A method for optimizing the CAE mesh of a battery pack

Abstract

The present invention discloses a method for optimizing the CAE mesh of a battery pack, comprising the following steps: S1. Open the battery pack model in 3D modeling software and perform structural feature simplification processing; S2. Complete the model with simplified features in the 3D modeling software in STP format, and open the model with simplified features in finite element software and perform segmentation on it; S3. Divide the battery pack model into surface meshes or hexahedron meshes; S4. Perform surface mesh generation processing; S5. Perform hexahedron mesh generation processing; S6. For a symmetric battery pack model, perform mirror processing on the battery pack model that has completed the mesh division; S7. Substitute the mesh of the overall battery pack model into a certain working condition for solution calculation. By combining 3D modeling software and finite element software and using the combination of hexahedron and surface elements for finite element mesh division, the present invention greatly reduces the number of meshes.

Description

Technical Field

[0001] The present invention relates to the technical field of CAE finite element, and particularly relates to a method for optimizing the CAE mesh of a battery pack. Technical Background

[0002] In the existing CAE simulation methods in the field of battery pack simulation, when CAE simulation engineers divide the mesh of components, they often directly divide the mesh of geometric bodies using software such as Hypermesh, and the same part only corresponds to one type of mesh. Shell elements are used for sheet metal parts, and tetrahedral or hexahedral elements are used for injection molded parts, etc.

[0003] When dealing with a battery pack model with a relatively complex structure, a combination of tetrahedral elements and surface meshes is generally used. Among them, the anti-twisting and deformation capabilities of tetrahedral meshes are not as good as those of hexahedral meshes or surface meshes, and usually, the quality of tetrahedrons is difficult to control. During the process of submitting to the computer for solution, problems such as mesh penetration and distortion are likely to occur, resulting in solution failure. Moreover, the existing method of dividing the mesh may cause a situation where the number of model meshes is large and the quality is poor. Poor mesh quality and excessive quantity will directly lead to problems such as slow simulation calculation speed and low output efficiency of CAE simulation engineers.

[0004] Therefore, it is very necessary to develop a new method for optimizing the CAE mesh of a battery pack. Summary of the Invention

[0005] The present invention provides a method for optimizing the CAE mesh of a battery pack to solve the problems of a large number of meshes and poor quality caused by the existing mesh division method, resulting in a relatively slow final simulation calculation speed.

[0006] The present invention discloses a method for optimizing the CAE mesh of a battery pack, including the following steps:

[0007] S1. Open the battery pack model in 3D modeling software, observe its overall structure, delete the redundant parts that are not involved in the simulation calculation, and perform structural feature simplification processing;

[0008] S2. Complete the model with simplified features in the 3D modeling software in STP format, open the model with simplified features in finite element software, observe whether each part is symmetric and perform simplification processing, then divide the single part after the simplification processing into a first part and a second part. The first part includes the parts with equal thickness, and the second part includes the parts with unequal thickness, and then cut the first part and the second part respectively;

[0009] S3. Divide the first part into surface meshes, divide the second part into hexahedron meshes, perform surface extraction on part of the surface meshes, and perform component grouping on the surface meshes according to the different thicknesses of each part;

[0010] S4. Perform surface mesh generation processing on each part divided into surface meshes;

[0011] S5. Perform hexahedron mesh generation processing on each part divided into hexahedron meshes;

[0012] S6. For a symmetric battery pack model, perform mirror processing on the battery pack model after completing the mesh division, complete the mesh of the other half of the battery pack model, and ensure that the meshes on the plane where the symmetry axis is located share the same nodes;

[0013] S7. Substitute the mesh of the overall battery pack model into a certain working condition for solution calculation.

[0014] Further, the structural feature simplification processing includes:

[0015] Remove non-essential structural areas that account for no more than 10% of the overall structure.

[0016] Further, in S2, the finite element software is Hypermesh finite element software, and the observation of whether each part is symmetric for simplification processing includes:

[0017] If the single part is symmetric, cut it along the symmetric central axis, and only retain half of the geometric entity;

[0018] If the single part is asymmetric, retain all the geometric entities.

[0019] Further, the division of the first part and the second part includes:

[0020] Divide the equal-thickness parts of the part model into surface meshes, cut them one by one, and classify the surface mesh parts with different thicknesses into different components;

[0021] Perform hexahedron mesh division on the remaining non-equal-thickness solid parts of the part and cut them one by one.

[0022] Further, the surface mesh generation processing for the part divided into surface meshes includes:

[0023] S41. Use the automesh part in the Hypermesh finite element software to perform mesh division on 2D surface mesh elements;

[0024] S42. Adjust the surface mesh quality according to different requirements for the surface mesh quality using the check elems option in the Hypermesh finite element software.

[0025] Further, perform hexahedral mesh generation processing on the part divided into hexahedral meshes:

[0026] S51. Select one end face of the part model as the reference plane, and project the cross-sectional feature lines other than the end face onto the reference plane;

[0027] S52. Perform surface mesh division on the reference plane: Use the check elements tool to check the surface mesh quality. For the unqualified reference plane mesh, re-divide the elems, smooth it, or use the quality index cleanup tools to optimize the reference plane mesh until it meets the requirements;

[0028] S53. Using different cross-sections and the parts intersecting with the middle plane as boundaries, use the drag command in the 3D panel to drag the surface mesh segment by segment until the hexahedral mesh division is completed;

[0029] S54. Perform co-node processing on the area where the hexahedral mesh and the surface mesh are in contact.

[0030] Further, performing surface mesh division on the reference plane further includes:

[0031] Using the density-adjust edge command in 2D-Automesh, click the left and right mouse keys on the shared edge to change the number of meshes at the shared edge, ensuring that the number of nodes in the part of the surface mesh boundary line intersecting with the reference plane mesh is the same.

[0032] Further, ensuring co-nodes of the mesh on the plane where the axis of symmetry is located includes:

[0033] S61. Use the preview equiv function in the Hypermesh finite element software and set a tolerance of 0.01;

[0034] S62. Click the equivalence command to perform co-node processing on the mesh on the plane where the axis of symmetry is located.

[0035] Compared with the prior art, the beneficial effects of the present invention are:

[0036] 1. By combining 3D modeling software such as Creo with the finite element software Hypermesh, and using a combination of hexahedrons and surface elements in Hypermesh to perform finite element mesh division on complex geometric parts, the number of meshes in the battery pack simulation model is greatly reduced, the mesh quality is better controlled, the computer calculation cost is significantly reduced, and the analysis efficiency of CAE simulation engineers is improved.

[0037] 2. Since hexahedron or surface meshes are used to replace tetrahedron meshes, the overall stability and anti-twist deformation ability of the meshes are increased, thereby reducing the risk of abnormal mesh elements during the solution process. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 - Schematic diagram of existing tetrahedral mesh division;

[0039] Figure 2 - Original geometric model diagram of the battery pack end plate part of the present invention;

[0040] Figure 3 - Model diagram of the battery pack end plate feature after cleaning of the present invention;

[0041] Figure 4 - Model diagram of retaining half of the geometric model of the battery pack end plate of the present invention;

[0042] Figure 5 - Schematic diagram of geometric segmentation;

[0043] Figure 6 - Figure 5 Segmentation effect diagram of

[0044] Figure 7 - Figure 5 Schematic diagram of the grouped model of

[0045] Figure 8 - Figure 5 Schematic diagram of the completed mesh division of

[0046] Figure 9 - Schematic diagram of the segmentation effect of the battery pack end plate of the present invention;

[0047] Figure 10 - Schematic diagram of the surface mesh division on the inner side of the battery pack end plate of the present invention;

[0048] Figure 11 - Partial schematic diagram of the battery pack end plate of the present invention;

[0049] Figure 12 - Figure 11 Schematic diagram after projection;

[0050] Figure 13 - Figure 12Schematic diagram of the completion of the benchmark surface meshing;

[0051] Figure 14 Schematic cross-sectional view of the segmented end plate of the battery pack of the present invention;

[0052] Figure 15 Schematic diagram of the completion of the hexahedron meshing on the outside of the end plate of the battery pack of the present invention;

[0053] Figure 16 Schematic diagram of the mirror-symmetric grid of the end plate of the battery pack of the present invention;

[0054] Figure 17 Schematic diagram of the removal of the middle black solid line of the end plate of the battery pack of the present invention;

[0055] Figure 18 Screenshot of the simulation calculation time of the existing tetrahedron meshing scheme;

[0056] Figure 19 Screenshot of the simulation calculation time of the meshing scheme of the present invention. Detailed implementation manners

[0057] The following further elaborates on the present invention in conjunction with the attached Figures 1 - 19 drawings and specific embodiments.

[0058] The present invention discloses a method for optimizing the CAE mesh of a battery pack, including the following steps:

[0059] S1. Open the battery pack model in 3D modeling software, observe its overall structure, delete the redundant parts that are not involved in the simulation calculation, and perform structural feature simplification processing;

[0060] S2. Complete the model with simplified features in the 3D modeling software in STP format, open the model with simplified features in finite element software, observe whether each part is symmetric and perform simplification processing, then divide the single part after simplification processing into a first part and a second part, the first part includes the parts with equal thickness, the second part includes the parts with unequal thickness, and then perform segmentation on the first part and the second part respectively;

[0061] S3. Divide the first part into surface meshes, divide the second part into hexahedron meshes, perform surface extraction processing on the surface mesh part, and perform component grouping processing on the surface mesh according to the different thicknesses of each part;

[0062] S4. Perform surface mesh generation processing on each part divided into surface meshes;

[0063] S5. Perform hexahedron mesh generation processing on each part divided into hexahedron meshes;

[0064] S6. For a symmetric battery pack model, perform mirror processing on the battery pack model after the mesh generation, complete the mesh of the other half of the battery pack model, and ensure that the meshes on the plane where the axis of symmetry is located share the same nodes.

[0065] S7. Substitute the mesh of the overall battery pack model into a certain working condition for solution calculation.

[0066] In the above step S6, the battery pack model in this embodiment is symmetric left and right, and mirror processing can be directly performed on the battery pack model after the mesh generation. For an asymmetric battery pack model, directly complete the mesh generation of the entire battery pack model.

[0067] In the present invention, by combining 3D modeling software such as Creo with the finite element software Hypermesh, and at the same time using the combination of hexahedron and surface elements in Hypermesh to perform finite element mesh generation on complex geometric parts, the number of meshes of the battery pack simulation model is greatly reduced, the mesh quality is better controlled, the computer calculation cost is largely reduced, and the analysis efficiency of CAE simulation engineers is improved; in addition, because hexahedron or surface meshes replace tetrahedron meshes, the overall stability and anti-twist deformation ability of the meshes are increased, thereby reducing the risk of abnormal mesh elements during the solution process.

[0068] In this embodiment, as Figure 2 and 3 shown, the structural feature simplification process includes:

[0069] Remove the non-essential structural areas that account for no more than 10% of the overall structure. The above non-essential structural areas include structural features such as chamfers, fillets, or steps in unnecessary and unconcerned areas. First, open the 3D drawing of the battery pack end plate with 3D modeling software and perform pre-processing of the part geometric features inside, such as chamfers, fillets, steps, etc. The principle for removing fillets should follow: fillets with a low proportion in the overall structure; unnecessary fillets, such as appearance fillets; fillets in unconcerned areas, etc. Chamfers and steps can be removed in a similar principle.

[0070] In this embodiment, observing whether each part is symmetric for simplification processing includes: if a single part is symmetric, cut it along the axis of symmetry center, and only retain half of the geometric entity; if a single part is asymmetric, retain all of the geometric entity. In this embodiment, as Figures 3 - 4As shown, the part model with features cleaned is exported in STP format in 3D modeling software and imported and opened in Hypermesh finite element software. By observing the part, it can be determined that this part is of left-right symmetric type. In Hypermesh, the solid is cut in half along the middle, temporary nodes in the middle are taken, and then cut along the vertical direction, and the right half of the solid is deleted. Since the end plate of the battery pack of the present invention is a left-right symmetric model, only half of the mesh needs to be divided, and finally the mirror operation is used to complete the overall mesh, saving the workload of mesh division for engineers.

[0071] In this embodiment, the splitting of the first part and the second part includes: dividing the equal-thickness part of the part model into surface meshes and cutting them one by one, and classifying the surface mesh parts with different thicknesses into different components for classification processing; dividing the remaining non-equal-thickness solid parts of the part into hexahedron meshes and cutting them one by one.

[0072] Specifically, as Figures 5 - 9 shown in the schematic diagram of geometric splitting listed, as Figure 5 shown, the upper part A is hemispherical, obviously not suitable for surface extraction for surface geometry, considering hexahedron division. The middle part B has a thickness of 4 mm, which can be cut off from the upper and lower parts for surface mesh processing. The left part C of the lower part is 3 mm, and the right part D is 2 mm, both of which are not of the same thickness as B. Another cut can be made along the lower surface of B to separate the C and D parts. At this time, the thickness of the D part is 2 mm, which is not of the same thickness as B and C, and surface mesh processing is done separately.

[0073] The specific process of this geometric splitting is as follows:

[0074] 1. Enter the Hypermesh software, import the STP model file, enter the Geom-solid edit sub-panel, and use the trim with plane / sur option to split the geometry. Select the geometric entity for solid and the upper and lower surfaces of the middle part B for surfs. Figure 6 is the model after splitting;

[0075] 2. Classify the split geometry into corresponding components according to different thicknesses and types. First, create several new components, and then use Organize to transfer them respectively;

[0076] 3. Divide the hemispherical part A into hexahedrons, and here use the 3D-solid map-one volume option;

[0077] 4. Perform a surface extraction on part B. Here, extract the upper surface and then assign a thickness offset attribute later. Taking the hemisphere as the object, find the outer surface of the face, and then divide the face mesh of part B with an element size of 2;

[0078] 5. Perform a mid-surface extraction on parts C and D. Use the Geom-surface edit-extend command to extend parts C and D until they intersect with B, and then divide the face mesh of parts C and D;

[0079] 6. Finally, use the Tool-faces-preview equiv command to check the non-common node areas of the mesh, and click equivalence to make the nodes common. Figure 8 This is the model with mesh division completed.

[0080] The battery pack model of the present invention is divided using a similar segmentation principle. After segmentation, the effect is as Figure 9 shown. Among them, the outer part of the end plate is divided using hexahedral meshes, and the inner part of the end plate is all divided using face meshes. Of course, some parts of the outer part of the end plate are of equal thickness, but for the sake of saving time, they are also all divided in the form of hexahedral meshes. And the inner part of the end plate is all composed of equal-thickness parts with different thicknesses, so it is all divided in the form of face meshes.

[0081] In this embodiment, after the physical segmentation of the battery pack model is completed, a surface extraction is performed on the middle equal-thickness part. In addition to the conventional operation of extracting the mid-surface, the upper and lower surfaces of the solid can also be directly used as the objects for face mesh division. Generally, the mid-surface extraction can be achieved using the midsurface in the software. Other finite element software also has similar processing methods. After the extraction of the face geometry is completed, a grouping of components is performed on it. The purpose of the grouping is, firstly, to facilitate the mesh division of different parts, and secondly, the classification processing of parts with different thicknesses as mentioned before.

[0082] In this embodiment, as Figure 10 shown, after the grouping is completed, the mesh division needs to start. Among them, the processing of generating the face mesh for the part divided into the face mesh includes:

[0083] S41. Use the automesh part in the Hypermesh finite element software to perform mesh division on the 2D face mesh elements;

[0084] S42. Use the check elems option in the Hypermesh finite element software to adjust the surface mesh quality according to different requirements for the surface mesh quality. The mesh quality can be checked using the check elems option in Tool. Generally, it is preferred that the jacobian is not less than 0.6, and the warpage and aspect are not greater than 5. Of course, there are also requirements such as the minimum mesh size, the maximum and minimum angles of the mesh, etc.

[0085] In this embodiment, as Figures 11 - 15 shown, perform hexahedral mesh generation processing on the part divided into hexahedral meshes:

[0086] S51. Select one end face as the reference plane. As Figure 15 shown, select its upper end face as the reference plane, and project the cross-sectional feature lines other than the end face onto the reference plane;

[0087] S52. Perform surface mesh division on the reference plane: Use the check elements tool to check the surface mesh quality. For the unqualified reference plane mesh, re-divide the elems, smooth it, or use the quality index cleanup tools to optimize the reference plane mesh until it meets the requirements; the number of nodes at the part of the edge where the reference plane mesh intersects with the intersecting surface mesh should be kept consistent to facilitate the co-noding of the solid mesh and the middle surface mesh in the subsequent process;

[0088] S53. Using different cross-sections and the parts intersecting with the middle surface as boundaries ( Figure 14 the dotted line part in), use the drag command in the 3D panel to drag the surface mesh in segments until the hexahedral division is completed. There are a total of 17 segments here; the purpose of segmental dragging is one to try to fit the shape of the solid geometry as much as possible, and the other is to ensure that there are nodes at the part where it intersects with the middle surface mesh, which is convenient for the co-noding of the hexahedral mesh and the surface mesh;

[0089] S54. Perform co-noding processing on the area where the hexahedral mesh and the surface mesh are joined. The surface mesh nodes can be dragged to coincide with the volume mesh nodes, or the nodes with a smaller equivalence distance can be used.

[0090] In this embodiment, performing surface mesh division on the reference plane further includes:

[0091] Using the density-adjust edge command in 2D-Automesh, click the left and right mouse buttons on the shared edge, and the number of meshes at the shared edge can be changed to ensure that the number of nodes at the part of the edge where the reference plane mesh intersects with the intersecting surface mesh is consistent.

[0092] In this embodiment, the co - nodal meshes of the plane where the symmetry axis lies include:

[0093] S61. Use the preview equiv function in the Hypermesh finite element software to set a tolerance of 0.01;

[0094] S62. Click the equivalence command to perform co - nodal processing on the meshes of the plane where the symmetry axis lies.

[0095] In this embodiment, as Figures 16 - 17 shown, perform a mirror processing on the sub - meshed model. Here, the sub - meshed model is the model that has completed the classification of surface meshes and hexahedron meshes. Complete the meshes of the other half of the model. Use the reflect function in the tool to perform mesh mirror processing along the middle section. Figure 16 is the complete mesh model after mirroring. There is a long black solid line in the middle. This is because the meshes on the left and right sides are not co - nodal. Next, use the preview equiv function in the software to set a relatively small tolerance, assume 0.01, and click equivalence to perform co - nodal processing on the model to eliminate the long black solid line.

[0096] In this embodiment, substitute the meshes of the overall battery pack model into a certain working condition for solution calculation. Compare the mesh generation method of the present invention with the existing tetrahedron mesh generation method. The number of meshes in the traditional sub - meshing method is 76164, and the number of meshes in the method described in the patent is 9601. The patent method significantly reduces the number of meshes. Additionally, it should be supplemented that in addition to the end - plate parts of the battery pack housing mentioned in this patent, it can also be applied to other parts suitable for sub - meshing by this method, and the applicability is relatively wide. As Figure 1 、 18 and as shown in 19, adopting the sub - meshing scheme combining hexahedron and surface meshes of the present invention, the calculation time is about 4 hours, while adopting the traditional tetrahedron calculation scheme, the time is about 6 hours. The calculation time cost is reduced by nearly 1 / 3, greatly reducing the running time.

[0097] The above - described invention only represents the implementation manners of the embodiments of the present invention, and thus cannot be construed as limiting the scope of the invention patent, nor does it impose any formal limitations on the structure of the embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the embodiments of the present invention, several changes and improvements can still be made, and these all fall within the protection scope of the embodiments of the present invention.

Claims

1. A method for optimizing the CAE mesh of a battery pack, characterized in that, It includes the following steps: S1. Open the battery pack model in 3D modeling software, observe its overall structure, delete the redundant parts that are not brought into the simulation calculation, and perform structural feature simplification processing; S2. Complete the model with simplified features in the 3D modeling software in STP format, and open the model with simplified features in finite element software. Observe whether each part is symmetric and perform simplification processing. Then divide the single part after simplification processing into a first part and a second part. The first part includes the parts with equal thickness, and the second part includes the parts with unequal thickness. Then perform segmentation on the first part and the second part respectively. Divide the parts with equal thickness of the part model into surface meshes, and cut them one by one. Incorporate the surface mesh parts with different thicknesses into different components for classification processing; Perform hexahedral mesh division on the remaining non-equal-thickness solid parts of the part and cut them one by one; S4. Divide the first part into surface meshes, divide the second part into hexahedral meshes, perform surface extraction on the part of the surface meshes, and perform component grouping processing on the surface meshes according to the different thicknesses of each part; S5. Perform surface mesh generation processing on each part divided into surface meshes; S6. Perform hexahedral mesh generation processing on each part divided into hexahedral meshes; S7. For the symmetric battery pack model, perform mirror processing on the battery pack model after mesh division, complete the mesh of the other half of the battery pack model, and ensure that the meshes on the plane where the axis of symmetry is located share the same nodes; S8. Substitute the mesh of the overall battery pack model into a certain working condition for solution calculation.

2. The method for optimizing the CAE mesh of a battery pack according to claim 1, wherein The structural feature simplification processing includes: Remove the non-essential structural areas that account for no more than 10% of the overall structure.

3. A method for optimizing the CAE mesh of a battery pack according to claim 1, characterized in that, In S2, the finite element software is Hypermesh finite element software. The observation of whether each part is symmetric and the simplification processing include: If a single part is symmetric, cut it along the axis of symmetry, and only retain half of the geometric entity; If a single part is asymmetric, retain all of the geometric entity.

4. A method for optimizing the CAE mesh of a battery pack according to claim 3, characterized in that, The surface mesh generation processing for the part divided into surface meshes includes: S41. Use the automesh part in the Hypermesh finite element software to perform mesh division on the 2D surface mesh elements; S42. Use the check elems option in the Hypermesh finite element software to adjust the surface mesh quality according to different requirements for the surface mesh quality.

5. A method for optimizing the CAE mesh of a battery pack according to claim 3, characterized in that, The hexahedral mesh generation processing for the part divided into hexahedral meshes: S51. Select one end face of the part model as the reference plane, and project the cross-sectional feature lines different from the end face onto the reference plane; S52. Perform surface mesh division on the reference plane: Use the check elements tool to check the surface mesh quality. For the unqualified reference plane mesh, re-mesh the elems, smooth them, or use the quality index cleanup tools to optimize the reference plane mesh until it meets the requirements; S53. Taking different cross-sections and the parts intersecting with the middle plane as boundaries, use the drag command under the 3D panel to segment and drag the surface mesh until the hexahedron meshing is completed; S54. Perform co-node processing on the area where the hexahedron mesh and the surface mesh are in contact.

6. A method for optimizing the CAE mesh of a battery pack according to claim 5, characterized in that, The surface meshing of the reference plane further includes: Using the density-adjust edge command in 2D-Automesh, click the left and right mouse buttons on the shared edge to change the number of meshes at the shared edge, ensuring that the number of nodes in the part of the surface mesh of the reference plane that intersects with the edge of the surface mesh is consistent.

7. A method for optimizing the CAE mesh of a battery pack according to claim 3, characterized in that The co-node processing of the mesh on the plane where the axis of symmetry is located includes: S61. Use the preview equiv function in the Hypermesh finite element software to set a tolerance of 0.01; S62. Click the equivalence command to perform co-node processing on the mesh on the plane where the axis of symmetry is located.

Images