www.ozeninc.com [email protected] (408) 732 4665 1210 E Arques Ave St 207 Sunnyvale, CA 94085
Reliable World Class Insights
Your Silicon Valley Partner in Simulation
ANSYS Sales, Consulting, Training & Support
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 1
A Total Li-Ion Battery Simulation Solution
Lewis Collins, Director of Software Development ANSYS Convergence Regional Conference Santa Clara, CA April 21, 2015
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 2
Advantages
Industry Solutions Value-Added Services Global Support
Unequalled Depth Unparalleled Breadth Comprehensive Multiphysics Engineered Scalability Adaptive Architecture
SDPD Vision Company Strength Independence
Engaged in electrochemical systems R&D for >15 years
• Primarily batteries and fuel cells
Confidence from usage in dozens of industries, hundreds of applications, thousands of organizations
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 3
Ongoing Collaborations
Scale-Bridging Models for Electrochemical Power Sources • U.S. Office of Naval Research, 2005-
• ANSYS, NRL, universities
• Metrology and resolved modeling of electrodes
• Upscaling methods for cell, pack, and system
Computer-Aided Engineering of Electric Drive Vehicle Batteries (CAEBAT)
• U.S. DoE Vehicle Technologies Office, 2011-
• ANSYS, GM, ESim, NREL, universities
• Improve CAE tools for battery cells and packs (usability, validation, interoperability)
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 4
In mobile devices, battery is connected with key design issues
Battery Considerations in Electronics
Power management (battery-to-chip)
Pressure on noise margins
Signal and power integrity
Electromagnetic interference (EMI)
Thermal
Hardware/software integration
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 5
Battery has become a limiting factor
Driving ultra-low power design methodologies
Demand Trend for Mobile Power
(for a specified form factor)
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 6
EV Everywhere Grand Challenge goal: make plug-in electric vehicles as affordable and convenient as today’s gasoline-powered vehicles by 2022
Battery is Key to Vehicle Electrification
Image credit: “FY 2013 Annual Progress Report”, Energy Storage R&D, Vehicle Technologies Office, U.S. Department of Energy
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 7
Cost
Performance (power and energy density)
Durability and service life (in disparate environments)
Safety (tolerance to abusive conditions)
Complex multi-scale, multi-physics system
Rapidly evolving materials and design concepts
Existing software tools not “tuned” for batteries
Engineering Challenges
Thermal
Elec-trical
Chemical
Fluid
Molecular Particle Electrode Cell Pack System
Need to account for interconnection
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 8
Methods Toolkit
Field simulation Electrode cell module scales
e.g. CFD: fluid, thermal, chemical, electrical
System simulation Module pack vehicle scales
Lumped-parameter models, controls
Reduced-order models (ROM) Small number of (linear or nonlinear) state eqns
Multidisciplinary design optimization
Cosimulation
Extraction
Expansion
Instantiation
Electrochemistry sub-models Homogenization (over particles, electrode layers)
DuCxy
BuAxdt
dx
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 9
Model-Based Systems Engineering
System Validation
Sub-System Integ. & Verification
Component Integration
& Verification
Requirements and Specifications
Sub-System Design
System Functional & Architectural Design
Mechanical Electrical Software
Detailed Design & Optimization
Functional Allocations
Detailed Architecture Architecture
Simplorer
Maxwell
Fluent
Mechanical
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 10
Balancing Speed and Resolution New techniques can reduce cost while preserving a sufficiently
accurate approximation of the responses of interest
“Selective use” of CFD, tailored to the unique objectives
fidelity
cost (log scale)
System Simulation
Field Simulation (e.g., CFD)
Spreadsheet/ Handbook Calcs
(orders of magnitude)
With ROM, cosimulation
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 11
Geometry: built-in parametric templates • Inputs available in Workbench Parameter Manager and DesignXplorer
Meshing: also templated, based on best practices
Cell Model: Field Simulation
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 12
Traditional approach supports maximum design creativity • SpaceClaim, DesignModeler, or CAD Interfaces
• Assumption: lithium transport is perpendicular to local electrode plane
Custom Geometry is Easily Handled
Image credits: www.apple.com/macbook/design
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 13
Electro-Chemical-Thermal Simulation
Battery Module a standard feature of Fluent
Single or multiple cells
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 14
Φ- Φ+
q, j
Φ- , Φ+ , T
No need to resolve individual electrode layers with the mesh
User-defined scalars represent electrical potentials F+ , F-
Chemical species are not explicit solution variables • Sub-grid model may track lithium ion concentration
Multiscale Approach
Ref: G-H Kim et al, “Multi-Domain Modeling of Lithium-Ion Batteries Encompassing Multi-Physics in Varied length Scales” J. Electrochem. Soc. 158(8) A955-A969 (2011).
)(tI
)(tV
1R2R
1C 2C
sR
)( socVocv
)(tI
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 15
“Single-point” (0-D) analysis option, from the same interface
Auto-merge separately-created cell and adjacent-structure models
Auto-fit properties from test data file • e.g. from calorimeter testing
Cell Model: Productivity Aids
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 16
Logical step for system simulation • Verification of material properties,
electrochemistry models, stoichiometry assumptions, etc.
Representative results from CAEBAT project:
Cell Model Validation
Images on this slide courtesy of General Motors LLC
HPPC @ 25 degC, 30% DOD
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 17
Goal: automate workflow while preserving tool generality
Exploit periodicity or symmetry when possible
System Simulation Templates and Scripts
Unit-Model Template
• Reusable building-block
• Define based on thermal
and electrical connectivity
• One-time manual creation
using standard Simplorer
components
• Store to User Library
Script
• Reusable procedure for a
specific pack architecture
• Automate connections, layout,
post-solution statistics
• One-time optional programming
task using highly-accessible
Python language
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 18
Automotive Examples
Image credits Left: General Motors LLC Right: Ricardo Strategic Consulting
Application PHEV BEV
Cell type 15 Ah pouch 3 Ah cylindrical
# cells 288 7104
Electrical configuration 8 x (12S3P) 16 x (6S74P)
Cooling configuration (cells/channel)
Totally parallel, 2
Series within module, 444
Unit definition (# cells) 6 2
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 19
Applicable to both rectangular and cylindrical types • Form-factor effects incorporated into lumped coefficients
Electrical • Equivalent-circuit model
• Coulomb-counter (SOC = state of charge)
Thermal • Includes adjacent part
of fluid cooling channel
System Hierarchy Starts with a Single Cell
Two-way Coupling
Drag-and-drop components
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 20
Created once and stored to Simplorer library file
(Shapes and layout for reference, not part of model)
Unit Model (Domain Decomposition)
f f f f c1
c2
c1 c2 c3 c4 c5 c6
– Hydraulic – Electrical
–
+ – +
– Thermal
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 21
Exploit periodicity, instancing templated models from library
Module Model
2S3P
Unit
2S3P
Unit
2S3P
Unit
2S3P
Unit
. . . (6x)
2P
Unit
2P
Unit
2P
Unit
2P
Unit . . . (37x)
Group (row)
– Electrical – Thermal – Hydraulic
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 22
Ready for coupling with BMS or full-vehicle models
Pack Model
12S3P
Module
12S3P
Module
12S3P
Module
12S3P
Module
. . . (8x)
LOAD
Outer
Case
Thermal
Mgmt System
6S74P
Module
6S74P
Module
6S74P
Module
6S74P
Module
. . . (16x)
LOAD
Outer
Case
Thermal
Mgmt System
– Electrical – Thermal – Hydraulic
AMBIENT AMBIENT
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 23
Script-generated form • Enter 3 integers
• Select unit from library
Process Automation
Click on any instance To drill into hierarchy
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 24
GM prototype 24-cell module • LG 24-Ah pouch cells (LMO/NCA cathode)
• Steady-state liquid cooling
• 32 thermocouples
Progressive model comparison: • “Brute force” CFD model
• Lumped-parameter system model
– Some inputs guided by CFD results
• CFD-derived ROM
Progressive case complexity: • Symmetric pulse charge/discharge
• Simulated drive-cycle load profile
Validation
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 25
Top View
Measurement CFD Prediction
No
dat
a
No
dat
a
Coolant in
Coolant out
Typical Comparison: Middle Cell
High-frequency ±3.5C pulse charge/discharge @50% SOC
Maximum difference between simulation and measurement < 1 C
Cooling fin Foam spacer
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 26
Typical Comparison: System Model
Volume-average cell temperature versus time
1 C
– Simplorer prediction – averaged measurements (other colors = individual measured locations on cell)
60 minutes (5 x US06 drive cycle + cooled rest, SOC 0.9-0.2)
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 27
Excellent thermal replication of cells in ordinary operation
ROM execution several orders of magnitude faster than CFD
Typical Comparison: ROM versus CFD
Ref: X. Hu and S. Stanton, “A Complete Li-Ion Battery Simulation Model,“ SAE 2014-01-1842
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 28
Temperatures transfer to Mechanical for analysis at any scale
Thermal Stress
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 29
Plug-and-play concept: several interfaces to support native, in-house, and third-party tool integration
• E.g., user-defined function (UDF) for custom electrochemistry models
• Other Tools can include micromechanics, cell-design apps, system simulators, battery cost models, …
Plug-In Hybrid Software
ANSYS
Workbench
CAD
Simplorer
UDF Fluent
Battery Model
Other Tools
Mechanical
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 30
Maintained by the non-profit Modelica Association
Supports simulation model exchange and tool coupling, including hardware-in-the-loop (HiL)
Functional Mockup Unit (FMU) = portable package containing • Interface description (XML schema)
• Model functionality (C code – source or binary)
Supported by ANSYS Simplorer, SCADE (and >60 other tools)
Functional Mockup Interface (FMI)
Ref: T. Blochwitz, et.al., “The Functional Mockup Interface for Tool independent Exchange of Simulation Models”, Modelica Association, Proceedings of the 2011 Modelica Conference, modelica.org
Tool
Solver
FMU
Model
FMI Wrapper
Master
Slave
Model
Solver
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 31
ANSYS is working with experts on simulation of abuse scenarios
Improving Battery Safety
Damage translation
Extensions for abuse kinetics
Common geometry
Structural Simulation
Quarter-symmetric indentation (image courtesy NREL)
Current and temperature after internal short
Electro-Chemical-Thermal
(image courtesy MIT)
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 32
Using Workbench Parameter Manager and DesignXplorer, parametric studies can explore battery trade-offs
• Robust design, service life and safety improvements, …
Design Optimization
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 33
Battery requires complex control system for • Cell SOC balancing (maximize lithium utilization)
• Charging protocols and dynamic power-limiting (maximize battery life)
• Safety isolation and cell protection
• Integration with power electronics and other systems
A model-based development environment for embedded software
Battery Management Systems
Image courtesy of General Motors LLC
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 34
Next frontier for simulation: efficient coupling with materials-science and -processing (ICME)
• Closing the loop on the real-world chain of dependencies
• Example issues:
– Optimal particle size, morphology, binder/conductor mixture, etc.
– New cathode / anode / electrolyte materials, beyond lithium-ion, etc.
• ANSYS collaborates with molecular-modeling software leaders
Supporting Materials Innovation
Atomistic Particles Electrodes Cell Pack Vehicle
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 35
Reconstructing Microstructure Geometry
To create a representative volume element (RVE) model
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 36
RVE considers more-fundamental physics while limiting cost
Resolved Electrode Approach
(Image courtesy R. Kee, Colorado School of Mines)
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 37
Advanced batteries present a challenging application for the ANSYS vision of Simulation-Driven Product Development
Key gaps involve domain interfaces and transitions
• Multiscale methods for electrode battery system
• Cyber-physical (software hardware) system optimization
• Materials & manufacturing process in-service performance
ANSYS is addressing these challenges using several technologies, to support battery innovation
Summary
Chevrolet Bolt - image courtesy of General Motors LLC
© 2015 ANSYS, Inc. Collins / Convergence Apr 2015 38
ANSYS contributors: Erik Ferguson, Xiao Hu, Genong Li, Shaoping Li, Sandeep Sovani, Dimitri Tselepidakis
Acknowledgments
This material is based in part upon work supported by the Alliance for Sustainable Energy LLC, Management and Operating Contractor for the National Renewable Energy Laboratory, under Award Number ZCI-1-40497-01, and by the Office of Naval Research, under Award Number N00014-05-1-0339.
This presentation includes an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.