Gull wing type pin welding spot shape simulation repair finite element modeling method

Abstract

The invention discloses a finite element modeling method for simulating and repairing the shape of a welding spot of a gull-wing-shaped pin, belongs to the field of finite element analysis of reliability of electronic packaging, and solves the problem that node information is lost in the materialization process of a surface evolution result file. According to the method, a solid pin area is firstly constructed and constrained, liquid domain parameters are set to evolve the shape of the welding spot, a file is output and materialized, a model is assembled and imported into software to be analyzed and solved, and finally the service life of the welding spot is predicted through a service life model. According to the method, the surface evolution result materialization problem is solved, the precision of the three-dimensional model is improved, the action description of infinitesimal elements is finer, the finite element calculation precision is improved, then the welding spot service life prediction accuracy is improved, and the reliability of key welding spots is enhanced.

Description

Technical Field

[0001] This invention relates to the field of finite element analysis of electronic packaging reliability, and in particular to a finite element modeling method for simulating and repairing the morphology of gull-wing pin solder joints. Background Technology

[0002] The reliability of electronic packaging devices is a key factor determining the stable operation of electronic systems. When the system faces different physical field environments such as thermal shock, thermal cycling, electrical loading, and vibration load, the reliability of the device solder joints has a significant impact on the reliability of electronic packaging devices. Experimentally setting the system in the actual physical environment has problems such as long experimental cycles, high costs, and inability to directly output parameters such as stress and energy. Therefore, finite element analysis has been introduced into the reliability analysis of electronic packaging devices. Mathematical equations are established and solved for the physical model. Based on the discretization method of mathematical processing, the system boundary conditions are applied to the discretized interrelated nodes and elements. Parallel computing is used to complete the solution, and the numerical property results of the system after applying boundary conditions are obtained. These results can be used in structural mechanics, life prediction and other analysis work.

[0003] In existing finite element analysis studies in the field of electronic packaging, the construction of complex solder joint morphologies often employs simplification, thus failing to accurately represent the morphological characteristics of the solder joints. The numerical calculation results obtained after numerical discretization of this model cannot precisely correspond to the actual solder joint structure of the device, resulting in systematic errors. Existing morphological simulation methods apply specific constraints to the liquid, defining parameters such as surface contact energy, free liquid surface tension, and gravity. The liquid surface shape evolves along the direction of lowest energy, ultimately calculating the surface morphology of the solid solder joint formed after welding. While the fitted potential function at the point of lowest energy is relatively accurate during the calculation process, calculation errors exist at the node positions of other curved surfaces. These errors lead to errors in the model solidification process, requiring correction. Currently, some researchers have used surface evolution methods to calculate the surface morphology of electronic packaging solder joints and applied them to the finite element analysis of some types of devices.

[0004] Currently, research on using surface evolution calculations to determine liquid surface morphology and incorporating this into finite element analysis generally follows two approaches: (1) Use the graphical module of the source program to output typical point parameters, and use these point values ​​to construct a three-dimensional model in the three-dimensional software; (2) Use third-party software to directly output geometric files to form composite surface element files that can be recognized by finite element software, and then solidify them in 3D modeling software.

[0005] Of the two approaches mentioned above, directly outputting feature point coordinates can directly form a solid model, but this process is simplified and suffers from surface shape distortion. Furthermore, complex surface source files output by third-party software cannot be solidified due to unclosed cross-sections, and surface geometry files cannot be directly used in finite element calculations. To address these issues, this invention proposes a surface source repair method, which solves the problem of unclosed surfaces being unable to be solidified for subsequent discretized finite element solutions. This method is particularly effective in simulating the morphology of gull-wing pin solder joints. The repaired solder joints have a higher degree of conformity to the actual structure's surface, improving the accuracy of finite element simulation results and consequently enhancing the accuracy of finite element predictions using lifetime models. Summary of the Invention

[0006] This invention proposes a finite element modeling method for morphological simulation and repair of gull-wing type pin solder joints. By using a surface repair method, it solves the problem of lost node information during the materialization process of surface evolution result files, improves the accuracy of the three-dimensional model in representing the actual solder joint structure, and enhances the accuracy of finite element simulation lifetime prediction.

[0007] A finite element modeling method for simulating and repairing the morphology of gull-wing type pin solder joints, the method comprising the following steps: S1. Construct a solid pin region in morphological simulation based on the parameters of the actual gull-wing pin device, and apply a fixed constraint "fixed" to all points, lines and surfaces in the solid domain; S2. Based on the structural matching parameters and filling process parameters, the spatial node position of the liquid domain is set as the initial position of the evolution, and the surface energy field control parameters of the liquid solder are set. After refinement and iteration, the shape of the liquid surface evolves along the direction of the lowest energy to obtain the surface morphology of the approximate gull-wing type pin solder joint. S3, Output the profile node geometry file; S4. Solidify the gull-wing pin profile, delete overlapping surfaces and lines, and generate curves attached to the outer surface of the pins through projection relationships. Construct a closed surface space through the curves, reconstruct the solder geometry based on the closed surface space, and solidify the reconstructed geometry into a pin and solder structure model. S5. Based on the solidified pin and solder structure model, assemble the three-dimensional model of the electronic device component after soldering, and import the three-dimensional model of the electronic device component into the finite element analysis software. In the finite element analysis software, perform material property binding, mesh discretization processing, boundary condition loading and multi-dimensional physical field load application operations in sequence, submit to the solver for solution, and obtain the structural mechanical response data required for life prediction through the post-processing module. The structural mechanical response data includes stress and strain amplitude. S6. Predict the service life of the weld joint using a life prediction model. Substitute the structural mechanical response data obtained from finite element calculations into the life model and use statistical methods to predict the expected life of each part of the weld joint.

[0008] Furthermore, in S2, the structure matching parameters include pad size and pin offset, and the filling process parameters include pin offset.

[0009] Furthermore, in S2, the control parameters of the liquid solder surface energy field include interface energy constraint, free liquid surface tension coefficient, and gravitational acceleration.

[0010] Furthermore, in S2, the initial geometry of the liquid domain is defined based on the pad length, pin offset distance, and solder volume parameters of the gull-wing type pin device. The interface energy constraint includes the contact angle between the pad and the solder, and the contact angle constraint between the pin outer surface and the solder. The liquid physical parameters include a surface tension coefficient of 0.45 kg / s. 2 Liquid density: 0.00846 kg / cm³ 3 And gravitational acceleration 980 cm / s 2 .

[0011] Furthermore, in S3, the file format of the profile node geometry file is a geometry file format that can be recognized by finite element analysis software.

[0012] Furthermore, in S4, the solidification repair process includes: identifying unclosed surface elements overlapping the solder portion and the pin; deleting the overlapping surfaces and lines; using the outer surface of the pin as a reference, projecting the undeleted surface element nodes onto the pin surface to generate a continuous closed curve; and reconstructing the solder surface based on this curve to form a closed solid model. The criterion for determining the overlapping surface elements is that the model gap is less than the preset tolerance value of 1×10⁻⁶ in the finite element software. -4 mm.

[0013] Furthermore, in S5, the mesh discretization process uses a patch conformal method to locally refine the fine structure of the solder joints, while the PCB substrate is divided into multiple regions to reduce the mesh density; the random vibration load spectrum is applied according to the QJ3086A-2016 standard, in which a 20-2000Hz power spectral density spectrum is applied perpendicular to the PCB direction, and the total root mean square acceleration is 28.5 grms.

[0014] Furthermore, in S6, the life prediction model adopts Miner's linear cumulative damage theory, including: extracting the 1σ, 2σ, and 3σ strain amplitudes at the weld joint and their corresponding cycle numbers ni from the finite element results; calculating the damage degree Di under each strain amplitude by combining the material's tensile strength and elastic modulus; and the total damage degree...

[0015] The result was obtained by weighted calculation of the three interval statistical distributions of 1σ strain amplitude (0.6831), 2σ strain amplitude (0.271), and 3σ strain amplitude (0.0433). When D ≥ 1, the weld joint was determined to be in failure.

[0016] A storage medium storing a computer program, which, when executed by a processor, implements the aforementioned finite element modeling method for simulating and repairing the morphology of gull-wing pin solder joints.

[0017] A computer device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-described finite element modeling method for simulating and repairing the morphology of gull-wing pin solder joints.

[0018] The beneficial effects of this invention are as follows: This invention provides a finite element modeling method for morphological simulation and repair of gull-wing type lead solder joints. It proposes a closed surface repair method, allowing the evolved composite surface element structure to be materialized and successfully applied to finite element calculations and lifetime prediction processes. This method can more accurately represent the three-dimensional structure of the device space, making the description of the interactions between micro-elements more refined. By transforming the complex solder surface into a three-dimensional model through morphological repair, the calculation accuracy of structural thermal and mechanical data is significantly improved. It is evident from the solder joint structure models before and after processing that the solder joint structure after morphological simulation is highly similar to the geometry of the actual solder joint after soldering, demonstrating good modeling accuracy. This significantly improves the accuracy of finite element calculations and solder joint lifetime prediction, effectively enhancing the reliability of critical solder joints. Attached Figure Description

[0019] Figure 1 It is an unrepaired gull-wing pin structure; Figure 2 The structure of the solid weld joint after processing; Figure 3 Simulation verification of gull-wing pin configuration; Figure 4 For assembling the physical model of the device; Figure 5 This refers to the grid division method. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Reference Figure 1-Figure 5 As shown, a finite element modeling method for simulating and repairing the morphology of gull-wing type pin solder joints is presented. The method includes the following steps: S1. Construct a solid pin region in morphological simulation based on the parameters of the actual gull-wing pin device, and apply a fixed constraint "fixed" to all points, lines and surfaces in the solid domain; S2. Based on the structural matching parameters and filling process parameters, the spatial node position of the liquid domain is set as the initial position of the evolution, and the surface energy field control parameters of the liquid solder are set. After refinement and iteration, the shape of the liquid surface evolves along the direction of the lowest energy to obtain the surface morphology of the approximate gull-wing type pin solder joint. S3, Output the profile node geometry file; S4. Solidify the gull-wing pin profile, delete overlapping surfaces and lines, and generate curves attached to the outer surface of the pins through projection relationships. Construct a closed surface space through the curves, reconstruct the solder geometry based on the closed surface space, and solidify the reconstructed geometry into a pin and solder structure model. S5. Based on the solidified pin and solder structure model, assemble the three-dimensional model of the electronic device component after soldering, and import the three-dimensional model of the electronic device component into the finite element analysis software. In the finite element analysis software, perform material property binding, mesh discretization processing, boundary condition loading and multi-dimensional physical field load application operations in sequence, submit to the solver for solution, and obtain the structural mechanical response data required for life prediction through the post-processing module. The structural mechanical response data includes stress and strain amplitude. S6. Predict the service life of the weld joint using a life prediction model. Substitute the structural mechanical response data obtained from finite element calculations into the life model and use statistical methods to predict the expected life of each part of the weld joint.

[0022] Specifically, this invention significantly improves the accuracy and practicality of solder joint morphology simulation and reliability analysis through innovative multi-step synergistic effects. Specifically, during the evolution of the liquid surface morphology, it employs coupled calculations based on pad length, pin offset, and solder volume parameters, combined with interface energy constraints and the liquid surface tension coefficient of 0.45 kg / s. 2 Liquid density: 0.00846 kg / cm³ 3 and gravitational acceleration 980 cm / s 2The combined effect of these factors ensures that the simulated liquid solder, evolving in the direction of lowest energy, achieves optimal consistency between the generated surface morphology and the actual solder joint structure. More importantly, through geometric repair operations on non-closed local surfaces, closed boundary curves are generated using projection mapping rules to reconstruct the solid model, successfully resolving the gap problem caused by node redundancy or missing nodes in traditional surface evolution results. This significantly reduces geometric description errors during finite element modeling. This high-fidelity 3D model, combined with the patch conformal method for local mesh refinement in finite element analysis and the loading of a random vibration spectrum with a total root mean square acceleration of 28.5 grms under the QJ3086A-2016 standard, generates stress-strain amplitude data with confidence levels improved to the 3σ confidence interval. Finally, the 3D statistically weighted model based on Miner's linear damage theory (1σ accounting for 68.31%, 2σ for 27.1%, and 3σ for 4.33%) successfully achieves accurate determination of the solder joint lifetime threshold (damage degree D≥1). Furthermore, in S2, the structure matching parameters include pad size and pin offset, and the filling process parameters include pin offset.

[0023] Specifically, firstly, the technical solution of this embodiment explicitly limits the surface tension coefficient of the liquid solder to 0.45 kg / s. 2 Liquid density: 0.00846 kg / cm³ 3 and gravitational acceleration 980 cm / s 2 The precise quantification of these parameters originates from experimental measurements of the actual solder physical properties, accurately reflecting the surface tension and hydrodynamic characteristics of liquid metal in microscale space. This ensures that the anisotropic properties of the free surface are accurately modeled during the energy minimization evolution process. Secondly, through precise matching of structural parameters such as pad length, pin offset distance, and solder volume, the geometric features of the initialized liquid domain more closely resemble the liquid-solid interface contact morphology under real-world conditions. Combined with active constraints on the contact angle parameters between the pad and solder, and the contact angle between the pin's outer surface and the solder, abnormal fluctuations in surface node deviation from the minimum energy curve are effectively suppressed. The resulting solder joint surface morphology not only fully preserves the characteristic gradient distribution of the wetting critical region but also avoids the contour distortion problem caused by the overall destruction of the geometric morphology in traditional methods, fundamentally solving the workload of subsequent surface element repair. This multi-layer parameter collaborative control strategy in this embodiment enables the technical solution defined in this claim to both meet the theoretical expectations of metallurgical processes for the solid-liquid interface and meet the strict closure requirements of finite element analysis for the solid model, providing a higher-precision benchmark model for solder joint fatigue life prediction.

[0024] Furthermore, in S2, the control parameters of the liquid solder surface energy field include interface energy constraint, free liquid surface tension coefficient, and gravitational acceleration.

[0025] Specifically, this embodiment establishes control parameters for the surface energy field of the liquid solder and specific contact angle constraints, enabling the physical field boundary conditions during liquid evolution to accurately express the actual wetting behavior and energy balance state of the liquid solder. This is achieved by defining the contact angle between the pad and the solder as 20 degrees, the contact angle of the lead sidewall as 15 degrees, the bottom contact angle as 15 degrees, and the surface tension coefficient as 0.45 kg / s. 2 Gravitational acceleration 980 cm / s² 2 By employing fixed numerical parameters, this invention eliminates the systematic deviation between contact angle settings and actual interface behavior in traditional simulations, and achieves a dynamic balance between the wetting potential energy of the solid surface and the gravity of the liquid solder through parameter coupling. Using this parameter system significantly reduces the cumulative error of node positions in the surface source file, thus forming a foundation for optimizing the quality of closed surfaces. In particular, the refined setting of the pad contact angle matches the actual contact angle data after Cu-OSP surface treatment. This technical detail greatly improves the spatial topological fit between the evolution boundary of the free surface and the actual solder spreading path, suppressing abnormal surface curvature defects caused by over-wetting or under-wetting from an energy balance perspective. Combined with the topology verification results after outputting the .stl format file through SE-FIT, the proportion of smooth transition areas on the solder surface optimized by this parameter system is greatly increased. This improvement in physical accuracy directly reduces redundant operations in surface repair in S4, thereby improving the mesh generation yield of the solidified 3D model.

[0026] Furthermore, in S2, the initial geometry of the liquid domain is defined based on the pad length, pin offset distance, and solder volume parameters of the gull-wing type pin device. The interface energy constraint includes the contact angle between the pad and the solder, and the contact angle constraint between the pin outer surface and the solder. The liquid physical parameters include a surface tension coefficient of 0.45 kg / s. 2 Liquid density: 0.00846 kg / cm³ 2 And gravitational acceleration 980 cm / s 2 .

[0027] Specifically, this embodiment explicitly defines the initial shape of the liquid domain based on the pad length, pin offset distance, and solder volume, while also incorporating specific interface contact angle constraints and precise liquid physical parameters, such as a surface tension coefficient of 0.45 kg / s. 2 Density 0.00846 kg / cm³ 3 Gravitational acceleration 980 cm / s² 2This creates a strong constraint system for the evolution of liquid solder morphology. The technical features of this embodiment precisely construct physical interface interactions through quantitative parameters, making the surface evolution process of liquid solder under the minimum energy criterion more closely resemble the fluid dynamics characteristics in actual welding processes. The constraints on the contact angle between the pad and the solder, and the contact angle of the pin surface, simulate real wetting behavior. Combined with the composite boundary effect formed by the geometric parameters of the solid pins, this effectively suppresses the calculation deviation of local surface curvature in the surface evolution algorithm, reduces the probability of unclosed profiles generated due to simulated phase deformation, and thus reduces the need for manual intervention in the subsequent solidification repair process. The matching setting of liquid density and gravitational acceleration ensures the calculation accuracy of the interaction between volume forces and surface tension, enabling the evolution results to reflect both the liquid column morphology dominated by microscopic surface tension and the solder collapse trend under macroscopic gravity. The multi-dimensional parameter coupling modeling method fundamentally optimizes the quality of the initial morphology generation, resulting in a dual improvement in geometric fidelity and physical rationality of the repaired closed solid model.

[0028] Furthermore, in S3, the file format of the profile node geometry file is a geometry file format that can be recognized by finite element analysis software.

[0029] Specifically, geometric file formats that can be recognized by finite element analysis software can be .stl, .obj, .stp, etc.

[0030] Furthermore, in S4, the solidification repair process includes: identifying unclosed surface elements overlapping the solder portion and the pin; deleting the overlapping surfaces and lines; using the outer surface of the pin as a reference, projecting the undeleted surface element nodes onto the pin surface to generate a continuous closed curve; and reconstructing the solder surface based on this curve to form a closed solid model. The criterion for determining the overlapping surface elements is that the model gap is less than the preset tolerance value of 1×10⁻⁶ in the finite element software. -4 mm.

[0031] Specifically, the physical repair method of the present invention accurately identifies and removes solder and pin overlap areas that are smaller than a preset tolerance threshold (1×10). -4By eliminating unclosed surface elements and interference lines (mm), gaps and defects in the 3D model caused by surface evolution errors are eliminated, allowing the originally non-closed composite surface element structure to form a continuous closed geometric entity. Based on this, the undeleted folded surface element nodes are projected onto the outer surface of the pin to generate a continuous closed curve that fits the pin's shape, ensuring a seamless connection between the reconstructed solder surface and the pin geometry. This process achieves accurate conversion of liquid surface evolution results to a solid model, overcoming the solidification failure problem caused by node misalignment or gaps in traditional surface element files. The solder geometry constructed using this method can realistically reproduce the complex details of the actual solder joint morphology and effectively avoid element distortion or solution divergence caused by model gaps during finite element mesh discretization, thus significantly improving the reliability of subsequent mechanical response data calculations and providing a high-fidelity geometric foundation for life prediction models based on stress-strain response.

[0032] Furthermore, in S5, the mesh discretization process uses a patch conformal method to locally refine the fine structure of the solder joints, while the PCB substrate is divided into multiple regions to reduce the mesh density; the random vibration load spectrum is applied according to the QJ3086A-2016 standard, in which a 20-2000Hz power spectral density spectrum is applied perpendicular to the PCB direction, and the total root mean square acceleration is 28.5 grms.

[0033] Specifically, the random vibration load spectrum application method mentioned in this embodiment is based on the QJ3086A-2016 standard and sets a power spectral density spectrum of 20-2000Hz and a total root mean square acceleration of 28.5 grms in the vertical direction of the PCB. This technical feature, by strictly adhering to aerospace industry standards and specifications, accurately reproduces the dynamic environmental load conditions borne by aerospace electronic equipment under actual service conditions. Applying random vibration excitation with a specific frequency band and power spectral density in the vertical direction can accurately characterize the load characteristics of PCB components in a real mechanical environment, avoiding the deviation between traditional load models and actual physical excitations, and effectively improving the simulation credibility of finite element dynamic response. By matching the spectral parameters and root mean square acceleration values ​​specified in the standard, this method ensures the consistency of the boundary conditions of the vibration fatigue life model with the spacecraft vibration test specifications, making the statistical distribution characteristics of stress and strain amplitudes more consistent with the experimental results, and providing a structural response data foundation consistent with the actual excitation spectrum for the life prediction model based on the three-interval damage accumulation theory. The standardization and completeness of the load spectrum directly ensures the supporting role of the finite element analysis results in the actual structural life assessment, and avoids systematic prediction errors caused by the simplification of external load conditions.

[0034] Furthermore, in S6, the life prediction model adopts Miner's linear cumulative damage theory, including: extracting the 1σ, 2σ, and 3σ strain amplitudes at the weld joint and their corresponding cycle numbers ni from the finite element results; calculating the damage degree Di under each strain amplitude by combining the material's tensile strength and elastic modulus; and the total damage degree...

[0035] The result was obtained by weighted calculation of the three interval statistical distributions of 1σ strain amplitude (0.6831), 2σ strain amplitude (0.271), and 3σ strain amplitude (0.0433). When D ≥ 1, the weld joint was determined to be in failure.

[0036] Specifically, the life prediction model involved in this embodiment is based on Miner's linear cumulative damage theory and incorporates statistical methods to perform a layered weighted evaluation of weld joint life, enabling multi-dimensional comprehensive calculation of damage under different stress levels. This embodiment extracts the strain amplitudes and corresponding cycle numbers of 1σ, 2σ, and 3σ strains at the weld joint from finite element simulation results, and dynamically calculates the damage contribution of each strain amplitude based on material mechanical properties. By introducing statistical distribution weights (a three-interval distribution with 1σ strain amplitude accounting for 68.31%, 2σ for 27.1%, and 3σ for 4.33%), this method overcomes the limitation of traditional single-extreme value evaluation methods that ignore low-probability high-damage events, effectively balancing the cumulative effects of daily and extreme operating conditions. Compared to conventional ultimate load life prediction models, this method better reflects the statistical characteristics of random vibration loads in actual service environments, avoiding prediction bias caused by oversimplification of stress distribution. Combined with accurate stress and strain data from the repaired model, this method achieves dynamic coupling of multiple stress intervals through damage weight allocation, enabling the life prediction results to reflect the true degradation process of the weld joint under different stress levels. The mechanism for determining failure when the total damage degree D≥1 not only follows the core principle of linear cumulative damage theory, but also enhances the engineering applicability of the prediction model through weighting. Ultimately, while ensuring the reliability of the prediction results, it provides a technical basis for the formulation of maintenance strategies that is both conservative and economical.

[0037] A storage medium storing a computer program, which, when executed by a processor, implements the aforementioned finite element modeling method for simulating and repairing the morphology of gull-wing pin solder joints.

[0038] Specifically, the computer program stored on the storage medium automatically executes the aforementioned finite element modeling method for morphological simulation and repair of gull-wing pin solder joints. It integrates the entire process of 3D model reconstruction, closed-surface repair, and lifetime prediction into a repeatable algorithm system, solving the problem of solidifying surface elements caused by unclosed geometric files in the morphological evolution process. The program can accurately identify gap deviations caused by calculation errors in the surface evolution output file, automatically delete overlapping surface elements, and reconstruct closed curves based on the projection relationship between the pads and the outer surfaces of the pins, achieving high-precision reconstruction of the solder geometry. By fully reproducing steps S1 to S6, the program ensures standardized execution of operations such as material property binding, conformal mesh generation, and multi-dimensional physical field loading, avoiding random errors caused by manual intervention in the extraction of solder joint structural mechanical response data and lifetime prediction. This storage medium deeply integrates morphological repair technology with numerical analysis processes, significantly improving the repeatability of model construction and simulation. It provides a universal, high-fidelity structural analysis foundation for gull-wing pin devices under different parameter combinations, effectively shortening the R&D cycle of electronic packaging design, reducing experimental verification costs, and strengthening the reliability verification capability of key solder joints.

[0039] A computer device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-described finite element modeling method for simulating and repairing the morphology of gull-wing pin solder joints.

[0040] Specifically, the computer device described in this invention provides an efficient and reliable hardware support platform for finite element modeling of gull-wing pin solder joint morphology simulation and repair through the synergistic effect of memory and processor. The dedicated computer program on the device transforms the step-by-step modeling process into executable digital tasks, achieving high-precision processing of dynamic geometric reconstruction, surface closure repair, and multi-physics coupling calculations during the evolution of complex solder joint morphology. Especially when handling iterative calculations of the surface energy field of large-scale liquid solder, its high-performance computing capability effectively supports the rapid convergence of the energy-optimized surface morphology generation algorithm. In the solidification repair stage, the processor automatically identifies and completes overlapping surface elements and unclosed curves in the model based on a preset tolerance threshold, solving the technical difficulty of solidifying non-closed geometric structures. Simultaneously, through parallel computing optimization, the device can efficiently complete the hybrid discretization processing of localized fine-grained mesh generation of the solder fine structure and multi-region partitioning of the PCB substrate during the finite element analysis stage, ensuring the rapid acquisition of high-resolution mechanical response data and making the cumulative calculation of structural fatigue damage after random vibration load spectrum loading more accurate and reliable. This computer device integrates the entire chain of algorithms for morphological simulation, geometric repair, finite element modeling, and lifetime prediction into a unified computing framework, significantly improving the automation capabilities and computational efficiency of 3D modeling and lifetime assessment of complex solder joint structures, and providing a precise digital technology platform for reliability analysis of electronic packaging devices.

[0041] The following is a specific example: There are many parameters that distinguish the solder joints of gull-wing type pins, such as solder amount, pin offset, pin height, etc. Example 1 illustrates the process of evolution, repair and solidification of the solder joint of a pin with a reference solder volume of 1, and applies vibration load in finite element analysis, and uses a three-interval model to predict the life of the solder joint structure.

[0042] The solder joint morphology simulation code was compiled using Surface Evolver software. First, geometric parameters were set, such as pin height, pad length and width, and pin curvature. The specific code is provided in the appendix. Constraints were then set for the evolving objects, such as pad boundaries, pin boundaries, interface energy constraints, gravity, and surface tension. Subsequently, the geometric positions of points, lines, surfaces, and solids were set, and boundary, energy, surface tension, and fixation constraints were applied to the corresponding objects to calculate the liquid morphology evolution process. Automated commands were set to complete the evolution process. The SE-FIT software was used to output a common .stl format file, which can be opened and processed using ANSYS series geometry processing software. In SpaceClaim, the cross-sections were processed to locate unclosed cross-section positions. Without a repair process, deviations exceeding the tolerance of the gaps would occur during solidification, preventing the model from forming a solid model that can be divided into elements. Therefore, it is necessary to repair the unclosed surface elements. Based on the projection relationship, the folded surface elements near the pin positions were formed into a curve attached to the outer surface of the pin. New surface elements were then constructed based on this curve to form a closed surface, completing the construction of the solder joint structure solid model.

[0043] A solid model of the solder joint assembly is built based on the obtained solder joint model. The model is built according to the geometric dimensions and other parameters in the chip specification file. The pin structure of the original chip is replaced with the morphologically optimized solder joint model formed in the previous specification. The solid is coherently processed by building the same parts.

[0044] Meshing is a mathematical process that discretizes the solid model numerically and obtains structural mechanical parameters through the calculation and processing of nodal displacement data. The quality of the mesh directly affects the convergence of the calculation. Ideally, the optimal mesh should be tetrahedral or hexahedral. Any other mesh form, such as face alignment, distortion, or non-coherent nodes, will reduce the mesh quality. In this example, patch conformal meshing is used for fine structures, while multi-region meshing is used to reduce the mesh size for structures with little impact on calculation accuracy, such as PCBs. The mesh size is controlled by element volume and linewidth to obtain a mesh with appropriate quantity and quality that ensures convergence. The final meshing effect is as follows. Figure 5 As shown.

[0045] This example uses random vibration loads as boundary conditions applied to the system. Vibration analysis requires first extracting the system's modes. In this case, the random vibration load spectrum is applied using the QJ3086A-2016 standard. Therefore, the modal analysis is set to retrieve the 10th natural frequency within the 20-2000Hz range for modal analysis, obtaining the natural frequency values ​​and corresponding mode shapes. According to the QJ3086A-2016 standard, a random vibration load spectrum is set, and random vibrations are applied parallel and perpendicular to the device directions to obtain the overall stress probability distribution of the system, which is used for lifetime prediction analysis. Table 1 shows the QJ3086A-2016 standard:

[0046] Table 1 When a material is subjected to stress exceeding its fatigue limit, a certain amount of damage will occur in each stress cycle (strain). This damage can accumulate, and failure will occur when the accumulated damage reaches a critical value. This is the theory of fatigue damage accumulation. Currently, the most representative and widely used linear fatigue damage accumulation theory is Miner's rule. Crandall and Mark proposed the Miner method based on random vibration results. Similar to the linear cumulative damage assumption of Miner's law, this method is based on a fixed amplitude, assuming that when the stress amplitude cycles n times, a certain percentage of the material's fatigue life is consumed. n / N In some cases, cycles at other stress levels also cause partial damage to the material in the same way.

[0047] This method assumes that stresses greater than 3σ occur only in 0.27% of the time period and that they do not cause any damage. Therefore, using Miner's law for fatigue calculations, the formula for calculating the total damage is:

[0048] In the formula: n 1σ ≤1 s The actual number of cycles at the level (0.6831) ); n 2σ ≤2 s The actual number of loops at the level (0.271) ); n 3σ ≤3 s The actual number of loops at the level (0.0433) ); N iσ For the corresponding equation (5-5) s iThe allowable number of cycles corresponding to each stress level ( i =1,2,3).

[0049] when D When the value is 1, it indicates that the fatigue life has been exhausted, that is, fatigue failure has occurred.

[0050] Therefore, using 1 s 2 s and 3 s Stress and statistically averaged frequency can be used to conveniently calculate the life of random vibrations. In ANSYS, the statistically averaged frequency is equal to 1. s Speed ​​divided by 1 s The quotient of displacement. The general process for calculating random vibration fatigue is as follows: (1) Calculate the estimated frequency of the mean statistical stress at possible failure locations. (Stress velocity / Stress); (2) Determine 1 s 2 s 3 s Cumulative vibration cycles of the weld joint at strain level n i ; (3) Determine the tensile strength and elastic modulus of the material; (4) Extract the strain values ​​at key points and calculate the cycle life under that strain. N i ; (5) Calculate the overall damage degree D Estimate the lifespan of the solder joint.

[0051] Reference Figure 1-Figure 2 As shown, this invention presents a finite element modeling method for morphological simulation and repair of gull-wing type lead solder joints. It proposes a closed surface repair method, enabling the solidification of evolved composite surface element structures and facilitating their application in finite element calculations and lifetime prediction processes. This method can more accurately represent the three-dimensional structure of the device space, allowing for a more refined description of the interactions between micro-elements. By transforming the complex solder surface into a three-dimensional model through morphological repair, the calculation accuracy of structural thermal and mechanical data is significantly improved. It is evident from the solder joint structure models before and after processing that the solder joint structure after morphological simulation closely approximates the geometry of the actual soldered lead solder portion, demonstrating excellent modeling accuracy. This significantly improves the accuracy of finite element calculations and solder joint lifetime prediction, effectively enhancing the reliability of critical solder joints.

[0052] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A gull wing pin solder joint morphology emulation repair finite element modeling method, characterized in that, The method comprises the following steps: S1, constructing a solid pin region in a morphological simulation according to parameters of an actual gull-wing pin device, and applying a fixed constraint "fixed" to all points, lines and surfaces of the solid region; S2, setting a spatial node position of a liquid region as an initial position of evolution according to structure matching parameters and filling process parameters, setting liquid solder surface energy field control parameters, and evolving a liquid surface shape in a direction of lowest energy through refinement and iteration to obtain an approximate surface topography of a gull-wing pin solder joint; S3, outputting a cross-sectional node geometry file; S4, solidifying the gull-wing pin cross section, deleting overlapping surfaces and lines, and generating curves attached to an outer surface of the pin through projection relationship, constructing a surface closed space through the curves, reconstructing a solder geometry structure based on the surface closed space, and solidifying the reconstructed geometry structure into a pin and solder structure model; S5, assembling a three-dimensional model of an electronic device assembly after welding based on the solidified pin and solder structure model, importing the three-dimensional model of the electronic device assembly into finite element analysis software, sequentially completing material attribute binding, mesh discretization processing, boundary condition loading and multi-dimensional physical field load application operation in the finite element analysis software, submitting a solver for solving, and obtaining structural mechanics response data required for life prediction through a post-processing module, wherein the structural mechanics response data includes stress and strain amplitude; S6, predicting the service life of the solder joint through a life prediction model, substituting structural mechanics response data obtained through finite element calculation into the life prediction model, and predicting the expected life of each part of the solder joint through a statistical method.

2. The gull wing shaped pin joint morphology emulation repair finite element modeling method of claim 1, wherein, In S2, the structure matching parameters include pad size and pin offset, and the filling process parameters include pin offset.

3. The gull wing shaped pin joint morphology emulation repair finite element modeling method of claim 2, wherein, In S2, the liquid solder surface energy field control parameters include interface energy constraint, free liquid surface tension coefficient and gravity acceleration.

4. The gull wing shaped pin joint morphology emulation repair finite element modeling method of claim 3, wherein, In S2, the initial geometry of the liquid domain is defined according to the pad length, pin offset distance and solder volume parameters of the gull-wing pin device, the interfacial energy constraints include the contact angle between the pad and the solder, the contact angle constraint between the pin outer surface and the solder, and the liquid physical parameters include the surface tension coefficient 0.45 kg / s 2 , the liquid density 0.00846 kg / cm 3 and the gravitational acceleration 980 cm / s 2 .

5. The gull wing shaped pin joint morphology emulation repair finite element modeling method of claim 1, wherein, In S3, the file format of the cross-sectional node geometry file is a geometry file format that can be recognized by finite element analysis software.

6. The gull wing shaped pin joint morphology emulation repair finite element modeling method of claim 1, wherein, In S4, the entity repair process includes: identifying the non-closed facet that overlaps the solder part and the pin, after deleting the overlapping facets and lines, taking the outer surface of the pin as the reference, projecting the non-deleted facet nodes to the pin surface to generate a continuous closed curve, and reconstructing the solder surface based on the curve to form a closed entity model. The determination standard of the overlapping facet is that the model gap is less than the preset tolerance value 1 × 10 -4 mm of the finite element software.

7. The gull wing shaped pin joint morphology emulation repair finite element modeling method of claim 1, wherein, In S5, patch conforming method is used for local encryption of the fine structure of the solder joint, and multi-region division method is used for reducing the mesh density of the PCB substrate; and the application of the random vibration load spectrum is based on the QJ3086A-2016 standard, wherein a power spectral density spectrum of 20-2000 Hz is loaded in the direction perpendicular to the PCB, and the total root mean square acceleration is 28.5 grms.

8. The gull wing shaped pin joint morphing repair finite element modeling method of claim 1, wherein, In S6, the life prediction model adopts the Miner linear cumulative damage theory, which includes: extracting 1σ, 2σ and 3σ strain amplitude values and their corresponding cycle numbers ni from the finite element results, calculating the damage degree Di under each strain amplitude by combining the material tensile strength and the elastic modulus, and obtaining the total damage degree D by weighted calculation according to the three-interval statistical distribution of 1σ strain amplitude accounting for 0.6831, 2σ accounting for 0.271 and 3σ accounting for 0.0433, and determining that the solder joint fails when D≥1. The computer program is executed by a processor to implement the gull-wing pin solder joint morphological simulation repair finite element modeling method of any one of claims 1-8.

9. A storage medium having stored thereon a computer program, characterized in that It comprises:

10. A computer device, comprising: ​ A memory, a processor, and a computer program stored on the memory and executable on the processor, the processor executing the program to implement the gull-wing shaped pin joint shape simulation repair finite element modeling method of any one of claims 1-8.

Images