (ICFD) solver in LS-DYNA® provides a powerful method to simulate flow interaction around a body and produce data to be used in understanding the aerodynamic behaviour. This paper shows some of the capabilities of LS-DYNA in visualising fluid flow around complex bodies when in platooning formation. Of interest in this paper is the variation of geometry of the vehicles within a platoon with a set separation distance. This paper considers the ability of LS-DYNA to analyse fluid flow interaction between multiple bodies and explores how this could be of use for future platooning research.

• Wear Analysis of Machinery Components in Buildings

  John Puryear, Ben Harrison (ABS Group), Lynsey Reese (NAVFAC Engineering and Expeditionary Warfare Center)

  Buildings incorporate a variety of mechanical systems for the convenience of occupants. These systems may include machinery for heavy lifting. Excessive wear in the machinery components can result in unanticipated maintenance and loss-of-service costs. The Wear Analysis utility in LS-PrePost® was used to quantify component wear over the service life for a vertical lift system. Candidate materials for a roller and jamb geometry were modeled. The wear predictions were then used to evaluate the candidate material combinations. The methodology provided by Jernberg and Borvall was followed for the analysis. Due to the relative low loading rate in the problem (i.e. much less than the loading rate in automotive impact), the contact force history between the roller and plate was calculated using implicit dynamics. This force history was the input into the Wear Analysis utility using the dynain file from *INTERFACE_SPRINGBACK_LSDYNA, and Archard’s wear law was used to calculate wear depth. The wear law returns wear volume as a function of surface hardness, normal force (as calculated in the implicit analysis), sliding distance and wear coefficient. The above methodology was applied to three combinations of materials for the roller and jamb plate selected from hardened steel, stainless steel and aluminum. *CONTACT_AUTOMATIC_SURFACE_TO_SURFACE_MORTAR was necessary for the implicit calculation to converge. Static and dynamic coefficients of friction (COFs) were obtained from the CRC Handbook; the decay constant from static to dynamic COF was from Stembalski. In this paper, the remaining details of the analysis are discussed, including the constitutive models used for the metals. The wear predictions are provided. Finally, recommendations are made for increasing confidence in the calculations and convenience of the calculation method.