Floating air ball icing flight performance high-fidelity prediction method

Abstract

The invention discloses a floating air ball icing flight performance high-fidelity prediction method and system based on multi-physical field dynamic strong coupling iteration. According to the method, flexible structure deformation, computational fluid dynamics, microscopic icing physics and aircraft macrodynamics are subjected to full-closed-loop dynamic coupling solution for the first time. According to the method, a simulation framework comprising four solvers including a structural deformation and mass updating module, a computational fluid dynamics and heat transfer module, a water drop impact and ice accumulation simulation module and a balloon flight mechanics and helium thermodynamics module and a dynamic coupling controller is constructed, and multiple rounds of data exchange and iteration are performed in each time step until a coupling system converges. Wherein ice load induced structural deformation and pneumatic grid real-time updating, pneumatic coefficient time-varying feedback and phase change latent heat bidirectional coupling are three major technical breakthroughs. According to the method, the problem of prediction distortion caused by the fact that a key feedback loop is ignored in traditional sequential simulation is thoroughly solved, complex phenomena such as height locking, oscillation lowering and asymmetric instability in the icing environment can be accurately predicted, and an unprecedented high-fidelity numerical tool is provided for safety design, task planning and real-time risk assessment of the floating air ball.

Description

Technical Field

[0001] This invention relates to the field of numerical simulation and flight safety technology for high-altitude airships, specifically to a high-fidelity numerical prediction method for the dynamic strong coupling effect between the surface ice accumulation process of stratospheric airships flying in clouds containing supercooled water droplets and the balloon's structural deformation, aerodynamic characteristics, and thermodynamic state. Background Technology

[0002] Stratospheric aerosol balloons are crucial platforms for long-endurance scientific missions such as astronomical observation, atmospheric sounding, and Earth observation. Their flight safety and mission success rate highly depend on accurate prediction of complex thermodynamic and dynamic environments. When a balloon passes through clouds with temperatures below 0°C (such as cirrus and altocumulus), encountering supercooled liquid water (SLW) can cause ice accumulation on the skin surface.

[0003] Existing technologies have fundamental limitations in addressing this problem, mainly in simplifying or decoupling the coupling process, as follows:

[0004] 1) The "icing-aerodynamic" decoupling assumption: Most existing methods employ a sequential simulation process, first calculating the flow and thermal fields under a pre-defined fixed geometry, then calculating the ice accumulation, and finally applying the ice as a static mass increment to the dynamic model. This method completely ignores the real-time changes in the balloon's local shape caused by ice growth (such as sinking or bulging), which immediately affect the surrounding flow field structure, surface pressure distribution, convective heat transfer coefficient, and overall aerodynamic drag coefficient C. d This is the key feedback loop.

[0005] 2) Neglecting the time-varying nature of "shape-drag": Traditional dynamic models generally adopt a constant C based on a clean shape. d Value. However, the increased surface roughness and intensified local flow field separation caused by ice accumulation lead to C d Becoming a time variable C d (t). Ignoring its variation would significantly overestimate the buoyancy surplus, leading to optimistic misjudgments about ascent speed, hang time, and cloud penetration capability.

[0006] 3) Unidirectional thermodynamic-icing transfer: Although existing thermodynamic models can calculate skin temperature, they are usually used as static boundary conditions for icing models. In reality, ice growth itself is a huge latent heat sink for phase change, which will drastically change the local skin temperature field, thereby affecting the freezing efficiency of subsequent water droplets and the morphology of ice (frost or clear ice). This bidirectional thermal coupling has not been fully considered.

[0007] 4) Lack of "Structural Response": Existing research has almost completely neglected the local structural deformation caused by ice loads. The skin of the aerosphere is a flexible thin film, and the added mass of local ice layers can lead to significant non-uniform deformation. This deformation not only changes the aerodynamic shape but may also cause stress concentration, affecting structural integrity. This crucial physical process is completely missing in existing simulations.

[0008] These simplifications result in significant biases in the existing technology when predicting balloon behavior in persistent or severe icing scenarios. They fail to accurately warn of dangerous conditions such as floating lock, uncontrolled descent, or aerodynamic instability that may be caused by coupling effects, severely restricting the reliable application and mission planning of floating balloons under complex weather conditions.

[0009] Therefore, developing a simulation method that can accurately characterize the dynamic and strongly coupled process of "ice accumulation-structural deformation-aerodynamic response-thermodynamic state" has become a key technical bottleneck for improving the flight safety and mission reliability of aerosol balloons. Summary of the Invention

[0010] The purpose of this invention is to break through the limitations of traditional sequential simulation and provide a high-fidelity, high-efficiency numerical prediction method and system for solving the multiphysics dynamic strongly coupled problem in the icing flight of airborne balloons. Specifically, this invention aims to achieve the following four objectives: 1) Establish a two-way coupled model between ice accumulation and the dynamic deformation of the balloon's flexible structure to simulate the skin deformation caused by ice load and its feedback on the flow field. 2) Achieve real-time feedback calculation of aerodynamic characteristics (especially drag coefficient) as the shape changes, introducing aerodynamic parameters as time-varying variables into the flight dynamics equations. 3) Construct a two-way thermally coupled solution for the latent heat of icing phase change and the skin thermodynamic field to achieve iterative balance between the surface temperature field and the icing rate. 4) Develop an efficient and robust numerical iterative solution strategy suitable for this strongly coupled problem, ensuring computational convergence and efficiency.

[0011] To achieve the above objectives, this invention proposes a high-fidelity prediction method and system for the icing flight performance of aerosol balloons based on multi-physics dynamic strong coupling iteration, comprising the following five core solution modules and a bidirectional data bus between them:

[0012] 1) The structural deformation and mass update module primarily calculates skin deformation caused by ice loads and internal pressure, and updates the computational mesh. Its inputs are the surface pressure distribution p(x) from the aerodynamic module and the local ice mass increment distribution Δm from the icing module. ice (x) and the total volume change ΔV from the flight mechanics module, output as the updated computational grid node coordinates of the balloon surface. Total mass m total The position of the centroid x cg .

[0013] 2) The Computational Fluid Dynamics and Heat Transfer module primarily solves the unsteady compressible Navier-Stokes equations and energy equations to obtain the flow and thermal fields. Its input is the updated surface mesh X. surf The current flight velocity vector V from the flight mechanics module ∞ The environmental atmospheric parameters are output as surface pressure distribution p(x), shear stress distribution τ(x), convective heat transfer coefficient distribution h(x), and surface temperature distribution T. surf (x), instantaneous aerodynamic drag coefficient C d Lift coefficient C l .

[0014] 3) The water droplet impact and ice accumulation simulation module primarily functions to solve for the water droplet trajectory, calculate the local water droplet collection efficiency β(x), and calculate the local ice growth rate and morphology based on a thermodynamic equilibrium model (such as the Messier model). Its inputs are flow field information (velocity, pressure) and surface temperature T from the CFD module. surf Cloud environment parameters (liquid water content LWC, water droplet size distribution MVD) are output as the local ice mass increment Δm. ice (x), ice thickness distribution δ ice (x), phase change heat flux released / absorbed during freezing

[0015] 4) The Balloon Flight Mechanics and Helium Thermodynamics module primarily solves the six-degree-of-freedom equations of motion for the balloon, the helium equation of state, and the energy equation. Its input is the total mass m from the structural module. total and centroid x cg Real-time aerodynamic coefficients C from the CFD module d C l and aerodynamic torque M aero The convective heat flux from the thermal module and the external radiative heat flux are output as the balloon's flight speed V. ∞ Position, attitude angle, helium temperature T He Pressure P He Volume V, valve status.

[0016] 5) The dynamic coupling controller's main functions are to manage the data exchange sequence between modules, monitor coupling residuals, control the internal iterative loop, determine time step convergence, and coordinate the dynamic mesh update strategy. This is the "brain" that enables strongly coupled solutions.

[0017] Furthermore, this invention proposes a dynamic strongly coupled solution framework centered on "prediction-correction-iterative convergence". Its core idea is to discretize the entire coupled system in time. Within each time step, a coupler containing multiple internal iterations coordinates the data exchange and state updates between the various physics solvers until the coupling convergence condition is met, then proceeds to the next time step. This dynamic strongly coupled solution method includes the following key technologies:

[0018] 1) Coupling method between ice-induced dynamic deformation and real-time update of aerodynamic mesh

[0019] Unlike traditional methods that treat ice formation as a single mass point, this invention treats ice formation as a distributed load causing deformation of the continuous medium. For ice load mapping and equivalent nodal force calculation, the ice mass increment on each element e of the surface mesh is represented. This will be converted into equivalent static nodal loads.

[0020]

[0021] Where Ω(i) is the set of cells containing node i. It is element e in local coordinates (ξ) e ,η e The value at () corresponds to the shape function value at node i, and g is the gravity vector. This mapping ensures the conservation and correct transmission of ice loads.

[0022] For solving the rapid deformation problem of the thin film structure, the skin of the aerosol balloon is modeled as a pre-tensioned thin film, and the bending stiffness is ignored. Its governing equation is:

[0023]

[0024] Where T is the membrane force tensor (related to the initial internal pressure and material properties), w is the normal displacement vector, and f is the normal displacement vector. ice The ice load intensity is given. After discretization using the finite element method, a linear system is obtained:

[0025] Kw = F ice

[0026] Where K is the equivalent stiffness matrix. Considering the geometric nonlinearity that may be caused by large deformation of the skin, this invention uses co-rotational membrane elements for linearization, which significantly improves computational efficiency while ensuring accuracy.

[0027] For the dynamic update strategy of the aerodynamic computational mesh, this invention employs a hybrid dynamic meshing technique. For the surface boundary layer mesh, a spring approximation method or an elastic body method is used to smoothly deform the prism or tetrahedral mesh within the boundary layer according to the surface node displacement w, strictly maintaining the quality of the boundary layer mesh (such as orthogonality and aspect ratio). For the far-field mesh, rigid motion or simple linear interpolation is used to reduce unnecessary computational overhead. When excessive deformation leads to mesh quality deterioration, the coupling controller triggers local mesh regeneration, performing local reconstruction only on the severely deformed regions and connecting them to the undeformed regions via a non-matching interface.

[0028] 2) Real-time feedback of aerodynamic parameters and coupling with flight mechanics

[0029] In each coupled iteration step, the flight mechanics module receives real-time aerodynamic coefficients from the CFD module:

[0030]

[0031] Among them, C d (t) and C m (t) is directly calculated from the shape and flow field of the current iteration step. The flight mechanics module then solves the equations of motion based on this.

[0032]

[0033] In this equation, the mass m total (t) and aerodynamic force F aero All of these are time-varying variables, fully reflecting the coupling effect of mass increase and aerodynamic deterioration.

[0034] 3) Phase change latent heat bidirectional thermal coupling

[0035] At the freezing interface, the energy balance equation can be written as:

[0036]

[0037] Among them, L f The latent heat of freezing of ice, This represents the localized freezing mass rate. Traditional methods incorporate the latent heat term. As a known heat sink, this equation is integrated into the surface energy boundary conditions of the CFD / heat transfer module in this invention.

[0038] A dynamic, strongly coupled solution framework of "prediction-correction-iterative convergence" is used for the solution:

[0039] a) Prediction step: The CFD module is based on the previous iteration step. Calculate surface temperature

[0040] b) Correction step: The icing module according to... Given the flow field conditions, new calculations are performed using a thermodynamic equilibrium model. And the corresponding latent heat release rate.

[0041] c) Iteration: Feedback is sent to the CFD module to resolve the energy equation and update T. surf This process iterates under the control of the coupler until the surface heat flux and icing rate reach equilibrium.

[0042] Furthermore, the strongly coupled iterative solution steps involving inner iterative loops in this invention are as follows:

[0043] For time step t n to t n+1 (Step size Δt), initialize state vector U (0) =U n , where U = [X surf V ∞ ,m ice ,T He ,...] T It contains all key coupling variables.

[0044] Perform an inner iteration loop (index k = 0, 1, 2, ...):

[0045] Step S1: Input is the ice mass increment distribution from the previous iteration. (from S3) and helium pressure (From S4) New surface coordinates are obtained by applying a coupled method of ice-induced dynamic deformation and real-time update of aerodynamic mesh. and total mass The solution steps are as follows:

[0046] Step S1.1: Apply unit ice load The equivalent nodal forces are mapped using finite element shape functions:

[0047]

[0048] Step S1.2: Solve the governing equations of the pre-tensioned membrane to obtain the skin deformation displacement field w. (k+1) :

[0049]

[0050] Step S1.3: Based on w (k+1) A hybrid dynamic mesh strategy is used to update the CFD computation mesh and obtain new surface coordinates. and total mass

[0051] Step S2: Input the updated mesh (From S1), Current Flight Speed (From the previous iteration S4), the real-time aerodynamic drag coefficient is solved by applying real-time feedback of aerodynamic parameters coupled with flight mechanics. Surface pressure distribution p (k+1) convective heat transfer coefficient h (k+1) (x) and surface temperature The solution steps are as follows:

[0052] Step S2.1: In the mesh Solve the unsteady flow field and energy equations.

[0053] Step S2.2: Integral calculation of real-time aerodynamic drag coefficient And other aerodynamic coefficients.

[0054] Step S2.3: [The text appears to be incomplete and contains several grammatical errors. A more accurate translation would require the full context.] As a time variable, it is related to the surface pressure distribution p (k+1) convective heat transfer coefficient h (k+1) (x), surface temperature Output them together.

[0055] Step S3: Input surface temperature convective heat transfer coefficient h (k+1) (x)(from S2) and flow field information, applying two-way thermal coupling of latent phase change heat, to solve for the ice mass increment. and latent heat flux of phase change The solution steps are as follows:

[0056] Step S3.1: Calculate the local ice growth rate based on the energy balance equation

[0057]

[0058] in:

[0059]

[0060] Step S3.2: Calculate the ice mass increment distribution for this iteration step. and the corresponding latent heat flux of phase change:

[0061]

[0062] Step S3.3: [The text appears to be incomplete and contains several grammatical errors. A more accurate translation would require the full context.] This serves as a new thermal boundary condition, providing a correction for solving the energy equation in step S2 of the next iteration. Output the ice mass increment. and latent heat flux of phase change

[0063] Step S4: Input the total mass (From S1), Real-time resistance coefficient (From S2), aerodynamic torque, and thermal flux boundary conditions are integrated for solution and state update, and the updated flight speed is output. Helium temperature pressure The solution steps are as follows:

[0064] Step S4.1: In the flight dynamics equations, use the real-time drag coefficient:

[0065]

[0066] Step S4.2: Solve the six-degree-of-freedom equations of motion for the balloon and the helium energy equation.

[0067] Step S4.3: Update the flight status and helium thermodynamic state, and output the updated flight speed. Helium temperature pressure

[0068] Step S5: Assemble the new global system state vector And determine whether convergence has occurred based on the convergence criterion. The formula for calculating the coupling residual is:

[0069]

[0070] If R (k) <∈ coupling (Preset convergence tolerance, such as 10) -4 If U ≠ U, then the internal iteration converges. n+1 =U (k+1) Complete this physical time step and advance to t. n+1 Otherwise, let k = k + 1, return to step S1, and proceed to the next inner iteration.

[0071] The above steps S1 to S5 form a complete and closed internal iteration step. Steps S1, S2 and S3 respectively apply three core coupling methods: ice-induced dynamic deformation and real-time aerodynamic grid update coupling method, real-time aerodynamic parameter feedback and flight mechanics coupling method, and phase change latent heat bidirectional thermal coupling method. These methods are integrated in step S4, and the decision on whether to continue the iteration is made in step S5.

[0072] Compared with existing technologies, this invention achieves a leap from "sequential approximation" to "dynamic coupling," bringing about significant technological progress. Its beneficial effects are as follows:

[0073] 1) Through fully coupled simulation, the prediction accuracy of key performance indicators such as maximum flight altitude, rate of ascent decay, and hang time of balloons under continuous icing conditions has been improved by more than 50%. It can accurately reproduce complex phenomena such as "altitude lock-in" and "oscillating descent", achieving a revolutionary improvement in prediction accuracy.

[0074] 2) It can simulate the fluid-ice interaction in which "local icing leads to increased flow field separation, which in turn changes the downstream icing distribution", quantitatively evaluate the impact of "skin stress redistribution caused by ice load" on structural fatigue, and reveal a new physical mechanism and failure mode.

[0075] 3) It can virtually test the effects of different anti-icing coating layouts and heating power strategies, formulate safer flight corridors and emergency descent procedures, and clearly give the critical LWC, MVD and other parameter thresholds for safe balloon flight under different cloud conditions. Attached Figure Description

[0076] Figure 1 This is a schematic diagram of the overall architecture of the multiphysics dynamic strongly coupled simulation system described in this invention. It shows four core solver modules: structural deformation, CFD / heat transfer, icing, and flight mechanics / thermodynamics, as well as a central dynamic coupling controller. The key data variables exchanged between modules are clearly marked with bidirectional arrows in the diagram, such as mesh coordinates X, surface pressure p, and ice mass m. ice Flight speed V, aerodynamic coefficient C d wait.

[0077] Figure 2 This is a flowchart of the strongly coupled iterative solution process (fixed-point iteration method) within a single time step as described in this invention. It details the complete internal iterative loop from initialization, structure / mesh update, CFD solution, icing calculation to flight mechanics solution, as well as the logic of coupled residual calculation and convergence judgment. Detailed Implementation

[0078] 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.

[0079] The initial parameters for this embodiment are: supercooled clouds exist at an altitude of 12.0-13.5km, with a maximum LWC of 0.8g / m³. 3 MVD = 25μm, ambient temperature -55℃. Balloon volume V0 = 7240m³. 3The balloon features a pumpkin-shaped design, a polyethylene skin (38 μm thick, elastic modulus E = 0.8 GPa, Poisson's ratio v = 0.4), an initial helium mass of 106 kg, and a mission payload of 124 kg. It launches from sea level with an initial climb rate of 5 m / s. The coupling solution parameters are configured as follows: physical time step Δt = 5 s (determined based on the balloon's dynamic characteristic time), and coupling convergence tolerance ∈ = 1 × 10⁻⁶. -4 Maximum number of inner iterations k max =10. Based on the above parameters, predict whether the balloon can penetrate the cloud layer and reach the target altitude of 25km, and analyze its dynamic behavior.

[0080] Please see Figure 1 This invention provides a high-fidelity prediction method and system for the icing flight performance of aerosol balloons based on multi-physics dynamic strong coupling iteration, comprising the following 5 core solution modules and a bidirectional data bus between them:

[0081] 1) The structural deformation and mass update module primarily calculates skin deformation caused by ice loads and internal pressure, and updates the computational mesh. Its inputs are the surface pressure distribution p(x) from the aerodynamic module and the local ice mass increment distribution Δm from the icing module. ice (x) and the total volume change ΔV from the flight mechanics module, output as the updated computational grid node coordinates of the balloon surface. Total mass m total The position of the centroid x cg .

[0082] 2) The Computational Fluid Dynamics and Heat Transfer module primarily solves the unsteady compressible Navier-Stokes equations and energy equations to obtain the flow and thermal fields. Its input is the updated surface mesh X. surf The current flight velocity vector V from the flight mechanics module ∞ The environmental atmospheric parameters are output as surface pressure distribution p(x), shear stress distribution τ(x), convective heat transfer coefficient distribution h(x), and surface temperature distribution T. surf (x), instantaneous aerodynamic drag coefficient C d Lift coefficient C l .

[0083] 3) The water droplet impact and ice accumulation simulation module primarily functions to solve for the water droplet trajectory, calculate the local water droplet collection efficiency β(x), and calculate the local ice growth rate and morphology based on a thermodynamic equilibrium model (such as the Messier model). Its inputs are flow field information (velocity, pressure) and surface temperature T from the CFD module. surf Cloud environment parameters (liquid water content LWC, water droplet size distribution MVD) are output as the local ice mass increment Δm. ice (x), ice thickness distribution δice (x), phase change heat flux released / absorbed during freezing

[0084] 4) The Balloon Flight Mechanics and Helium Thermodynamics module primarily solves the six-degree-of-freedom equations of motion for the balloon, the helium equation of state, and the energy equation. Its input is the total mass m from the structural module. total and centroid x cg Real-time aerodynamic coefficients C from the CFD module d C l and aerodynamic torque M aero The convective heat flux from the thermal module and the external radiative heat flux are output as the balloon's flight speed V. ∞ Position, attitude angle, helium temperature T He Pressure P He Volume V, valve status.

[0085] 5) The dynamic coupling controller's main functions are to manage the data exchange sequence between modules, monitor coupling residuals, control the internal iterative loop, determine time step convergence, and coordinate the dynamic mesh update strategy. This is the "brain" that enables strongly coupled solutions.

[0086] Please see Figure 2 Based on the five core solver modules and the bidirectional data bus between them according to the present invention, the implementation steps of the present invention are determined as follows:

[0087] Upon entering the cloud layer (H = 12.0 km) at a time step, the strongly coupled process begins. Taking a typical time step t... n For example, the inner iteration process will be explained in detail:

[0088] 1. Inner iteration k = 0 (prediction step):

[0089] Step S1: Enter the structural deformation and mass update module, using the ice shape from the previous time step. and current internal pressure With the mesh unchanged, new surface coordinates are obtained by applying a coupled method of ice-induced dynamic deformation and real-time aerodynamic mesh update. and total mass

[0090] Step S2: Enter the Computational Fluid Dynamics and Heat Transfer module. The calculations, using real-time feedback of aerodynamic parameters coupled with flight mechanics, yield the real-time aerodynamic drag coefficient. Surface pressure distribution p (0) convective heat transfer coefficient h (0) (x) and surface temperature

[0091] Step S3: Enter the water droplet impact and ice accumulation simulation module, based on (In some areas below -10℃) and LWC, using bidirectional thermal coupling of latent phase change heat, significant ice growth was calculated, and the output was...

[0092] Step S4: Enter the balloon flight mechanics and helium thermodynamics module, mass increased to use (From S2), aerodynamic torque, and thermal flux boundary conditions are integrated for solution and state update, and the updated flight speed is output. Helium temperature pressure

[0093] Step S5: Assemble the new global system state vector And determine whether convergence has occurred based on the convergence criterion. The formula for calculating the coupling residual is:

[0094]

[0095] If R (0) <∈=1×10 -4 (With a preset convergence tolerance), the inner iteration converges, completing the current physical time step, and progressing to t. n+1 Otherwise, let k = k + 1, return to step S1, and perform the next inner iteration until R... (k) <∈=1×10 -4 .

[0096] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

[0097] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A dynamic strongly coupled prediction method for the icing flight performance of aerosol balloons, characterized in that, Includes the following steps: Step S1: Establish a multiphysics coupled simulation system architecture that includes modules for structural deformation and mass update, computational fluid dynamics and heat transfer, water droplet impact and ice accumulation simulation, balloon flight mechanics and helium thermodynamics, and dynamic coupling controllers. Step S2: Within each physical time step Δt, the dynamic coupling controller coordinates the execution of a coupled solution process involving multiple internal iterations until a preset coupling convergence criterion is met; wherein, the internal iteration process includes: Step S2.1: The structural deformation and mass update module updates the local ice mass increment distribution Δm from the ice accumulation simulation module. ice Using (x) and the helium volume change from the flight mechanics and thermodynamics module, solve for the deformation displacement field w of the balloon skin, and update the surface node coordinates X of the aerodynamic calculation mesh accordingly. surf ; Step S2.2: The computational fluid dynamics and heat transfer module, based on the updated mesh X surf and the current flight speed V from the flight mechanics module ∞ Solve for the flow field and temperature field, and output the surface pressure distribution p(x), convective heat transfer coefficient distribution h(x), and surface temperature distribution T. surf (x) and the real-time aerodynamic drag coefficient C d (t); Step S2.3: The water droplet impact and ice accumulation simulation module is based on the flow field information and surface temperature T from the computational fluid dynamics and heat transfer module. surf Calculate the local water droplet collection efficiency β(x) and ice growth rate, and output the local ice mass increment distribution Δm within this time step. ice (x) and latent heat flux of phase change And The feedback is sent to the computational fluid dynamics and heat transfer module as part of the energy boundary conditions; Step S2.4: The balloon flight mechanics and helium thermodynamics module is based on the total mass m from the structural deformation module. total Real-time aerodynamic drag coefficient C from the computational fluid dynamics module d Given (t) and aerodynamic torque, as well as thermal boundary conditions, solve the equations of motion for the balloon and the energy equation for the helium gas, and update the flight speed V. ∞ Location, helium temperature T He and volume V; Step S2.5: The dynamic coupling controller calculates the residual R of the state vector change in the current inner iteration step. If R is less than the convergence tolerance ∈, the physical time step is determined to be converged; otherwise, the next inner iteration is started based on the updated physical field data. Step S3: After completing the coupled solution of one physical time step, proceed to the next time step and repeat step S2 until the preset total simulation time is reached.

2. The method according to claim 1, characterized in that, In step S2.1, solving for the deformation displacement field w of the balloon skin specifically involves: calculating the local ice mass increment. The finite element shape function is used to map the loads to equivalent nodal loads. And apply it as an external load to the prestressed thin film structure model. The solution is performed; the updated aerodynamic calculation mesh adopts a hybrid dynamic mesh strategy: the surface and attached boundary layer meshes are subjected to smooth deformation based on spring approximation or elastic body model, the local mesh is regenerated in areas with excessive deformation, and the far-field meshes are subjected to rigid motion or interpolation.

3. The method according to claim 1 or 2, characterized in that, In step S2.2, the real-time aerodynamic drag coefficient C d The calculation of (t) is based on the balloon shape and flow field updated in the current inner iteration step, and these are used as time variables input into the flight dynamics equations of step S2.

4. In this process, dynamic coupling between aerodynamic forces and shape changes is achieved.

4. The method according to claim 1, characterized in that, There is a bidirectional thermal coupling between step S2.3 and step S2.

2. The latent heat flux of the phase change... The updated surface temperature T calculated by the computational fluid dynamics and heat transfer module is added as a source term to the energy balance boundary conditions of the icy surface; surf (x) is then fed back to the water droplet impact and ice accumulation simulation module to calculate the ice growth rate for the next iteration. The surface heat flow and the phase change heat of icing are balanced through internal iteration.

Images