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

Abstract

The application discloses a microneedle fixation strength prediction and calculation method based on virtual simulation and mechanical modeling. The method obtains the geometric data of the connection part of the microneedle needle and handle, and establishes a three-dimensional geometric structure containing the connection interface characteristics. Based on the geometric structure, the elastic modulus, yield strength and fracture strength of the material are determined, and the friction coefficient and adhesive strength are set as the contact characteristic parameters. The finite element grid is constructed and the test working conditions of different pull-out speeds, angles and clamping positions are set. Through simulation calculation, the stress and displacement distribution is obtained, the maximum stress and its position are extracted, and the failure type is judged, and then the interface stress integral is obtained to obtain the pull-out force prediction curve. Combined with the actual pull-out test data, the deviation is calculated, the friction coefficient and the adhesive strength are adjusted repeatedly until the deviation meets the preset precision, and the corrected pull-out force prediction curve and the maximum pull-out force prediction value are output, so that the rapid and non-destructive calculation and evaluation of the microneedle fixation strength are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer-aided design and finite element data analysis, and particularly relates to a microneedle anchoring strength prediction calculation method based on virtual simulation and mechanical modeling. BACKGROUND

[0002] With the popularization of cosmetic medical technology, microneedle puncture equipment is widely used in the fields of skin beauty and drug transdermal delivery. In the prior art, the evaluation of the connection strength between the microneedle needle and the handle usually depends on a physical pull test, which measures the maximum force required during the pulling-out process by clamping the needle and applying a constant pulling force to judge the anchoring reliability of the needle. Some researchers try to use finite element analysis method to simulate the pulling-out process to study the influence of different materials or structural parameters on the anchoring performance, but they still mainly rely on idealized geometric models and empirical parameters, and cannot truly reflect the complexity of the actual assembly interface of the microneedle.

[0003] However, the existing test methods have many limitations in operation. Due to the small size of the cosmetic microneedle needle and the uneven thickness of the connecting adhesive layer, it is difficult to clamp the needle with a traditional tensile testing machine, and the test repeatability is poor. At the same time, the lack of real geometry and interface contact parameters in finite element analysis leads to a large deviation between simulation prediction and actual measurement. In addition, there is currently a lack of effective parameter correction mechanism, which cannot realize the adaptive optimization of the prediction model under different pulling-out speeds, angles and clamping conditions, affecting the accuracy and comparability of the anchoring strength evaluation.

[0004] In view of the above deficiencies in the prior art, it is necessary to propose a microneedle anchoring strength prediction calculation method combining real geometry reconstruction and mechanical simulation to realize non-destructive and high-precision strength evaluation. SUMMARY

[0005] The present application provides a microneedle anchoring strength prediction calculation method based on virtual simulation and mechanical modeling to realize non-destructive and high-precision strength evaluation.

[0006] The present application provides a microneedle anchoring strength prediction calculation method based on virtual simulation and mechanical modeling, comprising:

[0007] Obtain the geometric data of the connection part of the microneedle needle and the handle by a three-dimensional scanning device, extract the needle insertion depth, contact area, fitting gap and adhesive layer thickness, and establish a three-dimensional geometric structure containing the connection interface characteristics;

[0008] Based on the contact area and the adhesive layer thickness in the three-dimensional geometric structure, 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 the handle as the contact characteristic parameters;

[0009] According to the contact characteristic parameter, a three-dimensional finite element grid of a needle and a handle connection part is established, different pull-out speeds, pull-out angles and clamping positions are set to form a plurality of test working conditions, and grid encryption processing is performed in the connection interface area to improve the calculation precision;

[0010] According to the three-dimensional finite element grid after the encryption processing, stress field simulation calculation is performed on each test working condition, stress distribution data and displacement data of the connection part are obtained, the maximum stress value and its spatial position under each test working condition are extracted, and the failure type is judged by comparing the maximum stress value with the yield strength and the bonding strength, and a failure position mark is generated;

[0011] Based on the stress distribution data, integral operation is performed on the stress of the connection interface to obtain an initial prediction curve of the pull-out force changing with the pull-out displacement under each test working condition, pull-out force data of an actual microneedle pull-out test is collected, and a deviation between the initial prediction curve and the pull-out force data is calculated.

[0012] The three-dimensional finite element grid establishment and stress field simulation calculation process are re-executed by adjusting the friction coefficient and the bonding strength until the deviation meets the preset precision requirement, a corrected pull-out force prediction curve is output, and a microneedle fixation strength prediction result is calculated based on a maximum pull-out force prediction value in the pull-out force prediction curve.

[0013] The beneficial effects of the technical scheme provided in the application include:

[0014] (1) The application can obtain a high-precision fixation strength prediction result based on a virtual simulation model established based on real geometry and mechanical parameters, without the need of physically pulling and damaging the microneedle sample, thereby avoiding the problems of difficulty in clamping short needle heads, sample waste and large test error in the traditional tensile testing machine, and significantly improving the repeatability and standardization level of the detection process. (2) The contact area, fitting gap and adhesive layer thickness of the needle and handle connection part are extracted by three-dimensional scanning, and combined with the material elastic modulus, yield strength and bonding characteristic parameters, a finite element model conforming to the actual assembly state can be constructed. The method performs grid encryption in the interface area, so that the stress concentration and debonding evolution process are described with high resolution, thereby greatly improving the accuracy of the fixation strength simulation prediction. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 is a flowchart of a microneedle fixation strength prediction calculation method based on virtual simulation and mechanical modeling provided by the first embodiment of the application. DETAILED DESCRIPTION

[0016] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application. However, the present application can be practiced without the specific details. In other instances, well-known methods have not been described in detail in order not to unnecessarily obscure aspects of the present application.

[0017] The first embodiment of the present application provides a microneedle fixation strength prediction calculation method based on virtual simulation and mechanical modeling. Please refer to Figure 1 , which is a flowchart of the first embodiment of the present application. The following will be described in detail Figure 1 The first embodiment of the present application provides a microneedle fixation strength prediction calculation method based on virtual simulation and mechanical modeling.

[0018] Step S101: Obtain the geometric data of the connection part of the microneedle needle and handle by a three-dimensional scanning device, extract the needle insertion depth, contact area, fitting gap and adhesive layer thickness, and establish a three-dimensional geometric structure containing the connection interface characteristics.

[0019] Step S101 aims to obtain the real geometric information of the microneedle needle and handle in the actual assembly state without damaging the microneedle assembly to be tested, and to establish a three-dimensional geometric structure model that can be used for subsequent mechanical simulation analysis.

[0020] First, in the geometric data acquisition stage, a three-dimensional scanning device with micron-level resolution is used to scan and measure the connection area of the microneedle needle and handle. The three-dimensional scanning device can be a laser confocal scanning microscope, a white light interference surface scanner, or an industrial CT scanning system. For high-reflective materials such as metal needles or high-gloss polymers, a thin and thin diffuse reflection developing coating can be uniformly sprayed on the surface before scanning to reduce the interference of mirror reflection on the scanning quality, so that the scanning device can obtain stable surface reflection signals. When scanning, the same connection area to be tested should be collected from multiple angles, including at least the forward, inclined and radial directions, to ensure that the needle insertion area, the adhesive area and the inner hole wall area of the handle are completely covered. The original data after scanning is point cloud data or voxel density data, which represents the actual shape, position and mutual fitting relationship of the needle, adhesive layer and handle in space. In order to ensure the spatial consistency of subsequent analysis, the data from different scanning angles need to be aligned in coordinates. The commonly used method is to fit the common features of the overlapping areas, so that the point cloud data of each batch is coincided in the same coordinate system, so as to obtain a complete and continuous three-dimensional data set of the connection area.

[0021] After obtaining the integrated 3D dataset, it is necessary to extract geometric feature parameters related to the anchoring strength from the dataset. First, the needle insertion depth needs to be determined. The needle insertion depth refers to the actual embedded length of the needle in the handle insertion hole. To determine this length, the axis of the needle can be identified in the 3D data, and the bottom end of the needle (i.e., the position where the needle is in the deepest contact with the glue layer) and the handle insertion hole entrance position are marked. Along the axial direction of the needle, the distance between the two positions is measured, which is the needle insertion depth. If the needle is not inserted completely vertically into the handle, but has a certain inclination angle, the center axis direction of the needle itself should be taken as the reference direction for measurement, rather than the shape of the handle, to avoid measurement errors introduced by the inclined assembly. The reference direction can be determined by reading the center line direction vector obtained by fitting the needle shape, which is a normal 3D geometric post-processing operation.

[0022] The contact area needs to be identified from the scanned data as the area where the needle and the handle are in direct contact after assembly. The contact area refers to the area where the distance between the outer surface of the needle and the inner surface of the handle is close to zero or is completely filled by the glue layer. To identify this area, the needle outer surface and the handle inner surface can be compared point by point, and for each surface position, the shortest distance to the corresponding mating surface is calculated. When the distance is less than a predetermined contact judgment threshold, the area is judged to be an actual contact area. The contact judgment threshold can be determined according to the resolution of the scanning device, for example, when the scanning resolution is in the order of 2 microns, a common contact judgment threshold can be selected around 5 microns, i.e., if the closest distance between the two surfaces is less than about 5 microns, it is considered to be an effective contact area. By collecting all the local surface segments that are judged to be effective contact areas, the effective contact area between the needle and the handle can be obtained. The area of the effective contact area is defined as the contact area. The contact area reflects the force transmission range of the needle in the handle and is an important input for subsequent anchoring strength calculation.

[0023] The fit clearance refers to the space interval between the outer surface of the needle head and the inner hole surface of the handle at the position where they are not completely attached. Its physical meaning is the size of the gap that still exists between the two structures after assembly. The fit clearance needs to be obtained by distance measurement on the non-contact area. A series of distance values can be obtained by distance calculation between multiple representative measurement points on the outer surface of the needle head and the corresponding area on the inner surface of the handle. The average fit clearance can be obtained by statistical analysis of these values, and the maximum and minimum clearances can also be given as needed to reflect the uniformity of the clearance distribution. For example, if the clearance in most areas of a batch of microneedle assemblies is about 8 microns, and the clearance in individual areas reaches more than 20 microns, it indicates that the assembly has local uneven attachment, which will directly affect the stress concentration at the interface and the shape of the final pull-out force curve. Therefore, the fit clearance is not only a geometric measurement value, but also a boundary condition for subsequent interface contact modeling.

[0024] The thickness of the adhesive layer is used to describe the actual thickness distribution of the adhesive material between the needle head and the handle in the assembled state. The so-called adhesive layer refers to the adhesive medium used to achieve the fixed connection between the needle head and the handle, which is usually a medical adhesive or a cured resin. The thickness of the adhesive layer can be obtained in two ways. If industrial CT scanning is used, since the density contrast of different materials in scanning imaging is different, the needle body, handle body and intermediate adhesive layer can be distinguished in voxel data, and the thickness of the adhesive layer can be measured along the normal direction of the connecting interface to obtain the spatial distribution of the thickness. If an optical scanning device is used without distinguishing the density information of different materials, the thickness of the adhesive layer can be estimated by comparing the theoretical fit clearance of the components before and after assembly. Specifically, the nominal gap between the handle hole and the needle head shape in the theoretical assembly state can be recorded first, and then compared with the surface distance at the same position in the actual assembly state. The difference between the two can be regarded as the actual thickness of the adhesive layer at that position. This thickness can be repeatedly measured at multiple sampling positions along the connecting interface to obtain the average adhesive layer thickness and the distribution range of the adhesive layer thickness.

[0025] After the above-mentioned insertion depth, contact area, fit clearance and adhesive layer thickness are extracted, these measurement results need to be used for the reconstruction of the three-dimensional geometry. The so-called three-dimensional geometry refers to a three-dimensional entity model containing the relative positional relationship, envelope shape and interface form information of the needle, handle and the adhesive layer between them in space. The three-dimensional geometry is not only the idealized shape of the needle and handle, but also should maximize the actual assembly characteristics, such as the actual eccentric position of the needle inside the handle, the uneven thickness of the adhesive layer in the circumferential direction, the local chamfer or burr shape of the handle aperture, etc. This is an important point of difference from traditional ideal modeling. The specific reconstruction process can be completed in commercial three-dimensional modeling or simulation pre-processing software, for example, the aforementioned point cloud data can be imported into the modeling software, and the scattered point form data is converted into editable closed surface through surface fitting, and then the corresponding entity structure is generated based on the closed surface. The needle, handle body and adhesive layer should be defined as independent geometric bodies or parts, so that they can be given different material properties and different contact conditions in the subsequent finite element calculation stage. If the needle has a significant assembly inclination or eccentricity, the eccentric relationship should be retained in the model, and it should not be forcibly corrected to a coaxial state during modeling, because this small eccentricity will significantly affect the position and size of the local stress peak on the interface in the mechanical simulation.

[0026] For example, in a certain actual sample, the scanning results show that the total length of the microneedle needle is about 0.8 millimeters, the insertion depth of the needle into the handle is about 0.45 millimeters, the effective contact area between the needle and the handle is about 0.12 square millimeters, the average fit clearance is about 8 microns, and the average thickness of the adhesive layer is about 15 microns, and the adhesive layer is slightly thicker on one side and slightly thinner on the other side, showing obvious asymmetric distribution. After reflecting the above information to the three-dimensional geometry, a combined model containing the needle entity, the handle insertion hole entity and the uneven adhesive layer entity can be obtained, and the model clearly records the eccentric direction and eccentric amount of the needle in the handle. This combined model can be directly used as the basic geometry input in the subsequent finite element modeling, without the need for idealization and simplification, so that the stress distribution and displacement distribution obtained in the simulation process are more close to the stress state of the actual assembly.

[0027] Step S102: Based on the contact area and the adhesive layer thickness in the three-dimensional geometry, the elastic modulus, yield strength and breaking strength of the needle material, handle material and adhesive layer material are obtained, and the friction coefficient and adhesion strength between the needle and the handle are set as the contact characteristic parameters.

[0028] The content of step S102 is the key link of establishing the basis of mechanical calculation in the whole method, and the purpose is to provide complete and quantifiable physical parameter input for subsequent finite element simulation after obtaining the real three-dimensional geometric structure. This step determines the mechanical performance indicators of the material and the interface contact characteristic parameters, so that the virtual model can truly reflect the actual stress and deformation behavior between the microneedle needle and the handle, thereby ensuring that the simulation calculation result has physical rationality and prediction accuracy. In order to achieve this purpose, this step needs to determine the mechanical properties of the needle material, the handle material and the adhesive layer material respectively, and at the same time, the interface interaction between the needle and the handle needs to be parameterized, including the quantitative setting of the friction coefficient and the bonding strength.

[0029] In the process of determining the performance of the needle material, the specific material type of the needle should be identified first. The needle of the cosmetic microneedle is usually made of medical stainless steel, titanium alloy or hard polymer, and the elastic response and failure behavior of different materials are significantly different. The elastic modulus is a proportional constant that describes the stress and strain relationship of the material in the elastic deformation stage, which can usually be obtained by standard tensile test. In specific operation, a fine rod sample of about one millimeter in diameter can be taken, and a gradually increasing tensile force is applied and the elongation is measured. When the strain is in the linear range, the ratio of stress to strain is the elastic modulus. For example, the typical elastic modulus of medical stainless steel 304 material is about one hundred and ninety gigapascals, and the elastic modulus of polyether ether ketone (PEEK) material is about three point five gigapascals. According to the experimental data of the needle material under corresponding temperature and humidity conditions, the value can be directly used or modified to input the simulation model. The yield strength is used to characterize the critical stress value of the material from the elastic stage to the plastic stage, which 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 yield stage, and the yield strength can be identified by the inflection point of the stress-strain curve. The breaking strength is the maximum stress that the material can withstand when it is pulled or sheared to failure, which is usually the highest point of the stress-strain curve, and should be measured in the standard sample failure experiment. These three indicators together define the full process response characteristics of the needle material from linear elasticity, plastic yield to final fracture during drawing.

[0030] The method of obtaining the mechanical parameters of the handle material is similar to that of the needle, but since the handle is usually made of thermoplastic materials such as polycarbonate (PC), polypropylene (PP) or nylon (PA), the temperature should be controlled during testing to avoid the influence of thermal softening on the measurement results. For thermoplastic materials, uniaxial tensile testing in a constant temperature environment can be used, and the tensile speed should be consistent with the order of magnitude of the subsequent simulation pulling speed to ensure the comparability of the strain rate effect. The measured elastic modulus is usually 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, the parameters should be measured in different fiber directions to obtain the anisotropic material parameters.

[0031] The mechanical parameters of the adhesive layer are particularly important because this layer not only plays a bonding role in the connection of the needle and the handle, but also is the main channel for force transmission at the interface. The adhesive layer is usually a medical-grade epoxy resin or ultraviolet-cured acrylic glue. Its elastic modulus can be obtained by film tensile test or dynamic mechanical analysis (DMA). For the case where the thickness of the adhesive layer is less than twenty microns, a planar adhesive layer sample of the same formula can be prepared, and after curing, its stress-strain relationship is tested on a tensile machine to calculate the elastic modulus and yield strength of the material. The breaking strength can be determined by shearing test of a standard adhesive sample, i.e. sandwiching a layer of adhesive of the same thickness between two substrates of the same material, and the maximum stress value when the adhesive layer is destroyed by uniaxial tension is the breaking strength. The material data sheet provided by the manufacturer can also be referred to, and then corrected according to the curing conditions and the actual thickness of the adhesive layer.

[0032] After determining the basic performance parameters of the above three types of materials, the physical characteristics of the interface contact between the needle and the handle need to be established. Friction coefficient and adhesive strength are the most core parameters in the interface contact behavior. The friction coefficient is defined as the ratio of the tangential resistance between two surfaces to the normal pressure, which is used to describe the resistance when the interface slides during the pulling-out process. Since there is usually some mechanical fit or micro-roughness engagement between the needle and the inner wall of the handle, the friction coefficient is not only determined by the material type, but also related to the surface roughness, surface treatment state and the presence of residual adhesive layer. In specific determination, a micro-friction tester can be used for plane friction test, and the needle material and handle material sample are prepared into smooth contact surfaces, and the sliding friction force is measured under constant normal load, and then the average value is calculated to obtain the friction coefficient. For example, for the contact pair of untreated stainless steel and polypropylene, the friction coefficient is usually between zero point two and zero point three, while if a thin adhesive layer is coated on the surface, the friction coefficient can rise to more than zero point five.

[0033] The adhesive strength is used to describe the ability of the interface to resist separation during pulling, and its physical meaning is the maximum unit area tensile stress that the interface can withstand before separation. This parameter can be obtained by standardized pull-off test. During the test, a cylindrical adhesive sample similar to the actual microneedle assembly can be prepared, and the needle material cylinder and the handle material sleeve are bonded and cured through the same adhesive layer, an axial pulling load is applied on the tensile testing machine, and the maximum pulling force when the adhesive layer is destroyed is recorded, and then the adhesive strength can be obtained by dividing the adhesive area. For example, if the adhesive area of the sample is zero point one square millimeter, and the pulling force at the time of destruction is zero point three newtons, then the adhesive strength is three megapascals.

[0034] To ensure that the simulation input parameters are consistent with the actual state, the present step needs to map and correlate the above-mentioned physical parameters according to the contact area and the thickness of the adhesive layer extracted in the three-dimensional geometric structure. The greater the contact area, the more significant the influence of the interfacial bonding strength on the overall pull-out force; the thicker the adhesive layer, the longer the interfacial stress transfer path, and the stronger the local stress concentration effect. When modeling, these parameters should be input into the contact definition module of the simulation software, and the corresponding properties should be assigned to different contact areas. For example, for the edge contact area, a slightly lower bonding strength value can be set to reflect the early peeling tendency caused by stress concentration, while for the central contact area, the average value is maintained.

[0035] After determining all the parameters, a parameter verification mechanism should also be established to prevent the input values from deviating from the physical reality. Generally, the model can be simplified to a local interface unit through a small-scale simulation check method, and the sensitivity analysis of the pull-out force response under different friction coefficients and bonding strengths is performed. When the trend of the calculation results is consistent with the experimental expectation, it can be confirmed that the parameter setting is reasonable. For example, when the friction coefficient increases, the peak value of the pull-out force should rise, and when the bonding strength decreases significantly, the model should exhibit the characteristics of early separation of the interface.

[0036] Therefore, step S102 acquires the elastic modulus, yield strength and fracture strength of the needle, handle and adhesive layer through the system, and accurately defines the friction coefficient and bonding strength in combination with experimental or literature data, thereby constituting a complete contact characteristic parameter system. These parameters can accurately describe the interface bonding, sliding and failure behavior of the micro-needle connection part during the pulling-out process, thereby providing real and reliable physical inputs for subsequent finite element grid establishment and stress field simulation calculation.

[0037] Step S103: According to the contact characteristic parameters, a three-dimensional finite element grid of the connection part between the needle and the handle is established, different pulling-out speeds, pulling-out angles and clamping positions are set to form multiple test working conditions, and grid densification processing is performed in the connection interface area to improve the calculation accuracy.

[0038] Step S103 is the process of converting the geometric structure information and material parameters obtained in the previous two steps into a finite element model that can be subjected to numerical analysis, and is one of the core technical links in the implementation of the entire method. The purpose of this step is to establish a three-dimensional finite element grid model that can truly reflect the stress state of the connection part between the micro-needle needle and the handle, and to construct multiple different pulling-out working conditions based on this model to simulate the stress behavior under real test conditions. This step needs to include five levels of technical operations: geometric modeling, mesh division, boundary condition definition, loading condition setting and grid densification control, each of which directly affects the simulation accuracy and the reliability of the final prediction results.

[0039] Before establishing the finite element mesh, the three-dimensional geometry obtained in step S101 should be imported into the finite element pre-processing software. Common finite element modeling platforms include ANSYS, etc. The choice of software depends on the enterprise'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 there is no geometric gap or overlap at the interface. If the surface reconstructed from the scan data has small discontinuities, it can be repaired by surface reconstruction or local fairing algorithms to avoid meshing failure. At this time, the needle entity, the adhesive layer entity, and the handle entity should share the contact interface, and the geometric position of the interface surface node must strictly correspond to the contact area obtained in the original scan. For the adhesive layer, the thickness is usually only a few microns to a few tens of microns, so its geometric size is small relative to the overall model. In finite element modeling, the thickness of the adhesive layer should still have a true scale in the model so that the interface stress distribution and peeling behavior can be accurately captured during simulation.

[0040] After completing the geometry confirmation, enter the meshing stage. The so-called finite element mesh is to discretize the continuous three-dimensional geometry into a finite set composed of nodes and elements, so that the computer can solve the physical quantities such as stress, strain, and displacement for each element. The material of the needle and handle is usually isotropic metal or plastic, so tetrahedral or hexahedral solid elements can be used for meshing. For the needle tip area or the handle thin-walled area, adaptive tetrahedral elements can be used to avoid element shape distortion. The adhesive layer is thin and has a stress gradient in the thickness direction, so it is appropriate to use a multi-layer refined solid element or a laminated element structure to ensure that there are at least three layers of elements distributed in the thickness direction, so that the stress change of the adhesive interface along the thickness direction can be reflected. If the adhesive layer is too thin to divide the solid element, a geometric scale magnification factor can be set in the software, for example, the thickness of the adhesive layer is magnified by five times in the geometric layer for modeling, but the corresponding elastic modulus and adhesive strength values are reduced in the material parameters to maintain physical equivalence. For the contact area of the adhesive interface with the needle or handle, contact elements should be defined. Contact elements are a special form of elements that can transfer normal force and tangential friction when moving relative to each other. 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 the surface of the adhesive layer as the slave surface in the contact pair setting, and input the friction coefficient and adhesive strength parameters obtained in step S102.

[0041] Before setting up the load and boundary conditions, the simulation conditions need to be determined according to the physical process of the actual microneedle extraction experiment. The extraction speed refers to the linear motion speed of the needle head relative to the handle during the pulling-out process, the extraction angle refers to the angle between the pulling-out direction of the needle head and the axis of the handle, and the clamping position refers to the position of the handle or the needle head being fixed or constrained when the load is applied. In order to obtain the adhesion performance response under different working conditions, multiple calculation scenes need to be set in the model. For example, in a typical working condition combination, the extraction speed can be set to one hundred millimeters per minute, two hundred millimeters per minute and three hundred millimeters per minute respectively to simulate different operation speeds; the extraction angle can be set to zero degrees (i.e. completely axial extraction), fifteen degrees and thirty degrees to investigate the asymmetric distribution of interfacial stress when extracting obliquely; the clamping position can be fixed and constrained at the bottom of the handle, the middle of the handle or the tail end of the needle to analyze the influence of clamping method on the extraction force result. When the skilled person in the art establishes these working conditions in the software, the parameterized modeling function should be used to automatically generate the corresponding calculation model by changing the speed, angle and constraint position parameters, thereby improving the calculation efficiency and ensuring the geometric consistency between the working conditions.

[0042] In the definition of the load condition, the extraction speed is usually applied in the form of displacement load. A displacement boundary condition that changes linearly with time can be applied at the free end of the needle, and the direction is consistent with the extraction angle. For example, when extracting axially, a displacement control should be applied along the central axis of the needle, and when extracting at fifteen degrees, the displacement vector needs to be decomposed into axial and radial components according to the fifteen-degree direction. The size of the extraction speed should correspond to the speed range in the actual experiment, and the quasi-static condition should be maintained, that is, to ensure that the loading speed is small enough to avoid obvious influence of inertia effect on the result. The fixed end of the handle should be applied with complete constraint condition to prohibit its translational and rotational degrees of freedom in three directions to simulate the state that the handle is fixed by the clamp. In order to avoid rigid body motion of the overall model during numerical solving, the rotational degree of freedom of at least one node should be constrained.

[0043] In the aspect of mesh refinement, the element size of different regions should be controlled according to the stress concentration area and the complexity of the interface. The purpose of mesh refinement is to improve the accuracy of local stress calculation, especially at the edge of the adhesive layer, the insertion end of the needle, and the handle orifice, which are usually the areas with the highest stress gradient. The refinement strategy can be achieved by setting the local mesh size control or defining the "local refinement area" in the modeling software. For example, the element size of the adhesive layer region can be set to one-fifth to one-tenth of the adjacent region. If the geometric ratio of the needle and handle materials is significantly different, a local transition mesh should be used to avoid sudden changes in element shape. After mesh refinement, the entire model should be checked for quality to ensure that the element twist ratio, aspect ratio, and Jacobian matrix value are within the recommended reasonable range of the software, such as the minimum angle of the element being no less than twenty degrees and the aspect ratio being no more than five to one. Only when the mesh quality is qualified can the numerical stability of subsequent solving be ensured.

[0044] In this step, the input of contact characteristic parameters is an important link to ensure the physical reality of the model. The software usually provides a "friction-adhesion" composite contact model, which can consider the relationship between friction slip and adhesive bonding. The friction coefficient obtained in step S102 should be input in the contact setting, and the adhesion failure criterion should be enabled to automatically trigger separation when the adhesive layer reaches the adhesion strength. When defining the failure criterion, the standard based on the 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 input as the critical stress; if the energy release rate criterion is used, the fracture energy value per unit area should be input, which represents the energy required for complete peeling of the adhesive layer. For general cosmetic microneedle adhesive structures, the maximum tensile stress criterion can already meet the accuracy requirements.

[0045] After completing the above modeling, loading, and mesh division, the complete input file of each test condition should be saved in the software. For calculation verification, a typical condition can be selected for trial calculation to observe the distribution of stress field and displacement field during needle extraction. If the interface stress is too concentrated or there is unreasonable displacement mutation, the adhesive layer mesh density can be adjusted or the contact algorithm can be redefined. Through this verification step, it can be ensured that the model has good convergence and the calculation results are stable and reliable. Generally, the number of model nodes for a single condition is between one hundred thousand and three hundred thousand, which can ensure sufficient accuracy. If a higher density mesh is used, the calculation time will increase significantly but the accuracy will be limited, so the balance between accuracy and efficiency can be achieved according to the calculation resources.

[0046] In summary, step S103 realizes the construction and calculation preparation of the virtual simulation model by establishing a three-dimensional finite element grid model containing the needle, handle and adhesive layer, and combining multiple pull-out working conditions and contact parameter settings. 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 grid densification and multi-working condition loading strategy, thereby laying a reliable physical and numerical foundation for subsequent stress field simulation and pull-out force prediction.

[0047] Further, the three-dimensional finite element grid of the connection between the needle and the handle is established according to the contact characteristic parameters, different pull-out speeds, pull-out angles and clamping positions are set to form multiple test working conditions, and grid densification processing is performed in the connection interface area to improve the calculation accuracy, including:

[0048] The connection part of the needle, the adhesive layer and the handle is divided into regions using the contact characteristic parameters, the needle insertion area, the adhesive layer area and the handle constraint area are defined as independent grid division domains, and the element types and node constraint relationships corresponding to the material properties are assigned in each grid division domain to form an initial three-dimensional finite element grid with complete contact definition;

[0049] The initial three-dimensional finite element grid is subjected to a standardized pulling load, and a mechanical response analysis is performed in the adhesive layer area to obtain node stress distribution and element strain gradient. According to the stress gradient change amplitude, the stress concentration zone and the transition zone in the adhesive layer area are determined, and the spatial coordinates of the stress concentration zone are defined as the interface stress feature set;

[0050] According to the interface stress feature set, the stress concentration zone is proportionally and densely divided, the regional element size is adjusted to a predetermined proportion of the average element size of the original region, and linearly tapered element sizes are set in the adjacent transition zone in the thickness direction and the tangential direction to maintain node continuity and the smoothness of interface stress transmission, thereby forming a high-precision three-dimensional finite element grid after local densification;

[0051] Based on the high-precision three-dimensional finite element grid, a test working condition set is constructed, with pull-out speed, pull-out angle and clamping position as working condition variables, multiple representative test working conditions are generated according to the orthogonal combination principle, and the loading conditions of each test working condition and the high-precision three-dimensional finite element grid are established in a corresponding relationship to form a multi-working condition finite element grid set with complete boundary condition definition;

[0052] An interface continuity analysis is performed on the multi-working condition finite element mesh set before loading to determine whether the stress change of the nodes in the adhesive layer region meets the stability threshold value. When the local element stress change in any test working condition exceeds the predetermined stability limit value, the region is re-divided into local elements and secondary encryption processing is performed to generate a final finite element mesh set that has been stability corrected, and the final finite element mesh set is taken as the three-dimensional finite element mesh after encryption processing.

[0053] In the process of establishing a three-dimensional finite element mesh of the connecting part of the needle and the handle according to the contact characteristic parameters, it is necessary to first determine the object and scope of the finite element modeling. The key stress area of the microneedle system is located at the connecting part of the needle 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 refers to the part of the needle actually embedded in the handle, which is usually a conical or needle-shaped column structure in geometric shape; the adhesive layer area refers to the adhesive layer used to fix the needle and the handle, which is generally between 0.02 millimeters and 0.2 millimeters in thickness; the handle constraint area refers to the part that provides counter-supporting force, which is usually made of high-strength plastic or aluminum alloy. To ensure the accuracy of the finite element modeling, the three regions should be treated as independent mesh division domains to avoid non-physical stress concentration at the material transition interface.

[0054] After the region division is completed, appropriate element types and node connection relationships should be selected according to the mechanical properties of each structural material. For the needle insertion area, hexahedral elements are suitable due to its high rigidity and stress concentration near the root, which can provide high computational stability in the axial stress direction; for the adhesive layer area, thin layer elements are suitable due to the fact that the adhesive is usually a elastic-plastic or viscoelastic material, and its thickness is much smaller than its width, so that the nonlinear response of interface debonding and sliding can be accurately described; for the handle constraint area, tetrahedral or hexahedral elements can be selected according to the elastic modulus of the material to balance the calculation accuracy and resource consumption. The node constraint conditions should be set according to the actual assembly and test state, for example, the nodes at the bottom of the handle should be applied with full constraint conditions to prevent overall displacement, and the nodes at the outer end of the needle should be defined as loading nodes for applying the pull-out load. Through the above region division and node constraint definition, an initial three-dimensional finite element mesh with complete contact definition can be formed.

[0055] After obtaining the initial mesh, a standardization pull-off load analysis should be performed to extract the mechanical response characteristics. The so-called standardization pull-off load is the pulling force applied along the axial direction of the needle at a constant rate, usually set to 200 millimeters per minute in the virtual simulation environment to simulate the pull-off speed of the pulling force machine in the real test. During the loading process, the system calculates the stress (σ) and strain (ε) of each node, and further obtains the strain gradient, i.e. the rate of change of the strain within the element with respect to the spatial position. The strain gradient is an important basis for judging the degree of stress concentration. The gradient value can be calculated by numerical difference or finite volume approximation method, for example, the principal stress difference between adjacent element centers divided by the node distance can obtain the local stress gradient. If the gradient is significantly higher than twice the average value, it can be determined that there is a stress concentration phenomenon in this area. Extract and record the spatial coordinates of all stress concentration areas to form the so-called interface stress feature set, which is used to guide the subsequent encryption division.

[0056] After forming the interface stress feature set, proportional encryption division needs to be performed on the stress concentration area. Proportional encryption refers to dynamically adjusting the size of the mesh element according to the degree of local stress concentration, so that the unit size of the area with stronger stress concentration is smaller, so as to improve the simulation accuracy. For example, when the maximum stress of the stress concentration area is 3 times that of the adjacent area, the unit length of the stress concentration area can be adjusted to 1 / 3 of the original average size. This operation can be realized through the adaptive mesh division function in the finite element pre-processing software (such as ANSYS Workbench). In order to avoid the discontinuity of nodes caused by the sudden change of element size, a transition zone should be set outside the stress concentration area, and a linear gradient relationship of element size should be set along the thickness direction and tangential direction of the adhesive layer, so that the element size gradually increases from the concentration area to the outside. Linear gradient can be realized by specifying the gradient ratio (such as increasing the element size of each layer by 10%), so as to ensure that the force transmission path is continuous in geometry and smooth in mechanics. After the encryption processing is completed, a high-precision three-dimensional finite element mesh with local refinement and global continuity will be obtained.

[0057] After obtaining the high-precision mesh, it is necessary to construct a test condition set based on the mesh. The so-called test condition is the stress situation 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 head is pulled out, and three typical values of 0.1, 0.5 and 1.0 mm per second can be taken; the pull-out angle refers to the angle between the pulling direction and the center axis of the needle head, and three values of 0°, 15° and 30° can be taken; the clamping position refers to the position of the loading point relative to the length of the needle, and the positions of 1 / 3, 1 / 2 and 2 / 3 of the needle length can be selected. By using the orthogonal combination principle on the above parameters, 9 representative working conditions can be generated, so as to realize multi-variable coverage under a limited number of simulations. For each working condition, the corresponding boundary conditions need to be set, for example, in the working condition of the inclination angle, the vector form of the pulling direction should be defined, and the constraint nodes at the clamping position should be defined. Each test condition is in one-to-one correspondence with the high-precision finite element mesh, forming a multi-condition finite element mesh set with complete loading and boundary condition definition.

[0058] In order to ensure the stability of the encrypted mesh under each condition, it is necessary to perform interface continuity analysis before loading. Interface continuity refers to the smooth transmission of stress and displacement between adjacent elements. The method of judging continuity is to calculate the stress difference of adjacent nodes and compare it with the set threshold value. For example, when the principal stress difference of adjacent nodes in the bonding layer area is greater than 10%, it indicates that there is local instability or mismatch in element size at this position, and a local remeshing operation should be triggered. In actual implementation, an automatic detection script can be set in the finite element software, so that the system scans all node stress differences before analysis, and when the threshold value is exceeded, a secondary encryption process is performed in the corresponding area. Secondary encryption is usually limited to local areas, and the new element size is 0.8 times the original element size to avoid increasing the global calculation amount. After this process, the final finite element mesh set with stability correction can be obtained.

[0059] The final mesh set not only maintains the continuity of the nodes and the consistency of the thickness of the bonding layer in the geometric structure, but also realizes the smoothness of the stress transmission path in the mechanical performance. The final finite element mesh set after correction, that is, the encrypted three-dimensional finite element mesh, can be used for subsequent stress field simulation calculation steps. Its data structure includes node coordinates, element topology relationship, material property parameters and boundary condition definition, which can be directly recognized and called by the finite element solver.

[0060] The encrypted three-dimensional finite element mesh established by the above method can significantly improve the stress resolution and calculation stability at the micro-needle connection interface compared with the traditional fixed element division method.

[0061] Step S104: According to the stress field simulation calculation of the encrypted three-dimensional finite element grid for each test working condition, the stress distribution data and displacement data of the connecting part are obtained, the maximum stress value and its spatial position under each test working condition are extracted, and the failure type is judged by comparing the maximum stress value with the yield strength and the bonding strength, and the failure position mark is generated.

[0062] The content of step S104 is the core link from numerical solution to mechanical result analysis in the whole virtual simulation calculation process. The purpose is to obtain the stress and displacement distribution of the micro-needle needle head and handle connecting part under different working conditions through mechanical simulation calculation of the finite element model established in step S103, and identify the position and failure type of the interface where the damage may occur, so as to provide basic data for subsequent pull-out force prediction curve and anchorage strength calculation. This step must be executed under strict physical boundary conditions and material nonlinear conditions. The result of the calculation not only reflects the macro force-displacement relationship, but also truly reproduces the stress gradient distribution, strain concentration area and potential peeling or sliding behavior of the interface layer.

[0063] In specific implementation, the finite element grid model established in step S103 needs to be imported into the calculation module. Each model corresponds to a specific working condition parameter combination, including pull-out speed, pull-out angle and clamping position. The simulation calculation should select the steady-state nonlinear statics or quasi-static solution type, because the micro-needle pull-out process belongs to low-speed loading and the inertia effect can be ignored. Such solution requires considering 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 needle head, handle and adhesive layer respectively adopt the data such as elastic modulus, yield strength and fracture strength determined in the previous step; the friction coefficient and bonding strength are introduced as the contact properties of the interface contact part, so that the software can automatically determine whether the interface slips or separates in different loading stages. For adhesive materials such as adhesive layer, it is recommended to enable the "nonlinear bonding model" or "bond-slip model" in the calculation. This model can describe the mechanical response of the interface in the initial bonding stage, the sliding stage and the final debonding stage, ensuring that the simulation result can accurately reflect the real pull-out behavior.

[0064] Before simulation execution, reasonable convergence control conditions should be set. Due to the strong nonlinear effect of the adhesive interface, it is easy to appear non-convergence in numerical solution, therefore it is recommended to use step-by-step loading method. Specifically, the pull-out process can be divided into several sub-steps, and a small displacement increment is applied in each sub-step, for example, about zero point zero zero one millimeter of displacement change is applied in the direction of the needle head pull-out, so that the model can stably calculate the increment change of stress and strain in each step. In this way, the solution process can be smoothly converged, and the numerical oscillation caused by sudden debonding or sliding can be avoided.

[0065] 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 mechanical analysis. The stress field refers to the stress distribution at each position in the model under the force state, usually including three main types: normal stress, shear stress and equivalent stress. The equivalent stress is a comprehensive indicator that reflects the overall stress intensity of the material under multi-axial stress state. The commonly used calculation method is the von Mises criterion, which combines the principal stresses in different directions into a single value to determine whether the material has reached yield. For the adhesive layer region, more attention is paid to the distribution of normal tensile stress and tangential shear stress, as these two stresses represent the peeling and sliding tendencies of the adhesive layer, respectively, and are direct indicators of interface failure.

[0066] The displacement field reflects the movement of each point in the model relative to the initial position, and can show the overall deformation pattern of the interface during the needle extraction process. By observing the displacement vector distribution, it can be determined whether the extraction process is a pure pulling mode or accompanied by a certain angle of offset sliding, thereby verifying the rationality of the working condition setting.

[0067] After the stress field analysis is completed, the maximum stress value of the stress distribution under each test condition should be extracted. The maximum stress value refers to the stress peak value occurring at a certain point or region in the entire model, which usually occurs at geometric discontinuities, material interfaces or local regions with the highest grid density. When extracting, the "extreme value query" function of the finite element post-processing module can be used, and the adhesive layer region or the needle insertion end region can be selected as the monitoring range. The software will automatically output the maximum principal stress, maximum shear stress or maximum equivalent stress and their corresponding coordinate positions in this region. If multiple similar peaks are found, the most likely failure location should be determined based on the stress distribution gradient. Special attention should be paid to the edge of the adhesive layer, the vicinity of the needle port and the edge of the handle aperture, as these locations are often the first to fail due to structural discontinuity or stress concentration.

[0068] After obtaining the maximum stress value, it needs to be compared with the material yield strength and adhesive strength obtained in step S102 to determine the failure type of the model under the current loading condition. If the maximum stress value is less than the yield strength of the needle or handle material, but exceeds the adhesive 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 may undergo plastic deformation or fracture, which is material failure; if it approaches the critical value of both, it indicates that there is a 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, and different colors or symbols can be used to distinguish the failure regions, such as using red to indicate the interface debonding area and orange to indicate the plastic yield area, so as to intuitively identify the failure mode.

[0069] To ensure the results are repeatable, a numerical data file should also be outputted, recording the maximum stress value, corresponding coordinates, failure type and failure location under each test condition into a table. For practical engineering applications, these data can be used to analyze the influence of different pull-out speeds, angles or clamping methods on the failure mode. For example, when the pull-out angle increases, the calculation results often show that the shear stress at the edge of the adhesive layer increases significantly, while the normal tensile stress changes little, which indicates that oblique pulling is more likely to cause interfacial slip failure; when the pull-out speed increases, due to the strain rate effect, the equivalent stress of the adhesive layer may increase slightly, but the overall failure location remains unchanged. These rules can provide quantitative basis for the optimization of microneedle structure.

[0070] If the nonlinear contact model is enabled in the calculation model, the change of the interface bonding state with displacement can also be observed. Finite element software can usually output parameters such as "bonding state variable" or "contact separation ratio", whose values range from zero to one, with zero indicating complete bonding and one indicating complete separation. By analyzing the spatial distribution of these variables, the evolution process of interface peeling can be obtained, i.e. the extension path from local debonding to overall peeling. Those skilled in the art can determine the failure propagation direction and speed of the adhesive layer based on these results, so as to more accurately infer the stage at which the peak pull-out force occurs.

[0071] 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 about 12 MPa, slightly higher than the bonding strength of 11 MPa, and local peeling occurs at the interface; when the pull-out displacement further increases to 0.2 mm, the peeling area extends about one quarter of a circle in the circumferential direction, and the stress peak gradually shifts to the middle of the adhesive layer; when the pull-out displacement reaches 0.3 mm, most of the interface is debonded, and the pull-out force begins to decrease. Such simulation results clearly show the timing characteristics of interface failure, and also verify the reasonableness of the bonding parameter settings in the model.

[0072] Therefore, step S104 realizes the complete simulation of the stress mechanism of the microneedle connection part and the identification of the failure mode by solving the stress field and displacement field, extracting the maximum stress and determining the failure type. The stress distribution data, displacement data, maximum stress value and failure location identifier output by this step will be directly used as the core input data for the subsequent steps of pull-out force curve calculation and model correction.

[0073] Further, the stress field simulation calculation is performed for each test condition based on the encrypted three-dimensional finite element grid, the stress distribution data and displacement data of the connection part are obtained, the maximum stress value and its spatial position 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 the failure location identifier is generated, including:

[0074] The corresponding boundary conditions and load constraints are applied to each test condition on the three-dimensional finite element grid after encryption processing to obtain stress distribution data and displacement data of the needle insertion area, the adhesive layer area and the handle constraint area, forming a stress distribution data set for failure identification;

[0075] Based on the stress information of each finite element unit in the stress distribution data set, the maximum stress value and its spatial position of each finite element unit under the corresponding test condition are determined, forming a maximum stress value set containing the maximum stress value of each unit;

[0076] According to the maximum stress value set, each maximum stress value is compared with the yield strength and the bonding strength, the finite element units that meet the condition of exceeding the yield strength or exceeding the bonding strength are identified, and the finite element units determined to be failed are collected into failure region groups with spatial continuity according to the spatial position of the finite element units, forming preliminary failure region distribution information;

[0077] According to the continuity characteristics and spatial position relationship of each failure region group under different test conditions, the failure type of each failure region group is determined, and the corresponding failure position identifier is generated for each failure region group.

[0078] When simulating the stress field of each test condition on the three-dimensional finite element grid after encryption processing, the boundary conditions and load constraints consistent with the actual pulling process need to be applied first, so that the model can truly reflect the mechanical behavior of the needle insertion area, the adhesive layer area and the handle constraint area during the force process under each test condition. Boundary conditions refer to limiting the displacement or rotation freedom of some nodes, so that the simulation device is fixed in the experimental state by the clamp; load constraints are usually applied in the form of displacement loading or equivalent force loading in the pulling direction, which is used to simulate the actual test mechanical environment under different pulling speeds and pulling angles. According to the interface setting method of commonly used finite element software, the loading process can be defined as step loading, and the corresponding node displacement and node stress are recorded at each step. Through this process, stress distribution data and displacement data reflecting the stress condition of each finite element unit inside the model can be obtained, which constitutes a stress distribution data set for failure identification.

[0079] In the stress distribution dataset, each finite element unit corresponds to a set of data containing principal stress, shear stress and displacement in each direction. The first principal stress in the principal stress is usually taken as the key indicator to judge the risk of material stretching or debonding. In order to calculate the maximum stress value of each finite element unit, it is necessary to first query the principal stress value of the unit corresponding to several nodes under the working condition, and then select the maximum value from the principal stress values of these nodes. For example, if the principal stresses of the four nodes of a finite element unit are 12 MPa, 15 MPa, 14 MPa and 9 MPa, the maximum stress value of the unit is 15 MPa. The maximum stress value set is formed by storing the maximum stress value of all units and their spatial positions together. The spatial position in the maximum stress value set can be determined by the geometric center coordinates of the unit provided by the finite element software, and is usually described in a three-dimensional rectangular coordinate system.

[0080] After obtaining the maximum stress value set, it is necessary to judge whether each finite element unit fails. The yield strength is used to judge whether the material is in the tensile or compressive direction to occur plastic deformation, and the adhesive strength is used to judge whether the adhesive layer occurs interface debonding or sliding. Therefore, when the maximum stress value of a certain finite element unit exceeds the yield strength, it can be determined that the unit occurs failure in the material body direction; when the maximum stress value exceeds the adhesive strength, it can be determined that the unit occurs failure in the interface bonding direction. In actual use, these strength thresholds can be directly read according to the given material parameter table. For example, when 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 a certain unit reaches 9 MPa, the unit should be determined as an adhesive failure unit. All finite element units that meet the above conditions are classified according to their spatial positions, and by judging the distance and contact relationship between adjacent units, these units can be collected into a failure region group showing spatial continuity. Spatial continuity can be determined by judging whether the distance between the geometric centers of the units is less than a set threshold, for example, when the distance between the centers of adjacent units is less than 1.5 times the average edge length of the unit, it can be considered as continuous. In this way, the preliminary failure region distribution information can be formed.

[0081] After obtaining the preliminary failure region distribution information, it is also necessary to determine the failure type according to the continuity characteristics and spatial position relationship of each failure region group under different test conditions. If the failure of the units in a failure region group is caused by the maximum stress exceeding the bonding strength, the region group can be classified as an interface peeling failure region; if the failure reason is mainly the debonding slip caused by shear stress, the region group can be classified as a glue layer slip failure region. The main failure mode of the region can be identified by analyzing the principal stress direction of the units in each failure region group. For example, if the first principal stress direction of the units in the region is consistent with the pull-out force direction, it is usually a peeling failure; if the maximum shear stress is dominant, it is usually a slip failure. According to the failure type and the region position, the system generates a corresponding failure position identifier for each failure region group. The failure position identifier is usually represented in the form of region number and region spatial range combination, for example, using the "region number + coordinate envelope" method, so that the subsequent steps can accurately access each failure region in the three-dimensional structure. In this way, the failure type judgment and failure position identifier generation are completed, providing necessary basic data for the subsequent correction of the pull-out force prediction curve.

[0082] Further, the maximum stress value set is compared with the yield strength and the bonding strength, the finite element units that meet the condition of exceeding the yield strength or exceeding the bonding strength are identified, and the finite element units determined to be failed are collected into failure region groups with spatial continuity according to the spatial position of the finite element units, to form preliminary failure region distribution information, including:

[0083] The adjacency relationship between the finite element units is determined based on the spatial coordinates of the finite element units in the maximum stress value set, to obtain adjacency information for representing the spatial connectivity of the finite element units;

[0084] According to the adjacency information, the finite element units whose maximum stress values meet the condition of exceeding the yield strength or exceeding the bonding strength are divided into multiple candidate failure region groups according to the spatial connectivity, to obtain candidate failure region data containing each candidate failure region group;

[0085] The continuity of the candidate failure region data is checked, the regions consisting of only a single finite element unit and having no spatial connectivity with other finite element units are removed, and the candidate failure region groups with boundary connection and maximum stress value change satisfying a predetermined continuity condition are merged, to form a set of failure region groups after continuity correction;

[0086] Based on the spatial positions of the failure region groups in the set of failure region groups, the finite element units belonging to the same failure region group and the corresponding maximum stress values are recorded correspondingly, to generate preliminary failure region distribution information for representing the spatial distribution and stress distribution relationship of the failure regions.

[0087] In the process of identifying the failed finite element units according to the maximum stress value set and forming the preliminary failure region distribution information, the coordinate information of each finite element unit in the three-dimensional space needs to be obtained first. The spatial coordinates of the finite element unit refer to the positions of the nodes corresponding to the unit in the three-dimensional coordinate system. The representative coordinates of the unit can be obtained by calculating the geometric center of the unit formed by multiple nodes. For example, for a tetrahedral unit formed by four nodes, the geometric center is calculated by adding the three-dimensional coordinates of the four nodes one by one and then taking the average to obtain a specific three-dimensional coordinate point, which represents the position of the unit in space.

[0088] After obtaining the spatial coordinates of all finite element units by the above method, the spatial adjacency relationship between the finite element units can be determined. The adjacency relationship refers to whether two finite element units share at least one node or are in close contact in space. In the field, the adjacency relationship is usually determined by judging whether two units have common nodes. For example, when unit A and unit B share at least one node, it is determined that they have a direct adjacency relationship; if there is no common node, the distance between the geometric centers can be further calculated, and when the distance is less than a preset multiple of the grid element size, it is considered that the two units have connectivity in space. By the above method, an adjacency information set can be established, which uses unit number as index and records the unit numbers adjacent to each unit, so that the system can identify continuous spatial regions in subsequent steps.

[0089] After obtaining the adjacency information, the finite element units marked as exceeding the yield strength or exceeding the adhesive strength in the maximum stress value set can be used as initial input to determine whether these units form a continuous failure region in space. The way to form a candidate failure region group is to start from any finite element unit that meets the failure condition, add units that have adjacency relationship with it and also meet the failure condition to the same group, and continue to expand outward until all adjacent units that meet the condition are included in the same candidate region. For example, when the maximum stress values of three adjacent units all exceed the adhesive strength, and the three units share nodes in space, they are automatically grouped into a candidate failure region group; if any unit does not form an adjacency relationship with other failed units, it constitutes a candidate failure region alone.

[0090] For the above candidate failure region groups, further continuity check needs to be performed to ensure that the final failure region has clear spatial continuity and mechanical rationality. The purpose of continuity check is to eliminate those finite element units that are isolated in space although they meet the stress condition. Generally, a failure region should be composed of multiple finite element units that are close in space, and the appearance of isolated units is often due to local numerical fluctuations. For example, in actual data, if the maximum stress value of a certain finite element unit is slightly higher than the yield strength, but all the units around it are much lower than the strength value, the person skilled in the art can judge that this situation does not conform to the actual failure propagation law and should be eliminated. In addition, for two boundary contact candidate region groups, it is necessary to judge whether the maximum stress value change between them is smooth. When the maximum stress value difference of the boundary element of the two regions is maintained within a reasonable range (for example, lower than a certain proportion of the average stress value of the two regions), they can be merged into one failure region group to reflect the continuous damage characteristics in the real material.

[0091] After performing continuity check on the candidate failure region groups and completing the necessary merging operation, a set of failure region groups with continuity correction can be obtained. Subsequently, it is necessary to establish the correspondence between the failure region group set and the maximum stress value. The specific way is to record the spatial position and corresponding maximum stress value of all finite element units in each failure region group one by one, so that each position point in three-dimensional space can establish a clear mapping relationship with the corresponding failure region number and maximum stress value. This recording process should strictly maintain index consistency, for example, using finite element unit number as the only identifier, to ensure that these information can be accurately called for failure region analysis in subsequent steps.

[0092] After the above information is completely established, the preliminary failure region distribution information for representing the relationship between the spatial distribution of the failure region and the stress distribution can be formed. This information directly reflects the preliminary failure behavior of the micro-needle connection part under specific test conditions, including the spatial position, continuity range of the failure region and the maximum stress value distribution in the region.

[0093] Step S105: based on the stress distribution data, the stress of the connection interface is integrated to obtain the initial prediction curve of the pull-out force varying with the pull-out displacement under each test condition, the pull-out force data of the actual micro-needle pull-out test is collected, and the deviation between the initial prediction curve and the pull-out force data is calculated.

[0094] Step S105 is a process of converting the complex stress distribution data into quantifiable pull-out force curve after completing the simulation calculation of stress field and displacement field, and further comparing and analyzing with the actual experimental results. The purpose of this step is to establish the initial prediction curve of pull-out force varying with pull-out displacement by numerical processing of the interface stress data obtained by simulation, and to determine the difference between the simulation model and the actual situation by combining the pull-out force data measured by experiment, so as to provide a basis for subsequent model correction.

[0095] After the simulation calculation is completed, the finite element model will output the node stress distribution data on the connecting interface at each time step or displacement increment. The so-called "connecting interface" refers to the interaction region formed between the needle head and the handle through the adhesive layer or frictional contact, and the stress distribution of this region is a direct manifestation of the mechanical response in the pulling process. The stress data includes the normal stress, tangential stress and corresponding node coordinates and area information of each interface element. Since the pull-out force is a macroscopic quantity, and the stress belongs to a local quantity, it is necessary to integrate all the stress of the microelement units on the interface to convert it into the overall force. The integral operation is realized by discrete summation in numerical simulation, that is, the interface is divided into a number of elements, the average stress of each element is calculated and multiplied by its area, and then the results of all elements are accumulated to obtain the total force at a certain time. Those skilled in the art can understand that the interface is regarded as composed of many extremely small force pieces, the direction of the force on each piece is consistent with the direction of the needle pulling, and the overall pull-out force is the resultant force of all these local forces.

[0096] For example, in a certain displacement step of the simulation output, the interface is divided into one thousand elements, and the normal stress value of each element represents the local strength of the adhesive layer or the contact surface resisting pulling at that position. If the area of each element is about one square micrometer, the total pull-out force at that displacement can be obtained by multiplying these stresses by the corresponding areas and summing them up. It should be noted that only the normal stress component acts as resistance in the pulling direction, and the tangential stress mainly affects the interface slip characteristics and is not included in the direct component of the pull-out force, but can reflect the contribution of friction energy consumption in the subsequent energy analysis. By sequentially processing the stress data at different displacement increments, a discrete point sequence of the pull-out force varying with displacement can be obtained. In order to form a continuous prediction curve, spline interpolation or polynomial fitting method can be used to connect these discrete points into a smooth curve, which is the initial prediction curve of pull-out force-displacement.

[0097] The definition of pull-out displacement refers to the relative movement of the needle head with respect to the handle, which can be directly obtained from the loading displacement control parameter in the simulation model. Generally, the pull-out process can be divided into three stages: the initial stage, the pull-out force rises rapidly with the displacement, indicating that the adhesive layer at the interface is in overall elastic deformation state; the intermediate stage, the pull-out force reaches the peak value and slightly fluctuates, corresponding to the local adhesive layer peeling or slipping; the final stage, the pull-out force decreases significantly, and the interface is completely separated. The initial prediction curve should reflect this typical mechanical characteristic pattern. If the force-displacement relationship in the simulation result shows linear growth without obvious peak value, it means that the contact parameter setting is improper or the interface bonding model is not correctly enabled, which should be corrected in the subsequent steps.

[0098] After obtaining the simulation prediction curve, the actual pull-out experiment data need to be collected for comparison. The actual pull-out test is usually completed by a precision electronic tensile testing machine, which fixes the connection sample of the needle head and handle in the upper and lower clamps to pull at a constant speed (for example, two hundred millimeters per minute), and records the changes of force and displacement in real time. The test result is output as a force-displacement curve, and its peak value represents the maximum pull-out force measured in the experiment, and the curve shape reflects the real interface failure process. In order to facilitate comparison with the simulation result, the experimental curve and the simulation curve should be drawn in the same coordinate system, and the same unit and scale should be used. If the overall trend of the experimental curve and the prediction curve is consistent, it means that the model correctly reflects the stress law on the macro level; if there is a difference in the peak position or curve shape between the two, quantitative analysis is needed to calculate the deviation.

[0099] The deviation can be calculated in the form of mean absolute error or root mean square error to evaluate the difference between the two curves at different displacement points. Those skilled in the art can realize it by numerical method, that is, at each same displacement point, the difference between the experimental force and the predicted force is taken, and the average value or the square root of the sum of squares of all differences is calculated. The smaller the deviation value is, the better the simulation model fits the actual situation. For example, if the maximum pull-out force in the experiment is zero point four Newton, and the simulation prediction result is zero point three eight Newton, the relative error is five percent; if the overall experimental curve is slightly lower than the prediction curve, but the shape trend is the same, it can be considered that the model has high accuracy. If the deviation exceeds the preset threshold (for example, ten percent), it means that the input parameters such as friction coefficient or adhesive strength in the model still need to be adjusted, which will be completed in the subsequent steps.

[0100] In addition to the overall bias, local bias can also be analyzed, such as calculating the error at the initial stage, peak stage, and failure stage respectively to identify which stage the model has the worst fitting effect. Generally speaking, the bias at the initial stage reflects the accuracy of the elastic modulus setting, the bias at the peak stage reflects the rationality of the bonding strength and friction coefficient, and the bias at the failure stage is related to the fracture energy of the adhesive layer or the interface peeling criterion. Through such segmented comparison, the person skilled in the art can more targetedly judge the error source of the model.

[0101] For example, in a certain simulation, the predicted pull-out force at the initial stage is slightly higher than the experimental result, indicating that the stiffness of the adhesive layer is set too large; at the peak stage, the two are in good agreement, indicating that the bonding strength parameter is accurate; at the end stage, the simulation curve drops slowly, indicating that the threshold of the fracture criterion is too high or the debonding energy is set too large. According to 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.

[0102] After completing the bias calculation, the experimental data and the predicted results should be compared and saved as a structured data file for subsequent statistical analysis or automated calibration. For multiple test conditions, such as different pull-out angles or speeds, the bias of each condition can be calculated separately, and the weighted average value is taken as the overall model bias index. The weighting coefficient can be determined according to the importance or reliability of the experiment, for example, giving higher weight to the condition closest to the actual use condition.

[0103] In addition, in order to verify the stability of the initial prediction curve, sensitivity analysis can be performed. By applying a small perturbation to the key parameters such as friction coefficient and bonding strength, the change amplitude of the predicted curve is observed. If a small parameter change causes significant curve fluctuation, the model is sensitive to that parameter and needs to be optimized in subsequent correction; if the change amplitude is small, the parameter is stable and reliable. This analysis can help the technician to clarify the subsequent correction direction and improve the convergence efficiency of the simulation calculation.

[0104] Step S106: re-perform the three-dimensional finite element mesh establishment and stress field simulation calculation process by adjusting the friction coefficient and the bonding strength until the bias meets the preset accuracy requirement, 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.

[0105] Step S106 is the convergence and correction stage in the whole microneedle fixation strength prediction calculation method, and the core goal is to continuously optimize the interface contact parameters in the aforementioned simulation model through iterative method, so that the deviation between the simulation prediction result and the actual experimental data reaches the preset accuracy requirement, thereby outputting the stable, reliable and engineering significance of the pull-out force prediction curve and the fixation strength prediction result. This step is not only the parameter correction process of the model, but also the key link of the mutual fusion of virtual simulation and experimental results, and the result determines the reliability and applicability of the whole prediction method.

[0106] After the comparison of the initial prediction curve and the experimental data in step S105 is completed, two curves with similar shapes but different values can usually be obtained. The deviation is caused by the difference between the initial assumed value of the interface contact parameter and the actual bonding state, especially the values of the friction coefficient and the bonding strength. In actual application, these two parameters are greatly affected by factors such as material surface roughness, curing condition, glue layer thickness and temperature and humidity environment, so it is impossible to obtain the optimal result by one-time setting. This step adjusts these two key parameters repeatedly, so that the simulation model gradually approaches the real mechanical response.

[0107] Firstly, the preset accuracy requirement needs to be defined. The preset accuracy is a standard for evaluating the consistency of the simulation result and the experimental result, which can be expressed by the percentage deviation or the root mean square error. For example, if the maximum pull-out force difference between the two curves is not more than five percent of the experimental value, or the average error between the two curves in the whole displacement interval is less than three percent, it is considered to reach the preset accuracy. The threshold value can be set according to different application requirements, for example, for medical grade microneedle products, the prediction accuracy is usually required to be not less than five percent, so as to ensure that the simulation result is sufficient to guide the structure optimization.

[0108] Before starting the iterative correction, the direction and amplitude of the adjustment parameter should be determined. The friction coefficient and the bonding strength have different characteristics on the pull-out force curve: the friction coefficient mainly affects the downward trend of the middle and late stages of the curve, that is, the mechanical resistance in the interface slip stage during the pull-out process; the bonding strength mainly determines the peak value of the curve and the steepness of the initial rising stage. If the overall level of the simulation curve is significantly lower than that of the experimental curve, it means that the interface bonding force is set too weak, and the bonding strength value should be increased; on the contrary, if the simulation peak is too high and the curve descends too slowly, it means that the interface bonding is too strong or the friction coefficient is too large, and the corresponding parameter should be appropriately reduced. In order to prevent oscillation or excessive adjustment in the iteration process, the parameter adjustment amount is generally controlled within the range of five percent to ten percent of the previous set value. For example, if the initial bonding strength is set to ten megapascals, and the simulation result is about eight percent lower than the experimental result, the bonding strength can be increased to about ten point eight megapascals in the next calculation.

[0109] The parameter adjustment requires re-execution of the finite element simulation. This process includes re-establishing the contact characteristic parameter file, re-dividing the finite element mesh, and re-applying the boundary conditions. Since the geometric structure and loading mode remain unchanged in each iteration, the main change is reflected in the interface contact properties. To improve computational efficiency, the "parametric analysis" function of the software 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 with the experimental curve. By comparing the deviation changes between the results of two consecutive iterations, it can be determined whether the correction direction is correct. When the deviation decreases significantly, it indicates that the parameter adjustment is reasonable; if the deviation increases instead, it needs to be adjusted in the opposite direction or with a smaller adjustment amplitude.

[0110] In actual implementation, to prevent falling into a local optimal solution, a step-by-step iteration strategy can be adopted. First, fix the friction coefficient and only adjust the adhesive strength until the curve peak value coincides with the experimental value; then fix the adhesive strength and fine-tune the friction coefficient to make the curve tail trend coincide. This sequence can avoid mutual interference between parameters and improve convergence speed. For example, in a certain simulation, the initially set adhesive strength is 90 MPa and the friction coefficient is 0.3. After the first iteration adjustment, the adhesive strength is increased to 100 MPa, and the simulation peak value difference from the experimental peak value is reduced from 12% to 5%; at this time, the friction coefficient is adjusted from 0.3 to 0.28, and the tail deviation is further reduced to 2%, finally meeting the preset precision condition.

[0111] When the model deviation meets the preset accuracy requirement, the corrected pull-out force prediction curve can be output. The corrected curve should be smooth and continuous, and its peak value, rising rate, and descending trend should all coincide with the experimental curve. Under normal circumstances, the corrected prediction curve is slightly lower than the experimental curve at the initial stage, the peak value position coincides or differs by a very small amount, and the final stage is slightly higher than the experimental result. This difference is within an acceptable range and reflects the approximation of 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.

[0112] On this basis, the maximum pull-out force prediction value needs to be calculated. The maximum pull-out force is the force value corresponding to the peak point of the corrected prediction curve, which reflects the maximum load that the interface can withstand during the pull-out process and is a direct indicator of the evaluation of the anchoring strength. When calculating, the "peak value extraction" function in the post-processing module of the simulation software can be used, or the maximum point of the force value can be directly searched in the exported data. If the curve has multiple local peak values, the overall maximum value should be used. To verify the stability of the prediction results, the maximum pull-out force can also be calculated under multiple working conditions, such as at different pull-out angles or speeds, and the average value is taken as the final prediction result to reduce the influence of accidental errors under a single working condition.

[0113] After the maximum pull-out force prediction value is obtained, the microneedle fixation strength prediction result can be calculated accordingly. The fixation strength refers to the maximum pull-out stress that the needle head can withstand per unit contact area in the combined state with the handle. The calculation method is to divide the maximum pull-out force by the actual contact area between the needle head and the handle. The contact area has been extracted in step S101. For example, if the contact area is 0.1 square millimeters and the corrected maximum pull-out force prediction value is 0.4 Newton, the fixation strength prediction result is 4 megapascals. This result can be compared with the design standard or experimental average value of the material to verify the product assembly quality or the stability of the bonding process.

[0114] To ensure the universality and repeatability of the model, this step can be repeatedly performed on multiple samples to obtain the optimal combination range of the friction coefficient and the bonding strength through statistical analysis. If the final parameters of different samples tend to be consistent, it indicates that the method has good stability; if the difference is large, further research can be conducted on external factors that affect the parameters, such as temperature, humidity or curing time, and these factors can be introduced into the future parameter model for compensation.

[0115] Further, the method comprises:

[0116] The initial pull-out force prediction curve is compared with the pull-out force data obtained by the actual pull-out test under each test condition to obtain error data reflecting the distribution of prediction errors, and a deviation index representing the overall deviation degree is determined from the error data;

[0117] Based on the deviation index, the stress distribution state of the adhesive layer region is analyzed, and the correction direction and correction amplitude of the friction coefficient and the bonding strength are determined according to the offset relationship between the simulation stress peak position and the actual failure position, so as to obtain parameter correction information for updating the simulation input conditions;

[0118] The parameter correction information is weighted and integrated in different pull-out speed, pull-out angle and clamping position test conditions to obtain the corrected friction coefficient and bonding strength commonly applicable under multiple working conditions, and the three-dimensional finite element grid establishment and stress field simulation calculation are re-executed using the corrected parameters to generate a new pull-out force prediction curve;

[0119] The new pull-out force prediction curve is compared with the previous round of prediction curve in deviation, when the new pull-out force prediction curve meets the preset convergence condition in deviation index and keeps stable stress change trend in the adhesive layer region, the corrected pull-out force prediction curve is output, and the maximum pull-out force prediction value in the corrected pull-out force prediction curve is used to determine the microneedle fixation strength prediction result.

[0120] When the contact characteristic parameters are corrected to obtain a more accurate pull-out force prediction curve, systematic step-by-step adjustment is needed based on the difference between the initial simulation output and the actual pull-out test result. In order to ensure the executability of the adjustment process, it is necessary to first compare the initial pull-out force prediction curve with the pull-out force data obtained by the actual pull-out test one by one under each test condition. The so-called corresponding comparison is to compare the predicted pull-out force and the actual pull-out force in the same pull-out displacement position, so as to obtain error data representing the prediction deviation. The pull-out displacement refers to the displacement amount of the needle head from the initial contact position to the completely detached position during the pull-out process. In order to ensure the accuracy of the comparison, it is necessary to unify the displacement points under different test conditions, for example, dividing the pull-out displacement into several fixed sampling points, for example, 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, when the actual pull-out force at 0.20 mm displacement is 1.50 N and the predicted pull-out force is 1.20 N, the error at this point is 0.30 N; when the actual pull-out force at 0.30 mm displacement is 2.10 N and the predicted value is 2.25 N, the error is −0.15 N. By repeating the above calculation at each sampling point, an error data sequence covering the entire pull-out process can be obtained. Then the average deviation and the maximum deviation are calculated according to the error sequence, wherein the average deviation can be obtained by averaging all error values, and the maximum deviation can be obtained by selecting the error with the largest absolute value in the error sequence. The deviation index formed in this way can be used to quantitatively represent the overall deviation between the prediction and the actual value.

[0121] After obtaining the deviation index, the stress distribution state of the bonding layer region needs to be analyzed to determine the spatial deviation relationship between the stress peak position predicted by simulation and the actual failure position. The bonding layer region is the key region connecting the needle and handle through the bonding material, and is also the main part affecting the strength of the microneedle. The stress peak position obtained by simulation represents the spatial position of the maximum principal stress or equivalent stress calculated by finite element; the actual failure position can be determined by the position of interface debonding, sliding or material failure in the pull-out experiment. When the simulation stress peak position lags behind the actual failure position, that is, the stress concentration region predicted by simulation is more backward than the actual failure region, it means that the friction coefficient is too low, because the insufficient friction force leads to the reduction of interface resistance in simulation, which delays the failure trend; on the contrary, if the debonding range predicted by simulation is smaller than the actual debonding range, it means that the bonding strength is too high, because the bonding layer is set too strong in simulation, which is not conducive to forming a consistent failure mode with the actual one. Based on the above correspondence, the correction direction of the friction coefficient and the bonding strength can be determined, for example, the friction coefficient needs to be improved or the bonding strength needs to be reduced. The correction amplitude can be determined according to the size of the deviation index, for example, a larger correction amplitude is used when the average deviation is larger, and a smaller correction amplitude is used when the deviation is smaller, so as to avoid excessive correction leading to unstable calculation. Thus, the parameter correction information for updating the simulation input conditions is formed.

[0122] In order to ensure the stability of the modified parameters under different working conditions, the parameter correction information needs to be weighted and integrated in different test working conditions of pull-out speed, pull-out angle and clamping position. The significance of weighted integration is to consider the influence difference between different working conditions, and to avoid the correction result of a certain working condition dominating the final parameters. For example, if the deviation index in a certain working condition is larger, the working condition should be given a higher weight, and the working condition with smaller deviation should be given a lower weight. The value of the weighting coefficient can be obtained by calculating the proportion of the average deviation of each working condition in the total deviation, for example, when the average deviation of a certain working condition accounts for 40% of the total deviation of all working conditions, the weight of the working condition can be set to 0.4. Using the weighted information of each working condition, the modified friction coefficient and bonding strength suitable for all working conditions can be obtained. Using the modified parameters, a new three-dimensional finite element mesh is established and stress field simulation calculation is performed, so as to obtain a new pull-out force prediction curve.

[0123] In order to determine whether the above modification is sufficient accuracy, it is necessary to compare the deviation of the new pull-out force prediction curve with the previous prediction curve, that is, to recalculate the deviation index between the new prediction curve and the actual test data, and compare it with the deviation index of the last round. When the new prediction curve meets the preset convergence condition in the deviation index, for example, the new average deviation is less than 80% of the last round average deviation, and the stress change rate of adjacent finite element units in the adhesive layer area remains within a stable range in several continuous iteration steps, it can be considered that the modification result has reached a stable state. At this time, the parameter modification process is terminated, and the modified pull-out force prediction curve is output. The pull-out force value of the peak point of the pull-out force is extracted from the prediction curve, that is, the maximum value of the curve, as the maximum pull-out force prediction value. Then, combined with the stress distribution of the corresponding peak position in the adhesive layer area, the prediction result of the microneedle fixation strength can be finally determined.

[0124] The second embodiment of the present application provides an electronic device, comprising:

[0125] a processor;

[0126] a memory for storing a program, wherein the program is read and executed by the processor to perform the microneedle fixation strength prediction calculation method based on virtual simulation and mechanical modeling provided in the first embodiment of the present application.

[0127] The third embodiment of the present application provides a computer readable storage medium, which stores a computer program, and the program is executed by a processor to perform the microneedle fixation strength prediction calculation method based on virtual simulation and mechanical modeling provided in the first embodiment of the present application.

[0128] Although the above is disclosed in the preferred embodiment of the present application, it is not intended to limit the present application, and any person skilled in the art can make possible changes and modifications without departing from the spirit and scope of the present application. Therefore, the protection scope of the present application should be limited by the scope defined in the claims of the present application.

Claims

1. A method for predicting the adhesion strength of microneedles based on virtual simulation and mechanical modeling, characterized in that, The method comprises the following steps: Obtain the geometric data of the connection part of the microneedle needle and handle by a three-dimensional scanning device, extract the needle insertion depth, contact area, fitting gap and adhesive layer thickness, and establish a three-dimensional geometric structure containing the connection interface characteristics; Based on the contact area and adhesive layer thickness in the three-dimensional geometric structure, obtain the elastic modulus, yield strength and breaking strength of the needle material, handle material and adhesive layer material, and set the friction coefficient and adhesion strength between the needle and the handle as the contact characteristic parameters; According to the contact characteristic parameters, establish a three-dimensional finite element grid of the connection part of the needle and the handle, set different pull-out speeds, pull-out angles and clamping positions to form multiple test conditions, and perform grid encryption processing in the connection interface area to improve the calculation accuracy; According to the three-dimensional finite element grid after encryption, stress field simulation calculation is performed for each test condition to obtain stress distribution data and displacement data of the connection part, the maximum stress value and its spatial position under each test condition are extracted, and the failure type is judged by comparing the maximum stress value with the yield strength and adhesion strength, and a failure position mark is generated; Based on the stress distribution data, the stress of 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 adhesion strength, the three-dimensional finite element grid establishment and stress field simulation calculation process are re-executed until the deviation meets the preset accuracy requirement, and 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 of predicting and calculating the microneedle sticking strength based on virtual simulation and mechanical modeling according to claim 1, wherein, The method comprises the following steps: Divide the connection part of the needle, the adhesive layer and the handle into regions using the contact characteristic parameters, define the needle insertion zone, the adhesive layer region and the handle constraint zone as independent grid division domains respectively, and assign the element types and node constraint relationships corresponding to the material properties in each grid division domain to form an initial three-dimensional finite element grid with complete contact definition; Apply a standardized pull-out load to the initial three-dimensional finite element grid, perform a mechanical response analysis in the adhesive layer region, obtain the node stress distribution and element strain gradient, determine the stress concentration zone and the transition zone in the adhesive layer region according to the stress gradient change amplitude, and define the spatial coordinates of the stress concentration zone as the interface stress feature set; According to the interface stress feature set, perform proportional encryption division on the stress concentration zone, adjust the regional element size to a predetermined proportion of the average element size of the original region, and set linearly tapered element sizes in the adjacent transition zone in the thickness direction and the tangential direction to maintain the continuity of the nodes and the smoothness of the interface stress transmission, thereby forming a high-precision three-dimensional finite element grid after local encryption; construct a test working condition set based on the high-precision three-dimensional finite element grid, take the pull-out speed, the pull-out angle and the clamping position as the working condition variables, generate a plurality of representative test working conditions according to the orthogonal combination principle, and establish a corresponding relationship between the loading conditions of each test working condition and the high-precision three-dimensional finite element grid to form a multi-working condition finite element grid set with complete boundary condition definition; perform interface continuity analysis on the multi-working condition finite element grid set before loading to determine whether the node stress change in the adhesive layer region meets the stability threshold value, when the local element stress change under any test working condition exceeds the predetermined stability limit value, redivide the region into local elements and perform secondary encryption processing to generate a final finite element grid set corrected for stability, and take the final finite element grid set as the encrypted three-dimensional finite element grid.

3. The method of predicting the microneedle sticking strength based on virtual simulation and mechanical modeling according to claim 1, wherein, According to the encrypted three-dimensional finite element grid, stress field simulation calculation is performed on each test working condition to obtain stress distribution data and displacement data of the connecting part, the maximum stress value and its spatial position under each test working condition are extracted, the failure type is judged by comparing the maximum stress value with the yield strength and the bonding strength, and a failure position mark is generated, including: On the encrypted three-dimensional finite element grid, corresponding boundary conditions and load constraints are applied to each test working condition to obtain stress distribution data and displacement data of the needle insertion area, the adhesive layer region and the handle constraint area to form a stress distribution data set for failure identification; Based on the stress information of each finite element unit in the stress distribution data set, the maximum stress value and its spatial position of each finite element unit under the corresponding test working condition are determined to form a maximum stress value set containing the maximum stress value of each unit; According to the maximum stress value set, each maximum stress value is compared with the yield strength and the bonding strength, the finite element units that meet the conditions of exceeding the yield strength or exceeding the bonding strength are identified, and the finite element units determined to be failed are collected into failure region groups with spatial continuity according to the spatial position of the finite element units to form preliminary failure region distribution information; According to the continuity characteristics and spatial position relationship of each failure region group under different test working conditions, the failure type of each failure region group is determined, and a corresponding failure position mark is generated for each failure region group.

4. The method of predicting and calculating the microneedle sticking strength based on virtual simulation and mechanical modeling according to claim 3, wherein, According to the maximum stress value set, each maximum stress value is compared with the yield strength and the bonding strength, the finite element units that meet the conditions of exceeding the yield strength or exceeding the bonding strength are identified, and the finite element units determined to be failed are collected into failure region groups with spatial continuity according to the spatial position of the finite element units to form preliminary failure region distribution information, including: Determine the adjacency relationship between the finite element units based on the spatial coordinates of each finite element unit in the maximum stress value set to obtain adjacency information representing the spatial connectivity of the finite element units; According to the adjacency information, the finite element units that meet the conditions of exceeding the yield strength or exceeding the bonding strength are divided into a plurality of candidate failure region groups according to the spatial connectivity to obtain candidate failure region data containing each candidate failure region group; The continuity of the candidate failure region data is checked, regions composed of only a single finite element unit and having no spatial continuity with other finite element units are removed, and candidate failure region groups that are adjacent to each other and have a maximum stress value change satisfying a predetermined continuity condition are merged to form a set of failure region groups after continuity correction; Based on the spatial positions of the failure region groups in the set of failure region groups, the finite element units belonging to the same failure region group are recorded in correspondence with the corresponding maximum stress values to generate preliminary failure region distribution information for representing the relationship between the spatial distribution of the failure region and the stress distribution.

5. The method for predicting the microneedle sticking strength based on virtual simulation and mechanical modeling according to claim 1, wherein, The three-dimensional finite element grid establishment and stress field simulation calculation process are re-executed by adjusting the friction coefficient and the bonding strength until the deviation meets the preset accuracy requirement, and a corrected pull-out force prediction curve is output, and a microneedle fixation strength prediction result is calculated based on the maximum pull-out force prediction value in the pull-out force prediction curve, including: The initial pull-out force prediction curve is compared with the pull-out force data obtained by actual pull-out test under each test condition to obtain error data reflecting the prediction error distribution, and a deviation index representing the overall deviation degree is determined from the error data; Based on the deviation index, the stress distribution state of the adhesive layer region is analyzed, and the correction direction and correction amplitude of the friction coefficient and the bonding strength are determined according to the offset relationship between the simulation 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 in different pull-out speed, pull-out angle and clamping position test conditions to obtain the corrected friction coefficient and bonding strength commonly applicable under multiple working conditions, and the three-dimensional finite element grid establishment and stress field simulation calculation are re-executed using the corrected parameters to generate a new pull-out force prediction curve; The new pull-out force prediction curve and the previous prediction curve are compared for deviation, and when the new pull-out force prediction curve meets the preset convergence condition in the deviation index and maintains a stable stress change trend in the adhesive layer region, the corrected pull-out force prediction curve is output, and the maximum pull-out force prediction value in the corrected pull-out force prediction curve is used to determine the microneedle fixation strength prediction result.

Images