Micro-needle fixation strength prediction and calculation method based on virtual simulation and mechanical modeling

Abstract

The invention discloses a micro-needle fixation strength prediction and calculation method based on virtual simulation and mechanical modeling. The method comprises the following steps: acquiring geometric data of a connecting part of a microneedle head and a handle, and establishing a three-dimensional geometric structure containing connecting interface characteristics; the elasticity modulus, the yield strength and the breaking strength of the material are determined based on the geometric structure, and the friction coefficient and the bonding strength are set as contact characteristic parameters. A finite element grid is constructed, and test conditions of different pull-out speeds, angles and clamping positions are set. Stress and displacement distribution is obtained through simulation calculation, the maximum stress and the position of the maximum stress are extracted, the failure type is judged, and then an extraction force prediction curve is obtained through interface stress integration. Calculating deviation by combining actual pull-out test data, repeatedly iterating by adjusting the friction coefficient and the bonding strength until the deviation meets preset precision, and outputting a corrected pull-out force prediction curve and a maximum pull-out force prediction value, so as to realize rapid and non-destructive calculation and evaluation of the fixation strength of the microneedle.

Description

Technical Field

[0001] This invention relates to the fields of computer-aided design and finite metadata analysis, and in particular to a method for predicting and calculating the fixation strength of microneedles based on virtual simulation and mechanical modeling. Background Technology

[0002] With the popularization of cosmetic medical technology, microneedle puncture devices are widely used in fields such as skin rejuvenation and transdermal drug delivery. In existing technologies, the evaluation of the connection strength between the microneedle tip and the handle typically relies on physical pull-out tests. These tests involve clamping the needle tip and applying a constant pulling force, measuring the maximum force required during withdrawal to determine the needle's adhesion reliability. Some researchers have attempted to use finite element analysis to simulate the withdrawal process and study the influence of different materials or structural parameters on adhesion performance. However, these methods are still primarily based on idealized geometric models and empirical parameters, failing to truly reflect the complexity of the actual microneedle assembly interface.

[0003] However, existing testing methods have many operational limitations. Due to the tiny size of the cosmetic microneedle tip and the uneven thickness of the adhesive layer, traditional tensile testing machines are difficult to use and have poor test repeatability. Furthermore, the lack of realistic geometric and interface contact parameters in finite element analysis leads to significant discrepancies between simulation predictions and actual test results. In addition, the current lack of an effective parameter correction mechanism makes it impossible to adaptively optimize the prediction model under different withdrawal speeds, angles, and clamping conditions, affecting the accuracy and comparability of adhesion strength evaluation.

[0004] To address the shortcomings of the existing technologies, there is an urgent need to propose a method for predicting and calculating the microneedle fixation strength by combining real geometric reconstruction and mechanical simulation, so as to achieve non-destructive and high-precision strength assessment. Summary of the Invention

[0005] This application provides a method for predicting and calculating the fixation strength of microneedles based on virtual simulation and mechanical modeling, so as to achieve non-destructive and high-precision strength assessment.

[0006] This application provides a method for predicting and calculating the fixation strength of microneedles based on virtual simulation and mechanical modeling, including: Geometric data of the connection between the microneedle tip and the handle are obtained by using a 3D scanning device. The insertion depth, contact area, mating gap and adhesive layer thickness of the needle tip are extracted to establish a 3D geometric structure that includes the characteristics of the connection interface. Based on the contact area and adhesive layer thickness in the three-dimensional geometry, the elastic modulus, yield strength and fracture strength of the needle material, handle material and adhesive layer material are obtained, and the friction coefficient and adhesion strength between the needle and handle are set as contact characteristic parameters. A three-dimensional finite element mesh is established at the connection between the needle tip and the handle based on the contact characteristic parameters. Different withdrawal speeds, withdrawal angles, and clamping positions are set to form multiple test conditions. Mesh refinement processing is performed in the connection interface area to improve calculation accuracy. Stress field simulation calculations are performed on each test condition based on the encrypted three-dimensional finite element mesh to obtain stress distribution data and displacement data of the connection parts. The maximum stress value and its spatial location under each test condition are extracted. The failure type is determined by comparing the maximum stress value with the yield strength and bonding strength and a failure location identifier is generated. Based on the stress distribution data, the stress at the connection interface is integrated to obtain the initial prediction curve of the pull-out force changing with the pull-out displacement under each test condition. The pull-out force data of the actual microneedle pull-out test is collected, and the deviation between the initial prediction curve and the pull-out force data is calculated. By adjusting the friction coefficient and bonding strength, the three-dimensional finite element mesh establishment and stress field simulation calculation process is re-executed until the deviation meets the preset accuracy requirements. The corrected pull-out force prediction curve is output, and the microneedle fixation strength prediction result is calculated based on the maximum pull-out force prediction value in the pull-out force prediction curve.

[0007] The beneficial effects of the technical solution provided in this application include: (1) This invention establishes a virtual simulation model based on real geometric and mechanical parameters, which can obtain high-precision fixation strength prediction results without physically pulling and damaging the microneedle sample. This avoids the problems of short needles being difficult to clamp, sample waste, and large test errors in traditional tensile testing, and significantly improves the repeatability and standardization of the testing process. (2) By extracting the contact area, mating gap, and adhesive layer thickness of the needle and handle connection part through three-dimensional scanning, and combining the material elastic modulus, yield strength, and bonding characteristic parameters, a finite element model that conforms to the actual assembly state can be constructed. This method performs mesh refinement in the interface area, so that the stress concentration and debonding evolution process can be described at high resolution, thereby greatly improving the accuracy of fixation strength simulation prediction. Attached Figure Description

[0008] Figure 1 This is a flowchart of a method for predicting and calculating the fixation strength of microneedles based on virtual simulation and mechanical modeling, provided in the first embodiment of this application. Detailed Implementation

[0009] Many specific details are set forth in the following description to provide a full understanding of this application. However, this application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this application; therefore, this application is not limited to the specific embodiments disclosed below.

[0010] The first embodiment of this application provides a method for predicting and calculating the fixation strength of microneedles based on virtual simulation and mechanical modeling. Please refer to... Figure 1 This figure is a flowchart of the first embodiment of this application. The following is in conjunction with... Figure 1 The first embodiment of this application provides a detailed description of a method for predicting and calculating the fixation strength of microneedles based on virtual simulation and mechanical modeling.

[0011] Step S101: Obtain geometric data of the connection between the microneedle tip and the handle using a 3D scanning device, extract the tip insertion depth, contact area, mating gap, and adhesive layer thickness, and establish a 3D geometric structure including the connection interface features.

[0012] Step S101 aims to obtain the actual geometric information of the microneedle tip and handle in the actual assembly state without damaging the microneedle component under test, and to establish a three-dimensional geometric structure model that can be used for subsequent mechanical simulation analysis.

[0013] First, in the geometric data acquisition stage, a 3D scanning device with micron-level resolution is required to scan and measure the connection area between the microneedle tip and the handle. This 3D scanning device can be a laser confocal scanning microscope, a white light interferometric surface scanner, or an industrial CT scanning system. For highly reflective materials such as metal needles or high-gloss polymers, to reduce the interference of specular reflection on the scanning quality, a thin, uniformly sprayed diffuse reflection developing coating can be applied to the surface before scanning, enabling the scanning device to obtain a stable surface reflection signal. During scanning, the same connection area to be measured should be acquired from multiple angles, including at least forward, oblique, and radial views, to ensure complete coverage of the needle insertion area, the adhesive bonding area, and the inner wall area of ​​the handle. The raw data after scanning are point cloud data or voxel density data, which represent the actual shape, position, and interrelationship of the needle tip, adhesive layer, and handle in space. To ensure spatial consistency in subsequent analysis, coordinate alignment processing is required for data from different scanning angles. A common method is to fit common features of overlapping regions so that point cloud data from different batches coincide in the same coordinate system, thereby obtaining a complete and continuous three-dimensional dataset of connected regions.

[0014] After obtaining the integrated 3D dataset, it is necessary to extract geometric feature parameters related to adhesion strength from this dataset. First, the needle insertion depth needs to be determined. The needle insertion depth is the actual length of the needle embedded in the handle socket. To determine this length, the needle axis can be identified in the 3D data, and the bottom of the needle (i.e., the position where the needle makes the deepest contact with the adhesive layer) and the entrance position of the handle socket can be marked. The distance between these two positions along the axial direction of the needle is the needle insertion depth. If the needle is not inserted perpendicularly into the handle but at a certain angle, the direction of the needle's own central axis should be used as the measurement reference direction, rather than the shape of the handle, to avoid measurement errors introduced by the tilted assembly. This reference direction can be determined by reading the centerline direction vector obtained by fitting the needle shape; this is a normal 3D geometric post-processing operation.

[0015] The contact area needs to be identified from the scanned data to determine the region where the needle and handle are in direct contact after assembly. The contact area is defined as the region where the distance between the outer surface of the needle and the inner surface of the handle is close to zero, or where the adhesive layer completely fills the gap. To identify this area, the outer surface of the needle and the inner surface of the handle are compared point by point. For each surface location, the shortest distance to the corresponding mating surface is calculated. When this distance is less than a preset contact threshold, the area is considered a valid contact area. This contact threshold can be determined based on the resolution of the scanning device. For example, when the scanning resolution is at the 2-micrometer level, a common contact threshold is around 5 micrometers; that is, if the closest distance between the two surfaces is less than approximately 5 micrometers, it is considered a valid contact area. By summing up all the local surface segments identified as valid contact areas, the effective contact area between the needle and the handle can be obtained. The area of ​​this effective contact area is defined as the contact area. The contact area reflects the range of force transmission of the needle within the handle and is an important input for subsequent fixation strength calculations.

[0016] The mating clearance is the spatial interval between the outer surface of the needle tip and the inner surface of the handle at the point where they are not fully mated. Physically, it represents the size of the gap that remains between these two structures after assembly. The mating clearance is obtained by measuring the distance between non-contact areas. Multiple representative measurement points on the outer surface of the needle tip and corresponding areas on the inner surface of the handle can be used to calculate the distance, resulting in a series of spacing values. These values ​​can be statistically analyzed to obtain the average mating clearance. Maximum and minimum clearances can also be provided as needed to reflect the uniformity of the clearance distribution. For example, in a batch of microneedle assemblies, if the clearance in most areas is around 8 micrometers, while the clearance in some areas reaches over 20 micrometers, it indicates localized uneven mating. This unevenness directly affects the stress concentration at the interface and the shape of the final pull-out force curve. Therefore, the mating clearance is not only a geometric measurement but also a boundary condition for subsequent interface contact modeling.

[0017] Adhesive layer thickness describes the actual thickness distribution of the bonding material between the needle and the handle in the assembled state. The adhesive layer refers to the bonding medium used to secure the connection between the needle and the handle, typically a medical adhesive or curing resin. Adhesive layer thickness can be obtained in two ways. If industrial CT scanning is used, the needle body, handle body, and the intermediate adhesive layer can be distinguished in voxel data due to the different density contrasts of different materials in the scanned image. In this case, the thickness of the adhesive layer can be measured cross-sectionally along the normal direction of the connection interface to directly obtain the spatial distribution of the thickness. If optical scanning equipment is used and the density information of different materials is not distinguished, the adhesive layer thickness can be estimated by comparing the theoretical mating distances of the components before and after assembly. Specifically, the nominal gap between the handle hole and the needle shape in the theoretical assembly state can be recorded first, and then compared with the surface distance at the same location in the actual assembly state. The difference between the two can be considered as the actual thickness of the adhesive layer at that location. This thickness can be repeatedly measured at multiple sampling locations along the connection interface to obtain the average adhesive layer thickness and the distribution range of the adhesive layer thickness.

[0018] After extracting the insertion depth, contact area, mating clearance, and adhesive layer thickness, these measurements need to be used to reconstruct the three-dimensional geometry. The three-dimensional geometry refers to a three-dimensional solid model containing information about the relative spatial positions, envelope shape, and interface morphology of the needle, handle, and the adhesive layer between them. This three-dimensional geometry is not merely an idealized shape of the needle and handle, but should retain actual assembly features to the greatest extent possible, such as the actual eccentric position of the needle inside the handle, the circumferential thickness variation of the adhesive layer, and the chamfered or burr morphology of the handle orifice. This is a key difference from traditional ideal modeling. The specific reconstruction process can be completed in commercial 3D modeling or simulation preprocessing software. For example, the aforementioned point cloud data can be imported into modeling software, and the scattered data can be converted into editable closed surfaces through surface fitting, then the corresponding solid structure can be generated based on the closed surfaces. The needle, handle body, and adhesive layer should be defined as independent geometric bodies or parts so that they can be assigned different material properties and different contact conditions in the subsequent finite element calculation stage. If the needle has a significant assembly tilt or eccentricity, this eccentricity should be preserved in the model, rather than being forcibly corrected to a coaxial state during modeling, because such a small eccentricity will significantly affect the location and magnitude of the local stress peak at the interface in mechanical simulation.

[0019] For example, in a real sample, the scanning results showed that the total length of the microneedle tip was approximately 0.8 mm, the insertion depth of the needle into the handle was approximately 0.45 mm, the effective contact area between the needle tip and the handle was approximately 0.12 square millimeters, the average mating gap was approximately 8 micrometers, and the average thickness of the adhesive layer was approximately 15 micrometers. Furthermore, the adhesive layer was slightly thicker on one side and slightly thinner on the other, exhibiting a significant asymmetrical distribution. Reflecting this information into a three-dimensional geometry, a combined model is obtained, comprising the needle tip entity, the handle insertion hole entity, and the non-uniform adhesive layer entity. This model clearly records the eccentricity direction and amount of the needle tip within the handle. This combined model can be directly used as the basic geometric input in subsequent finite element modeling without further idealization or simplification, thus making the stress and displacement distributions obtained during simulation closer to the stress state of the actual assembly.

[0020] Step S102: Based on the contact area and adhesive layer thickness in the three-dimensional geometry, obtain the elastic modulus, yield strength, and fracture strength of the needle material, handle material, and adhesive layer material, and set the friction coefficient and adhesion strength between the needle and handle as contact characteristic parameters.

[0021] Step S102 is a crucial step in establishing the mechanical calculation foundation of the entire method. Its purpose is to provide complete and quantifiable physical parameter inputs for subsequent finite element simulations after obtaining the realistic three-dimensional geometric structure. This step determines the mechanical properties and interfacial contact characteristics of the materials, enabling the virtual model to realistically reflect the actual force and deformation behavior between the microneedle tip and the handle, thereby ensuring the physical rationality and predictive accuracy of the simulation results. To achieve this, this step requires measuring the mechanical properties of the tip material, handle material, and adhesive layer material, and also requires parameterizing the interfacial interaction between the tip and handle, including quantifying the friction coefficient and adhesive strength.

[0022] In determining the properties of needle materials, the specific material type of the needle should be identified first. Cosmetic microneedles are typically made of medical-grade stainless steel, titanium alloys, or hard polymers; different materials exhibit significant differences in elastic response and failure behavior. The elastic modulus is a proportionality constant describing the stress-strain relationship of a material during elastic deformation, and is usually obtained through a standard tensile test. In practice, a thin rod sample of the same material with a diameter of about one millimeter is taken, and gradually increasing tensile force is applied while measuring its elongation. When the strain is within a linear range, the ratio of stress to strain is the elastic modulus. For example, the typical elastic modulus of medical-grade 304 stainless steel is approximately 190 gigapascals, while the elastic modulus of polyetheretherketone (PEEK) is approximately 3.5 gigapascals. This value can be directly adopted or modified and input into the simulation model based on experimental data of the needle material under corresponding temperature and humidity conditions. Yield strength is used to characterize the critical stress value at which a material transitions from the elastic stage to the plastic stage, and can be obtained from the yield point of the tensile curve. For metal needles, a small stress increment should be applied before and after the yielding stage, and the yield strength should be identified by the inflection point of the stress-strain curve. Fracture strength is the maximum stress a material can withstand under tension or shear until failure, typically the highest point of the stress-strain curve, and should be determined in a standard specimen failure test. These three indicators together define the overall response characteristics of the needle material during the drawing process, from linear elasticity and plastic yielding to final fracture.

[0023] The methods for obtaining the mechanical parameters of the handle material are similar to those for the needle. However, since handles are often made of thermoplastic materials such as polycarbonate (PC), polypropylene (PP), or nylon (PA), temperature control is necessary during testing to avoid thermal softening affecting the measurement results. For thermoplastic materials, uniaxial tensile testing under constant temperature conditions can be used. The tensile speed should be on the same order of magnitude as the pull-out speed in subsequent simulations to ensure the comparability of strain rate effects. The measured elastic modulus is typically in the range of two to three gigapascals, and the yield strength is generally between fifty and one hundred megapascals. For glass fiber reinforced plastics or composite materials, measurements should be taken separately in different fiber directions to obtain anisotropic material parameters.

[0024] The mechanical parameters of the adhesive layer material are particularly important because this layer serves both as an adhesive bond and a primary channel for interfacial stress transmission in the connection between the needle and the handle. The adhesive layer is typically made of medical-grade epoxy resin or UV-curable acrylic adhesive. Its elastic modulus can be obtained through thin film tensile testing or dynamic mechanical analysis (DMA). For adhesive layer thicknesses less than 20 micrometers, planar adhesive layer samples with the same formulation can be prepared, cured, and their stress-strain relationship tested on a tensile testing machine to estimate the material's elastic modulus and yield strength. Fracture strength can be determined through shear testing of standard bonded specimens; that is, a layer of adhesive of the same thickness is sandwiched between two substrates of the same material, and the maximum stress value at which the adhesive layer fails under uniaxial tension is the fracture strength. Alternatively, the material data sheet provided by the manufacturer can be consulted, and adjustments can be made based on curing conditions and the actual adhesive layer thickness.

[0025] After determining the basic performance parameters of the three types of materials mentioned above, it is necessary to establish the physical characteristics of the interface contact between the needle and the handle. The coefficient of friction and adhesive strength are the most critical parameters in interface contact behavior. The coefficient of friction is defined as the ratio of tangential resistance to normal pressure between two surfaces, used to describe the magnitude of resistance during interface sliding during withdrawal. Since there is usually some mechanical fit or micro-roughness meshing between the needle and the inner wall of the handle, the coefficient of friction is not only determined by the type of material, but also by surface roughness, surface treatment state, and the presence of residual adhesive layer. In specific measurements, a micro-friction tester can be used to conduct a plane friction test. Samples of the needle material and handle material are prepared as smooth contact surfaces, and their sliding friction force is measured under a constant normal load. The average value is then used to calculate the coefficient of friction. For example, for an untreated stainless steel and polypropylene contact pair, the coefficient of friction is usually between 0.2 and 0.3, while if a thin adhesive layer is coated on the surface, the coefficient of friction may rise to above 0.5.

[0026] Adhesive strength describes the ability of an interface to resist separation during a pull-out process. Physically, it represents the maximum tensile stress per unit area that the interface can withstand before separation. This parameter can be obtained through a standardized pull-out test. During the test, a cylindrical adhesive sample similar to the actual microneedle assembly is prepared. The cylinder of the needle material and the sleeve of the handle material are bonded and cured with the same adhesive layer. An axial pull-out load is applied on a tensile testing machine, and the maximum tensile force at which the adhesive layer fails is recorded. Dividing this maximum tensile force by the adhesive area yields the adhesive strength. For example, if the adhesive area of ​​the sample is 0.1 square millimeters, and the tensile force at failure is 0.3 Newtons, then the adhesive strength is 3 MPa.

[0027] To ensure the simulation input parameters match the actual conditions, this step requires mapping and correlating the aforementioned physical parameters based on the contact area and adhesive layer thickness extracted from the 3D geometry. A larger contact area significantly impacts the overall pull-out force due to the interfacial adhesive strength; a thicker adhesive layer results in a longer stress transmission path and a stronger local stress concentration effect. During modeling, these parameters should be input into the contact definition module of the simulation software, and corresponding attributes should be assigned to different contact areas. For example, a slightly lower adhesive strength value can be set for the edge contact area to reflect the early peeling trend caused by stress concentration, while the average value should be maintained for the central contact area.

[0028] After determining all parameters, a parameter verification mechanism should be established to prevent input values ​​from deviating from physical reality. This can typically be achieved through small-scale simulations, simplifying the model to local interface elements and performing sensitivity analysis on the pull-out force response under different friction coefficients and bond strengths. When the trend of the calculated results matches the experimental expectations, the parameter settings are considered reasonable. For example, as the friction coefficient increases, the peak pull-out force should also increase, while when the bond strength decreases significantly, the model should exhibit premature interface separation.

[0029] Therefore, step S102 obtains the elastic modulus, yield strength, and fracture strength of the three materials—needle tip, handle, and adhesive layer—through the system, and accurately defines the friction coefficient and bonding strength by combining experimental or literature data, thus forming a complete system of contact characteristic parameters. These parameters can accurately describe the interfacial adhesion, slippage, and failure behavior of the microneedle connection during the removal process, thereby providing realistic and reliable physical input for subsequent finite element mesh establishment and stress field simulation calculations.

[0030] Step S103: Establish a three-dimensional finite element mesh for the connection between the needle tip and the handle based on the contact characteristic parameters, set different pull-out speeds, pull-out angles and clamping positions to form multiple test conditions, and perform mesh refinement processing in the connection interface area to improve calculation accuracy.

[0031] Step S103 involves converting the geometric information and material parameters obtained in the first two steps into a finite element model suitable for numerical analysis. This is one of the core technical aspects of the entire method. The purpose of this step is to establish a three-dimensional finite element mesh model that accurately reflects the stress state at the connection between the microneedle tip and the handle. Based on this model, various pull-out conditions are constructed to simulate the stress behavior under real test conditions. This step requires five levels of technical operations: geometric modeling, mesh generation, boundary condition definition, loading condition setting, and mesh refinement control. Each step directly affects the simulation accuracy and the reliability of the final prediction results.

[0032] Before creating the finite element mesh, the 3D geometry obtained in step S101 should be imported into the finite element preprocessing software. Commonly used finite element modeling platforms include ANSYS, and the choice of software depends on the company's usage habits and the degree of support for nonlinear contact calculations. When importing the geometric model, the topological integrity of the model should be checked to ensure that the needle, handle, and adhesive layer are independent and identifiable entities, and that there are no geometric gaps or overlaps at the interfaces. If there are small discontinuities on the surface reconstructed from the scan data, they can be repaired by surface reconstruction or local smoothing algorithms to avoid mesh generation failure. In this case, the needle entity, adhesive layer entity, and handle entity should share the same contact interface, and the geometric position of the interface surface nodes must strictly correspond to the contact area obtained in the original scan. For the adhesive layer, the thickness is usually only a few micrometers to tens of micrometers, so its geometric dimensions are relatively small compared to the overall model. In finite element modeling, it should be ensured that the adhesive layer thickness still has a true proportion in the model so as to accurately capture the interface stress distribution and peeling behavior during the simulation.

[0033] After geometric verification, the mesh generation stage begins. A finite element mesh discretizes a continuous three-dimensional geometry into a finite set of nodes and elements, allowing the computer to solve for physical quantities such as stress, strain, and displacement in each element. The needle and handle are typically made of isotropic metal or plastic, so tetrahedral or hexahedral solid elements can be used for meshing. For the needle tip region or the thin-walled region of the handle, adaptive tetrahedral elements can be used to avoid element shape distortion. Due to the thinness of the adhesive layer and the stress gradient along its thickness, multi-layered refined solid elements or laminated element structures are recommended to ensure at least three layers of elements are distributed along the thickness direction, thus reflecting the stress changes at the adhesive interface along the thickness direction. If the adhesive layer is too thin, making solid element meshing difficult, a geometric scaling factor can be set in the software. For example, the adhesive layer thickness can be scaled up five times in the geometric model, but the elastic modulus and adhesive strength values ​​can be correspondingly reduced in the material parameters to maintain physical equivalence. For the contact area between the adhesive interface and the needle or handle, contact units should be defined. A contact unit is a special type of unit that can transmit normal force and tangential friction force during relative motion. Those skilled in the art should define the outer surface of the needle as the master surface and the inner surface of the handle or adhesive layer as the slave surface in the contact pair configuration, and input the friction coefficient and adhesive strength parameters obtained in step S102.

[0034] Before setting loading and boundary conditions, the simulation conditions need to be determined based on the physical process of the actual microneedle extraction experiment. Extraction speed is the linear velocity of the pointer relative to the handle during the extraction process; extraction angle is the angle between the pointer's extraction direction and the handle's axis; and clamping position refers to the position where the handle or needle is fixed or constrained when a load is applied. To obtain the fixation performance response under different conditions, multiple calculation scenarios need to be set in the model. For example, in a typical combination of conditions, extraction speeds of 100 mm / min, 200 mm / min, and 300 mm / min can be set to simulate different operating speeds; extraction angles can be set to zero degrees (i.e., fully axial extraction), 15 degrees, and 30 degrees to examine the asymmetric distribution of interface stress during oblique extraction; and clamping positions can be fixed at the bottom of the handle, the middle of the handle, or the tail end of the needle to analyze the influence of the clamping method on the extraction force results. When these working conditions are established in the software, those skilled in the art should utilize the parametric modeling function so that different working conditions can automatically generate corresponding calculation models by simply changing the speed, angle and constraint position parameters, thereby improving calculation efficiency and ensuring geometric consistency between various working conditions.

[0035] In the definition of loading conditions, the pull-out speed is usually applied in the form of a displacement load. A displacement boundary condition that varies linearly with time can be applied to the free end of the needle, with its direction consistent with the pull-out angle. For example, during axial pull-out, displacement control should be applied along the central axis of the needle, while during 15-degree oblique pull-out, the displacement vector needs to be decomposed into axial and radial components along the 15-degree direction. The magnitude of the pull-out speed should correspond to the speed range in the actual experiment and maintain quasi-static conditions, i.e., ensure that the loading speed is small enough to avoid significant influence of inertial effects on the results. Full constraints should be applied to the fixed end of the handle, prohibiting its translational and rotational degrees of freedom in three directions to simulate the state of the handle being fixed by the clamp. To avoid rigid body motion in the overall model during numerical solution, at least one node's rotational degree of freedom should be constrained.

[0036] In mesh refinement, the element size needs to be controlled for different regions based on stress concentration areas and interface complexity. The purpose of refinement is to improve the accuracy of local stress calculations, especially at the edges of the adhesive layer, the needle insertion end, and the handle orifice, where stress gradients are typically highest. Refinement strategies can be implemented by setting local mesh size controls or defining "local refinement regions" in the modeling software. For example, the element size for the adhesive layer can be set to one-fifth to one-tenth of that of adjacent regions. If the geometric proportions of the needle and handle materials differ significantly, a local transition mesh should be used to avoid abrupt changes in element shape. After mesh refinement, the entire model should be quality checked to ensure that the element twist rate, aspect ratio, and Jacobian matrix values ​​are within the reasonable range recommended by the software; for example, the minimum element angle should not be less than 20 degrees, and the aspect ratio should not exceed 5:1. Only when the mesh quality is satisfactory can the numerical stability of subsequent solutions be guaranteed.

[0037] In this step, inputting the contact characteristic parameters is crucial for ensuring the physical realism of the model. The software typically provides a "friction-adhesion" composite contact model, which can simultaneously consider the interaction between frictional slippage and adhesive bonding. The friction coefficient obtained in step S102 should be entered in the contact settings, and the adhesive failure criterion should be enabled so that the simulation can automatically trigger separation when the adhesive layer reaches its adhesive strength. When defining this failure criterion, a standard based on maximum tensile stress or energy release rate can be selected. If the maximum tensile stress criterion is selected, the adhesive layer fracture strength value should be entered as the critical stress; if the energy release rate criterion is used, the fracture energy value per unit area should be entered, representing the energy required for complete adhesive layer peeling. For general cosmetic microneedle adhesive structures, the maximum tensile stress criterion is sufficient to meet the calculation accuracy requirements.

[0038] After completing the modeling, loading, and meshing, the complete input file for each test case should be saved in the software. For computational verification, a typical test case can be selected first to observe the distribution of stress and displacement fields during needle removal. If excessive stress concentration or unreasonable displacement abrupt changes are found at the interface, the mesh density of the adhesive layer can be adjusted appropriately or the contact algorithm can be redefined. This verification step ensures good model convergence and stable and reliable calculation results. Generally, a model with 100,000 to 300,000 nodes for a single test case is sufficient to guarantee adequate accuracy. Using a higher density mesh will significantly increase computation time but only slightly improve accuracy; therefore, accuracy and efficiency should be balanced according to computational resources.

[0039] In summary, step S103, by establishing a three-dimensional finite element mesh model including the needle, handle, and adhesive layer, and combining various pull-out conditions and contact parameter settings, fully realizes the construction and calculation preparation of the virtual simulation model. This model not only accurately describes the stress and displacement characteristics of the connection between the needle and handle under different conditions, but also ensures the accuracy and applicability of the calculation results through local mesh refinement and multi-condition loading strategies, thus laying a reliable physical and numerical foundation for subsequent stress field simulation and pull-out force prediction.

[0040] Furthermore, the step of establishing a three-dimensional finite element mesh for the connection between the needle tip and the handle based on the contact characteristic parameters, setting different withdrawal speeds, withdrawal angles, and clamping positions to form multiple test conditions, and performing mesh refinement processing in the connection interface area to improve calculation accuracy includes: The contact characteristic parameters are used to divide the connection area of ​​the needle, adhesive layer and handle into regions. The needle insertion area, adhesive layer area and handle constraint area are defined as independent mesh domains. In each mesh domain, element types and node constraint relationships corresponding to material properties are assigned to form an initial three-dimensional finite element mesh with complete contact definition. A standardized pull-out load is applied to the initial three-dimensional finite element mesh, and a mechanical response analysis is performed in the adhesive layer region to obtain the nodal stress distribution and element strain gradient. The stress concentration area and transition area in the adhesive layer region are determined based on the stress gradient variation amplitude, and the spatial coordinates of the stress concentration area are defined as the interface stress feature set. Based on the interface stress feature set, the stress concentration area is proportionally densified and divided. The size of the regional unit is adjusted to a predetermined ratio of the average size of the original regional unit. Linearly gradient unit sizes are set in the adjacent transition area along the thickness direction and the tangential direction to maintain the continuity of nodes and the smoothness of interface stress transmission, thereby forming a high-precision three-dimensional finite element mesh after local densification. Based on the high-precision three-dimensional finite element mesh, a set of test conditions is constructed. The pull-out speed, pull-out angle, and clamping position are used as the test condition variables. Multiple representative test conditions are generated according to the orthogonal combination principle. The loading conditions of each test condition are then associated with the high-precision three-dimensional finite element mesh to form a multi-condition finite element mesh set with complete boundary condition definitions. Before loading, an interface continuity analysis is performed on the multi-condition finite element mesh set to determine whether the stress change of the nodes in the adhesive layer area meets the stability threshold. When the stress change of the local element under any test condition exceeds the predetermined stability limit, the local element of the region is re-divided and a secondary densification process is performed to generate the final finite element mesh set after stability correction. The final finite element mesh set is then used as the densified three-dimensional finite element mesh.

[0041] In establishing a 3D finite element mesh for the connection between the needle tip and the handle based on contact characteristic parameters, it is first necessary to define the object and scope of the finite element modeling. The key stress area of ​​the microneedle system is located at the connection between the needle tip and the handle, which includes three main structural regions: the needle insertion area, the adhesive layer area, and the handle constraint area. The needle insertion area is the part where the needle tip is actually embedded inside the handle, and its geometry is usually a conical or needle-shaped cylindrical structure; the adhesive layer area refers to the adhesive layer used to fix the needle tip and the handle, and its thickness is generally between 0.02 mm and 0.2 mm; the handle constraint area is the part that provides reverse support force, usually made of high-strength plastic or aluminum alloy. To ensure the accuracy of the finite element modeling, these three areas need to be treated as independent meshing domains to avoid non-physical stress concentrations at the material transition interface.

[0042] After the region is partitioned, appropriate element types and node connections should be selected based on the mechanical properties of each structural material. For the needle insertion area, due to its high rigidity and stress concentration near the root, hexahedral elements are suitable, as they provide high computational stability in the axial force direction. For the adhesive layer area, since the adhesive is usually an elasto-plastic or viscoelastic material, and its thickness is much smaller than its width, cohesive elements are suitable, as they accurately describe the nonlinear response of interface debonding and slippage. For the handle constraint area, tetrahedral or hexahedral elements can be selected based on the elastic modulus of the material to balance computational accuracy and resource consumption. Node constraints should be set according to the actual assembly and testing conditions. For example, nodes at the bottom of the handle should be fully constrained to prevent overall displacement, while nodes at the outer end of the needle should be defined as loading nodes to apply pull-out loads. Through the above region partitioning and node constraint definitions, an initial three-dimensional finite element mesh with complete contact definitions can be formed.

[0043] After obtaining the initial mesh, a standardized pull-out load analysis should be performed to extract mechanical response characteristics. The standardized pull-out load refers to a tensile force applied at a constant rate along the needle's axial direction, typically set to 200 mm / min in a virtual simulation environment to simulate the pull-out speed of a tensile testing machine in a real test. During loading, the system calculates the stress (σ) and strain (ε) at each node and further determines the strain gradient, i.e., the rate of change of strain within an element with spatial location. The strain gradient is a crucial indicator of stress concentration. This gradient value can be calculated using numerical difference or finite volume approximation methods; for example, dividing the principal stress difference between adjacent element centers by the node distance yields the local stress gradient. If this gradient is significantly higher than twice the average value, stress concentration can be identified in that region. The spatial coordinates of all stress concentration regions are extracted and recorded, forming the so-called interface stress feature set, which guides subsequent mesh refinement.

[0044] After forming the interface stress feature set, proportional mesh refinement needs to be performed on the stress concentration areas. Proportional refinement refers to dynamically adjusting the mesh element size according to the degree of local stress concentration, so that the element size is smaller in areas with stronger stress concentration, thereby improving simulation accuracy. For example, when the maximum stress in the stress concentration area is three times that of the adjacent area, its element side length can be adjusted to 1 / 3 of the original average size. This operation can be achieved through the adaptive meshing function in finite element preprocessing software (such as ANSYS Workbench). To avoid node discontinuities caused by abrupt changes in element size, a transition zone should be set around the stress concentration area, and a linear gradient relationship of element size should be set along the adhesive layer thickness direction and tangential direction, so that the element size gradually increases from the concentration area outward. The linear gradient can be achieved by specifying the gradient ratio (such as increasing the element size by 10% per layer), thereby ensuring that the force transmission path is geometrically continuous and mechanically smooth. After the refinement process is completed, a high-precision three-dimensional finite element mesh with local refinement and global continuity will be obtained.

[0045] After obtaining the high-precision mesh, a test case set needs to be constructed based on this mesh. A test case refers to the force conditions under different combinations of pull-out speed, pull-out angle, and clamping position. The pull-out speed refers to the speed at which the needle tip is pulled out, and can take three typical values: 0.1, 0.5, and 1.0 mm / s. The pull-out angle refers to the angle between the pulling force direction and the central axis of the needle tip, and can take three values: 0°, 15°, and 30°. The clamping position refers to the position of the loading point relative to the length of the needle tip, and can be selected at 1 / 3, 1 / 2, and 2 / 3 of the needle length. By applying the orthogonal combination principle to the above parameters, nine representative test cases can be generated, thus achieving multivariate coverage with a limited number of simulations. For each test case, corresponding boundary conditions need to be set. For example, in the case of an inclined angle, the pulling force direction in vector form should be defined, and constraint nodes should be limited at the clamping position. Each test case establishes a one-to-one correspondence with the high-precision finite element mesh, forming a multi-case finite element mesh set with complete loading and boundary condition definitions.

[0046] To ensure the stability of the refined mesh under various working conditions, an interface continuity analysis needs to be performed before loading. Interface continuity refers to whether the stress and displacement transfer between adjacent elements is smooth and without abrupt changes. The method for determining continuity is to calculate the stress difference between adjacent nodes and compare it with a set threshold. For example, if the principal stress difference between adjacent nodes in the adhesive layer region is greater than 10%, it indicates local instability or element size mismatch, and a local re-meshing operation should be triggered. In practice, an automatic detection script can be set in the finite element software to allow the system to scan all node stress differences before analysis. When the stress difference exceeds the threshold, a secondary refinement process is performed in the corresponding region. Secondary refinement is usually limited to local areas, and the new element size is 0.8 times that of the original elements to avoid increasing the global computational load. After this process, a final finite element mesh set with stability correction can be obtained.

[0047] This final mesh set not only maintains node continuity and adhesive layer thickness consistency in terms of geometry, but also achieves smooth stress transfer paths in terms of mechanical properties. The corrected final finite element mesh set can be considered a refined three-dimensional finite element mesh for subsequent stress field simulation calculations. Its data structure includes node coordinates, element topology relationships, material property parameters, and boundary condition definitions, which can be directly recognized and used by the finite element solver.

[0048] The encrypted 3D finite element mesh established by the above method can significantly improve the stress resolution and computational stability at the microneedle connection interface compared with the traditional fixed element division method.

[0049] Step S104: Perform stress field simulation calculations for each test condition based on the encrypted three-dimensional finite element mesh to obtain stress distribution data and displacement data of the connection parts, extract the maximum stress value and its spatial location under each test condition, determine the failure type by comparing the maximum stress value with the yield strength and bond strength, and generate a failure location identifier.

[0050] Step S104 is the core step in the entire virtual simulation calculation process, from numerical solution to mechanical result analysis. Its purpose is to obtain the stress and displacement distribution at the connection between the microneedle tip and the handle under different working conditions by performing mechanical simulation calculations on the finite element model established in step S103. Based on these calculation results, the location and failure type of potential interface failure can be identified, thus providing fundamental data for subsequent pull-out force prediction curves and adhesion strength calculations. This step must be performed under strict physical boundary conditions and material nonlinearity. The calculation results must not only reflect the macroscopic force-displacement relationship but also realistically reproduce the stress gradient distribution, strain concentration areas, and potential peeling or slippage behavior between interface layers.

[0051] In practical implementation, the finite element mesh model established in step S103 needs to be imported into the calculation module. Each model corresponds to a specific combination of working parameters, including pull-out speed, pull-out angle, and clamping position. The simulation calculation should select steady-state nonlinear statics or quasi-static solution type, because the microneedle pull-out process is a case of low-speed loading and negligible inertial influence. This type of solution requires consideration of the nonlinear yield characteristics of the material and the bonding, friction, and debonding behavior of the contact interface in the calculation. In the input parameters, the elastic modulus, yield strength, and fracture strength data determined in the previous step are used for the needle, handle, and adhesive layer, respectively; for the interface contact part, the friction coefficient and adhesive strength are introduced as contact properties, so that the software can automatically determine whether the interface slips or separates at different loading stages. For adhesive materials such as adhesive layers, it is recommended to enable the "nonlinear bonding model" or "bond-slip model" in the calculation. This model can simultaneously describe the mechanical response of the initial bonding stage, slip stage, and final debonding stage of the interface, ensuring that the simulation results can accurately reflect the real pull-out behavior.

[0052] Before simulation, reasonable convergence control conditions should be set. Due to the strong nonlinear effects at the adhesive interface, non-convergence is prone to occur in numerical solutions; therefore, a step-by-step loading method is recommended. Specifically, the pull-out process can be divided into several sub-steps, with a very small displacement increment applied in each sub-step. For example, a displacement change of approximately 0.001 mm can be applied each time in the needle pull-out direction, ensuring that the model can stably calculate the incremental changes in stress and strain in each step. This allows the solution process to converge smoothly, avoiding numerical oscillations caused by sudden debonding or slippage.

[0053] After the solution is completed, the finite element software will output a data file containing the stress components, strain components, and displacement distribution of each node and element. This step requires extracting these results and performing a mechanical analysis. The stress field refers to the stress distribution at various locations in the model under stress, typically including three main types: normal stress, shear stress, and equivalent stress. Equivalent stress is a comprehensive index used to reflect the overall strength of a material under multiaxial stress. A commonly used calculation method is the von Mises criterion, which combines principal stresses in different directions into a single value according to a certain relationship to determine whether the material has reached yield. For the adhesive layer region, the distribution of normal tensile stress and tangential shear stress is of greater concern, as these two stresses represent the peeling and slipping tendencies of the adhesive layer, respectively, and are direct indicators of interface failure.

[0054] The displacement field reflects the amount of movement of each point in the model relative to its initial position, and can display the overall deformation of the interface during needle removal. By observing the distribution of displacement vectors, it is possible to determine whether the removal process is a pure pulling mode or accompanied by a certain angle of offset sliding, thereby verifying the rationality of the working condition settings.

[0055] After stress field analysis, the maximum stress value should be extracted for each test condition. The maximum stress value refers to the stress peak value at a certain point or region in the entire model. It usually appears at geometric abrupt changes, material interfaces, or local areas with the highest mesh density. The extraction can be performed using the "Extreme Value Query" function of the finite element post-processing module. Select the adhesive layer area or the needle insertion end area as the monitoring range, and the software will automatically output the maximum principal stress, maximum shear stress, or maximum equivalent stress of that area and its corresponding coordinates. If multiple similar peak values ​​are found, the most likely failure location needs to be determined based on the stress distribution gradient. Special attention should be paid to locations such as the edge of the adhesive layer, near the needle port, and the edge of the handle orifice, as these locations are often the first to fail due to structural discontinuities or stress concentrations.

[0056] After obtaining the maximum stress value, it needs to be compared with the material yield strength and bond strength obtained in step S102 to determine the failure type of the model under the current loading conditions. If the maximum stress value is less than the yield strength of the needle or handle material but exceeds the bond strength of the adhesive layer, it indicates that the failure mainly occurs at the interface, i.e., interface debonding failure. If the maximum stress value exceeds the yield strength of the needle or handle material, it indicates that the material body may undergo plastic deformation or fracture, which is material failure. If both are close to the critical values, it indicates that there is coupled failure, i.e., the interface and the material yield simultaneously. After the judgment is completed, the failure location should be marked in the post-processing diagram of the model. Different colors or symbols can be used to distinguish the failure area, for example, red is used to mark the interface debonding area and orange is used to mark the plastic yielding area, so as to intuitively identify the failure mode.

[0057] To ensure the repeatability of the results, a numerical data file should be output, recording the maximum stress value, corresponding coordinates, failure type, and failure location for each working condition in a table. For practical engineering applications, this data can be used to analyze the influence of different pull-out speeds, angles, or clamping methods on failure modes. For example, when the pull-out angle increases, the calculation results often show a significant increase in shear stress at the edge of the adhesive layer, while the normal tensile stress changes relatively little, indicating that oblique pull-out is more likely to induce interface slip failure. When the pull-out speed increases, the equivalent stress of the adhesive layer may increase slightly due to the strain rate effect, but the overall failure location remains unchanged. These patterns can provide a quantitative basis for the optimization of microneedle structures.

[0058] If a nonlinear contact model is enabled in the computational model, the change in the interfacial adhesion state with displacement can also be observed. Finite element software typically outputs parameters such as "adhesion state variables" or "contact separation ratio," with values ​​ranging from zero to one, where zero represents complete adhesion and one represents complete separation. By analyzing the spatial distribution of these variables, the evolution process of interfacial debonding can be obtained, i.e., the expansion path from local debonding to overall debonding. Those skilled in the art can use these results to determine the failure propagation direction and speed of the adhesive layer, thereby more accurately inferring the stage at which the peak pull-out force occurs.

[0059] For example, in a typical simulation case, when the pull-out displacement reaches 0.1 mm, the equivalent stress at the edge of the adhesive layer reaches approximately 12 MPa, slightly higher than the bond strength of 11 MPa, and localized delamination occurs at the interface. When the pull-out displacement further increases to 0.2 mm, the delamination area expands circumferentially by about a quarter circle, and the stress peak gradually shifts to the middle of the adhesive layer. At 0.3 mm, most of the interface debonds, and the pull-out force begins to decrease. Such simulation results clearly demonstrate the temporal characteristics of interface failure and verify the rationality of the bonding parameters set in the model.

[0060] Therefore, step S104, through solving the stress and displacement fields, extracting the maximum stress, and determining the failure type, achieves a complete simulation of the stress mechanism and identification of the failure mode at the microneedle connection. The stress distribution data, displacement data, maximum stress value, and failure location marker output in this step will directly serve as the core input data for subsequent steps of pull-out force curve calculation and model calibration.

[0061] Furthermore, the stress field simulation calculation is performed on each test condition based on the encrypted three-dimensional finite element mesh to obtain stress distribution data and displacement data of the connection part, extract the maximum stress value and its spatial location under each test condition, and determine the failure type and generate a failure location identifier by comparing the maximum stress value with the yield strength and bond strength, including: On the encrypted three-dimensional finite element mesh, corresponding boundary conditions and load constraints are applied to each test condition to obtain stress distribution data and displacement data of the needle insertion area, adhesive layer area and handle constraint area, forming a stress distribution dataset for failure identification. Based on the stress information of each finite element in the stress distribution dataset, the maximum stress value and its spatial location of each finite element under the corresponding test condition are determined, forming a set of maximum stress values ​​containing the maximum stress values ​​of each element. Based on the set of maximum stress values, each maximum stress value is compared with the yield strength and bond strength to identify finite element elements that meet the conditions of exceeding the yield strength or bond strength. Based on the spatial location of the finite element elements, the finite element elements that are determined to be failed are grouped into a group of failure regions with spatial continuity to form preliminary failure region distribution information. Based on the continuity characteristics and spatial relationship of each failure region group under different test conditions, the failure type of each failure region group is determined, and a corresponding failure location identifier is generated for each failure region group.

[0062] When performing stress field simulation calculations for various test conditions on the encrypted 3D finite element mesh, it is necessary to first apply boundary conditions and load constraints consistent with the actual pull-out process to the 3D finite element mesh. This ensures that the model can realistically reflect the mechanical behavior of the needle insertion area, adhesive layer area, and handle constraint area under stress in each test condition. Boundary conditions restrict the displacement or rotational degrees of freedom of certain nodes, simulating the state of the device being fixed by a clamp in the experiment. Load constraints are usually applied by displacement loading in the pull-out direction or by equivalent force loading to simulate the actual test mechanical environment under different pull-out speeds and angles. The loading process can be defined as step-by-step loading according to the commonly used finite element software interface settings, recording the corresponding nodal displacement and nodal stress at each step. Through this process, stress distribution data and displacement data reflecting the stress state of each finite element within the model can be obtained, forming a stress distribution dataset for failure identification.

[0063] In the stress distribution dataset, each finite element corresponds to a set of data including principal stresses, shear stresses, and displacements in various directions. The first principal stress is typically used as a key indicator for assessing the risk of material tensile stress or debonding. To calculate the maximum stress value for each finite element, the principal stress values ​​for that working condition are first queried at several nodes corresponding to the element, and then the maximum value is selected from these node principal stress values. For example, if the principal stresses at the four nodes of a finite element are 12 MPa, 15 MPa, 14 MPa, and 9 MPa, then the maximum stress value for that element is 15 MPa. Storing the maximum stress values ​​of all elements along with their spatial locations forms the maximum stress value set. The spatial locations in the maximum stress value set can be determined using the element geometric center coordinates provided by the finite element software, typically described using a three-dimensional Cartesian coordinate system.

[0064] After obtaining the set of maximum stress values, it is necessary to determine whether each finite element element has failed. Yield strength is used to determine whether the material undergoes plastic deformation in the tensile or compressive direction, while adhesive strength is used to determine whether interfacial debonding or slippage has occurred in the adhesive layer. Therefore, when the maximum stress value of a finite element exceeds the yield strength, the element can be considered to have failed in the material bulk direction; when the maximum stress value exceeds the adhesive strength, the element can be considered to have failed in the interfacial bonding direction. In practical applications, these strength thresholds can be directly read from the given material parameter table. For example, if the adhesive strength of the adhesive material is 8 MPa, the yield strength of the plastic material is 20 MPa, and the maximum stress of an element reaches 9 MPa, then that element should be considered an adhesive failure element. All finite element elements that meet the above conditions are classified according to their spatial location. By judging the distance and contact relationship between adjacent elements, these elements can be grouped into a failure region group exhibiting spatial continuity. Spatial continuity can be determined by judging whether the distance between the geometric centers of the elements is less than a set threshold. For example, when the distance between the centers of adjacent elements is less than 1.5 times the average side length of the element, it can be considered continuous. This will allow us to generate preliminary information on the distribution of failure areas.

[0065] After obtaining preliminary information on the distribution of failure areas, it is necessary to determine the failure type based on the continuity characteristics and spatial relationships of each failure area group under different test conditions. If the failure of all elements in a failure area group is caused by the maximum stress exceeding the bond strength, the group can be classified as an interface peeling failure area; if the main cause of failure is debonding and slippage caused by shear stress, the group can be classified as an adhesive slip failure area. The main failure mode of each area can be identified by analyzing the principal stress direction of the elements in each failure area group. For example, if the first principal stress direction of the elements in the area is consistent with the pull-out force direction, it is usually a peeling failure; if the maximum shear stress dominates, it is usually a slip failure. Based on the failure type and area location, the system generates a corresponding failure location identifier for each failure area group. The failure location identifier is usually represented by a combination of area number and area spatial range, for example, using the "area number + coordinate envelope" method, so that subsequent steps can accurately access each failure area in the three-dimensional structure. In this way, the determination of the failure type and the generation of failure location identifiers are completed, providing the necessary basic data for the subsequent correction of the pull-out force prediction curve.

[0066] Furthermore, based on the set of maximum stress values, each maximum stress value is compared with the yield strength and bond strength to identify finite element elements that meet the conditions exceeding the yield strength or bond strength. Then, based on the spatial location of the finite element elements, the finite element elements determined to be failed are grouped into a failure region group with spatial continuity, forming preliminary failure region distribution information, including: The adjacency relationship between finite element elements is determined based on the spatial coordinates of each finite element element in the set of maximum stress values, and adjacency information used to represent the spatial connectivity of finite element elements is obtained. Based on the adjacency information, finite element elements whose maximum stress values ​​satisfy the conditions of exceeding the yield strength or exceeding the bond strength are divided into multiple candidate failure region groups according to spatial connectivity, thus obtaining candidate failure region data containing each candidate failure region group. The continuity of the candidate failure region data is checked. Regions consisting of only a single finite element and not spatially connected to other finite element elements are eliminated. Candidate failure region groups with connected boundaries and whose maximum stress value changes meet the predetermined continuity conditions are merged to form a set of failure region groups after continuity correction. Based on the spatial location of each failure region group in the failure region group set, the finite element elements belonging to the same failure region group are recorded with their corresponding maximum stress values ​​to generate preliminary failure region distribution information to characterize the relationship between the spatial distribution of failure regions and stress distribution.

[0067] In the process of identifying failed finite element elements based on the set of maximum stress values ​​and forming preliminary failure region distribution information, it is necessary to first obtain the coordinate information of each finite element in three-dimensional space. The spatial coordinates of a finite element refer to the position of the corresponding node in a three-dimensional coordinate system. For an element composed of multiple nodes, the representative coordinates of the element can be obtained by calculating its geometric center. For example, for a tetrahedral element composed of four nodes, its geometric center is calculated by adding the three-dimensional coordinates of the four nodes one by one and taking the average, resulting in a specific three-dimensional coordinate point that represents the position of the element in space.

[0068] After obtaining the spatial coordinates of all finite element elements using the above method, the spatial adjacency relationships between the finite element elements can be determined. Adjacency relationship refers to whether two finite element elements share at least one node, or whether they are in close contact in space. In this field, adjacency relationships are typically determined by whether two elements share a common node. For example, if element A and element B share at least one node, they are considered to have a direct adjacency relationship. If they do not share a common node, the distance between their geometric centers can be further calculated. If this distance is less than a preset multiple of the mesh element size, they are considered to be spatially connected. This method allows the establishment of an adjacency information set, indexed by element number, recording the numbers of other adjacent elements for each element, enabling the system to identify continuous spatial regions in subsequent steps.

[0069] After obtaining the adjacency information, finite element elements marked as exceeding the yield strength or bond strength from the maximum stress value set can be used as initial input to determine whether these elements constitute a continuous failure region in space. The method for forming a candidate failure region group is to start with any finite element that meets the failure condition, add adjacent elements that also meet the failure condition to the same group, and continue expanding outwards until all adjacent elements that meet the condition are classified into the same candidate region. For example, when the maximum stress values ​​of three adjacent elements all exceed the bond strength, and the three share a node in space, they are automatically classified into a candidate failure region group; if any element does not form an adjacency relationship with other failed elements, it constitutes a separate candidate failure region.

[0070] For the aforementioned candidate failure region group, further continuity verification is required to ensure that the final failure region has clear spatial continuity and mechanical rationality. The purpose of continuity verification is to eliminate finite element elements that, although satisfying the stress conditions, exist in spatial isolation. Typically, a failure region should consist of multiple finite element elements spatially close to each other, while the appearance of isolated elements is often due to noise caused by local numerical fluctuations. For example, in actual data, if the maximum stress value of a finite element is slightly higher than the yield strength, but all surrounding elements are much lower than that strength value, those skilled in the art can determine that this situation does not conform to the actual failure propagation law and should be eliminated. Furthermore, for a candidate region group with two contacting boundaries, it is necessary to determine whether the change in maximum stress value between them is smooth. When the difference in maximum stress values ​​between the boundary elements of the two regions remains within a reasonable range (e.g., below a certain proportion of the average stress value of the two regions), they can be merged into a single failure region group to reflect the continuous failure characteristics in real materials.

[0071] After performing continuity checks and necessary merging operations on the candidate failure region groups, a set of failure region groups after continuity correction can be obtained. Subsequently, it is necessary to establish the correspondence between the set of failure region groups and the maximum stress value. Specifically, the spatial location and corresponding maximum stress value of each finite element in each failure region group are recorded one by one, ensuring that every point in three-dimensional space can be clearly mapped to the corresponding failure region number and maximum stress value. This recording process should strictly maintain index consistency, for example, using the finite element number as a unique identifier, to ensure that this information can be accurately retrieved for failure region analysis in subsequent steps.

[0072] Once the above information is fully established, preliminary failure area distribution information can be generated to characterize the relationship between the spatial distribution and stress distribution of the failure area. This information directly reflects the preliminary failure behavior of the microneedle connection under specific test conditions, including the spatial location of the failure area, its continuity range, and the distribution of the maximum stress value within the area.

[0073] Step S105: Based on the stress distribution data, perform integral calculation on the stress at the connection interface to obtain the initial prediction curve of the pull-out force changing with the pull-out displacement under each test condition, collect the pull-out force data of the actual microneedle pull-out test, and calculate the deviation between the initial prediction curve and the pull-out force data.

[0074] Step S105 involves transforming the complex stress distribution data into a quantifiable pull-out force curve after completing the simulation calculations of the stress and displacement fields, and then comparing and analyzing the results with actual experimental data. The purpose of this step is to establish an initial predicted curve of pull-out force versus pull-out displacement through numerical processing of the simulated interface stress data, and then, combined with experimentally measured pull-out force data, determine the degree of difference between the simulation model and the actual situation, thereby providing a basis for subsequent model correction.

[0075] After the simulation calculation is completed, the finite element model outputs the nodal stress distribution data on the interface at each time step or displacement increment. The "interface" is the interaction area formed between the pointer and the handle through adhesive or frictional contact; the stress distribution in this area directly reflects the mechanical response during the pull-out process. The stress data includes the normal stress, tangential stress, and corresponding nodal coordinates and area information for each interface element. Since the pull-out force is a macroscopic quantity, while stress is a local quantity, it is necessary to integrate the stresses of all micro-elements on the interface to transform them into an overall force. In numerical simulation, this integration is achieved through discrete summation: the interface is divided into several elements, the average stress of each element is calculated and multiplied by its area, and then the results of all elements are summed to obtain the total force at a given moment. Those skilled in the art can understand this as considering the interface as composed of many tiny force-bearing plates, each with a force in the same direction as the needle pull-out direction; the overall pull-out force is the resultant force of all these local forces.

[0076] For example, in a certain displacement step of the simulation output, the interface is divided into one thousand elements. The normal stress value of each element represents the local strength of the adhesive layer or contact surface at that location against pull-out. If the area of ​​each element is approximately one square micrometer, the total pull-out force at that displacement can be obtained by multiplying these stresses by the corresponding area and then summing them. It is important to note that only the normal stress component along the pull-out direction contributes to resistance; the tangential stress mainly affects the interface slip characteristics and is not included in the direct component of the pull-out force, but it can reflect the contribution of frictional energy consumption in subsequent energy analysis. By sequentially processing the stress data under different displacement increments, a discrete sequence of pull-out force as a function of displacement can be obtained. To form a continuous prediction curve, spline interpolation or polynomial fitting methods can be used to connect these discrete points into a smooth curve, which is the initial prediction curve of pull-out force versus displacement.

[0077] The pull-out displacement is defined as the relative movement of the pointer head relative to the handle, which can be directly obtained from the loading displacement control parameters in the simulation model. The pull-out process typically consists of three stages: Initial stage: the pull-out force rises rapidly with displacement, indicating that the adhesive layer at the interface is in a state of overall elastic deformation; Intermediate stage: the pull-out force reaches its peak and then fluctuates slightly, corresponding to local peeling or slippage of the adhesive layer; Final stage: the pull-out force decreases significantly, and the interface completely separates. The initial prediction curve should reflect this typical mechanical characteristic. If the force-displacement relationship shows a linear increase without a significant peak in the simulation results, it indicates that the contact parameters are improperly set or the interface bonding model is not correctly enabled, and corrections should be made in subsequent steps.

[0078] After obtaining the simulation prediction curve, data from actual pull-out experiments need to be collected for comparison. Actual pull-out tests are typically performed using a precision electronic tensile testing machine. The sample connecting the needle and handle is fixed in upper and lower clamps and pulled out at a constant speed (e.g., 200 mm per minute), with changes in force and displacement recorded in real time. The test results are output as a force-displacement curve, where the peak value represents the maximum pull-out force measured experimentally, and the curve shape reflects the actual interface failure process. To facilitate comparison with simulation results, the experimental and simulation curves should be plotted on the same coordinate system, using the same units and scale. If the overall trend of the experimental and prediction curves is consistent, it indicates that the model correctly reflects the force law at the macroscopic level; if there are differences in the peak position or curve shape, quantitative analysis is needed to calculate the deviation.

[0079] The deviation can be calculated using the mean absolute error or root mean square error to assess the degree of difference between the two curves at different displacement points. Those skilled in the art can achieve this numerically, by taking the difference between the experimental force and the predicted force at each identical displacement point and calculating the average or square root of the sum of all differences. A smaller deviation indicates a better fit between the simulation model and the actual situation. For example, if the maximum pull-out force in the experiment is 0.4 Newtons, while the simulation prediction is 0.38 Newtons, the relative error is 5%. If the experimental curve is slightly lower than the predicted curve overall, but the shape and trend are the same, the model can be considered to have high accuracy. If the deviation exceeds a preset threshold (e.g., 10%), it indicates that input parameters such as the friction coefficient or adhesive strength in the model still need adjustment, which will be completed in subsequent steps.

[0080] In addition to overall deviation, local deviations can also be analyzed. For example, errors can be calculated separately at the initial pull-out stage, peak stage, and failure stage to identify the stage where the model's fit is worst. Generally, the deviation at the initial stage reflects the accuracy of the elastic modulus setting, the deviation at the peak stage reflects the rationality of the bond strength and friction coefficient, while the deviation at the failure stage is related to the adhesive layer fracture energy or interface peeling criterion. Through this segmented comparison, those skilled in the art can more effectively determine the sources of error in the model.

[0081] For example, in one simulation, the predicted pull-out force in the initial stage was slightly higher than the experimental result, indicating that the adhesive layer stiffness was set too high; in the peak stage, the two matched well, indicating that the bond strength parameters were accurate; in the final stage, the simulation curve decreased slowly, indicating that the fracture criterion threshold was too high or the debonding energy was set too large. Based on this analysis, the equivalent elastic modulus or fracture energy of the adhesive layer material can be adjusted in subsequent steps to gradually optimize the accuracy of the model.

[0082] After completing the deviation calculation, the experimental data and predicted results should be compared in a table and saved as a structured data file for use in automated calibration or subsequent statistical analysis. For multiple test conditions, such as different pull-out angles or speeds, the deviation for each condition can be calculated separately, and then the weighted average can be taken as the overall model deviation index. The weighting coefficients can be determined based on the importance of the experiment or the reliability of the data; for example, the condition closest to actual usage conditions can be given a higher weight.

[0083] Furthermore, sensitivity analysis can be performed to verify the stability of the initial prediction curve. By applying small perturbations to key parameters such as the friction coefficient and bond strength, the magnitude of the changes in the prediction curve is observed. If a small parameter change causes significant curve fluctuations, it indicates that the model is sensitive to that parameter and needs to be optimized in subsequent corrections; if the change is small, the parameter is stable and reliable. This analysis helps engineers clarify the direction of subsequent corrections and improve the convergence efficiency of simulation calculations.

[0084] Step S106: By adjusting the friction coefficient and bonding strength, re-execute the three-dimensional finite element mesh establishment and stress field simulation calculation process until the deviation meets the preset accuracy requirements, output the corrected pull-out force prediction curve, and calculate the microneedle fixation strength prediction result based on the maximum pull-out force prediction value in the pull-out force prediction curve.

[0085] Step S106 is the convergence and correction stage in the entire microneedle adhesion strength prediction calculation method. Its core objective is to continuously optimize the interface contact parameters in the aforementioned simulation model through iteration, so that the deviation between the simulation prediction results and the actual experimental data reaches the preset accuracy requirements, thereby outputting a stable, reliable, and engineering-significant pull-out force prediction curve and adhesion strength prediction results. This step is not only a parameter correction process for the model, but also a crucial link in integrating virtual simulation and experimental results. Its result determines the reliability and applicability of the entire prediction method.

[0086] After comparing the initial predicted curve with the experimental data in step S105, two curves with roughly similar shapes but different values ​​are usually obtained. This deviation stems from the difference between the initial assumed values ​​of the interface contact parameters and the actual bonding state, especially the values ​​of the friction coefficient and bond strength. In practical applications, these two parameters are significantly affected by factors such as material surface roughness, curing conditions, adhesive layer thickness, and temperature and humidity environment; therefore, optimal results cannot be obtained by setting them all at once. This step involves repeatedly adjusting these two key parameters to gradually bring the simulation model closer to the actual mechanical response.

[0087] First, a preset accuracy requirement needs to be defined. Preset accuracy is a standard for evaluating the consistency between simulation results and experimental results, and can be expressed as a percentage of deviation or root mean square error. For example, if the difference in the maximum pull-out force between the two curves does not exceed five percent of the experimental value, or if the average error between the two curves over the entire displacement range is less than three percent, the preset accuracy is considered to have been achieved. This threshold can be set according to different application requirements. For example, for medical-grade microneedle products, a prediction accuracy of no less than five percent is typically required to ensure that the simulation results are sufficient to guide structural optimization.

[0088] Before starting iterative corrections, the direction and magnitude of parameter adjustments should be determined. The coefficient of friction and adhesive strength have different characteristics in influencing the pull-out force curve: the coefficient of friction mainly affects the downward trend in the latter part of the curve, i.e., the mechanical resistance during the interface slippage stage of the pull-out process; adhesive strength mainly determines the peak value of the curve and the steepness of the initial upward stage. If the overall level of the simulation curve is significantly lower than the experimental curve, it indicates that the interface bonding force setting is too weak, and the adhesive strength value should be increased. Conversely, if the simulation peak value is too high and the curve declines too slowly, it indicates that the interface adhesion is too strong or the coefficient of friction is too large, and the corresponding parameters need to be appropriately reduced. To prevent oscillations or over-adjustment during iteration, the parameter adjustment amount is generally controlled within the range of 5% to 10% of the previous setting value. For example, if the initial adhesive strength is set to 10 MPa, and the simulation result is about 8% lower than the experimental result, the adhesive strength can be increased to about 10.8 MPa in the next calculation.

[0089] After parameter adjustment, the finite element simulation calculation needs to be re-executed. This process includes re-establishing the contact characteristic parameter file, re-meshing the finite element mesh, and reapplying boundary conditions. Since the geometry and loading method remain unchanged in each iteration, the main changes are reflected in the interface contact properties. To improve computational efficiency, the software's "parametric analysis" function can be enabled, allowing the model to automatically update parameters and run a new round of simulation in each iteration. After each simulation, the system outputs a new pull-out force-displacement curve and calculates the difference between it and the experimental curve. By comparing the deviation changes between two consecutive iterations, it can be determined whether the correction direction is correct. When the deviation shows a clear decreasing trend, it indicates that the parameter adjustment is reasonable; if the deviation increases instead, it is necessary to reverse the adjustment or reduce the adjustment range.

[0090] In practical implementation, to prevent getting trapped in local optima, a step-by-step iterative strategy can be adopted. First, the friction coefficient is fixed, and only the bonding strength is adjusted until the peak value of the curve matches the experimental value. Then, the bonding strength is fixed, and the friction coefficient is fine-tuned until the tail of the curve coincides with the experimental value. This sequence avoids mutual interference between parameters and improves the convergence speed. For example, in a simulation, the initial bonding strength was set to 9 MPa, and the friction coefficient was 0.3. After the first iteration, the bonding strength increased to 10 MPa, and the difference between the simulated peak value and the experimental peak value decreased from 12% to 5%. Then, the friction coefficient was adjusted from 0.3 to 0.28, further reducing the tail deviation to 2%, ultimately meeting the preset accuracy conditions.

[0091] Once the model deviation meets the preset accuracy requirements, the corrected pull-out force prediction curve can be output. The corrected curve should be smooth and continuous, and its peak value, rate of rise, and downward trend should all be consistent with the experimental curve. Typically, the corrected prediction curve is slightly lower than the experimental curve in the initial stage, with peak values ​​coinciding or differing only slightly, and slightly higher than the experimental result in the final stage. This difference is within an acceptable range and reflects the approximation of the numerical calculation. The curve data can be exported as a standard data file, containing displacement, predicted pull-out force, and stress distribution information at each stage for subsequent analysis.

[0092] Based on this, the predicted maximum pull-out force needs to be calculated. The maximum pull-out force refers to the force value corresponding to the peak point of the corrected prediction curve. It reflects the maximum load the interface can withstand during the pull-out process and is a direct indicator of adhesion strength. During calculation, the "peak extraction" function can be used in the post-processing module of the simulation software, or the maximum force point can be searched directly in the exported data. If the curve has multiple local peaks, the overall maximum value should be used. To verify the stability of the prediction results, the calculation can be repeated under multiple working conditions, such as obtaining the maximum pull-out force at different pull-out angles or speeds, and taking the average value as the final prediction result to reduce the influence of random errors under a single working condition.

[0093] Once the predicted maximum pull-out force is obtained, the predicted microneedle adhesion strength can be calculated. Adhesion strength is the maximum pull-out stress that a unit contact area can withstand when the needle tip is engaged with the handle. It is calculated by dividing the maximum pull-out force by the actual contact area between the needle tip and the handle. The contact area has already been extracted in step S101. For example, if the contact area is 0.1 square millimeters and the corrected predicted maximum pull-out force is 0.4 Newtons, then the predicted adhesion strength is 4 MPa. This result can be compared with the material's design standards or experimental averages to verify the product assembly quality or bonding process stability.

[0094] To ensure the model's universality and repeatability, this step can be repeated on multiple samples, and the optimal combination range of friction coefficient and adhesive strength can be obtained through statistical analysis. If the final parameters among different samples tend to be consistent, it indicates that the method has good stability; if the differences are large, external factors affecting the parameters, such as temperature, humidity, or curing time, can be further investigated and incorporated into future parametric models for compensation.

[0095] Furthermore, the process of re-executing the three-dimensional finite element mesh establishment and stress field simulation calculation by adjusting the friction coefficient and bonding strength until the deviation meets the preset accuracy requirements, outputting the corrected pull-out force prediction curve, and calculating the microneedle adhesion strength prediction result based on the maximum pull-out force prediction value in the pull-out force prediction curve, includes: Under each test condition, the initial pull-out force prediction curve is compared with the pull-out force data obtained from the actual pull-out test to obtain error data that reflects the distribution of prediction error, and the error data is used to determine the deviation index that characterizes the overall deviation. The stress distribution in the adhesive layer area is analyzed based on the deviation index. The correction direction and correction magnitude of the friction coefficient and adhesive strength are determined based on the offset relationship between the simulated stress peak position and the actual failure position, so as to obtain parameter correction information for updating the simulation input conditions. The parameter correction information is weighted and integrated under test conditions with different pull-out speeds, pull-out angles and clamping positions to obtain the corrected friction coefficient and bonding strength applicable under multiple conditions. The corrected parameters are then used to re-execute the three-dimensional finite element mesh establishment and stress field simulation calculation to generate a new pull-out force prediction curve. The new pull-out force prediction curve is compared with the previous prediction curve. When the new pull-out force prediction curve meets the preset convergence condition in terms of deviation index and maintains a stable stress change trend in the adhesive layer area, the corrected pull-out force prediction curve is output, and the microneedle fixation strength prediction result is determined by the maximum pull-out force prediction value in the corrected pull-out force prediction curve.

[0096] When correcting contact characteristic parameters to obtain a more accurate pull-out force prediction curve, a systematic and gradual adjustment is required based on the difference between the initial simulation output and the actual pull-out test results. To ensure the feasibility of this adjustment process, it is necessary to first compare the initial pull-out force prediction curve with the pull-out force data obtained from the actual pull-out test under each test condition. This comparison refers to numerically comparing the predicted pull-out force and the actual pull-out force at the same pull-out displacement position to obtain error data representing the prediction deviation. The pull-out displacement is the amount of displacement experienced by the pointer tip from the initial contact position to the fully disengaged position during the pull-out process. To ensure the accuracy of the comparison, the displacement points under different test conditions need to be standardized, for example, by dividing the pull-out displacement into several fixed sampling points, such as calculating the pull-out force value every 0.01 mm. Then, the difference between the predicted value and the actual value is calculated at each sampling point. For example, if the actual pull-out force at a displacement of 0.20 mm is 1.50 N and the predicted pull-out force is 1.20 N, then the error at that point is 0.30 N; if the actual pull-out force at a displacement of 0.30 mm is 2.10 N and the predicted value is 2.25 N, then the error at that point is −0.15 N. By repeating the above calculation for each sampling point, an error data sequence covering the entire pull-out process can be obtained. Then, the average deviation and maximum deviation are calculated based on this error sequence. The average deviation can be obtained by averaging all error values, while the maximum deviation can be obtained by selecting the error with the largest absolute value in the error sequence. The resulting deviation index can be used to quantitatively characterize the overall deviation between the prediction and the actual value.

[0097] After obtaining the deviation index, it is necessary to analyze the stress distribution in the adhesive layer area to determine the spatial offset between the simulated stress peak position and the actual failure position. The adhesive layer area is the key area where the needle and handle are connected by the adhesive material, and it is also the main part affecting the adhesion strength of the microneedle. The simulated stress peak position represents the spatial location of the maximum principal stress or equivalent stress obtained from the finite element calculation; the actual failure position can be determined by the location of interface debonding, slippage, or material failure in the pull-out experiment. When the simulated stress peak position lags behind the actual failure position, that is, the simulated stress concentration area is further back than the actual failure area, it indicates that the friction coefficient is too low, because insufficient friction reduces the interface resistance in the simulation, thus delaying the failure trend; conversely, if the simulated debonding range is smaller than the actual debonding range, it indicates that the adhesive strength is too high, because the adhesive layer in the simulation is set too strong, which is not conducive to forming a failure mode consistent with reality. Based on the above correspondence, the correction direction of the friction coefficient and adhesive strength can be determined, such as increasing the friction coefficient or decreasing the adhesive strength. The correction magnitude can be determined based on the size of the deviation index. For example, a larger correction magnitude is used when the average deviation is large, and a smaller correction magnitude is used when the deviation is small, to avoid over-correction that could lead to computational instability. This generates parameter correction information for updating simulation input conditions.

[0098] To ensure the stability of the corrected parameters under various working conditions, the parameter correction information needs to be weighted and integrated under test conditions with different pull-out speeds, pull-out angles, and clamping positions. The significance of weighted integration is to account for the differences in influence between different working conditions and to prevent the correction result of any one working condition from excessively dominating the final parameters. For example, if the deviation index is large under a certain working condition, it should be given a higher weight, while working conditions with smaller deviations should be given a lower weight. The weighting coefficient can be calculated by taking the average deviation of each working condition as a percentage of the total deviation. For example, if the average deviation of a certain working condition accounts for 40% of the total deviation of all working conditions, then the weight of that working condition can be set to 0.4. Using the weighted information of each working condition, the corrected friction coefficient and adhesive strength applicable to all working conditions can be obtained. A new three-dimensional finite element mesh is then rebuilt using these corrected parameters, and stress field simulation calculations are performed to obtain a new pull-out force prediction curve.

[0099] To determine whether the above correction has achieved sufficient accuracy, it is necessary to compare the deviation between the new pull-out force prediction curve and the previous prediction curve. This involves recalculating the deviation index between the new prediction curve and the actual test data, and comparing it with the deviation index from the previous round. When the new prediction curve meets the preset convergence condition in terms of the deviation index, such as the new average deviation being less than 80% of the previous average deviation, and the stress change rate of adjacent finite element units in the adhesive layer region remaining within a stable range over several consecutive iterations, the correction result can be considered to have reached a stable state. At this point, the parameter correction process is terminated, and the corrected pull-out force prediction curve is output. The pull-out force value at the peak point of this prediction curve, i.e., the maximum value of the curve, is extracted as the maximum predicted pull-out force value. Then, combined with the stress distribution at the corresponding peak position in the adhesive layer region, the predicted result of the microneedle adhesion strength can be finally determined.

[0100] A second embodiment of this application provides an electronic device, the electronic device comprising: processor; The memory is used to store a program, which, when read and executed by the processor, executes a method for predicting and calculating the fixation strength of microneedles based on virtual simulation and mechanical modeling, as provided in the first embodiment of this application.

[0101] The third embodiment of this application provides a computer-readable storage medium storing a computer program thereon. When the program is executed by a processor, it executes a microneedle fixation strength prediction and calculation method based on virtual simulation and mechanical modeling provided in the first embodiment of this application.

[0102] Although this application discloses preferred embodiments as described above, it is not intended to limit this application. Any person skilled in the art can make possible changes and modifications without departing from the spirit and scope of this application. Therefore, the scope of protection of this application should be determined by the scope defined in the claims of this application.

Claims

1. A method for predicting and calculating the fixation strength of microneedles based on virtual simulation and mechanical modeling, characterized in that, include: Geometric data of the connection between the microneedle tip and the handle are obtained by using a 3D scanning device. The insertion depth, contact area, mating gap and adhesive layer thickness of the needle tip are extracted to establish a 3D geometric structure that includes the characteristics of the connection interface. Based on the contact area and adhesive layer thickness in the three-dimensional geometry, the elastic modulus, yield strength and fracture strength of the needle material, handle material and adhesive layer material are obtained, and the friction coefficient and adhesion strength between the needle and handle are set as contact characteristic parameters. A three-dimensional finite element mesh is established at the connection between the needle tip and the handle based on the contact characteristic parameters. Different withdrawal speeds, withdrawal angles, and clamping positions are set to form multiple test conditions. Mesh refinement processing is performed in the connection interface area to improve calculation accuracy. Stress field simulation calculations are performed on each test condition based on the encrypted three-dimensional finite element mesh to obtain stress distribution data and displacement data of the connection parts. The maximum stress value and its spatial location under each test condition are extracted. The failure type is determined by comparing the maximum stress value with the yield strength and bonding strength and a failure location identifier is generated. Based on the stress distribution data, the stress at the connection interface is integrated to obtain the initial prediction curve of the pull-out force changing with the pull-out displacement under each test condition. The pull-out force data of the actual microneedle pull-out test is collected, and the deviation between the initial prediction curve and the pull-out force data is calculated. By adjusting the friction coefficient and bonding strength, the three-dimensional finite element mesh establishment and stress field simulation calculation process is re-executed until the deviation meets the preset accuracy requirements. The corrected pull-out force prediction curve is output, and the microneedle fixation strength prediction result is calculated based on the maximum pull-out force prediction value in the pull-out force prediction curve.

2. The method for predicting and calculating the fixation strength of microneedles based on virtual simulation and mechanical modeling according to claim 1, characterized in that, The process involves establishing a three-dimensional finite element mesh for the connection between the needle tip and the handle based on the contact characteristic parameters, setting different withdrawal speeds, withdrawal angles, and clamping positions to form multiple test conditions, and performing mesh refinement processing in the connection interface area to improve calculation accuracy, including: The contact characteristic parameters are used to divide the connection area of ​​the needle, adhesive layer and handle into regions. The needle insertion area, adhesive layer area and handle constraint area are defined as independent mesh domains. In each mesh domain, element types and node constraint relationships corresponding to material properties are assigned to form an initial three-dimensional finite element mesh with complete contact definition. A standardized pull-out load is applied to the initial three-dimensional finite element mesh, and a mechanical response analysis is performed in the adhesive layer region to obtain the nodal stress distribution and element strain gradient. The stress concentration area and transition area in the adhesive layer region are determined based on the stress gradient variation amplitude, and the spatial coordinates of the stress concentration area are defined as the interface stress feature set. Based on the interface stress feature set, the stress concentration area is proportionally densified and divided. The size of the regional unit is adjusted to a predetermined ratio of the average size of the original regional unit. Linearly gradient unit sizes are set in the adjacent transition area along the thickness direction and the tangential direction to maintain the continuity of nodes and the smoothness of interface stress transmission, thereby forming a high-precision three-dimensional finite element mesh after local densification. Based on the high-precision three-dimensional finite element mesh, a set of test conditions is constructed. The pull-out speed, pull-out angle, and clamping position are used as the test condition variables. Multiple representative test conditions are generated according to the orthogonal combination principle. The loading conditions of each test condition are then associated with the high-precision three-dimensional finite element mesh to form a multi-condition finite element mesh set with complete boundary condition definitions. Before loading, an interface continuity analysis is performed on the multi-condition finite element mesh set to determine whether the stress change of the nodes in the adhesive layer area meets the stability threshold. When the stress change of the local element under any test condition exceeds the predetermined stability limit, the local element of the region is re-divided and a secondary densification process is performed to generate the final finite element mesh set after stability correction. The final finite element mesh set is then used as the densified three-dimensional finite element mesh.

3. The method for predicting and calculating the fixation strength of microneedles based on virtual simulation and mechanical modeling according to claim 1, characterized in that, The process involves performing stress field simulation calculations on each test condition based on the encrypted three-dimensional finite element mesh to obtain stress distribution data and displacement data at the connection points. The maximum stress value and its spatial location under each test condition are extracted. The failure type is determined by comparing the maximum stress value with the yield strength and bond strength, and a failure location identifier is generated. This includes: On the encrypted three-dimensional finite element mesh, corresponding boundary conditions and load constraints are applied to each test condition to obtain stress distribution data and displacement data of the needle insertion area, adhesive layer area and handle constraint area, forming a stress distribution dataset for failure identification. Based on the stress information of each finite element in the stress distribution dataset, the maximum stress value and its spatial location of each finite element under the corresponding test condition are determined, forming a set of maximum stress values ​​containing the maximum stress values ​​of each element. Based on the set of maximum stress values, each maximum stress value is compared with the yield strength and bond strength to identify finite element elements that meet the conditions of exceeding the yield strength or bond strength. Based on the spatial location of the finite element elements, the finite element elements that are determined to be failed are grouped into a group of failure regions with spatial continuity to form preliminary failure region distribution information. Based on the continuity characteristics and spatial relationship of each failure region group under different test conditions, the failure type of each failure region group is determined, and a corresponding failure location identifier is generated for each failure region group.

4. The method for predicting and calculating the fixation strength of microneedles based on virtual simulation and mechanical modeling according to claim 3, characterized in that, The method involves comparing each maximum stress value with the yield strength and bond strength based on the set of maximum stress values ​​to identify finite element elements that meet the conditions of exceeding the yield strength or bond strength. Furthermore, based on the spatial location of the finite element elements, those deemed to have failed are grouped into a spatially continuous group of failure regions, forming preliminary failure region distribution information, including: The adjacency relationship between finite element elements is determined based on the spatial coordinates of each finite element element in the set of maximum stress values, and adjacency information used to represent the spatial connectivity of finite element elements is obtained. Based on the adjacency information, finite element elements whose maximum stress values ​​satisfy the conditions of exceeding the yield strength or exceeding the bond strength are divided into multiple candidate failure region groups according to spatial connectivity, thus obtaining candidate failure region data containing each candidate failure region group. The continuity of the candidate failure region data is checked. Regions consisting of only a single finite element and not spatially connected to other finite element elements are eliminated. Candidate failure region groups with connected boundaries and whose maximum stress value changes meet the predetermined continuity conditions are merged to form a set of failure region groups after continuity correction. Based on the spatial location of each failure region group in the failure region group set, the finite element elements belonging to the same failure region group are recorded with their corresponding maximum stress values ​​to generate preliminary failure region distribution information to characterize the relationship between the spatial distribution of failure regions and stress distribution.

5. The method for predicting and calculating the fixation strength of microneedles based on virtual simulation and mechanical modeling according to claim 1, characterized in that, The process of re-executing the three-dimensional finite element mesh establishment and stress field simulation calculation by adjusting the friction coefficient and adhesion strength until the deviation meets the preset accuracy requirements is completed. A corrected pull-out force prediction curve is output, and the microneedle adhesion strength prediction result is calculated based on the maximum pull-out force prediction value in the curve, including: Under each test condition, the initial pull-out force prediction curve is compared with the pull-out force data obtained from the actual pull-out test to obtain error data that reflects the distribution of prediction error, and the error data is used to determine the deviation index that characterizes the overall deviation. The stress distribution in the adhesive layer area is analyzed based on the deviation index. The correction direction and correction magnitude of the friction coefficient and adhesive strength are determined based on the offset relationship between the simulated stress peak position and the actual failure position, so as to obtain parameter correction information for updating the simulation input conditions. The parameter correction information is weighted and integrated under test conditions with different pull-out speeds, pull-out angles and clamping positions to obtain the corrected friction coefficient and bonding strength applicable under multiple conditions. The corrected parameters are then used to re-execute the three-dimensional finite element mesh establishment and stress field simulation calculation to generate a new pull-out force prediction curve. The new pull-out force prediction curve is compared with the previous prediction curve. When the new pull-out force prediction curve meets the preset convergence condition in terms of deviation index and maintains a stable stress change trend in the adhesive layer area, the corrected pull-out force prediction curve is output, and the microneedle fixation strength prediction result is determined by the maximum pull-out force prediction value in the corrected pull-out force prediction curve.

Images