Underwater lithium-ion battery multi-physics field coupling modeling method

Abstract

The application relates to a kind of underwater lithium ion battery multi-physical field coupling modeling methods, belong to battery simulation technical field.It includes: constructing Thevenin equivalent circuit model, based on Thevenin equivalent circuit model to underwater lithium battery carries out HPPC mixed pulse power characteristic test, parameter identification obtains battery parameter;The battery parameter includes the ohmic internal resistance of battery, equivalent polarization capacitance and equivalent polarization internal resistance;Based on LS-DYNA and Ansys WorkBench, a battery model is established;Mechanical structure solver, thermodynamic solver and electromagnetic solver are respectively constructed for the established battery model, and the electromagnetic solver is solved based on the battery parameter;Mechanical structure solver, thermodynamic solver and electromagnetic solver are coupled to generate stress nephogram, displacement nephogram, temperature distribution graph and current density nephogram, obtain the multi-physical field change process of underwater lithium ion battery under stress, and better used in actual working condition by analyzing the multi-physical field change process.

Description

Technical Field

[0001] This invention relates to the field of underwater lithium-ion battery simulation technology, specifically to a multiphysics coupling modeling method for underwater lithium-ion batteries. Background Technology

[0002] The propulsion system is crucial in determining the endurance of a UUV. With the rapid development and application of UUVs in various fields, higher demands are being placed on their underwater endurance. Lithium-ion batteries, with their advantages of high energy density, high cycle life, and low self-discharge rate, have become the mainstream power source, meeting the power requirements of unmanned underwater vehicles for deep diving and long-range travel. Multiphysics coupling analysis of lithium-ion batteries is beneficial for the safe testing and optimized design of underwater vehicle energy systems.

[0003] The operating conditions of underwater lithium-ion batteries differ significantly from those on land, which are more favorable. The working environment is harsh, characterized by strong vibrations and poor environmental stability. Therefore, clarifying the distribution laws and mechanisms of the multi-physics fields of the battery's mechanical structure, electric field, and thermal field under underwater conditions is a critical issue that urgently needs to be addressed.

[0004] Currently, models for mechanical abuse, thermal runaway, and thermoelectric coupling of underwater lithium-ion batteries are well-developed, but research on multi-physics coupling models involving force, electricity, and heat is relatively weak. However, mechanical abuse of lithium batteries often involves more than one physics domain, frequently accompanied by mechanical damage, abnormal electrochemical responses, and thermal runaway. Focusing solely on mechanical behavior or electrical and thermal responses cannot meet the requirements for battery safety management. LS-DYNA is a mature explicit kinetic analysis software with a highly accurate mechanical solver, and it includes mature electromagnetic and thermal solvers for selection. In particular, the equivalent circuit model in the electromagnetic solver can effectively estimate the battery's electrochemical response. Compared to other finite element analysis software, LS-DYNA excels in explicit kinetic calculations, improving computational accuracy by achieving multi-physics coupling, making it more suitable for the operating conditions of underwater lithium batteries. Summary of the Invention

[0005] The technical problem to be solved by this invention is:

[0006] To clarify the multiphysics field distribution and mechanism of underwater lithium-ion batteries under underwater operating conditions—comprising battery mechanical structure, electric field, and thermal field—and to better apply them in practical applications, this invention provides a multiphysics field coupling modeling method for underwater lithium-ion batteries.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0008] A multiphysics coupling modeling method for underwater lithium-ion batteries, characterized by comprising:

[0009] A Thevenin equivalent circuit model was constructed, and HPPC hybrid pulse power characteristic tests were conducted on the underwater lithium battery based on the Thevenin equivalent circuit model. The battery parameters were identified, including the battery's ohmic internal resistance, equivalent polarization capacitance, and equivalent polarization internal resistance.

[0010] A battery model was built using LS-DYNA and Ansys WorkBench;

[0011] Mechanical structure solvers, thermodynamic solvers, and electromagnetic solvers are constructed for the established battery model. The electromagnetic solver solves based on the battery parameters. The mechanical structure solver, thermodynamic solver, and electromagnetic solver are coupled to generate stress cloud maps, displacement cloud maps, temperature distribution maps, and current density cloud maps, thus obtaining the multi-physics field change process of the underwater lithium-ion battery under stress.

[0012] A further technical solution of the present invention: the battery model established based on LS-DYNA and Ansys WorkBench includes:

[0013] Import or draw the lithium battery geometry model in Ansys Workbench;

[0014] Create an LS-DYNA module in Ansys Workbench and perform mesh generation based on LS-DYNA.

[0015] A further technical solution of the present invention: The mechanical structure solver establishes the constitutive relationship of the solid by calculating the momentum conservation equation of the battery assembly, and predicts the structural deformation of the battery under external or internal load; the thermodynamic solver obtains the three-dimensional temperature distribution of the battery by solving the energy conservation equation; the electrochemical analysis is completed by the electromagnetic solver, and the Randle equivalent circuit model obtains the potential and current density in the current collector by solving the charge balance equation. The EM solver detects whether potential contacts occur at the distribution points of the Randle equivalent circuit, thereby predicting battery short circuits.

[0016] A further technical solution of the present invention: the mechanical structure solver is constructed by including:

[0017] Apply constraints to the LS-DYNA module and define boundary conditions;

[0018] Apply loads to the LS-DYNA module to construct a mechanical force field model.

[0019] A further technical solution of the present invention: The construction of the electromagnetic solver specifically includes the following steps:

[0020] 1. On the right side, enter EM in Model→Keywrd→Model→Edit to add the keyword EM to the model;

[0021] 2. Click the plus sign on the left side of EM to expand all electrochemical types, select the CONTACT keyword, and select "Pick" in the GUI command interface to automatically generate the internal contact type of the lithium-ion battery;

[0022] 3. Select CONTROL_TIMESTEP and set the calculation time step. In the GUI command page, set the parameters: DTINIT - initial time step (e.g., 0.0, DYNA will automatically determine the initial step); TSSFAC - time step scaling factor, used to determine the new time step; ISDO - time step for calculating 4-node shell elements. Optional values ​​are as follows: 0: Feature length = area / min {longest side, longest diagonal}; 1: Feature length = area / longest diagonal; 2: Time step depends on stripe velocity and MAX {shortest side, area / min (longest side, longest diagonal)}; 3: Time step depends on the maximum eigenvalue.

[0023] 4. Select EM→MAT_001;

[0024] 5. Select the EM_BATTERY_RANDLES keyword to define the Randles equivalent circuit model of the lithium battery;

[0025] Where: CATHODE: Specifies the node number of the positive terminal, which is the positive terminal connection node of the battery;

[0026] ANODE: Specifies the node number of the negative terminal. The negative terminal is the negative terminal connection node of the battery.

[0027] SERIES_RESISTANCE: Defines the series resistance in the battery model; this is a parameter representing the internal resistance of the battery.

[0028] CAPACITANCE: Defines the capacitance in a battery model, which is used to describe the electrochemical reactions and charging / discharging processes of the battery.

[0029] SOCIT is the initial SOC of the battery;

[0030] SOCTOU gives the value of OCV source U, set to a constant value; R0DIS, R10DIS, and C10DIS give the values ​​of circuit elements R0, R10, and C10 when the current flows in the discharge direction; these are the battery parameters obtained from the HPPC hybrid pulse power characteristic test.

[0031] FRTHERM is used to determine whether the temperature of a parameter is a constant temperature or a temperature driven by a thermal solver.

[0032] R0TOTHERM is used to determine whether to add Joule heating caused by internal resistance R0 to the thermal solver.

[0033] A further technical solution of the present invention: the construction of the thermodynamic solver specifically includes the following steps:

[0034] 1. Add CONTROL_THERMAL_SOLVER to control parameters related to the thermal solver.

[0035] The relevant parameters are explained as follows: ATYPE: Classification of thermal analysis: 0: steady-state analysis, 1: transient analysis; PTYPE: Whether the thermal problem is linear or nonlinear: 0: linear problem, 1 or 2: nonlinear problem, requiring the definition of the *CONTROL_THERMAL_NONLINEAR keyword; FWORK: Percentage of plastic deformation energy converted into heat energy, allowing this parameter to consider the conversion of plastic deformation energy into heat energy during stamping or forging; TS: Time step setting: 0: fixed time step, 1: variable time step;

[0036] 2. Add CONTROL_THERMAL_TIMESTEP to control the time step of the solution. The time step is the length of each finite element integration step. When calculating the required time step, all elements should be checked. For stability reasons, use 0.9 to determine the minimum time step: Δt = 0.9l / c, where the characteristic length l and the wave propagation speed c are related to the element type. DTINIT: Initial time step, determined by DYNA. TSSFAC: Time step scaling factor, used to determine the new time step, default is 0.9. ISDO: Time step for calculating 4-node shell elements. Optional values ​​are as follows: 0: Characteristic length = area / min{longest side, longest diagonal}, 1: Characteristic length = area / longest diagonal, 2: Time step depends on the wave velocity and the longest side; this option provides a relatively large time step, which may lead to unstable calculations, especially when using triangular elements, 3: Time step depends on the maximum eigenvalue.

[0037] A computer system is characterized by comprising: one or more processors, and a computer-readable storage medium for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method described above.

[0038] A computer-readable storage medium is characterized by storing computer-executable instructions, which, when executed, are used to implement the above-described method.

[0039] The beneficial effects of this invention are as follows:

[0040] This invention constructs a complete multiphysics coupling modeling method for underwater lithium-ion batteries using experiments, modeling, and simulation. First, HPPC hybrid pulse power characteristic experiments are conducted on the underwater lithium battery to identify battery parameters based on the Thevenin model, laying the foundation for electromagnetic field solutions using LS-DYNA. Then, based on underwater operating conditions, mechanical structure solvers, thermodynamic solvers, and electromagnetic solvers for the battery are established using LS-DYNA and Ansys Workbench. Coupled calculations generate stress contour maps, displacement contour maps, thermal distribution maps, and current density contour maps, revealing the multiphysics changes in the underwater lithium-ion battery under stress. Attached Figure Description

[0041] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts.

[0042] Figure 1 Flowchart of the modeling method of this invention.

[0043] Figure 2 Relationship diagram of force, electro, and heat solvers.

[0044] Figure 3 LS-DYNA keyword structure relationship diagram.

[0045] Figure 4 Thevenin equivalent circuit diagram.

[0046] Figure 5 HPPC sampling data graph. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0048] like Figure 1As shown, this invention provides a multiphysics coupling modeling method for underwater lithium-ion batteries. A complete multiphysics coupling modeling method for underwater lithium-ion batteries is constructed through experimentation, modeling, and simulation. First, HPPC hybrid pulse power characteristic experiments are conducted on underwater lithium batteries to achieve battery parameter identification based on the Thevenin model, laying the foundation for electromagnetic field solutions using LS-DYNA. Then, based on underwater operating conditions, mechanical structure solvers, thermodynamic solvers, and electromagnetic solvers for the battery are established using LS-DYNA and Ansys WorkBench. Coupled calculations generate stress contour maps, displacement contour maps, temperature distribution maps, and current density contour maps, revealing the multiphysics changes of the underwater lithium-ion battery under stress.

[0049] This study utilizes multiple solvers in LS-DYNA to perform multiphysics simulation modeling and analysis of the mechanical, thermal, and electrochemical responses of lithium-ion batteries. Specifically, it employs a geometric model of an underwater lithium-ion pouch battery, combined with the Randles equivalent circuit model, and uses the mechanical solver, thermal solver, and EM electromagnetic solver in LS-DYNA to numerically simulate the battery, effectively predicting its thermoelectric response behavior during operation. The EM solver accepts the Thevenin equivalent circuit model; therefore, parameter identification methods are needed to obtain the battery's ohmic internal resistance, equivalent polarization capacitance, and equivalent polarization internal resistance.

[0050] In LS-DYNA, the mechanical solver, thermal solver, and EM electromagnetic solver are coupled. The mechanical solver establishes the constitutive relationship of the solid by calculating the momentum conservation equation of the battery assembly, predicting the structural deformation of the battery under external or internal loads; the thermal solver obtains the three-dimensional temperature distribution of the battery by solving the energy conservation equation; electrochemical analysis is performed by the electromagnetic solver. The Randles equivalent circuit model obtains the potential and current density within the current collector by solving the charge balance equation, and the EM solver can detect potential contacts at the distribution points of the Randles equivalent circuit, thereby predicting battery short circuits. Each solver and the equivalent model provides term sources to each other, such as... Figure 2 As shown.

[0051] The mechanical solver transmits displacement and stress calculations to the electromagnetic solver and transfers the heat generated by the work done to the thermal solver. The electromagnetic solver calculates the Lorentz force reaction to the mechanical solver, measures the damping effect of the circuit, and provides Joule heat to the thermal solver. Temperature, as a result of the thermal solver, can also serve as a term source for the electromechanical solver. In the electromagnetic solver, the current and the heat it generates are modeled using the Randle equivalent circuit to describe the electrochemical response of the current collector. The Randle equivalent circuit is related to temperature, current direction, and battery state of charge (SOC).

[0052] During numerical calculations, due to potential deformation of the mechanical structure, the electromagnetic and thermal solvers must periodically recalculate their matrix systems. Therefore, the step size of the mechanical solver is smaller than that of the electromagnetic and thermal solvers, while the step size of the electromagnetic and thermal solvers is similar. When traversing the parameters of the mechanical solver, the thermoelectric parameters can be estimated through numerical calculation methods using function interpolation to update the thermoelectric solver. This allows the mechanical solver to predict the occurrence of internal or external short circuits in the initial stage, and then the thermal, electrical, and electrochemical solvers capture the evolution of temperature and current distribution after the short circuit.

[0053] 1. Configure the mechanical solver

[0054] An explicit mechanical solver seeks solutions to the momentum conservation equation:

[0055]

[0056] Where σ represents stress, x represents coordinate, u represents displacement, ρ is the module density, f represents the external load applied to the battery cell, and t represents the loading time. This is the Lorentz force generated by electromagnetic interaction.

[0057] The constitutive equation for solid-state lithium-ion batteries is:

[0058] σ=C kl δ kl

[0059] Among them, C kl This is the stiffness matrix of the battery module, defined as δ. kl For variables related to displacement and direction, the expressions are as follows:

[0060]

[0061] 2. Configure the thermodynamic solver

[0062] The governing equation for heat conduction in a three-dimensional solid is given by the following equation.

[0063]

[0064] Where T is temperature, c is specific heat, and k is... ij Q is the thermal conductivity, and Q is the heat generated per unit volume per unit time by the component.

[0065]

[0066] W plastic E is the work, provided by the mechanical solver, that induces plastic deformation. joule It is Joule heat generated by electromagnetic interaction, and V is the volume of the component.

[0067] 3. Configure the electromagnetic solver

[0068] The electromagnetic solver is used to describe the electrochemical response of a battery and consists of a Randle equivalent circuit and an EM solver.

[0069] (1) EM solver

[0070] The premise of using the EM solver is that eddy current effects, no magnetic effects or any other electrostatic effects, no divergent current density and no charge accumulation are not considered. Its main purpose is to study the heat generated by current passing through a conductor and observe its effect on temperature.

[0071] The EM solver solves the two equations that constitute the system response:

[0072]

[0073]

[0074] in, It is a vector magnetic potential. The scalar potential and vector magnetic potential are the unknowns that need to be solved. κ is the conductivity, μ is the permeability, and j is the vector magnetic potential. s It is the current density of the power supply.

[0075] The EM solver calculates the Lorentz force generated by electromagnetic interaction. Return to the mechanical solver.

[0076]

[0077] in, B is the current density, B is the magnetic field strength, and E is the Joule heat generated by the electromagnetic interaction. joule Return to the thermal solver.

[0078]

[0079] in, κ is the current density, and κ is the conductivity.

[0080] (2) Equivalent circuit model

[0081] This method proposes to take several symmetrical nodes on average between the two current collectors of the battery and connect them with the Thevenin equivalent circuit to accurately estimate the current density of the battery.

[0082] Thevenin equivalent circuit is a first-order RC equivalent circuit. It exhibits good nonlinear characteristics and can quickly and conveniently simulate the dynamic and static voltage characteristics of lithium batteries. The Thevenin equivalent circuit consists of an ideal voltage source, an internal resistance, and two RC circuits representing the battery polarization. Before numerical calculations, the parameters of the equivalent circuit need to be identified using the HPPC method.

[0083] The equivalent circuit of a Randle has multiple nodes evenly distributed, therefore the electrochemical parameters of the component are m times that of a single transmission node, i.e.

[0084] R 0total =mR0

[0085] R total =m(R1+R2)

[0086]

[0087] For a single transmission node, the Thevenin equivalent circuit diagram is as follows: Figure 4 As shown.

[0088] The specific steps described above are as follows:

[0089] 1. Battery parameter identification based on the Thevenin model

[0090] 1.1 Constructing the Thevenin equivalent circuit model

[0091] Thevenin equivalent circuit diagram is as follows: Figure 4 As shown.

[0092] Thevenin equivalent circuit is mainly composed of components R0, C10, R10, and Em constant voltage source, where: R0 is the ohmic internal resistance, C10 is the equivalent polarization capacitance, R10 is the equivalent polarization internal resistance, Em is the open circuit voltage, and Uoc is the terminal voltage.

[0093] 1.2 Conduct HPPC battery testing

[0094] The hybrid pulse power characteristic (HPPC) test method can be used to calculate the parameters of various components. For specific HPPC tests, please refer to the "Freedom CAR Test Manual".

[0095] The HPPC method applies a pulsed current of t seconds to the battery, analyzes the voltage change, and calculates the battery's peak power based on a given formula. The pulsed current value is determined by fitting and inferring from multiple test data. The constant power test method, on the other hand, continuously charges or discharges the battery at a constant power for t seconds until the battery's operating voltage reaches its upper or lower limit. This constant power is obtained by adjusting the power setting and conducting multiple tests to gradually approach t seconds.

[0096] The specific experimental steps are as follows:

[0097] 1.2.1 Charge the battery to full charge using the constant current and constant voltage method;

[0098] 1.2.2 Discharge the battery to a certain SOC point using the constant current method;

[0099] 1.2.3 Let stand for one hour;

[0100] 1.2.4 Discharge with pulsed current for 10 seconds, rest for 40 seconds, charge with pulsed current for 10 seconds, and record the current and voltage values;

[0101] 1.2.5 Set the SOC in step 1.2.2 to decrease from 90% to 10%, with each decrease in SOC at 10% intervals, and repeat steps 1.2.2 to 1.2.4 nine times.

[0102] The following principles can be followed when collecting sampling data:

[0103] 1.3 Parameter Identification

[0104] 1.3.1 Based on the voltage change during the pulse discharge process, the ohmic internal resistance of the battery can be calculated as follows:

[0105]

[0106] Among them, R o U1 is the internal resistance of the battery in ohms, U2 is the voltage at time t1, U1 is the voltage at time t2, I1 is the current at time t1, and I2 is the current at time t2.

[0107] In the Thevenin circuit, let the time constant be...

[0108] τ=R 10 ·C 10

[0109]

[0110]

[0111] Therefore, it can be deduced from formulas (1) and (2)

[0112] time constant

[0113]

[0114] Where U1 is the voltage value at t1, U4 is the voltage value at t4, U3 is the voltage value at t3, and t3 and t4 are the initial and final times when the battery is stationary.

[0115] U3=U OC +iR0+U RC ……(3)

[0116]

[0117] Therefore, from formulas (3) and (4) we can deduce...

[0118]

[0119] Where I2 is the current value at time t2, U OC This indicates the terminal voltage.

[0120] The ohmic internal resistance R0, the equivalent polarization capacitance C10, and the equivalent polarization internal resistance R10 can be calculated from the above steps.

[0121] 2. Model preprocessing based on Ansys Workbench: Import or draw the geometric model of the lithium battery in Ansys Workbench to describe the physical system of the battery.

[0122] 2.1 In the Ansys Workbench toolbox on the left, select the "Geometry" module in the "Component System" and drag it into the project schematic area on the right to generate the geometry module for geometric modeling of the battery cells.

[0123] 2.2 For a completed battery geometry model, the import function can be used. Right-click on "Geometry", select "Import Geometry Model", and proceed to step 3.

[0124] 2.3 Double-click the "Geometry" module to enter the Ansys modeling interface. The modeling tools you can choose are SpaceClaim, Discovery, and DesignModerler.

[0125] 2.4 Draw the geometric model of the lithium battery in the "Structure" module. Refer to the *ANSYS Basic Procedure Guide* and *ANSYS Modeling and Meshing Guide* for drawing instructions. The structure of a single battery cell should include an aluminum-plastic film, a positive electrode, a separator, and a negative electrode.

[0126] 3. Create an LS-DYNA module in Ansys Workbench for mesh generation.

[0127] 3.1 In the toolbox on the left side of Ansys Workbench, select the "LS-DYNA" module, drag it into the schematic area on the right side of the project, generate the "LS-DYNA" module, and long press and drag to connect the geometry of the "Geometric Model" module with the geometry of the "LS-DYNA" module to achieve data sharing.

[0128] 3.2 To open the geometry of the "LS-DYNA module", the geometric model in step 1 needs to be simplified to improve its accuracy and reduce the solver's workload. Specifically, delete small holes, chamfers, threaded surfaces, gaps, grooves, etc., that do not affect the feature structure, making the model a regular body.

[0129] 3.3 Mesh Generation Based on LS-DYNA

[0130] 3.3.1 Open the model in the "LS-DYNA module", click Mesh → Insert → Method, the system will automatically generate a mesh, but the mesh is coarse at this time and needs to be refined.

[0131] 3.3.2 Click the mesh module, and in the properties box at the bottom left, select Range → Geometry → check the entire battery cell structure, then define → Method → ​​Hexahedral Dominant. Since underwater lithium-ion pouch batteries are usually square structures, a regular hexahedral mesh structure can effectively enhance the stability of the mesh and improve the calculation accuracy.

[0132] 4. Apply constraints to the geometric model and define boundary conditions.

[0133] 4.1 Building upon step 3, continue defining connection types in the "LS-DYNA Module" model. In the left-hand project pane, select Connection → Insert → Geometry Interaction. In the properties pane, select All Geometry as the scope and Frictionless as the type.

[0134] 4.2 Based on step 4.1, continue to define constraint types in the model of the “LS-DYNA module”.

[0135] 4.2.1 In the left-hand project bar, select LS-DYNA, select Analysis Settings, and in the Properties box, select Step Control → End Time → Enter 0.01s. If the geometric model is thick, the end time can be increased appropriately.

[0136] 4.2.2 In the left-hand project bar, select LS-DYNA → Insert → Fixed Support, and in the property bar, select the aluminum-plastic film base plate as the range and apply it.

[0137] 4.3 In the upper toolbar, select "LSDYNA Pre" → Rigid Body Tools → Rigid Body Constraints → Define → Select "Free in Z-direction" for the person, allowing the battery to deform only in the Z-axis direction.

[0138] 5. Apply loads to the geometric model to construct a mechanical force field.

[0139] 5.1 Continue by selecting Analysis Settings → Insert → Force. Based on the underwater working conditions, select Range in the Properties bar → check the entire battery cell structure, set the force magnitude to 1000kN, select the point of application of the force as the center of the aluminum-plastic film cover, and set the direction of the force to be perpendicular to the cover and towards the inside of the battery.

[0140] 5.2 Continue by selecting LD-DYNA → Insert → Displacement, set the range to aluminum-plastic film cover, and define the x component and y component to 0mm.

[0141] 5.3 Click "Solve" in the left-hand project panel, then click the solve button above to generate the model. Right-click the solver and select the input.k file, which is the mechanical field modeling file for the battery.

[0142] 6. Establish an electromagnetic solution model in LS-PrePost software.

[0143] 6.1 Define the unit type

[0144] 6.1.1 Import the mechanical force field model into LS-PrePost software: File → Open → LS-DYNA Keyword File → Open the input.k file from step 4.3 → Open

[0145] 6.1.2 Implement LS-DYNA control modeling through GUI commands. First, go to Main Menu → Preference → “LS-DYNA Explicit”. This will completely filter the menu into explicit and dynamic input options.

[0146] 6.1.3 Each individual structure (positive electrode, negative electrode, separator, aluminum-plastic film) has a component keyword "part" for setting the structure. After selecting the unit type of the structure, the software automatically links the physical, electrochemical and thermodynamic properties of the structure.

[0147] 6.1.3.1 On the right side, enter PART in Model→Keywrd→Model→Edit to add the PART keyword to the model.

[0148] 6.1.3.2 Double-click the part to expand the GUI command page for settings. Enable the pick interface to allow data to be picked directly from the model. (Where EOSID is the equation of state for the material *EOS keyword; HGID: hourglass control for this component, referencing a card of type HourGlass; GRAV: gravity initialization, 0—initialization of all parts; 1—initialization only for the current part; ADPOPT: mesh adaptation, 0—no mesh adaptation; 1—3D mesh adaptation; 2—adaptation to both 2D and 3D meshes; TMID: thermodynamic properties defined by *MAT_THEMAL.)

[0149] 2. In the Model→Keywrd→Model→Edit section on the right, enter SECTION to add element types to the model. Click the plus sign to the left of SECTION to expand all element types. In SECTION, select the SHELL keyword and choose shell element as the element type for both the positive and negative poles. In the GUI command page, select the type under the corresponding keyword. ELFORM: The integration algorithm for solving the problem; the default is 1, 2—single-point integration algorithm, fast but cannot accurately handle warping and cannot be used in coarse meshes; SHRF: Shear factor, the default is 1, 5 / 6 is recommended. T1, T2, T3, T4: The mesh thickness at the four nodes; often defining T1 is sufficient, and T2, T3, T4 are usually consistent with T1 by default. Note: For shell elements, usually only the integration algorithm ELFORM and the thickness need to be modified; the rest can be left as default.

[0150] 6.1.3.3 Similarly, in SECTION, select the SOLID keyword and choose solid elements as the element type for both the positive and negative electrode materials; ELFORM: solid element type selection; AET: ambient environment type option. For collision, extrusion, and impact simulations, the default values ​​can be used.

[0151] 6.2 Define the material model

[0152] 6.2.1 In the Model→Keywrd→Model→Edit section on the right, enter MAT to add the MAT keyword to the model. Click the plus sign to the left of MAT to expand all material types. In the SECTION, select the SHELL keyword and set the shell element to the isotropic piecewise linear plastic LS-DYNA material model MAT_24_PIECEWISE_LINEAR_PLASTICITY to define it as an elastoplastic material. Define its material constitutive model using the following parameters: Rho—density; E—Young's modulus; Nu—Poisson's ratio; SIGY—yield strength; ETAN—uniform modulus; TOEL—minimum time step for deleting mesh elements; C, P—strain rate coefficients (used to characterize the rate of expansion, contraction, and shear deformation of the material); LCSS: stress-strain curve of the material; LCSR—yield limit curve at different strain rates. After defining the LCSS stress-strain curve, the SIGY and ETAN parameters will be ignored because they can be calculated from the stress-strain curve; after defining the LCSR, the C and P values ​​will also be ignored.

[0153] 6.2.2 Set the solid element to the MAT_126_Modified_Honeycomb honeycomb material model (model number 126). In the GUI material interface, input the material's density (DENS), elastic modulus (EX), tensile cutoff stress (TC), hysteresis unloading factor (HU) (1 for no energy dissipation, 0 for all energy dissipation, usually between 0 and 1), retardation constant (BEA), viscosity coefficient (DAMP) (recommended value between 0.05 and 0.5), shape unloading factor (SHAPE), failure selection at cutoff stress (FAIL) (0 for tensile stress holding cutoff value, 1 for tensile stress zero value), and volumetric viscosity effect flag (BVFLAG) (0 for no volumetric viscosity, 1 for volumetric viscosity effect).

[0154] 6.3 Define contact parameters

[0155] 6.3.1 On the right side, enter CONTACT in Model→Keywrd→Model→Edit to add the CONTACT keyword to the model.

[0156] 6.3.2 Click the plus sign to the left of CONTACT to expand all contact types. In CONTACT, select the keyword SURFACE_TO_SUIFACE to generate face-to-face contacts between structures. Double-click SURFACE_TO_SUIFACE to set parameters in the GUI interface:

[0157] Among them, SSID is used to specify the slave face; MSID is used to specify the master face; SPR and MPR control whether to calculate and output interface forces; 0 - do not calculate; 1 - calculate and output in RCFORC.

[0158] FS and FD specify the static and dynamic friction coefficients, respectively; DC: calculates the friction coefficient; VC: viscous friction coefficient; VDC: viscous damping coefficient; PENCHK: initial contact check; O—no check; 1—check; 2—check enabled, performing minimum diagonal search for contact; BT and DT: contact start and end times; SFS and SFM: penalty function coefficients for the master and slave surfaces, default value is 1; FSF: Coulomb friction coefficient; VSF: viscous friction coefficient.

[0159] 6.4 Add an electromagnetic solution module

[0160] 6.4.1 On the right side, enter EM in Model→Keywrd→Model→Edit to add the keyword EM to the model.

[0161] 6.4.2 Click the plus sign on the left side of EM to expand all electrochemical types, select the CONTACT keyword, and select "Pick" in the GUI command interface to automatically generate the internal contact type of the lithium-ion battery.

[0162] 6.4.3 Select CONTROL_TIMESTEP to set the calculation time step. In the GUI command page, set the parameters: DTINIT - Initial time step. If set to 0.0, DYNA will automatically determine the initial step. TSSFAC - Time step scaling factor, used to determine the new time step. The default is 0.9. When the calculation is unstable, this value can be reduced, but this will increase the calculation time. ISDO - Time step for calculating 4-node shell elements (different values ​​correspond to different algorithms for feature length; 2 is recommended because this option provides the largest time step, but the presence of triangular elements can lead to calculation instability). Optional values ​​are as follows: 0: Feature length = Area / min {longest side, longest diagonal}, 1: Feature length = Area / longest diagonal, 2: Time step depends on bar wave speed and MAX {shortest side, Area / min (longest side, longest diagonal)}. This option provides a relatively large time step, which may lead to calculation instability, especially when applying triangular elements, 3: Time step depends on the maximum eigenvalue. This option is suitable for structures where the sound propagation speed of the material gradually changes. The computational overhead for calculating the largest eigenvalue is significant, but increasing the time step usually takes into account shorter computation cycles without quality scaling.

[0163] 6.4.4 Select EM→MAT_001. When temperature changes, causing a change in conductivity (i.e., resistance), the EM solver must recalculate its matrix. To account for this effect, the user must define the state equations associated with *EM_MAT by entering the corresponding numerical values ​​for the keywords. The parameters for the MAT_001 keyword include permeability (MU), conductivity (SIGMA), and dielectric constant (EPSILON), etc., and the specific parameters depend on the type of material and electromagnetic properties you are simulating.

[0164] 6.4.5 Select the EM_BATTERY_RANDLES keyword to define the Randles equivalent circuit model of the lithium battery.

[0165] Where: CATHODE: Specifies the node number of the positive terminal. The positive terminal is the positive connection node of the battery. ANODE: Specifies the node number of the negative terminal. The negative terminal is the negative connection node of the battery.

[0166] SERIES_RESISTANCE: Defines the series resistance in the battery model. This is a parameter representing the internal resistance of the battery. CAPACITANCE: Defines the capacitance in the battery model. Capacitance is used to describe the electrochemical reactions and charge / discharge processes of the battery. SOCIT is the initial SOC of the battery.

[0167] SOCTOU gives the value of the OCV source U, set to a constant value. R0DIS, R10DIS, and C10DIS give the values ​​of circuit elements R0, R10, and C10 when the current flows in the discharge direction. These are the values ​​from the Matlab run in step 1, set to normal values. FRTHERM is used to determine whether the parameter temperature is a constant temperature or a temperature driven by the thermal solver. R0TOTHERM is used to determine whether to add Joule heating generated by the internal resistance R0 to the thermal solver; this step should be filled with a constant 1 to couple the electrothermal solver.

[0168] The values ​​of R0DIS, R10DIS, C10DIS, SOCIT, and SOCTOU here are all derived from the parameter identification conclusions in step 1 and the HPPC test conditions.

[0169] 7. Establish a thermal solution model in LS-PrePost software.

[0170] 7.1 Add CONTROL_THERMAL_SOLVER to control relevant parameters of the thermal solver. Parameter descriptions: ATYPE: Classification of thermal analysis: 0: Steady-state analysis, 1: Transient analysis; PTYPE: Whether the thermal problem is linear or nonlinear: 0: Linear problem, 1 or 2: Nonlinear problem (requires defining the *CONTROL_THERMAL_NONLINEAR keyword); FWORK: Percentage of plastic deformation energy converted into heat energy, allowing this parameter to consider the conversion of plastic deformation energy into heat energy during stamping or forging; TS: Time step setting: 0: Fixed time step, 1: Variable time step (can be increased or decreased, controlled by the program).

[0171] 7.2 Add CONTROL_THERMAL_TIMESTEP to control the time step size of the solution. The time step size is the length of each finite element integration step. All elements must be checked when calculating the required time step. For stability reasons, use 0.9 (default) to determine the minimum time step: Δt = 0.9l / c, where the characteristic length l and wave propagation speed c are related to the element type. DTINIT: Initial time step size, determined automatically by DYNA. TSSFAC: Time step scaling factor, used to determine the new time step size, defaults to 0.9. This value can be decreased when the calculation is unstable, but this will increase the calculation time. ISDO: Time step size for calculating 4-node shell elements (different values ​​correspond to different algorithms for the characteristic length; 2 is recommended because this option provides the largest time step size, but the presence of triangular elements can lead to calculation instability). The possible values ​​are as follows: 0: Feature length = Area / min{longest side, longest diagonal}; 1: Feature length = Area / longest diagonal; 2: Time step depends on bar wave speed and the longest side. This option provides a relatively large time step, which may lead to computational instability, especially when applying triangular elements; 3: Time step depends on the maximum eigenvalue. This option is suitable for structures where the sound propagation speed of the material varies gradually. The computational cost of calculating the maximum eigenvalue is significant, but increasing the time step usually considers shorter computation cycles without mass scaling.

[0172] 7.3 Click the "Run" button on the toolbar to perform the calculation using LS-DYNA software and export the k file.

[0173] 8. Open the LS-DYNA model in Ansys Workbench and perform post-processing on the model calculation results.

[0174] 8.1 Clicking the "Solve" button in the toolbar above allows you to observe the stress, displacement, temperature distribution, and current density cloud maps of an underwater lithium-ion battery under underwater operating conditions. The dynamic changes in these cloud maps reveal the multi-physics distribution of the battery, enabling optimization of the battery's mechanical structure and environmental conditions based on parameter variations. For the stress cloud map, the areas of highest stress are where stress concentration occurs and structural reinforcement is needed. The displacement cloud map allows you to observe battery deformation; if mechanical failure exists, further modifications to the environmental conditions are required to minimize stress at the failure points. The temperature distribution map shows the battery's temperature distribution during operation; areas with high temperatures require optimization of heat dissipation methods and structures. If the highest temperature exceeds the heat resistance range of the casing material, the material must be replaced or a more effective heat dissipation solution must be implemented. The current density distribution map reveals areas of high current concentration; these areas require reducing pressure on the battery under environmental conditions to prevent mechanical abuse or to improve the battery's manufacturing process for balanced discharge.

[0175] 8.2 Optimize the solution results: Select "Solve" in the toolbar above to choose to view the maximum and minimum values, use a probe to detect the parameter value of a point, whether to display the grid, only display the border lines, and display the part of the parameter value that is less than or greater than a fixed value.

[0176] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the scope of the technology disclosed in the present invention, and such modifications or substitutions should all be covered within the scope of protection of the present invention.

Claims

1. A multiphysics coupling modeling method for underwater lithium-ion batteries, characterized in that, include: A Thevenin equivalent circuit model was constructed, and HPPC hybrid pulse power characteristic tests were conducted on the underwater lithium battery based on the Thevenin equivalent circuit model. The battery parameters were identified, including the battery's ohmic internal resistance, equivalent polarization capacitance, and equivalent polarization internal resistance. A battery model was built using LS-DYNA and Ansys WorkBench; the battery model built using LS-DYNA and Ansys WorkBench includes: Import or draw the lithium battery geometry model in Ansys Workbench; Create an LS-DYNA module in Ansys Workbench and perform mesh generation based on LS-DYNA; Mechanical structure solvers, thermodynamic solvers, and electromagnetic solvers are constructed for the established battery model, with the electromagnetic solver solving based on battery parameters. The mechanical structure solver, thermodynamic solver, and electromagnetic solver are coupled to generate stress cloud maps, displacement cloud maps, temperature distribution maps, and current density cloud maps, thus obtaining the multi-physics field change process of the underwater lithium-ion battery under stress. The mechanical structure solver establishes the constitutive relationship of the solid by calculating the momentum conservation equation of the battery assembly, and predicts the structural deformation of the battery under external or internal loads; the thermodynamic solver obtains the three-dimensional temperature distribution of the battery by solving the energy conservation equation; the electrochemical analysis is completed by the electromagnetic solver, and the Randle equivalent circuit model obtains the potential and current density in the current collector by solving the charge balance equation. The EM solver detects whether potential contacts have occurred at the distribution points of the Randle equivalent circuit, thereby predicting battery short circuits.

2. The multiphysics coupling modeling method for underwater lithium-ion batteries according to claim 1, characterized in that, The mechanical structure solver construction includes: Apply constraints to the LS-DYNA module and define boundary conditions; Apply loads to the LS-DYNA module to construct a mechanical force field model.

3. The multiphysics coupling modeling method for underwater lithium-ion batteries according to claim 1, characterized in that, The construction of the electromagnetic solver specifically includes the following steps: 3.

1. On the right side, enter EM in Model→Keywrd→Model→Edit to add the keyword EM to the model; 3.2 Click the plus sign on the left side of EM to expand all electrochemical types, select the CONTACT keyword, and select "Pick" in the GUI command interface to automatically generate the internal contact type of the lithium-ion battery; 3.3 Select CONTROL_TIMESTEP and set the calculation time step. In the GUI command page, set the parameters: DTINIT - initial time step (e.g., 0.0, DYNA will automatically determine the initial step); TSSFAC - time step scaling factor, used to determine the new time step; ISDO - time step for calculating 4-node shell elements. Optional values ​​are as follows: 0: Feature length = area / min {longest side, longest diagonal}, 1: Feature length = area / longest diagonal, 2: Time step depends on strip velocity and MAX {shortest side, area / min (longest side, longest diagonal)}, 3: Time step depends on the maximum eigenvalue. 3.

4. Select EM→MAT_001; 3.5 Select the EM_BATTERY_RANDLES keyword to define the Randles equivalent circuit model of the lithium battery; Where: CATHODE: Specifies the node number of the positive terminal, which is the positive terminal connection node of the battery; ANODE: Specifies the node number of the negative terminal. The negative terminal is the negative terminal connection node of the battery. SERIES_RESISTANCE: Defines the series resistance in the battery model; this is a parameter representing the internal resistance of the battery. CAPACITANCE: Defines the capacitance in a battery model, which is used to describe the electrochemical reactions and charging / discharging processes of the battery. SOCIT is the initial SOC of the battery; SOCTOU gives the value of OCV source U, set to a constant value; R0DIS, R10DIS, and C10DIS give the values ​​of circuit elements R0, R10, and C10 when the current flows in the discharge direction; these are the battery parameters obtained from the HPPC hybrid pulse power characteristic test. FRTHERM is used to determine whether the temperature of a parameter is a constant temperature or a temperature driven by a thermal solver. R0TOTHERM is used to determine whether to add Joule heating caused by internal resistance R0 to the thermal solver.

4. The multiphysics coupling modeling method for underwater lithium-ion batteries according to claim 1, characterized in that, The construction of the thermodynamic solver specifically includes the following steps: 4.1 Add CONTROL_THERMAL_SOLVER to control the parameters related to the thermal solver. The relevant parameters are explained as follows: ATYPE: Classification of thermal analysis: 0: steady-state analysis, 1: transient analysis; PTYPE: Whether the thermal problem is linear or nonlinear: 0: linear problem, 1 or 2: nonlinear problem, requiring the definition of the *CONTROL_THEMAL_NONLINEAR keyword; FWORK: Percentage of plastic deformation energy converted into heat energy, allowing this parameter to consider the conversion of plastic deformation energy into heat energy during stamping or forging; TS: Time step setting: 0: fixed time step, 1: variable time step. 4.2 Add CONTROL_THERMAL_TIMESTEP to control the time step of the solution. The time step is the length of each finite element integration step. When calculating the required time step, all elements should be checked. For stability reasons, use 0.9 to determine the minimum time step: Δt = 0.9 l / c, where the characteristic length l and the wave propagation speed c are related to the element type. DTINIT: Initial time step, determined by DYNA. TSSFAC: Time step scaling factor, used to determine the new time step, default is 0.

9. ISDO: Time step for calculating 4-node shell elements. Optional values ​​are as follows: 0: Characteristic length = area / min{longest side, longest diagonal}, 1: Characteristic length = area / longest diagonal, 2: Time step depends on the wave velocity and the longest side; this option provides a relatively large time step, which may lead to unstable calculations, especially when applying triangular elements, 3: Time step depends on the maximum eigenvalue.

5. A computer system, characterized in that... include: One or more processors, a computer-readable storage medium for storing one or more programs, wherein, when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to perform the method of any one of claims 1-4.

6. A computer-readable storage medium, characterized in that... The device stores computer-executable instructions, which, when executed, are used to implement the method described in any one of claims 1-4.

Images