Catheter balloon rupture and release simulation system based on a physics engine

Abstract

This invention relates to the field of medical device simulation and computer physics simulation technology, specifically a physics engine-based catheter balloon rupture and fluid release simulation system, comprising: a balloon stress calculation module for receiving external input data and initial fluid volume, and outputting stress distribution data of each grid cell on the balloon membrane surface containing the maximum principal stress; a rupture processing module for reading stress distribution data to calculate energy release rate, and calculating and outputting rupture status data containing the current total effective rupture area; a fluid simulation module for receiving rupture status data and writing back the updated balloon internal pressure to the balloon stress calculation module; and a display monitoring module for monitoring the sum of the volumes of balloon internal fluid, released fluid, and fluid in transmission within each rendering frame and comparing it with the initial fluid volume, and outputting a display screen and status indicator. This invention ensures that the rupture location and fluid release process are consistent with actual operating parameters.

Description

Technical Field

[0001] This invention relates to the field of medical device simulation and computer physics simulation technology, specifically a physics engine-based system for simulating catheter balloon rupture and fluid release. Background Technology

[0002] In existing catheter training systems, it is usually necessary to simulate and display the balloon inflation, compression, removal, and rupture release processes. The system generally includes a balloon deformation calculation part, a rupture handling part, a fluid release part, and an interface display part. These parts work together to form the balloon state changes and release effects during the training process.

[0003] In the existing technology, when the balloon is subjected to the combined effects of water injection pressure, external pressure displacement and tube removal force, the stress state of the balloon membrane will continuously change, and stress concentration is likely to occur in local areas. Whether a rupture forms and the release process after the rupture forms are closely related to the internal pressure of the balloon, the stress distribution of the membrane surface and the change of the rupture area. Therefore, it is necessary to take into account the correlation between balloon deformation, fracture judgment, release calculation and volume statistics in the simulation process.

[0004] However, existing catheter training simulation schemes often struggle to establish a stable and continuous linkage mechanism between balloon stress changes, tear evolution, and fluid release. This can easily lead to problems such as inaccurate tear determination, inconsistencies between the fluid release process and actual operating parameters, and decreased continuity when the computational load increases, thus affecting the realism and stability of the training system. Summary of the Invention

[0005] The purpose of this invention is to provide a physics engine-based simulation system for urinary catheter balloon rupture and fluid release, which avoids the simulation distortion caused by directly displaying the rupture according to the preset animation, and can effectively control the volume error caused by particle dispersion and the calculation time under extreme fluid release conditions, ensuring the real-time consistency of balloon rupture evolution, fluid release and graphic display before and after complex operations.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] The physics engine-based simulation system for urinary catheter balloon rupture and fluid release runs on the physics engine platform and includes the following modules that all call the platform's solution interface:

[0008] The balloon stress calculation module is used to receive external input data and initial water injection volume, combine balloon membrane material parameters including critical energy release rate, discretize the balloon model into a triangular mesh model and solve its deformation, and output the stress distribution data of each mesh unit of the balloon membrane surface containing the maximum principal stress.

[0009] The crack processing module is used to read the stress distribution data to calculate the energy release rate, make a fracture judgment based on the maximum principal stress and the critical energy release rate, split the vertices of the triangular mesh model, update the crack connection relationship and geometric boundary, and calculate and output crack status data containing the current total effective crack area.

[0010] The fluid simulation module is used to receive the crack state data, calculate the instantaneous release volume flow rate based on the current total effective crack area, the preset liquid density and the pressure difference inside and outside the balloon, drive the release of fluid particles and output the release state of fluid particles, and write back the updated internal pressure of the balloon to the balloon force calculation module.

[0011] The display monitoring module is used to monitor the sum of the volumes of the fluid inside the balloon, the released fluid, and the fluid in transit within each rendering frame based on stress distribution data, crack status data, and fluid particle release status, and compare it with the initial water injection volume. It also monitors the calculation time of a single frame and outputs the display screen and status indicators.

[0012] As a further aspect of the present invention, the peripheral input data is data collected by the input peripheral, including the water injection rate, the withdrawal force acting on the catheter, and the pressure displacement acting on the balloon area.

[0013] The parameters of the balloon membrane include the parameters of the Mooney-Rivlin two-parameter constitutive model, film thickness, Young's modulus, critical energy release rate, and initial defect size;

[0014] The balloon stress calculation module assigns the initial inflation volume and initial internal pressure of the balloon to the triangular mesh model.

[0015] As a further aspect of the present invention, the balloon stress calculation module generates stress distribution data based on peripheral input data and balloon membrane material parameters, specifically including:

[0016] A damping matrix is ​​introduced into the structural mechanics solver of the physics engine platform to calculate the deformation results caused by the displacement of each node in the triangular mesh model; Rayleigh damping based on the mass matrix and stiffness matrix is ​​adopted.

[0017] The stress tensor of each mesh element is generated based on the deformation results;

[0018] Extract the maximum principal stress and the principal direction vector corresponding to the maximum principal stress from the stress tensor;

[0019] The maximum principal stress and principal direction vector are written as stress distribution data into a preset shared buffer.

[0020] As a further aspect of the present invention, the crack treatment module generates crack state data based on stress distribution data, specifically including:

[0021] The energy release rate of each grid cell is calculated based on the maximum principal stress, Young's modulus, and the projected equivalent half-length of the existing crack in the preset neighborhood centered on each grid cell. If the region is an intact membrane surface, the projected equivalent half-length is taken as the initial defect size.

[0022] Set a delayed damage counter for each grid cell, with an initial value of zero; determine the relationship between the energy release rate and the critical energy release rate; when the energy release rate is greater than or equal to the critical energy release rate, increment the delayed damage counter by one; when the energy release rate is less than the critical energy release rate, clear the delayed damage counter to zero.

[0023] If the delayed damage counter satisfies the energy release rate being greater than or equal to the critical energy release rate within a preset number of consecutive calculation substeps, then the fracture judgment is triggered.

[0024] After triggering the fracture detection, vertex splitting is performed along the direction perpendicular to the principal direction vector in the local membrane tangent plane of the mesh element, the fracture topology geometry is updated, the current total effective fracture area is calculated, and the current total effective fracture area is output as fracture status data.

[0025] As a further aspect of the present invention, the fluid simulation module generates the fluid particle release state based on the crack state data, specifically including:

[0026] The instantaneous release volumetric flow rate is calculated based on the current total effective rupture area, the preset discharge coefficient, the liquid density, and the pressure difference between the inside and outside of the balloon determined by the difference between the internal pressure of the balloon and the preset external environmental pressure.

[0027] Based on the preset fluid particle spacing and the preset single-particle representative volume, the instantaneous liquid release volume flow rate is discretized into a corresponding number of fluid particles according to the single-particle representative volume, and emitted along the local average normal direction of the crack opening surface.

[0028] The cumulative volume of released fluid particles is counted, and the current internal pressure of the balloon is calculated by combining the initial filling volume and the initial internal pressure of the balloon.

[0029] The updated current internal pressure of the balloon is written back to the balloon force calculation module as the updated internal pressure of the balloon until the energy release rate is lower than the critical energy release rate, at which point the vertex splitting stops and the fluid particle release state is generated.

[0030] As a further aspect of the present invention, the display monitoring module performs volume error correction based on the fluid particle release state, specifically including:

[0031] The total volume of liquid is calculated in each rendered frame. The total volume of liquid consists of the current internal volume of the balloon, the cumulative volume that has been released, and the volume of liquid being transported.

[0032] Compare the total liquid volume with the initial injection volume and calculate the residual.

[0033] If the residual is greater than 0.5% of the initial water injection volume, the volume compensation coefficient is output; if the residual is less than or equal to 0.5% of the initial water injection volume, the volume compensation coefficient is set to one.

[0034] The fluid simulation module receives the volume compensation coefficient, multiplies the density of fluid particles by the volume compensation coefficient, and performs volume compensation.

[0035] As a further aspect of the present invention, the display monitoring module performs particle merging and grid data processing based on the crack state data and the fluid particle release state, specifically including:

[0036] When the current number of fluid particles reaches 90% of the preset particle limit, scan the fluid particle field;

[0037] The relationship between the spatial distance between adjacent fluid particle pairs and the preset fluid particle spacing, as well as the relationship between the velocity vector difference and the preset velocity threshold, is determined. If the spatial distance is less than the preset fluid particle spacing and the velocity vector difference is less than the preset velocity threshold, the adjacent fluid particle pairs are merged into a single particle. The mass of the new particle is obtained by summing the masses of the two particles, and the velocity of the new particle is calculated according to the conservation of momentum. The position of the new particle is determined by mass weighting. Otherwise, the merging is not performed.

[0038] Determine the number of mesh cells that meet the breakage criteria within a single frame. If the number is greater than 10, sort them by the maximum principal stress and perform vertex splitting only on the top three mesh cells, leaving the remaining mesh cells for the next rendering frame. If the number is less than or equal to ten, perform vertex splitting on all mesh cells that meet the breakage criteria.

[0039] As a further aspect of the present invention, the display monitoring module performs a simplified calculation process based on the single-frame calculation time, specifically including:

[0040] The system monitors the single-frame computation time in real time and determines the relationship between the single-frame computation time and the preset single-frame computation time threshold. If the single-frame computation time is greater than the preset single-frame computation time threshold, the system performs the first simplification process, which increases the target spacing of newly released particles and resamples the existing fluid particle field to amplify the fluid particle spacing. If the single-frame computation time is less than or equal to the preset single-frame computation time threshold, the system maintains the current computation accuracy unchanged.

[0041] If the single-frame computation time is still greater than the preset single-frame computation time threshold after the first simplification process is performed, the second simplification process is performed, the sub-step iteration of the crack processing module is paused, and the current crack topology geometry remains unchanged.

[0042] If the second simplification process is executed for three consecutive frames, the third simplification process is executed, and real-time physics calculation is paused.

[0043] The current inflation volume, film thickness, and rupture location are extracted as the current balloon state parameters. The matching rupture release curve is retrieved from the release data table that is generated and stored in advance based on different inflation volumes, film thicknesses, and rupture locations. The approximate release flow rate-time data corresponding to the current balloon state parameters is output through interpolation.

[0044] As a further aspect of the present invention, the system sets different calculation time steps for each module, specifically including:

[0045] The time step of the balloon force calculation module is set to the first time step and synchronized with the system rendering frame rate.

[0046] The time step of the crack processing module is set to the second time step, and multiple sub-step iterations are performed within a single rendering frame.

[0047] The time step of the fluid simulation module is set to the third time step and matched with fluid stability conditions;

[0048] The first time step is larger than the third time step, and the third time step is larger than the second time step;

[0049] Each module reads and writes cached data through a shared cache area.

[0050] As a further aspect of the present invention, the display monitoring module is presented through a graphical interface, specifically including:

[0051] When the energy release rate reaches 80% of the critical energy release rate, the corresponding grid cell is marked as an early warning cell and displayed in the graphical interface by stress highlighting.

[0052] When the third simplification process is performed, an approximate calculation status indicator is displayed in the graphical interface;

[0053] The graphical interface simultaneously displays the internal pressure curve of the balloon, the maximum principal stress change curve of the membrane surface, and the balloon rupture and fluid release animation.

[0054] The beneficial effects of this invention are:

[0055] 1) This invention introduces a delayed damage counter into the crack processing module, combines the relationship between the energy release rate of the current mesh cell and the critical energy release rate to perform multi-substep fracture judgment, and triggers vertex splitting and crack topology update only after the continuous substep condition is met; at the same time, the internal pressure of the balloon updated by the fluid simulation module is written back to the force calculation module to update the stress state of the membrane surface; this design establishes a closed-loop linkage between balloon stress change, crack evolution and liquid release, solves the problems of inaccurate crack judgment and inconsistency between the liquid release process and actual operating parameters in the prior art, and improves the realism of simulation training;

[0056] 2) This invention monitors the single-frame computation time in real time through a display monitoring module, and triggers the first, second and third simplification processes step by step when the computation times out; combined with the merging mechanism for adjacent particles in the fluid particle field, and the upper limit sorting control of the number of grid cells that perform vertex splitting within a single frame, this scheme effectively controls the computation scale, solves the problem of decreased system display continuity when the computation load of balloon rupture and fluid release surges, and ensures the stability of the system.

[0057] 3) In the display monitoring module, the total volume of liquid is counted for each rendering frame and compared with the initial water injection volume to calculate the residual. When the residual exceeds the set percentage threshold, the system calculates and outputs the volume compensation coefficient. The fluid simulation module receives the coefficient and dynamically adjusts the density of fluid particles to perform volume compensation. This design effectively corrects the cumulative volume error caused by discretizing the continuous fluid into particles, ensuring the consistency of liquid volume during the simulation process. Attached Figure Description

[0058] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0059] Figure 1 This is a diagram of a catheter balloon rupture and fluid release simulation system according to an embodiment. Detailed Implementation

[0060] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0061] Please see Figure 1 The physics engine-based simulation system for urinary catheter balloon rupture and fluid release runs on the physics engine platform and includes the following modules that all call the platform's solution interface:

[0062] The balloon stress calculation module is used to receive external input data and initial water injection volume, combine balloon membrane material parameters including critical energy release rate, discretize the balloon model into a triangular mesh model and solve its deformation, and output the stress distribution data of each mesh unit of the balloon membrane surface containing the maximum principal stress.

[0063] The crack processing module is used to read stress distribution data to calculate energy release rate, make fracture judgment based on maximum principal stress and critical energy release rate, split the vertices of the triangular mesh model, update the crack connection relationship and geometric boundary, and calculate and output crack status data containing the current total effective crack area.

[0064] The fluid simulation module is used to receive the crack state data, calculate the instantaneous release volume flow rate based on the current total effective crack area, the preset liquid density and the pressure difference inside and outside the balloon, drive the release of fluid particles and output the release state of fluid particles, and write back the updated internal pressure of the balloon to the balloon force calculation module.

[0065] The display monitoring module is used to monitor the sum of the volumes of the fluid inside the balloon, the released fluid, and the fluid in transit within each rendering frame based on stress distribution data, crack status data, and fluid particle release status, and compare it with the initial water injection volume. It also monitors the calculation time of a single frame and outputs the display screen and status indicators.

[0066] The peripheral input data is data collected by the input peripheral, including the water injection rate, the withdrawal force applied to the catheter, and the pressure displacement applied to the balloon area;

[0067] The parameters of the balloon membrane include the parameters of the Mooney-Rivlin two-parameter constitutive model, film thickness, Young's modulus, critical energy release rate, and initial defect size; the balloon stress calculation module assigns the initial filling volume and initial internal pressure of the balloon to the triangular mesh model;

[0068] The balloon stress calculation module generates stress distribution data based on peripheral input data and balloon membrane material parameters. Specifically, it includes: introducing a damping matrix into the structural mechanics solver of the physics engine platform to calculate the deformation results caused by the displacement of each node in the triangular mesh model; and using Rayleigh damping based on the mass matrix and stiffness matrix.

[0069] The stress tensor of each mesh element is generated based on the deformation results; the maximum principal stress and the principal direction vector corresponding to the maximum principal stress are extracted from the stress tensor; the maximum principal stress and the principal direction vector are written as stress distribution data into a preset shared buffer.

[0070] For example, in practical applications: the system is used to simulate balloon inflation, compression, tube removal, and rupture and fluid release during catheter training;

[0071] The physics engine platform keeps the balloon mesh, fluid particle pool and display interface running simultaneously. The input interface continuously collects the water injection rate, the catheter removal force acting on the catheter and the pressure displacement acting on the balloon area given by the input peripheral as peripheral input data. The main control part writes the peripheral input data into the preset memory input buffer of the current calculation cycle and sends the initial water injection volume and balloon membrane parameters into the balloon force calculation module.

[0072] The parameters of the balloon membrane material include at least the parameters of the Mooney-Rivlin two-parameter constitutive model consisting of the first hyperelastic material constant and the second hyperelastic material constant, the film thickness, Young's modulus, critical energy release rate and initial defect size, which are used to limit the deformation of the balloon membrane surface under load and the subsequent fracture determination conditions.

[0073] During system initialization, the balloon force calculation module discretizes the balloon model into a triangular mesh model, and simultaneously allocates the initial inflation volume and initial internal pressure of the balloon as the base state for multiple subsequent calculation cycles.

[0074] This implementation mainly illustrates how the balloon stress calculation module generates stress distribution results that can be directly called upon for subsequent crack processing; the structural mechanics solver reads peripheral input data from the current preset memory input buffer in each calculation cycle, updates the nodal loads and boundary constraints of the triangular mesh model, and the specific macro-micro mapping rules are as follows:

[0075] Extract the extubation force vector acting on the urinary catheter, and divide the magnitude of the vector equally into nodes. Apply the concentrated force as a nodal force to each node in the preset neck grid node set at the bottom of the balloon.

[0076] The pressure displacement acting on the balloon region is extracted, and the set of contact mesh nodes that currently coincide with the bounding box of the external pressure surface model is identified by geometric interferometry. The pressure displacement is directly applied to the three-dimensional coordinate components of the contact mesh node set as a Dirichlet strong boundary condition, and Rayleigh damping is introduced in the solution process.

[0077] The damping matrix is ​​obtained by adding the mass matrix scaled by the mass proportional damping coefficient and the stiffness matrix scaled by the stiffness proportional damping coefficient. It is used to suppress high-frequency vibrations on the balloon membrane surface during rapid loading or local deformation.

[0078] During the system initialization phase, the steps for establishing the mass proportional damping coefficient and the stiffness proportional damping coefficient are as follows: perform pre-sampling modal analysis on the balloon model with initial filling volume, extract the first-order inherent low-frequency parameters that dominate the overall deformation of the balloon, and the higher cutoff frequency parameters that represent the high-frequency numerical fluctuations of the local membrane surface.

[0079] Based on the above two target frequencies and the preset target damping ratio, a Rayleigh damping simultaneous equation system is constructed. By solving this equation system, the mass proportional damping coefficient and the stiffness proportional damping coefficient are determined in reverse.

[0080] The specific construction and solution logic is as follows: Set the target damping ratio as a uniform constant. According to Rayleigh damping theory, the target damping ratio is equal to the quotient of the mass proportional damping coefficient divided by twice the frequency, plus the quotient of the product of the stiffness proportional damping coefficient and the frequency divided by two.

[0081] Substituting the first-order inherent low-frequency parameter and the higher cutoff frequency parameter as frequency variables into the relational formula, a system of two linear equations is formed. Solving the equations, we find that the mass proportional damping coefficient is equal to twice the target damping ratio multiplied by the product of the two target frequencies and then divided by the sum of the two frequencies. The stiffness proportional damping coefficient is equal to twice the target damping ratio divided by the sum of the two frequencies. By constructing and solving the algebraic relationship, we can enable the system to absorb non-physical high-frequency numerical oscillations and maintain the physical deformation of the fracture.

[0082] After the solver outputs the displacement of each node, it forms a stress tensor at the mesh element level, and then extracts the maximum principal stress and the corresponding principal direction vector from the stress tensor.

[0083] The maximum principal stress and principal direction vector are written into a preset shared buffer to form the stress distribution data required for subsequent fracture calculations;

[0084] Based on this, the rupture processing module reads the maximum principal stress and principal direction vector of each grid cell on the balloon membrane from the shared buffer area, calculates the energy release rate corresponding to the current stress state for each cell, and compares it with the critical energy release rate.

[0085] When the fracture condition is met, the relevant vertices of the triangular mesh model are split to update the fracture connection relationship and geometric boundary; the total effective fracture area is recalculated according to the current fracture boundary and written into the fracture status data area.

[0086] The fluid simulation module takes the data area of ​​the crack state as input, combines the preset liquid density and the pressure difference inside and outside the balloon to calculate the current instantaneous release volume flow rate, and then drives the fluid particles to be released according to the flow rate;

[0087] As the release process proceeds, the module synchronously updates the internal pressure of the balloon and writes the updated pressure value back to the balloon force calculation module, so that the membrane force in the next calculation cycle can be solved again according to the new internal pressure conditions.

[0088] The display monitoring module reads stress distribution data, crack status data and fluid particle release status in parallel, sums up the volume of liquid inside the balloon, the volume of liquid released and the volume of liquid in transmission in each rendering frame, and compares it with the initial water injection volume. At the same time, it monitors and records the calculation time of a single frame and outputs the display screen and status indicators in the graphical interface.

[0089] To ensure the continuity of the calculation process, if the peripheral input data of the current calculation cycle has not arrived completely, the system retains the peripheral input data of the previous valid cycle as temporary input and continues to perform force calculation and display update for the results of that cycle; after the new peripheral input data arrives, it is replaced with the real-time value.

[0090] If the shared cache has not completed the previous round of writing at the current read time, the crack processing module reads the stress distribution data of the most recently completed write and continues to perform the fracture determination, thereby avoiding sudden changes in crack area due to read-write competition.

[0091] If no effective rupture area is formed in the current cycle, the fluid simulation module keeps the number of particles released at zero, and only updates the internal pressure and display results of the balloon in the unruptured state to prevent premature fluid release before the rupture is formed.

[0092] Under the above training, when the operator continuously injects water, the pressure inside the balloon rises, and the stress distribution on the balloon membrane changes accordingly; if a pull-out force or external displacement of the balloon area is applied at the same time, the local principal stress on the membrane surface will be concentrated in the corresponding grid cell.

[0093] The system does not directly display the rupture according to the preset animation. Instead, it first forms the stress distribution and then hands it over to the crack processing module to judge the fracture according to the critical energy release rate. Only after the judgment conditions are met will the crack boundary and the release flow rate be updated.

[0094] Therefore, the location of the tear and the fluid release process shown in the display are consistent with the actual operating parameters, and the changes in the pressure inside the balloon can continue to affect subsequent stress calculations.

[0095] In a preferred embodiment of the present invention, the crack treatment module generates crack state data based on stress distribution data, specifically including: calculating the energy release rate of each grid cell based on the maximum principal stress, Young's modulus, and the projected equivalent half-length of existing cracks in a preset neighborhood centered on each grid cell; if the region is an intact membrane surface, the projected equivalent half-length is taken as the initial defect size.

[0096] Set a delayed damage counter for each grid cell, with an initial value of zero; determine the relationship between the energy release rate and the critical energy release rate; when the energy release rate is greater than or equal to the critical energy release rate, increment the delayed damage counter by one; when the energy release rate is less than the critical energy release rate, clear the delayed damage counter to zero.

[0097] If the delayed damage counter satisfies the energy release rate being greater than or equal to the critical energy release rate within a preset number of consecutive calculation substeps, then the fracture judgment is triggered.

[0098] After triggering the fracture detection, vertex splitting is performed along the direction perpendicular to the principal direction vector in the local membrane tangent plane of the mesh element, the fracture topology geometry is updated, the current total effective fracture area is calculated, and the current total effective fracture area is output as fracture status data.

[0099] The fluid simulation module generates the fluid particle release state based on the crack state data, specifically including: calculating the instantaneous release volume flow rate based on the current total effective crack area, preset discharge coefficient, liquid density, and the pressure difference inside and outside the balloon determined by the difference between the pressure inside the balloon and the preset external environmental pressure;

[0100] Based on the preset fluid particle spacing and the preset single-particle representative volume, the instantaneous liquid release volume flow rate is discretized into a corresponding number of fluid particles according to the single-particle representative volume, and emitted along the local average normal direction of the crack opening surface.

[0101] The cumulative volume of released fluid particles is counted, and the current internal pressure of the balloon is calculated by combining the initial filling volume and the initial internal pressure of the balloon.

[0102] The updated current internal pressure of the balloon is written back to the balloon force calculation module as the updated internal pressure of the balloon until the energy release rate is lower than the critical energy release rate, at which point the vertex splitting stops and the fluid particle release state is generated.

[0103] The display monitoring module performs volume error correction based on the fluid particle release status, specifically including: calculating the total liquid volume in each rendering frame, which consists of the current internal volume of the balloon, the cumulative volume released, and the volume of liquid in transmission; comparing the total liquid volume with the initial injection volume and calculating the residual.

[0104] If the residual is greater than 0.5% of the initial water injection volume, the volume compensation coefficient is output; if the residual is less than or equal to 0.5% of the initial water injection volume, the volume compensation coefficient is set to one.

[0105] The fluid simulation module receives the volume compensation coefficient, multiplies the density of fluid particles by the volume compensation coefficient, and performs volume compensation.

[0106] For example, in practical applications: "When the catheter training process enters a situation where the balloon is under local pressure or the force during catheter removal exceeds the preset force threshold, the stress distribution data in the previous embodiment alone is not enough to reliably determine whether the tear has actually formed; because under short-cycle force fluctuations, the local grid cells on the balloon membrane may briefly experience high principal stress. If vertex splitting is performed directly at this moment, the tear topology may change frequently in adjacent cycles, affecting the subsequent calculation of the release volume."

[0107] Therefore, this embodiment further defines the connection method between crack treatment and liquid release calculation, so that fracture determination, particle release and volume compensation are completed under the same set of state transfer relationships;

[0108] After reading the maximum principal stress and principal direction vector in the shared buffer, the crack processing module extracts the projected equivalent half length of existing cracks in the preset neighborhood with the current mesh element as the center, and participates in the calculation of the energy release rate of the element together with Young's modulus.

[0109] The spatial range of the preset neighborhood is dynamically determined by the side length of the current grid cell. Specifically, a spherical space is constructed with the centroid of the current grid cell as the center and three times the maximum side length of the cell as the search radius. All crack segments falling into this spherical space are considered to belong to the preset neighborhood.

[0110] The energy release rate is the product of the square of the maximum principal stress, pi, and the projected equivalent half-length, then divided by Young's modulus; if the region is a good membrane surface, the projected equivalent half-length is taken as the initial defect size.

[0111] Set a delay damage counter for each grid cell, with the initial value kept at zero; determine the relationship between the energy release rate and the critical energy release rate, and increment the delay damage counter by one when the energy release rate is greater than or equal to the critical energy release rate;

[0112] When the energy release rate is less than the critical energy release rate, the delayed damage counter is cleared to zero; if the delayed damage counter satisfies the energy release rate being greater than or equal to the critical energy release rate within a preset number of consecutive calculation substeps, then the fracture judgment is triggered, and vertex splitting is performed along the direction perpendicular to the main direction vector in the local membrane tangent plane of the mesh element.

[0113] After the split is completed, the topological geometry of the split is updated, the current total effective split area is calculated, and the current total effective split area is output as the split status data.

[0114] The fluid simulation module does not directly use the mesh splitting results, but instead receives the crack state data, uses the difference between the pressure inside the balloon and the preset external environmental pressure as the pressure difference inside and outside the balloon, and calculates the instantaneous release volume flow rate based on the current total effective crack area, preset discharge coefficient, liquid density, and pressure difference inside and outside the balloon.

[0115] The specific calculation logic is as follows: The system calculates twice the difference between the pressure inside the balloon and the preset external environmental pressure, divides it by the liquid density to obtain the ratio and takes the square root, and then multiplies the square root result with the preset discharge coefficient and the current total effective rupture area to obtain the instantaneous liquid release volume flow rate.

[0116] The preset discharge coefficient in the above calculation is used to correct the flow loss of real viscous fluid when passing through irregular breaks. Its calibration basis is: in the offline physical hydraulic calibration experiment, the ratio of steady-state release flow rate to the theoretical flow rate calculated by the ideal Bernoulli equation under the same pressure difference is taken, and its preset value range is limited to between 0.60 and 0.65.

[0117] Based on the above fluid release calculation method, fluid release is controlled by the dynamic pressure gradient;

[0118] Once the instantaneous liquid release volumetric flow rate is determined, the system uses the cube of the preset fluid particle spacing as the representative volume of a single particle, discretizes the instantaneous liquid release volumetric flow rate into the corresponding number of fluid particles according to the representative volume of a single particle, and emits them along the local average normal direction of the crack opening surface.

[0119] After the particles enter the fluid field, the cumulative volume of the released fluid particles is counted, and the current internal pressure of the balloon is calculated by combining the initial filling volume and the initial internal pressure of the balloon.

[0120] The updated current internal pressure of the balloon is written back to the balloon force calculation module as the updated internal pressure of the balloon until the energy release rate is lower than the critical energy release rate, at which point the vertex splitting stops and the fluid particle release state is generated.

[0121] In terms of volume statistics, the display monitoring module calculates the total liquid volume in each rendered frame. The total liquid volume consists of the current internal volume of the balloon, the cumulative volume that has been released, and the volume of liquid being transmitted. The total liquid volume is compared with the initial injection volume, and the residual is calculated.

[0122] If the residual is greater than 0.5% of the initial water injection volume, the volume compensation coefficient is output; if the residual is less than or equal to 0.5% of the initial water injection volume, the volume compensation coefficient is set to one.

[0123] To clarify the feedback correction mechanism of the compensation coefficient, the system suppresses error expansion through ratio convergence: the volume compensation coefficient is the ratio of the current total liquid volume to the initial injection volume;

[0124] The fluid simulation module receives the volume compensation coefficient and, while keeping the physical liquid density used to calculate the instantaneous release volume flow rate constant, multiplies the apparent density of the fluid particles used for rendering by the volume compensation coefficient to perform volume compensation.

[0125] The single-particle representative volume of a fluid particle within the rendering system is obtained by dividing its allocated mass by the current apparent density.

[0126] When the calculated total liquid volume is unexpectedly greater than the initial injection volume, a positive volume error occurs due to particle dispersion. The compensation coefficient will be greater than one, which passively amplifies the calculated density of newly launched particles. Ultimately, this reduces the volume represented by a single particle of the same mass, thereby automatically reducing the volume increment during subsequent physical release and achieving negative feedback suppression.

[0127] This compensation coefficient only applies to the particle state update after the current state and is not written back to the crack handling module;

[0128] If a grid cell cannot obtain an effective neighborhood crack projection length in the current substep, but the cell already has a continuous count record, the system will still recalculate the energy release rate of the current substep based on the initial defect size of the intact membrane surface, and will not directly use the projection length of the previous substep, so as to avoid the crack length being amplified when the data is interrupted.

[0129] If the crack state data area is not updated within a certain rendering frame, the fluid simulation module continues to use the most recent effective total crack area to calculate particle release, but does not add new vertex splitting results;

[0130] If the difference between the current internal pressure of the balloon and the preset external environmental pressure is insufficient to form an effective release of fluid, the number of particles emitted will remain zero, while the geometry of the slit will remain unchanged, and only the released volume will continue to be accumulated and the display results will be refreshed.

[0131] If the volume compensation coefficient has been output but no new particles are added in the current frame, the system will use the compensation coefficient to update the existing particle density without generating additional compensation particles.

[0132] During catheterization training, as the operator continuously increases the amount of water injected or applies displacement near the balloon, the maximum principal stress of the local grid cell increases. However, a tear is only confirmed to have formed when this stress level remains at the fracture threshold for several consecutive substeps.

[0133] Once a tear is formed, the total effective tear area is immediately used in the calculation of the release flow rate. The fluid inside the balloon is released from the tear direction, and the internal pressure change caused by the release then affects the next round of balloon stress calculation.

[0134] The total fluid volume check in the display interface simultaneously constrains the statistical consistency of fluid inside and outside the balloon, ensuring that the fluid release volume displayed during training corresponds to the initial injection volume.

[0135] In a preferred embodiment of the present invention, the display monitoring module performs particle merging and grid data processing based on the crack state data and the fluid particle release state, specifically including: scanning the fluid particle field when the current number of fluid particles reaches 90% of the preset particle upper limit;

[0136] The relationship between the spatial distance between adjacent fluid particle pairs and the preset fluid particle spacing, as well as the relationship between the velocity vector difference and the preset velocity threshold, is determined. If the spatial distance is less than the preset fluid particle spacing and the velocity vector difference is less than the preset velocity threshold, the adjacent fluid particle pairs are merged into a single particle. The mass of the new particle is obtained by summing the masses of the two particles, and the velocity of the new particle is calculated according to the conservation of momentum. The position of the new particle is determined by mass weighting. Otherwise, the merging is not performed.

[0137] Determine the number of mesh cells that meet the breakage criteria within a single frame. If the number is greater than 10, sort them by the maximum principal stress, and only perform vertex splitting on the top three mesh cells, leaving the remaining mesh cells for the next rendering frame for further evaluation. If the number is less than or equal to ten, perform vertex splitting on all mesh cells that meet the breakage criteria.

[0138] The display monitoring module performs calculation simplification processing based on the single frame calculation time. Specifically, it includes: real-time monitoring of the single frame calculation time, determining the relationship between the single frame calculation time and the preset single frame calculation time threshold, and if the single frame calculation time is greater than the preset single frame calculation time threshold, then the first simplification processing is performed, which increases the target spacing of newly released particles and resamples the existing fluid particle field to amplify the fluid particle spacing.

[0139] If the calculation time for a single frame is less than or equal to the preset single frame calculation time threshold, the current calculation accuracy will remain unchanged.

[0140] After performing the first simplification process, if the single-frame computation time is still greater than the preset single-frame computation time threshold, then the second simplification process is performed, the substep iteration of the crack processing module is paused, and the current crack topology geometry remains unchanged; if the second simplification process is performed for three consecutive frames, then the third simplification process is performed, and real-time physical computation is paused.

[0141] The current inflation volume, film thickness, and rupture location are extracted as the current balloon state parameters. The matching rupture release curve is retrieved from the release data table that is generated and stored in advance based on different inflation volumes, film thicknesses, and rupture locations. The approximate release flow rate-time data corresponding to the current balloon state parameters is output by interpolation.

[0142] The system sets different calculation time steps for each module. Specifically, the time step of the balloon force calculation module is set as the first time step and synchronized with the system rendering frame rate.

[0143] The time step of the crack handling module is set to the second time step, and multiple sub-step iterations are performed within a single rendering frame; the time step of the fluid simulation module is set to the third time step, and is matched with fluid stability conditions.

[0144] The first time step is larger than the third time step, and the third time step is larger than the second time step; the modules read and write cached data through a shared cache area;

[0145] The monitoring module is displayed through a graphical interface, specifically including: when the energy release rate reaches 80% of the critical energy release rate, the corresponding grid cell is marked as an early warning cell and displayed in the graphical interface by stress highlighting;

[0146] When the third simplification process is performed, an approximate calculation status indicator is displayed in the graphical interface; the internal pressure curve of the balloon, the maximum principal stress change curve of the membrane surface, and the balloon rupture and fluid release animation are displayed simultaneously in the graphical interface.

[0147] For example, in practical applications: when the catheter training enters the stage of rapid balloon rupture and continuous fluid release, the system load is mainly concentrated on fluid particle renewal, rupture topology changes and graphical interface refresh.

[0148] If the complete calculation path of the previous implementation method is still executed, when the number of fluid particles reaches the preset particle upper limit threshold or the number of units to be split in the same rendering frame is greater than the preset number threshold, the calculation time of a single frame may continue to increase, affecting the continuity of the training screen.

[0149] The display monitoring module continuously reads the current number of fluid particles; when the current number of fluid particles reaches 90% of the preset particle limit, it scans the fluid particle field.

[0150] The relationship between the spatial distance between adjacent fluid particle pairs and the preset fluid particle spacing, as well as the relationship between the velocity vector difference and the preset velocity threshold, is determined. If the spatial distance is less than the preset fluid particle spacing and the velocity vector difference is less than the preset velocity threshold, the adjacent fluid particle pairs are merged into a single particle. The mass of the new particle is obtained by summing the masses of the two particles, and the velocity of the new particle is calculated according to the conservation of momentum. The position of the new particle is determined by mass weighting. Otherwise, the merging is not performed.

[0151] Through this process, the system prioritizes merging particles with similar motion states and spatial proximity, avoiding premature compression of particle numbers in liquid regions near cracks where velocity changes significantly.

[0152] For mesh data processing, the display monitoring module also needs to determine the number of mesh cells that meet the fracture criteria within a single frame. When the number is greater than 10, the mesh cells are sorted by the maximum principal stress, and vertex splitting is performed only on the top three mesh cells. The remaining mesh cells are retained until the next rendering frame for further judgment. When the number is less than or equal to ten, vertex splitting is performed on all mesh cells that meet the fracture criteria.

[0153] By limiting the amount of crack topology update within a single frame, crack propagation is made to adhere to a preset number of regions with the highest maximum principal stress value.

[0154] In terms of single-frame calculation time control, the display monitoring module monitors the single-frame calculation time in real time and judges the relationship between the single-frame calculation time and the preset single-frame calculation time threshold. If the single-frame calculation time is greater than the preset single-frame calculation time threshold, the first simplification process is executed, which increases the target spacing of newly released particles and resamples the existing fluid particle field to amplify the fluid particle spacing.

[0155] The specific execution rules for resampling are as follows: construct a uniform three-dimensional voxel grid with the current fluid field bounding box of the system as the boundary, and set the increased target spacing as the side length of the voxel grid;

[0156] Iterate through all fluid particles and accumulate the mass and momentum of all particles that fall into the same voxel to calculate the total mass and total momentum of that voxel.

[0157] A unique equivalent particle is generated at the geometric center of the voxel. The calculated total mass is used to generate the equivalent particle, and the average velocity vector obtained by dividing the total momentum by the total mass is used as the velocity of the equivalent particle. All the original discrete particles in the voxel are deleted. In this way, the resampling is completed in strict accordance with the conservation law and the total number of particles is effectively reduced.

[0158] If the single-frame calculation time is less than or equal to the preset single-frame calculation time threshold, the current calculation accuracy remains unchanged. After performing the first simplification process, if the single-frame calculation time is still greater than the preset single-frame calculation time threshold, the second simplification process is performed, the substep iteration of the crack processing module is paused, and the current crack topology geometry remains unchanged, so that the system will not continue to refine the crack expansion in the subsequent frames, and only retain the liquid release calculation corresponding to the existing crack area.

[0159] If the second simplification process is performed for three consecutive frames, the third simplification process is performed, and real-time physical calculation is paused. The current inflation volume, film thickness and tear location are extracted as the current balloon state parameters. The matching rupture release curve is retrieved from the release data table that is generated and stored in advance according to different inflation volumes, film thicknesses and tear locations. The approximate release flow rate-time data corresponding to the current balloon state parameters is output through interpolation.

[0160] The construction and interpolation rules of the pre-generated and stored liquid release data table are as follows: During the non-operation period, the system uses offline high-precision fluid-solid coupled finite element simulation to take the filling amount, film thickness and crack location as orthogonal combination input items, traverses various typical working conditions, and records the discrete sequence points of instantaneous flow rate decaying with time after reaching stable rupture and liquid release, thereby constructing a multi-dimensional gridded lookup table.

[0161] When the system's real-time computing is limited and it enters the third simplified processing stage, the technical motivation shifts to using table lookup and interpolation strategies to avoid excessive consumption of computing power. In the specific interpolation calculation, the system locks the nearby known sample grid nodes that include the current state parameters in the data table and calculates the normalized parameter distance of the current state from each sample node.

[0162] The specific normalization conversion logic is as follows: For each single dimension of filling volume, film thickness and tear location, calculate the absolute value of the difference between the current balloon state parameter value and the lower limit value of the locked sample grid node interval, and then divide it by the difference between the upper limit value and the lower limit value of the sample grid node interval in that dimension, so as to map the physical absolute difference of each dimension to the dimensionless [0,1] interval, ensuring that the milliliter-level filling volume and the millimeter-level film thickness have equivalent contribution weights in the distance measurement;

[0163] A trilinear interpolation weight equation is constructed for the three dimensions of filling amount, film thickness and crack location. The product of the relative coordinate components of the current state parameters in the envelope interval of each dimension is used as the weight of each corresponding grid node. The rupture release curve of each node is multilinearly weighted and summed to obtain the approximate flow rate value at the current moment.

[0164] The algorithm pre-sets a node overlap determination condition before calculating weights: it detects the calculated total distance in real time, and if this total distance is less than or equal to a preset minimum threshold for machine precision, then... If the weight of the overlapping node is locked to one, the weight of all other envelope nodes is set to zero to avoid the calculation error of the reciprocal denominator of the distance being zero, and the rupture and release curve of the precisely matched node is directly output.

[0165] To illustrate this, consider the following quantitative example: Assume the current balloon inflation volume is 12 mL and the membrane thickness is 0.85 mm, while the release data table only has pre-stored discrete grid nodes for inflation volumes of 10 mL and 15 mL and membrane thicknesses of 0.8 mm and 0.9 mm.

[0166] The system acquires the flow time series data corresponding to these four envelope boundaries, calculates a multivariate weight matrix based on the distance of the anisotropic normalized parameter, and performs a bilinear smooth transition based on this matrix, finally synthesizing a flow curve without discontinuities that conforms to the current 12mL and 0.85mm state; the graphical interface displays an approximate calculation status indicator at this stage to distinguish between real-time calculation and table lookup output;

[0167] To ensure that the above simplified processing corresponds to the original calculation chain, the system sets different calculation time steps for each module; the time step of the balloon force calculation module is set as the first time step and synchronized with the system rendering frame rate to ensure that the balloon deformation on the interface is consistent with the operation.

[0168] The time step of the crack processing module is set to the second time step, and multiple sub-step iterations are performed within a single rendering frame in order to obtain finer fracture determination during the crack initiation stage.

[0169] The time step of the fluid simulation module is set to a third time step, which is strictly controlled by the Courant-Friedrichs-Lyuvi fluid stability condition. Its specific quantitative constraint is: third time step... It must be less than or equal to the preset safety factor. With respect to the current preset fluid particle spacing The product divided by the maximum instantaneous velocity modulus in the fluid particle field The merchants, among which a preset safety factor The range of values ​​is limited to ;

[0170] The first time step is greater than the third time step, and the third time step is greater than the second time step. The modules read and write cached data through a shared cache area. Therefore, when any module enters simplified processing, the other modules can still read the most recent valid state and continue to perform their current duties.

[0171] To address the timing differences between the first and second time steps and the issue of the static maximum principal stress not being refreshed within a single cycle, the system establishes a cross-frequency domain stress interpolation compensation mechanism.

[0172] The specific execution strategy is as follows: During the calculation output of each large cycle, the balloon stress calculation module additionally estimates the local sensitivity coefficient of the maximum principal stress of each grid element to the internal pressure of the balloon using the forward finite difference method. The specific estimation logic is as follows:

[0173] Current internal pressure of the balloon Apply a preset small pressure increment on top of the existing pressure. The structural mechanics solver is invoked to obtain the maximum principal stress of the perturbation in the corresponding mesh element under this increment. The maximum principal stress of the disturbance is compared with the current maximum principal stress of the static state. The difference divided by the small pressure increment :

[0174]

[0175] The resulting quotient is used as the local sensitivity coefficient of the mesh element, and it is written into the shared buffer along with the stress distribution data.

[0176] In each substep iteration of the rupture treatment module, the algebraic change in pressure inside the balloon caused by the update of the fluid simulation module in the previous tiny time step is extracted. This algebraic change is multiplied by the corresponding sensitivity coefficient as the local stress correction amount. The algebraic sum is added to the static maximum principal stress to obtain the real-time equivalent principal stress of the current substep.

[0177] The equivalent principal stress is used in real time to calculate the energy release rate; this solves the data isolation between multi-rate modules and ensures that the crack handling model can still effectively calculate the stress relaxation phenomenon caused by fluid release during the inter-frame gap when the stress calculation module is not active, thus maintaining the causal self-consistency of the underlying physics engine.

[0178] In addition to displaying the balloon rupture and fluid release animation, the graphical interface also simultaneously displays the internal pressure curve of the balloon and the maximum principal stress change curve of the membrane surface;

[0179] When the energy release rate reaches 80% of the critical energy release rate, the corresponding grid cell is marked as a warning cell and displayed in the graphical interface by stress highlighting to indicate that the area is approaching the fracture condition.

[0180] When the third simplification process is performed, an approximate calculation status indicator is displayed in the graphical interface, allowing the operator to distinguish whether the current release animation is still obtained from real-time physical calculations.

[0181] Since the pressure curve inside the balloon and the maximum principal stress change curve are both derived from valid results in the shared buffer, the interface can still maintain continuous display even if the tear processing module temporarily stops substep iteration, without interrupting the training process.

[0182] To ensure a smooth state transition, if the display monitoring module fails to complete the particle field scan in the current frame, the particle merging result of the previous frame is retained, and the particle mass and position are not partially updated without completing the scan.

[0183] If the number of mesh cells that meet the fracture criteria in a single frame is greater than ten, but the sorting data is not fully generated, the system will not perform the new vertex splitting for that frame, will maintain the fracture topology geometry of the previous frame, and will re-count the data in the next rendering frame.

[0184] If the system has entered the third simplified process, and there are missing current balloon state parameters in the shared buffer, then the release data table retrieval result corresponding to the most recent complete state parameters is retained until new valid parameters are written before interpolation update is performed.

[0185] During catheter training, when the volume of fluid released after balloon rupture exceeds a preset volume threshold, both the number of particles and the number of rupture updates increase rapidly.

[0186] Through the above particle merging and mesh processing, the system prioritizes retaining regions whose differences from the current principal stress distribution and fluid motion state are greater than a preset difference threshold.

[0187] Through the first, second, and third simplification processes, the system gradually reduces the amount of real-time solution required when computation time is insufficient; furthermore, by using different time steps and shared buffer arrangements, the balloon deformation, rupture evolution, release particles, and interface display can still maintain consistency throughout the same training process.

[0188] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A physics engine-based system for simulating catheter balloon rupture and fluid release, characterized in that, Running on the physics engine platform, including the following modules that all call the platform's solution interface: The balloon stress calculation module is used to receive external input data and initial water injection volume, combine balloon membrane material parameters including critical energy release rate, discretize the balloon model into a triangular mesh model and solve its deformation, and output the stress distribution data of each mesh unit of the balloon membrane surface containing the maximum principal stress. The crack processing module is used to read the stress distribution data to calculate the energy release rate, make a fracture judgment based on the maximum principal stress and the critical energy release rate, split the vertices of the triangular mesh model, update the crack connection relationship and geometric boundary, and calculate and output crack status data containing the current total effective crack area. The fluid simulation module is used to receive the rupture state data, calculate the instantaneous release volume flow rate based on the current total effective rupture area, the preset liquid density and the pressure difference inside and outside the balloon, drive the release of fluid particles and output the release state of fluid particles, and write back the updated internal pressure of the balloon to the balloon force calculation module. The display monitoring module is used to monitor the sum of the volumes of the liquid inside the balloon, the released liquid, and the liquid in transmission within each rendering frame based on the stress distribution data, crack status data, and fluid particle release status, and compare it with the initial water injection volume. It also monitors the calculation time of a single frame and outputs the display screen and status indicators.

2. The physics engine-based catheter balloon rupture and fluid release simulation system according to claim 1, characterized in that, The peripheral input data is data collected by the input peripheral, including the water injection rate, the withdrawal force acting on the catheter, and the pressure displacement acting on the balloon area; The parameters of the balloon membrane include the parameters of the Mooney-Rivlin two-parameter constitutive model, the film thickness, Young's modulus, the critical energy release rate, and the initial defect size; The balloon force calculation module assigns an initial inflation volume and an initial internal pressure to the triangular mesh model.

3. The physics engine-based catheter balloon rupture and fluid release simulation system according to claim 2, characterized in that, The balloon stress calculation module generates the stress distribution data based on the peripheral input data and the balloon membrane parameters, specifically including: A damping matrix is ​​introduced into the structural mechanics solver of the physics engine platform to calculate the deformation results caused by the displacement of each node in the triangular mesh model; Rayleigh damping based on the mass matrix and stiffness matrix is ​​adopted. The stress tensor of each mesh element is generated based on the deformation results; Extract the maximum principal stress and the principal direction vector corresponding to the maximum principal stress from the stress tensor; The maximum principal stress and the principal direction vector are written as the stress distribution data into a preset shared buffer.

4. The physics engine-based catheter balloon rupture and fluid release simulation system according to claim 3, characterized in that, The crack treatment module generates the crack state data based on the stress distribution data, specifically including: The energy release rate of each grid cell is calculated based on the maximum principal stress, the Young's modulus, and the projected equivalent half-length of the existing crack in the preset neighborhood centered on each grid cell; if the region is an intact membrane surface, the projected equivalent half-length is taken as the initial defect size. Set a delay damage counter for each grid cell, with an initial value of zero; determine the relationship between the energy release rate and the critical energy release rate; when the energy release rate is greater than or equal to the critical energy release rate, increment the delay damage counter by one; when the energy release rate is less than the critical energy release rate, clear the delay damage counter to zero. If the delayed damage counter satisfies the energy release rate being greater than or equal to the critical energy release rate within a preset number of consecutive calculation substeps, then a fracture judgment is triggered. After triggering the fracture detection, vertex splitting is performed along the local membrane tangent plane of the mesh unit and in a direction perpendicular to the main direction vector, the fracture topology geometry is updated, the current total effective fracture area is calculated, and the current total effective fracture area is output as the fracture state data.

5. The physics engine-based catheter balloon rupture and fluid release simulation system according to claim 4, characterized in that, The fluid simulation module generates the fluid particle release state based on the crack state data, specifically including: The instantaneous release volumetric flow rate is calculated based on the current total effective rupture area, the preset discharge coefficient, the liquid density, and the pressure difference inside and outside the balloon determined by the difference between the internal pressure of the balloon and the preset external environmental pressure. Based on the preset fluid particle spacing and the preset single-particle representative volume, the instantaneous release volume flow rate is discretized into a corresponding number of fluid particles according to the single-particle representative volume, and emitted along the local average normal direction of the crack opening surface. The cumulative volume of released fluid particles is counted, and the current internal pressure of the balloon is calculated by combining the initial inflation volume with the initial internal pressure of the balloon. The updated current internal pressure of the balloon is written back to the balloon force calculation module as the updated internal pressure of the balloon until the energy release rate is lower than the critical energy release rate, at which point vertex splitting stops and the fluid particle release state is generated.

6. The physics engine-based catheter balloon rupture and fluid release simulation system according to claim 5, characterized in that, The display monitoring module performs volume error correction based on the fluid particle release state, specifically including: The total volume of liquid is calculated in each rendered frame, which consists of the current internal volume of the balloon, the cumulative volume that has been released, and the volume of liquid being transported. The total volume of the liquid is compared with the initial injection volume, and the residual is calculated. If the residual is greater than 0.5% of the initial water injection volume, a volume compensation coefficient is output; if the residual is less than or equal to 0.5% of the initial water injection volume, the volume compensation coefficient is set to one. The fluid simulation module receives the volume compensation coefficient, multiplies the density of the fluid particles by the volume compensation coefficient, and performs volume compensation.

7. The physics engine-based catheter balloon rupture and fluid release simulation system according to claim 6, characterized in that, The display monitoring module performs particle merging and grid data processing based on the crack state data and the fluid particle release state, specifically including: When the current number of fluid particles reaches 90% of the preset particle limit, scan the fluid particle field; The relationship between the spatial distance between adjacent fluid particle pairs and the preset fluid particle spacing, and the relationship between the velocity vector difference and the preset velocity threshold are determined. If the spatial distance is less than the preset fluid particle spacing and the velocity vector difference is less than the preset velocity threshold, the adjacent fluid particle pairs are merged into a single particle. The mass of the new particle is obtained by summing the masses of the two particles, and the velocity of the new particle is calculated according to the conservation of momentum. The position of the new particle is determined by mass weighting. Otherwise, no merging is performed. Determine the number of mesh cells that meet the breakage criteria within a single frame. If the number is greater than 10, sort them according to the maximum principal stress, and perform vertex splitting only on the top three mesh cells, leaving the remaining mesh cells for the next rendering frame for further judgment. If the number is less than or equal to ten, perform vertex splitting on all mesh cells that meet the breakage criteria.

8. The physics engine-based catheter balloon rupture and fluid release simulation system according to claim 7, characterized in that, The display monitoring module performs a simplified calculation based on the single-frame calculation time, specifically including: The calculation time of a single frame is monitored in real time, and the relationship between the calculation time of a single frame and a preset single frame calculation time threshold is determined. If the calculation time of a single frame is greater than the preset single frame calculation time threshold, a first simplification process is performed, which increases the target spacing of newly released particles and resamples the existing fluid particle field to amplify the spacing between fluid particles. If the calculation time of a single frame is less than or equal to the preset single frame calculation time threshold, the current calculation accuracy is kept unchanged. If the single-frame calculation time is still greater than the preset single-frame calculation time threshold after the first simplification process is performed, the second simplification process is performed to pause the sub-step iteration of the crack processing module and keep the current crack topology geometry unchanged. If the second simplification process is executed for three consecutive frames, the third simplification process is executed, and real-time physics calculation is paused. The current inflation volume, the film thickness, and the rupture location are extracted as the current balloon state parameters. The matching rupture release curve is retrieved from the release data table that is generated and stored in advance based on different inflation volumes, film thicknesses, and rupture locations. The approximate release flow rate-time data corresponding to the current balloon state parameters is output by interpolation calculation.

9. The physics engine-based catheter balloon rupture and fluid release simulation system according to claim 8, characterized in that, The system sets different computation time steps for each module, specifically including: The time step of the balloon force calculation module is set to the first time step and synchronized with the system rendering frame rate. The time step of the crack processing module is set to the second time step, and multiple sub-step iterations are performed within a single rendering frame. The time step of the fluid simulation module is set to a third time step and matched with fluid stability conditions; The first time step is larger than the third time step, and the third time step is larger than the second time step; The modules read and write cached data through the shared cache area.

10. The physics engine-based catheter balloon rupture and fluid release simulation system according to claim 9, characterized in that, The display and monitoring module is presented through a graphical interface, specifically including: When the energy release rate reaches 80% of the critical energy release rate, the corresponding grid cell is marked as an early warning cell and displayed in the graphical interface using stress highlighting. When the third simplification process is performed, an approximate calculation status indicator is displayed in the graphical interface; The graphical interface simultaneously displays the internal pressure curve of the balloon, the maximum principal stress change curve of the membrane surface, and the balloon rupture and fluid release animation.

Images