An automatic finite element simulation construction method for an integrated circuit package structure

CN122924760APending Publication Date: 2026-10-16BEIJING UNIV OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611144113.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-30
Publication Date
2026-10-16

AI Technical Summary

Technical Problem

[0004]然而,现有技术在工程应用中仍存在以下不足:人工建模方式的效率与精确性难以控制

Benefits of technology

[0023] (1) The geometric dimensions of the RVE finite element model are automatically determined based on the resolution of the integrated circuit package structure image, so as to reduce the cumbersome process of repeatedly modifying the modeling script under the input of integrated circuit package structure images with different resolutions;

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122924760A_ABST
    Figure CN122924760A_ABST
Patent Text Reader

Abstract

The application discloses an integrated circuit package structure finite element simulation automatic construction method, relates to the technical field of integrated circuit package structure multiscale analysis and finite element automatic modeling, and comprises the following steps: processing and coordinate mapping of an integrated circuit package structure image, obtaining a target material region coordinate file and a path; an automatic modeling script construction and calling method of the integrated circuit package structure image, constructing an automatic modeling script according to the resolution of the integrated circuit package structure image, the set material parameters and the grid density parameters, automatically completing geometric creation, grid division, material assignment and assembly body establishment of an RVE finite element model, and finally outputting a finite element analysis file consistent with the topology of the integrated circuit package structure. The integrated circuit package structure image processing method, the automatic modeling script and the corresponding software platform of the application are deeply fused, and the efficiency, adaptability and automation degree of the integrated circuit package structure homogenization analysis are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of multi-scale analysis and automated finite element modeling of integrated circuit packaging structures, and in particular to a method and software platform for automatically constructing representative volume elements by pixel coordinate mapping for finite element simulation of typical integrated circuit packaging structures. Background Technology

[0002] The high-density interconnects and low linewidth / spacing characteristics of advanced integrated circuit packaging structures pose a significant challenge to the computational cost of finite element analysis of their local microstructures. Finite element analysis for the reliability of integrated circuit packaging structures typically requires establishing a representative volume element (RVE) finite element model to characterize the equivalent mechanical properties of local microstructures. Establishing a homogenized model of the local microstructure to replace the real structure can effectively reduce numerical costs while ensuring the accuracy of finite element analysis.

[0003] In existing technologies, there are two main methods for modeling the local microstructures of integrated circuit packages corresponding to Rectangular Variations (RVEs): manual modeling and script-based modeling. Manual modeling involves researchers creating a geometric model in a finite element analysis platform based on the topological information of the integrated circuit package structure, dividing the geometric regions one by one, assigning material properties, meshing, and building the assembly. Script-based modeling, on the other hand, involves pre-writing corresponding modeling scripts for specific integrated circuit package structures, specific image resolutions, and specific dimensional parameters, thus achieving the semi-automatic creation of the RVE finite element model.

[0004] However, existing technologies still have the following shortcomings in engineering applications: The efficiency and accuracy of manual modeling are difficult to control. This method is highly dependent on the operator's experience and involves numerous geometric processing and material assignment steps. Especially when the topological information of the integrated circuit package structure is complex, differences in manual operation can easily affect the consistency between the corresponding RVE model and the topological structure. The script-based modeling method lacks versatility. Existing script-based modeling methods typically need to correspond to specific package structures, image resolutions, and size parameters. When the resolution or size of the input integrated circuit package structure image changes, the script content needs repeated adjustments. Furthermore, there are differences in the origin and direction between the pixel coordinate system of the integrated circuit package structure image and the coordinate system of the finite element model. An effective coordinate transformation and pixel mapping mechanism is needed to accurately map the target material region in the integrated circuit package structure to the target material region in the RVE finite element model, thereby ensuring the accuracy of target / non-target material region identification and attribute assignment. In addition, the existing processes for preprocessing integrated circuit package structure images, identifying target material regions, and generating modeling scripts are usually disconnected from the calling process of the finite element analysis platform. An automatic connection from integrated circuit package structure images to their corresponding RVE finite element model output has not yet been formed, resulting in low efficiency, poor adaptability, and high barriers to entry for its engineering applications.

[0005] Therefore, how to provide a finite element modeling method that ensures the consistency of the generated RVE finite element model with the topological structure of the integrated circuit package structure image is a problem that urgently needs to be solved by those skilled in the art. This would improve the adaptability of existing technologies to different integrated circuit package structure image inputs, enhance the automation of target material region pixel coordinate extraction, coordinate transformation, and modeling script generation, and strengthen the continuity between the integrated circuit package structure image and its corresponding RVE finite element model. Summary of the Invention

[0006] In existing technologies, RVE modeling based on integrated circuit packaging structures often suffers from problems such as low manual efficiency, difficulty in adapting the image resolution of integrated circuit packaging structures to the size of the RVE finite element model, ambiguity in the mapping mechanism between the pixels of the target material region and the attribute assignment of the target material region, fragmentation of the process from image processing of integrated circuit packaging structures to preprocessing and modeling of RVE finite element models, and high barriers to entry for engineering applications.

[0007] To address the aforementioned technical problems, this invention provides a method and software platform for automatically constructing representative volume elements based on pixel coordinate mapping in finite element simulation of typical integrated circuit packaging structures, comprising:

[0008] A preprocessing and coordinate mapping method for integrated circuit package structure images includes grayscale conversion and threshold segmentation of integrated circuit package structure images, extraction of pixel coordinate sets of target material regions, coordinate transformation of pixel coordinates of integrated circuit package structure images to coordinate system of finite element model, and generation of pixel coordinate files of target material regions.

[0009] The method for constructing and calling an automatic modeling script for an integrated circuit package structure image is used to write an automatic modeling script and output a finite element analysis file consistent with the topology of the integrated circuit package structure based on the resolution of the integrated circuit package structure image and the input information of the graphical user interface of the software platform.

[0010] Specifically, the above methods are connected using a graphical user interface and implemented as a whole through the following steps:

[0011] Step 1: Collect the core information of the above method through the graphical user interface, including: the path of the finite element analysis platform call file, the storage path of the RVE finite element model, the image path of the integrated circuit packaging structure, two types of material parameters and mesh density parameters; the material parameters include at least the elastic modulus, Poisson's ratio and coefficient of thermal expansion.

[0012] Step 2: Based on the imported integrated circuit package structure image, determine the length, width, and mesh seed of the corresponding RVE finite element model according to its resolution.

[0013] Step 3: Perform grayscale conversion and threshold segmentation on the imported integrated circuit packaging structure image, identify the pixel coordinates corresponding to the target material region and the non-target material region, and establish a set; wherein, the target material region is preferably the metal wiring region in the packaging structure, and the non-target material region is preferably the dielectric material region or other substrate material region.

[0014] Step 4: Based on the correspondence between the pixel coordinate system of the integrated circuit package structure image and the coordinate system of the finite element model, perform coordinate transformation on the established pixel coordinate set to generate a pixel coordinate file of the target material region and record the path; wherein, since the origin of the pixel coordinate of the integrated circuit package structure image is located at the upper left corner of the image, while the origin of the coordinate in the RVE finite element model plane is usually located at the lower left corner of the plane, the coordinate transformation includes the flip correction of the vertical coordinate to at least obtain the pixel coordinate set of the target material region.

[0015] Step 5: After completing the preprocessing, coordinate transformation and pixel mapping process of the integrated circuit package structure image, receive instructions through the graphical user interface to trigger the construction and calling process of the automatic modeling script for the integrated circuit package structure image.

[0016] Step 6: Write the collected two types of material parameters, mesh density parameters, the determined length, width, and mesh seed of the RVE finite element model, and the recorded target material region pixel coordinate file path into the preset modeling script template, and generate an automatic modeling script for the integrated circuit package structure in the specified storage path of the RVE finite element model.

[0017] Step 7: Call the finite element analysis platform to execute the automatic modeling script in a way that does not use a graphical user interface.

[0018] Step 8: Invoke the automatic modeling script to automatically complete the geometry creation and mesh generation of the RVE model; wherein, the mesh generation is based on the determined mesh generation seed and the collected mesh density parameters to generate a regularly distributed finite element mesh.

[0019] Step 9: Based on the target material region coordinate file, determine the region corresponding to the target material region in the RVE finite element model with the finite element mesh already divided, and assign the target material parameters to the region, while assigning non-target material parameters to the remaining regions.

[0020] Step 10: After completing the material assignment, create the assembly of the RVE finite element model, and output the RVE finite element model and its finite element analysis file that are consistent with the topology of the integrated circuit packaging structure in the collected model saving path, so as to realize the connection from the integrated circuit packaging structure to its finite element computable model.

[0021] Preferably, the threshold segmentation process in step 3 uses a preset grayscale threshold of 127 to distinguish the integrated circuit packaging structure image into target material regions and non-target material regions.

[0022] Compared with the prior art, the present invention has at least the following beneficial effects:

[0023] (1) The geometric dimensions of the RVE finite element model are automatically determined based on the resolution of the integrated circuit package structure image, so as to reduce the cumbersome process of repeatedly modifying the modeling script under the input of integrated circuit package structure images with different resolutions;

[0024] (2) By automatically mapping the pixel coordinates of the target material region in the integrated circuit packaging structure image to the coordinates of the target material region in the finite element model, the time cost of relying on manual establishment of its RVE finite element model is effectively reduced, and the degree of automation and accuracy of target material region identification and assignment is improved.

[0025] (3) By calling the automatic modeling script and the finite element analysis platform without a graphical user interface, the automatic connection from the input of the integrated circuit package structure image to the output of the RVE finite element model is realized, which significantly improves the modeling efficiency while reducing the threshold of use.

[0026] (4) By setting the mesh density parameter, the mesh density of the RVE finite element model can be adjusted, so that the modeling efficiency and simulation accuracy can be flexibly adjusted. Attached Figure Description

[0027] To more clearly illustrate the technical solutions and specific embodiments of the present invention, the accompanying drawings are briefly described below. Obviously, the drawings described below are only a part of the embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings without creative effort.

[0028] Figure 1 This is a flowchart illustrating the overall process of the method of the present invention.

[0029] Figure 2 This is a schematic diagram of a graphical user interface for implementing the method of the present invention;

[0030] Figure 3 This is a schematic diagram illustrating the data flow between the integrated circuit packaging structure image, the automatic modeling script, and the finite element analysis platform in the method of this invention.

[0031] Figure 4 This is a schematic diagram of the automatic modeling results in Example 1, wherein, Figure 4 (a) is an image of an integrated circuit package structure with a resolution of 50×50. Figure 4 (b) is the automatically constructed RVE finite element model;

[0032] Figure 5 This is a schematic diagram of parameter input and graphical user interface in Example 1;

[0033] Figure 6 This is a schematic diagram of the automatic modeling results in Example 2, wherein, Figure 6 (a) is an image of an integrated circuit package structure with a resolution of 200×200. Figure 6 (b) is the automatically constructed RVE finite element model;

[0034] Figure 7 This is a schematic diagram of parameter input and graphical user interface in Example 2;

[0035] Figure 8 This is a schematic diagram of the modeling results under different mesh density parameters in Example 3, where, Figure 8 (a) is an image of an integrated circuit package structure with a resolution of 100×100. Figure 8 (b) to Figure 8 (d) are the RVE finite element models automatically constructed when the mesh density parameters are 1, 2, and 4, respectively.

[0036] in, Figure 2 In the middle: 1 is the menu button, 2 is the finite element analysis platform file call setting button, 3 is the RVE finite element model save path setting button, 4 is the view instruction document button, 5 is the select integrated circuit package structure image button, 6 is the integrated circuit package structure image path display area, 7 is the flowchart of the automatic modeling method, 8 is the target material region attribute input area, 9 is the non-target material region attribute input area, 10 is the integrated circuit package structure image display area, 11 is the mesh density control area, 12 is the automatic modeling method running status display area, and 13 is the run automatic modeling method button. Detailed Implementation

[0037] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the following examples are for illustrative purposes only and are not intended to limit the scope of protection of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0038] To make the above-mentioned objectives, features and advantages of the present invention more readily understood, the present invention will be further described in detail with reference to the accompanying drawings and specific embodiments.

[0039] like Figures 1 to 3 As shown, this invention provides a method and software platform for automatically constructing representative volume elements by pixel coordinate mapping for finite element simulation of typical integrated circuit package structures. This method acquires the finite element analysis platform's call file path, RVE finite element model storage path, integrated circuit package structure image path, material parameters, and mesh density parameters through a graphical user interface. It then achieves automatic construction and storage of the integrated circuit package structure into its corresponding RVE finite element model through an automatic modeling script and a non-graphical user interface call method to the finite element analysis platform.

[0040] In practice, users through Figure 2 In the graphical user interface shown, button 2 acquires the file path of the finite element analysis platform; button 3 acquires the storage path of the RVE finite element model; button 5 acquires the image of the integrated circuit package structure; material parameters are acquired in the target material region attribute input area 8 and the non-target material region attribute input area 9; mesh density parameters are acquired through the mesh density control area 11; finally, button 13 triggers the entire process. The automatic modeling method running status display area 12 displays the current operation status and method execution results.

[0041] Specifically, after receiving data acquired from the graphical user interface, the selected integrated circuit package structure image is read and subjected to grayscale conversion and threshold segmentation to obtain a binary image for characterizing the target material region. In one embodiment, the target material region is preferably a metal wiring region in the integrated circuit package structure, and the non-target material region is preferably a dielectric material region or other substrate material region. The pixel coordinates corresponding to the target material region are extracted from the binary image and formed into a set.

[0042] This invention further performs coordinate transformation on the pixel coordinate set of the target material region, including vertical inversion correction of the pixel coordinates in the integrated circuit package structure image, so that the pixel coordinates of the target material region in the integrated circuit package structure image correspond one-to-one with the coordinates of the target material region in the RVE finite element model. The transformed pixel coordinate set is written into a target material region coordinate file and the path is recorded, preferably a .csv file, to reflect the topological information of the integrated circuit package structure.

[0043] After generating the target material region coordinate file, the dimensional parameters of the RVE finite element model are determined based on the resolution of the integrated circuit package structure image and preset geometric parameters. Simultaneously, these parameters, along with the material parameters, mesh density parameters, and the target material region coordinate file path, are written into a preset modeling script template. After configuring the automatic modeling script corresponding to the input data, the finite element analysis platform calls the file and executes the automatic modeling script in a graphical user interface-free manner. This automatically completes the geometric creation, mesh generation, material creation, assignment of target and non-target material regions, assembly establishment, and finite element model saving processes for the RVE finite element model.

[0044] In summary, this invention establishes an automated workflow for inputting integrated circuit package structure images into RVE finite element models. This method can quickly, accurately, and automatically construct the topology of metal wiring in integrated circuit package structures within a finite element analysis platform, improving modeling efficiency and ensuring the accuracy of the RVE finite element model.

[0045] Example 1

[0046] like Figure 4 and Figure 5 As shown, this example selects an integrated circuit package structure image with a resolution of 50×50 as input to verify the basic automatic modeling function of the method of the present invention.

[0047] Specifically, users in Figure 5 In the parameter input and graphical user interface shown, after setting the finite element analysis platform call file path and RVE finite element model storage path in sequence, import as follows: Figure 4(a) shows an integrated circuit package structure image with a resolution of 50×50; input material parameters in the target material region attribute input area 8 and the non-target material region attribute input area 9, and set the mesh density parameter to 1 in the mesh density control area 11. Finally, click button 13 to trigger the automatic construction process of the RVE finite element model for this integrated circuit package structure image.

[0048] The automatic modeling method first performs grayscale conversion and threshold segmentation on a 50×50 resolution image of an integrated circuit package structure, identifies the target material region, and extracts its pixel coordinate set. Then, based on the correspondence between the pixel coordinate system of the integrated circuit package structure image and the coordinate system of the finite element model, it performs coordinate transformation on the pixel coordinate set and generates a target material region coordinate file. Finally, the collected parameters are written into a preset modeling script template to generate a modeling script for... Figure 4 (a) After the automatic modeling script of the integrated circuit package structure image is obtained, the finite element analysis platform is called in a user-free form to complete the finite element preprocessing process and obtain its RVE finite element model.

[0049] like Figure 4 As shown in (b), the topology of the RVE finite element model obtained by the automatic modeling method is consistent with the topology of the integrated circuit package structure image input in this embodiment, indicating that the present invention can realize the automatic construction process from the integrated circuit package structure image to the RVE finite element model without the need for manual writing or modification of the modeling script.

[0050] Example 2

[0051] like Figure 6 and Figure 7 As shown, this example selects an integrated circuit package structure image with a resolution of 200×200 as input to illustrate that when inputting integrated circuit package structure images with different resolutions, the present invention can still achieve automatic modeling function without the need for manual modification of the preset modeling logic and script.

[0052] Specifically, users through Figure 7 The graphical user interface shown imports an integrated circuit package structure image with a resolution of 200×200 and sets relevant parameters as needed. Compared with Example 1, this example only differs in the resolution of the input integrated circuit package structure image; the rest of the operation process, modeling script logic, and finite element analysis platform invocation method remain unchanged. After importing integrated circuit package structure images of different resolutions, the process of grayscale conversion, binarization, pixel coordinate extraction and conversion, generation of target material region coordinate files, automatic modeling script generation, and finite element analysis platform invocation is automatically completed, generating images as shown in the figure. Figure 6 (b) shows the RVE finite element model.

[0053] This embodiment demonstrates that when the resolution of the input integrated circuit package structure image changes, the present invention does not require manual modification of the main content of the preset modeling script; after configuring the corresponding parameters, it can still automatically generate the corresponding RVE finite element model, verifying the adaptability and reusability of the present invention when dealing with integrated circuit package structure images of different resolutions.

[0054] Example 3

[0055] like Figure 8 As shown, in this embodiment, an integrated circuit package structure image with a resolution of 100×100 is selected as input. While keeping the other configurations unchanged, the corresponding RVE finite element models are generated by comparing the mesh density parameters 1, 2, and 4 respectively, to illustrate that the model established by the method of the present invention has the ability to adjust the finite element mesh density.

[0056] Specifically, users import an integrated circuit package structure image with a resolution of 100×100 through the graphical user interface, and then set the file path for the finite element analysis platform, the storage path for the RVE finite element model, and the material parameters in sequence. Under the conditions of mesh density parameters of 1, 2, and 4, RVE finite element models with different mesh densities are generated for the same integrated circuit package structure image.

[0057] like Figure 8 (b)-(d) show the RVE finite element models generated under different mesh density parameters. Under the premise of keeping the target material region assignment of the RVE finite element model consistent with the topology of the integrated circuit packaging structure image in this embodiment, the mesh density of the corresponding RVE finite element model gradually increases as the mesh density parameter increases.

[0058] This embodiment further illustrates that the present invention not only has the function of automatically constructing the corresponding RVE finite element model based on the integrated circuit packaging structure image, but also can flexibly control the mesh density of the RVE finite element model by adjusting the mesh density parameter, thereby meeting different simulation analysis needs.

Claims

1. An automatic construction method for finite element simulation of integrated circuit packaging structures, characterized in that, include: The user interface in the diagram provides the file path for the finite element analysis platform, the storage path for the RVE finite element model, the image path for the integrated circuit packaging structure, and the parameters for two types of materials and the mesh density. The image of the integrated circuit package structure is grayscaled and thresholded to extract the set of pixel coordinates of the target material region. Based on the correspondence between the pixel coordinates of the integrated circuit package structure image and the coordinate system of the finite element model, coordinate transformation is performed to obtain the coordinate file and path of the target material region for reading in the finite element analysis platform. An automatic modeling script is constructed based on the resolution of the integrated circuit package structure image, the set material parameters, and the mesh density parameters. The script is then executed by calling the graphical interface of the finite element analysis platform. The script automatically completes the geometric creation, mesh generation, material assignment, and assembly establishment of the RVE finite element model, and finally outputs a finite element analysis file that is consistent with the topology of the integrated circuit package structure.

2. The automatic construction method for finite element simulation of integrated circuit packaging structure according to claim 1, characterized in that, The information obtained through the graphical user interface includes: The finite element analysis platform calls the file path, the storage path of the RVE finite element model, the image path of the integrated circuit package structure, the material parameters of the target material region, the material parameters of the non-target material region, and the mesh density parameters.

3. The automatic construction method for finite element simulation of integrated circuit packaging structure according to claim 1, characterized in that, The grayscale conversion and threshold segmentation of the integrated circuit packaging structure image includes: The imported integrated circuit package structure image is converted into a grayscale image, and a preset grayscale threshold is used to binarize the grayscale image of the integrated circuit package structure image to distinguish the target material area from the non-target material area by grayscale value.

4. The automatic construction method for finite element simulation of integrated circuit packaging structure according to claim 1, characterized in that, The coordinate transformation based on the correspondence between the pixel coordinates of the integrated circuit package structure image and the coordinate system of the finite element model includes: The pixel coordinates in the image, with the top left corner as the reference, are converted into coordinates in the RVE finite element model plane with the bottom left corner as the reference, and only the vertical coordinates of the pixel coordinates in the integrated circuit package structure image are flipped and corrected.

5. The automatic construction method for finite element simulation of integrated circuit packaging structure according to claim 1, characterized in that, The automatic modeling script includes: Write the resolution of the integrated circuit packaging structure image, the two types of material parameters, the mesh density parameters, and the coordinate path of the target material region into a preset modeling script and save it to the RVE finite element model storage path.

6. The automatic construction method for finite element simulation of integrated circuit packaging structure according to claim 1, characterized in that, The material assignment includes: The finite element analysis platform is invoked to read the automatic modeling script and the target material region coordinate file. Based on the mapping relationship between the pixel coordinates in the target material region coordinate file and the mesh coordinates in the RVE finite element model plane, the finite element mesh corresponding to the target material region in the integrated circuit package structure image in the RVE finite element model is identified, and the material parameters of the target material region are assigned to the finite element mesh of that region. The material parameters of the non-target material region are assigned to the finite element mesh of the remaining regions.

7. The automatic construction method for finite element simulation of integrated circuit packaging structure according to claim 1, characterized in that, The length, width, and mesh seed of the RVE finite element model are determined based on the resolution of the integrated circuit package structure image.

8. The automatic construction method for finite element simulation of integrated circuit packaging structure according to claim 1, characterized in that, The density of the RVE finite element mesh is controlled according to the mesh division seed and mesh density parameters to generate an adjustable density finite element mesh.

9. The automatic construction method for finite element simulation of integrated circuit packaging structure according to claim 5, characterized in that, The preset modeling script includes: Based on the parametric programming language of the finite element analysis platform preprocessing stage, it is used to realize the automatic process of geometry creation, material creation, mesh generation, material assignment, assembly creation, and finite element analysis file saving of RVE finite element model.