A method and system for predicting the combustion performance of solid propellants considering pore structure characteristics
By establishing a quantitative mapping relationship between the actual particle diameter of the binder and the equivalent characteristic length of the pores, the problem of the lack of physical basis for the value of the binder particle size was solved, and accurate prediction and efficient calculation of the combustion performance of solid propellants were achieved.
Patent Information
- Application Number
- CN202611484354.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-22
- Publication Date
- 2026-11-10
AI Technical Summary
Existing technologies have failed to effectively establish a quantitative correlation between binder particle size and the characteristic length of pores inside the propellant, resulting in distorted combustion performance predictions and an inability to accurately distinguish differences in burning rate under different porosities or pore size distributions.
By establishing a quantitative mapping relationship between the actual particle diameter of the binder and the equivalent characteristic length of the pores inside the propellant, the physical basis for determining the binder particle size value is established. Furthermore, the computational efficiency is improved by using the coarse-grained discrete element method, ensuring that the binder particle size distribution matches the pore characteristic length distribution, thereby predicting the burning rate and pressure index.
It enables accurate prediction of the combustion performance of solid propellants under different pore structures, improving the accuracy of combustion performance prediction and computational efficiency.
Smart Images

Figure CN123069709A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of solid propellant numerical simulation technology, specifically to a method and system for predicting the combustion performance of solid propellants considering pore structure characteristics. Background Technology
[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.
[0003] Solid propellants are particulate-filled energetic composite materials composed of oxidizer particles, metallic fuel particles, and a polymer binder matrix. Their macroscopic combustion performance (burning rate and pressure index) is highly dependent on the microscopic pore structure characteristics. Existing combustion performance prediction techniques typically first use numerical methods such as the Discrete Element Method (DEM) to establish a microscopic structure model of the propellant, extract microscopic structural parameters such as binder particle size and specific surface area, and then substitute them into a semi-empirical burning rate model based on the Arrhenius form, combined with the thermal feedback mechanism of the combustion surface, to predict the burning rate and pressure index.
[0004] However, the characteristic lengths of solidification shrinkage pores, bubble pores, and intergranular gap pores inside the propellant are on the same order of magnitude as the bridging scale of the binder in the intergranular gaps. Existing technologies have not established a quantitative correlation between binder particle size and pore characteristic length. The binder particle size is usually set as a fixed value or an empirical value without physical basis. As a result, the microscopic model cannot truly reflect the filling morphology of the binder in the pores, and the particle size distribution cannot reflect the distribution of pore characteristic length. This leads to distorted combustion performance prediction results, an inability to distinguish the differences in burning rate under different porosities or pore size distributions, poor prediction accuracy, and insufficient ability to identify pore structures. Summary of the Invention
[0005] This invention provides a method and system for predicting the combustion performance of solid propellants that takes into account pore structure characteristics. It establishes a quantitative mapping relationship between the actual particle diameter and distribution of binder and the equivalent characteristic length and distribution of pores inside the propellant. This gives the binder particle size value a clear physical basis and truly reflects the microscopic pore structure, thereby accurately predicting the burning rate and pressure index of solid propellants under different pore structures. This overcomes the shortcomings of existing technologies, such as poor prediction accuracy and insufficient ability to identify pore structures.
[0006] To achieve the above objectives, the present invention adopts the following technical solution: The first aspect of the present invention provides a method for predicting the combustion performance of solid propellants considering pore structure characteristics, comprising the following steps: Oxidant particles and / or metal fuel particles are generated and filled in the computational domain to form a solid particle skeleton structure. Pore identification is performed on the solid particle skeleton structure, and the equivalent feature length of the interstitial space or pores between skeleton particles is extracted to obtain the distribution of the equivalent feature length of pores. Based on the equivalent characteristic length distribution, the true particle diameter distribution of the binder phase is determined so that the true particle diameter of each binder is matched with the equivalent characteristic length of the corresponding pore within a preset tolerance range. Based on the actual particle diameter distribution, the volumetric surface area of the binder phase is calculated to determine the thermal feedback parameters of the combustion surface. Furthermore, combined with the burning rate model, the burning rate and pressure index of the solid propellant are predicted.
[0007] Furthermore, the equivalent characteristic length of the interstitial space or pore between the skeletal particles is extracted. Specifically, the equivalent diameter, hydraulic diameter, or maximum inscribed sphere diameter of each interstitial space or pore is calculated, and the equivalent diameter, hydraulic diameter, or maximum inscribed sphere diameter is used as the equivalent characteristic length.
[0008] Furthermore, the actual diameter of each binder particle and the equivalent characteristic length of the corresponding pore are matched within a preset tolerance range. Specifically, the equivalent diameter of the actual binder particle and the equivalent characteristic length of the pore satisfy a preset tolerance relationship, as shown in the following formula: ; in, This represents the equivalent diameter of the actual binder particles. The equivalent characteristic length of the pores, with a preset tolerance. e This represents the upper limit of the relative deviation between the equivalent diameter and the equivalent feature length.
[0009] Furthermore, the actual particle diameter distribution of the binder phase is consistent with the equivalent characteristic length distribution of the pores. Specifically, the particle size distribution function of the actual binder particles is consistent with the characteristic length distribution function of the pores. Consistency includes that the expected value and variance of the two are the same, or that the Kolmogorov-Smirnov test statistic of the cumulative distribution function of the two is less than a preset threshold.
[0010] Furthermore, after determining the actual particle diameter of the binder, the actual binder particles are equivalently mapped to coarse-grained particle elements using a scaling factor to improve computational efficiency. The equivalent calculated particle size of the coarse-grained particle elements is related to the actual particle diameter of the binder and satisfies the following equation: ; in, The equivalent calculated particle size for coarse-grained particle units. Equivalent diameter of the actual binder particles; scaling factor l The value of is determined based on the computational accuracy requirements, computational efficiency requirements, particle number control targets, and numerical stability conditions.
[0011] Furthermore, based on the principle of mass conservation, under the condition that the actual binder particles and the coarse-grained particle units have the same equivalent density and are both represented by a spherical equivalent, the mass of the coarse-grained particle unit is scaled according to a scaling factor, as shown in the following equation: ; in, The quality of coarse-grained particle units, For the actual mass of the binder particles, l This is the scaling factor.
[0012] Furthermore, the time step of the numerical model using the coarse-grained discrete element method is determined based on the scaled particle mass, contact stiffness, and stability criterion. Under the adopted contact parameter scaling strategy, the time step is amplified or corrected according to the scaling factor, and verified according to the Rayleigh time step or the contact vibration period stability criterion, as shown in the following equation: ; in, The time step of the numerical model using the coarse-grained discrete element method. The time step is the numerical simulation time using the real particle-scale discrete element method. l This is the scaling factor.
[0013] Furthermore, the number of coarse-grained binder particles generated is determined based on the target binder mass and a preset mass tolerance. For coarse-grained binder particles with a particle size distribution, the cumulative mass of each coarse-grained particle unit matches the target binder mass calculated based on the actual mass fraction of binder in the solid propellant within the preset mass tolerance range, satisfying:
[0014] in, The total number of coarse-grained particle units of the binder. The target binder mass is determined based on the actual mass fraction of binder in the solid propellant. For the density of the adhesive, Let be the equivalent particle size of the i-th coarsened binder particle unit. This refers to the allowable relative deviation between the actual total mass of the coarse-grained binder particles and the target binder mass.
[0015] Furthermore, the particle size distribution of the actual binder particles is used to characterize the thermal conduction boundary conditions and mass transport boundary conditions during the combustion process of solid propellants, so as to simulate the changes in the burning rate and pressure index of solid propellants under different actual particle size distributions.
[0016] A second aspect of the invention provides a solid propellant combustion performance prediction system that takes into account pore structure characteristics, comprising: The solid skeleton generation module is configured to generate and fill oxidant particles and / or metal fuel particles in the computational domain to form a solid particle skeleton structure. The pore identification module is configured to: identify pores in the solid particle skeleton structure, extract the interstitial space between skeleton particles or the equivalent feature length of the pores, and obtain the distribution of the equivalent feature length of the pores. The particle size mapping module is configured to: determine the true particle diameter distribution of the binder phase based on the equivalent characteristic length distribution, so that the true particle diameter of each binder is matched with the equivalent characteristic length of the corresponding pore within a preset tolerance range; The combustion performance prediction module is configured to: calculate the volumetric surface area of the binder phase based on the actual particle diameter distribution, determine the thermal feedback parameters of the combustion surface, and predict the burning rate and pressure index of the solid propellant in conjunction with the burning rate model.
[0017] Compared with existing technologies, one or more of the above technical solutions have the following beneficial effects: By establishing a quantitative correlation between the determination of the actual particle diameter of the binder and the equivalent characteristic length of the pores inside the propellant, the binder particle size is determined by the pore structure rather than being artificially set. This fundamentally solves the problem in existing technologies where the binder particle size lacks a physical basis and cannot truly reflect the filling morphology of the binder in the pore gaps. At the same time, it matches the distribution of the actual particle diameter of the binder with the distribution of the equivalent characteristic length of the pores, ensuring that the pore structure information is completely transmitted to the particle size distribution. Then, through the quantitative transmission chain of "volume specific surface area - thermal feedback parameters - burning rate model", the predicted results of burning rate and pressure index can change accordingly with changes in pore structure. This overcomes the shortcomings of existing technologies, which cannot distinguish the differences in burning rate under different porosities or pore size distributions due to the fixed binder particle size, and have poor prediction accuracy. It achieves accurate prediction of the combustion performance of solid propellants under different pore structures. Attached Figure Description
[0018] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.
[0019] Figure 1 This is a flowchart illustrating the method for predicting the combustion performance of solid propellants considering pore structure characteristics provided by the present invention. Figure 2 This is a schematic diagram of a model of the binder in a solid propellant provided by the present invention; Figure 3This is a schematic diagram comparing the traditional DEM model and the pore scale mapping CGDEM model of this invention. (a) shows the traditional DEM model where binder particles are generated using a fixed particle size; (b) shows the CGDEM model of this invention where the actual diameter of the binder particles is based on the pore size characteristic length between adjacent particles. A schematic diagram of adaptive determination. Detailed Implementation
[0020] The present invention will be further described below with reference to the accompanying drawings and embodiments.
[0021] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.
[0022] Discrete element method (DEM) is a numerical method that discretizes discontinuous media into independent particle elements and calculates the interaction forces and motion states between particles one by one through a contact model. It is the basic means of constructing microscopic models of particulate materials.
[0023] The coarse-grained discrete element method (CGDEM) is a numerical method that, based on the discrete element method, merges multiple real-scale granular elements into a significantly reduced number of coarse-grained elements according to a scaling factor. This method reduces the computational scale and improves computational efficiency while maintaining the overall mass conservation. It is the core means of this scheme to achieve efficient simulation while maintaining physical fidelity.
[0024] The equivalent characteristic length of pores is a characteristic scale parameter obtained by converting irregularly shaped gaps or pores in the solid particle skeleton through measurement methods such as equivalent diameter, hydraulic diameter, or the diameter of the largest inscribed sphere. It is used to quantitatively characterize the geometric size of pores and is the physical basis for determining the true particle diameter of the binder in this scheme.
[0025] The scaling factor λ is the ratio between the equivalent calculated particle size of the coarse-grained particle unit and the actual particle diameter of the binder during the coarse-graining equivalence process. It is used to determine the particle merging factor and to scale the particle mass and calculation time step accordingly.
[0026] The coarse-grained particle unit is a computational unit formed by equivalent mapping of multiple real binder particles according to the scaling factor. It represents the binder material assembly within a preset volume range and is the basic object that actually participates in the calculation in the CGDEM mesoscopic model.
[0027] The preset tolerance ε is the maximum allowable relative deviation between the actual particle diameter of the binder and the equivalent characteristic length of the pores, used as a quantitative criterion to define whether the two are "substantially equal".
[0028] The binder phase, a polymer matrix (such as HTPB) that fills the gaps between oxidizer particles and metal fuel particles in solid propellants and acts as a binder, is the direct object of the particle size mapping relationship and modeling process in this scheme.
[0029] Volumetric surface area, the surface area per unit volume of binder phase, is inversely proportional to the particle size of the binder and is a key intermediate parameter connecting the microscopic particle size distribution with the macroscopic combustion heat feedback process.
[0030] Burning rate and pressure index are the core characterization indicators of solid propellant combustion performance: burning rate is the distance the burning surface is propelled per unit time, and pressure index characterizes the sensitivity of burning rate to combustion chamber pressure. The two constitute the output results of the combustion performance prediction in this scheme.
[0031] The Sottle mean diameter, also known as the volume-surface area mean diameter, is the equivalent average particle size used to characterize the specific surface area of a non-uniform particle system. For spherical particle systems, it can be determined based on the ratio of the total particle volume to the total surface area. In existing technologies, mesoscopic modeling and burning rate prediction are sequentially linked: first, binder particles are generated based on empirical particle sizes to fill the gaps in the framework, and then parameters such as particle size and specific surface area are statistically analyzed for use by the burning rate model. The root of its defect lies in the disconnect between the particle size determination stage and the pore identification stage. The binder particle size does not change with the pore scale, while the characteristic length of the pores inside the propellant is on the same order of magnitude as the bridging scale of the binder. Once the particle size is decoupled from the pores, the specific surface area, local packing density, and heat conduction path of the binder / pore interface will be distorted. Specific surface area is also a key input for the thermal feedback calculation of the burning rate model, ultimately leading to inaccurate predictions of burning rate and pressure index.
[0032] Therefore, this solution provides a method and system for predicting the combustion performance of solid propellants that takes into account the characteristics of pore structure. The actual particle diameter of the binder is determined by the characteristic length of the pores, so that the two are substantially equal and have consistent distribution within a preset tolerance. A quantitative transmission chain of "pore structure → particle size distribution → specific surface area → thermal feedback → burning rate / pressure index" is established, and coarse-graining equivalence is used to ensure computational efficiency, so as to achieve high-fidelity and high-efficiency combustion performance prediction.
[0033] It should be noted that this scheme, which models the binder phase in solid propellants, can also be applied to non-solid propellant materials such as concrete, mortar, resin-bonded granular materials, and ceramic matrix granular composites, as long as their microstructure can be summarized as "particle skeleton + pore / gap space + binder phase", and the particles are mainly bonded together by the binder to form an integral structure.
[0034] This scheme establishes a quantitative mapping relationship between the actual diameter of binder particles and the characteristic length of pores within the propellant, thus clarifying the physical meaning of the actual binder particle size and accurately reflecting the filling morphology of the binder in the interstic space of the skeletal framework. Simultaneously, it correlates the particle size distribution of the actual binder particles with the characteristic length distribution of the pores, enabling the model to accurately characterize the influence mechanism of particle size distribution on propellant combustion performance (including burning rate and pressure index). Based on this, a coarse-grained discrete element method is introduced. Through a reasonable scaling factor determination strategy, the actual binder particles are equivalently mapped and the computational elements are reduced. While maintaining the matching relationship between the actual particle size and the characteristic length of the pores, this significantly reduces the computational complexity of the model and improves the computational efficiency of large-scale mesoscopic simulations. This approach can be applied in engineering to the formulation design, combustion performance prediction, and mechanical behavior analysis of solid propellants.
[0035] A method for predicting the combustion performance of solid propellants considering pore structure characteristics includes the following steps: Obtain the equivalent characteristic length of the internal pores of the solid propellant. Based on the equivalent characteristic length, determine the actual particle diameter and its distribution of the binder phase, so that the actual particle diameter and the equivalent characteristic length are substantially equal (matched) within a preset tolerance range. Obtain the true particle diameter distribution of the binder phase in the solid propellant, and construct the binder phase into multiple coarse-grained particle units, each representing a binder material aggregate within a preset volume range; Based on the actual particle diameter distribution, the volumetric surface area of the binder phase is calculated, and the thermal feedback parameters of the combustion surface are determined according to the volumetric surface area. Based on the thermal feedback parameters and the burning rate model, the predicted results of the burning rate and pressure index of the solid propellant are output.
[0036] Furthermore, ensuring that the actual particle diameter and the equivalent feature length are substantially equal within a preset tolerance range specifically includes: Equivalent diameter of actual binder particles Equivalent characteristic length of pores Satisfying the relation:
[0037] Among them, the preset tolerance e The range of values is 0 < e ≤ 0.1; Preferably, a preset tolerance is used. e The range of values is 0 < e ≤ 0.05.
[0038] Furthermore, this makes the actual particle diameter substantially equal to the equivalent feature length. It also includes: The particle size distribution function of the actual binder particles is consistent with the characteristic length distribution function of the pores; Consistency refers to the mathematical expectation of both. m and variance s 2 The Kolmogorov-Smirnov test statistic for the cumulative distribution functions of the two functions are the same, or the cumulative distribution functions of the two functions are less than the preset threshold. Preferably, the particle size distribution and characteristic length distribution are either normal distribution, log-normal distribution, or Weibull distribution.
[0039] Furthermore, the equivalent characteristic length of the pores is obtained as follows: Oxidant particles and / or metal fuel particles are generated and filled in the computational domain to form a solid particle skeleton structure; pores are identified in the solid particle skeleton structure, the gap space or pores between the skeleton particles are extracted, and the equivalent diameter, hydraulic diameter or maximum inscribed sphere diameter of each pore is calculated, with the equivalent diameter, hydraulic diameter or maximum inscribed sphere diameter as the equivalent feature length.
[0040] Furthermore, after determining the actual particle diameter of the binder, a scaling factor is used. l Real particles are equivalently mapped to coarse-grained particle units to improve computational efficiency; the equivalent computational particle size of the coarse-grained particle units is... Related to the actual particle diameter of the binder, satisfying:
[0041] in, l The scaling factor is determined based on computational accuracy requirements, computational efficiency needs, particle number control targets, and numerical stability conditions. l >0. When λ>1, the model's particle elements are coarse-grained, used to reduce the number of particles and improve computational efficiency; when... l When = 1, the model particle units maintain the true particle size of the binder; when 0 < l When the value is less than 1, the model particle unit is used for local fine modeling or model verification.
[0042] Furthermore, based on the principle of mass conservation, the mass of the coarse-grained particle unit is determined; preferably, under the condition that the actual binder particles and the coarse-grained particle units have the same equivalent density and are both represented by spherical equivalents, the mass of the coarse-grained particle unit is determined. Scaling is performed according to the scaling factor, satisfying the following:
[0043] in, This refers to the actual mass of the binder particles.
[0044] Furthermore, the time step of the numerical model using the coarse-grained discrete element method is determined based on the scaled particle mass, contact stiffness, and stability criteria; under the adopted contact parameter scaling strategy, the time step can be determined according to the scaling factor. l Enlarge or correct to satisfy:
[0045] in, The time step of the numerical model using the coarse-grained discrete element method. The time step is the numerical simulation time step of the real particle-scale discrete element method. The time step amplification or correction relationship can be checked according to the Rayleigh time step or the contact vibration period stability criterion to improve computational efficiency while maintaining numerical stability.
[0046] like Figure 1 As shown, as a preferred embodiment of this solution, the solution specifically includes the following steps: S1, oxidant particles and / or metal fuel particles are generated and filled in the computational domain to form a solid particle skeleton structure; S2, perform pore identification on the solid particle skeleton structure formed in S1, extract the equivalent feature length distribution of the gap space or pores between skeleton particles, and determine the actual particle diameter distribution of the binder. S3, based on the equivalent feature length distribution extracted in S2, generates a group of real binder particles with the same statistical distribution characteristics, so that the diameter of the real binder particles corresponds to the corresponding equivalent feature length value of the pores within a preset tolerance range. The relationship between pores and real binder particles is based on a statistical correspondence according to the equivalent feature length distribution; the number of pores is not required to be equal to or one-to-one with the number of real binder particles. When the two numbers are inconsistent, the diameter of the real binder particles is repeatedly sampled or resampled according to the equivalent feature length distribution to maintain the consistency between the distribution of the real binder particle diameter and the equivalent feature length distribution of the pores. Furthermore, a scaling factor is used to equivalently map the group of real binder particles to a group of coarse-grained binder particles. S4. The coarse-grained particles of the binder generated in S3 are filled into the gaps or pores of the solid particle skeleton to complete the construction of the coarse-grained discrete element method micro-model of the solid propellant.
[0047] S5. Based on the actual particle diameter distribution of the binder determined in S2, calculate the volumetric surface area of the binder phase, determine the thermal feedback parameters of the combustion surface, and combine with the burning rate model to predict the burning rate and pressure index of the solid propellant.
[0048] Furthermore, in S3, the number of coarse-grained binder particles generated is determined based on the target binder mass and a preset mass tolerance. The cumulative mass of each coarse-grained particle unit matches the target binder mass within the preset mass tolerance range, satisfying: ; in, The total number of coarse-grained particle units of the binder. The target binder mass is determined based on the actual mass fraction of binder in the solid propellant. For the density of the adhesive, For the first i The equivalent particle size of each coarsened binder particle unit. This refers to the allowable relative deviation between the actual total mass of the coarse-grained binder particles and the target binder mass. Preferably, a preset quality tolerance is used. satisfy: ; More preferably, ; Furthermore, the particle size distribution of the actual binder particles and the coarse particle group obtained by its equivalent mapping are used to characterize the thermal conduction boundary conditions and mass transport boundary conditions during the propellant combustion process, so as to simulate the changes in propellant burning rate and pressure index under different actual particle size distributions.
[0049] Taking a typical ternary solid propellant formulation as an example, this paper illustrates the complete implementation process of the binder modeling method in solid propellants based on the coarse-grained discrete element method. The propellant formulation is as follows: ammonium perchlorate (AP) oxidant mass fraction 68%, average particle size 50 μm; aluminum powder (Al) metallic fuel mass fraction 18%, average particle size 20 μm; hydroxyl-terminated polybutadiene (HTPB) binder mass fraction 14%, density 0.92 g / cm³.
[0050] The mass fraction of each component is used to determine the propellant formulation; during particle formation and space filling, the volume fraction of each component is obtained by converting the mass fraction and density, and the conversion relationship is as follows:
[0051] in It is the quality score. It's density. This is the volume fraction. This conversion avoids directly equating mass fraction with volume fraction.
[0052] S1: Generate and fill oxidant particles and metal fuel particles in the computational domain to form a solid framework structure.
[0053] The computational domain size was set to 1.024 mm × 1.024 mm × 1.024 mm. Particle groups were generated according to the particle size distribution and volume fraction of the oxidizer and metallic fuel in the propellant formulation: AP particles were generated according to a normal distribution (mean 50 μm, standard deviation 15 μm), and Al particles were generated according to a log-normal distribution (mean 20 μm, standard deviation 5 μm), ensuring no significant overlap between particles during generation. A strategy combining stepwise deposition and vibrational compaction was employed for filling: AP particles were first deposited to form a large particle framework, then Al particles were deposited to fill the mesoscale gaps, and finally, low-frequency harmonic vibrations were applied to promote particle rearrangement and compaction, forming a stable solid particle framework structure.
[0054] The framework is the physical carrier of the pores. Only by filling oxidizer particles and / or metallic fuel particles with the particle size distribution and volume fraction of the actual formulation can a gap space consistent with the geometry of the actual propellant be generated. The stacking method of the framework directly determines the equivalent characteristic length distribution of the pores. This step is the physical starting point for the establishment of the entire mapping relationship, avoiding the defect of unclear pore geometry in the prior art.
[0055] S2: Extract the equivalent characteristic length distribution of pores in the skeleton to determine the true particle diameter distribution of the binder.
[0056] The pores of the solid particle skeleton structure formed in S1 are identified, the gaps and pores between the skeleton particles are extracted, the equivalent characteristic length of each pore is calculated using the equal volume sphere equivalent method, and the distribution function of the pore characteristic length is obtained statistically.
[0057] In this embodiment, the characteristic length of the pores can be fitted using a log-normal distribution, with its characteristic scale mainly concentrated around 15 μm; the specific distribution parameters are determined by fitting the pore identification results. Based on limited sample statistics, the average diameter of the actual binder particles used for subsequent combustion performance calculations is approximately 13.8 μm.
[0058] True particle size of binder This represents the equivalent particle size that matches the pore space of the binder at a mesoscale. Based on the equivalent pore characteristic length obtained above... The actual particle diameter of the binder is determined according to the principle of pore size matching, so that... and They are essentially equal within the preset tolerance range.
[0059] In this embodiment, the pore feature length is mainly concentrated around 15 μm. Therefore, the actual particle diameter of the binder is determined based on the distribution of the pore feature length. Its Sottle average diameter is about 13.8 μm, which can reflect the distribution characteristics of the actual particle size of the binder as the pore size changes.
[0060] It should be noted that the equivalent characteristic length of the pores is related to the actual particle diameter of the binder. Establish scale matching relationship, coarse particle diameter It is used as the computational equivalent scale for multiple real particles, and does not require a one-to-one correspondence with the equivalent feature length of a single pore.
[0061] The characteristic lengths of solidification shrinkage pores, air bubbles, and interparticle packing pores within the propellant are on the same order of magnitude as the bridging scale of the binder in the interparticle gaps. The characteristic pore length is an objective physical measure of the binder's filling morphology. Since particle size is determined by the characteristic pore length rather than being arbitrarily set, this fundamentally solves the problem of a lack of physical basis for binder particle size selection.
[0062] S3: Generate a group of coarse-grained binder particles based on the pore length distribution and perform coarse-graining equivalent.
[0063] Based on the pore characteristic length distribution function obtained in S2, the Monte Carlo sampling method is used to generate the actual particle group of the binder, so that the particle size distribution of the actual particles has consistent statistical characteristics (including but not limited to mathematical expectation and variance) with the pore characteristic length distribution.
[0064] It should be noted that the correspondence between pores and actual binder particles is based on the statistical distribution of pore characteristic lengths. It does not require the number of pores to be equal to the number of actual binder particles, nor does it require a one-to-one correspondence. When the number of actual binder particles is greater than the number of identified pores, repeated sampling can be performed according to the pore characteristic length distribution; when the number of actual binder particles is less than the number of pores, random sampling can be performed according to the pore characteristic length distribution, ensuring that the final generated group of actual binder particles maintains statistical characteristics consistent with the pore characteristic length distribution.
[0065] For the pore characteristic length obtained through sampling The actual particle diameter of the generated binder The following tolerance requirements must be met between the sample and the pore characteristic length sample: ; This embodiment is set e = 0.05. For outliers that significantly deviate from the preset pore characteristic length range, methods such as rejection followed by resampling, boundary truncation, or replacement with neighboring values can be used to avoid unreasonable deviations in the actual particle size of the generated binder due to abnormal pore sizes.
[0066] The total mass of the binder is first determined by the volume fraction of the binder obtained from the conversion of the mass fraction of the formulation and the density of each component, and then by... r binder Calculated; according to Based on an estimated 15μm size, the equivalent mass of a single real binder particle is approximately 1.626 × 10⁻⁶. 12 kg, the required geometrically equivalent number of real particles is approximately 1.6 × 10⁻⁶. 5 indivual.
[0067] To reduce computational complexity, a scaling factor is introduced after determining the actual particle diameter. l In this embodiment, considering the requirements of comprehensive calculation accuracy, particle number, and numerical stability, we take... l = 5, then each coarse-grained particle unit approximately represents l 3 = 125 actual binder particles; the actual number of coarse binder particles involved in the calculation can be reduced to approximately 1.3 × 10⁻⁶. 3 indivual.
[0068] Meanwhile, based on the CGDEM scaling relationship, the equivalent calculated particle size of the coarse-grained particles satisfies Under the condition of consistent equivalent density and spherical equivalent representation, the mass of coarse-grained particles is determined according to... Scaling is performed.
[0069] The time step is determined based on the scaled mass, contact stiffness, and stability criteria. Under the contact parameter scaling strategy adopted in this embodiment, it can be approximately calculated as follows: The scale is increased to improve computational efficiency while maintaining numerical stability.
[0070] Particle size distribution directly determines the specific surface area, local packing density, and heat conduction path of the binder / pore interface, thus affecting the burning rate and pressure index. This step requires that the particle size distribution be consistent with the characteristic length distribution of the pores to ensure that the pore structure information is completely transferred to the particle size distribution, thus solving the problem of inaccurate predictions caused by particle size distribution distortion. Based on this, a scaling factor is used to equivalently map real particles to coarse-grained particle units and scale the mass according to mass conservation. This reduces the number of calculated particles while maintaining the pore-particle size matching relationship, solving the problems of low computational efficiency and lack of physical criteria for scaling factors in traditional methods.
[0071] S4: Fill the skeleton pores with coarse-grained binder particles to complete the construction of the CGDEM microstructure model.
[0072] The coarse binder particles generated in S3 are filled according to the pore space of the skeleton identified in S2. The available pore area between the skeleton particles is used as the initial arrangement space of the coarse binder particles. The particle center position is randomly generated according to the spatial scale of the pore area, so that the coarse binder particles are distributed in the gap area between the oxidant particles and / or metal fuel particles.
[0073] During the filling process, overlap detection is performed on the newly generated coarse-grained binder particles. When the overlap between these particles and the skeleton particles or already filled coarse-grained binder particles exceeds a preset allowable range, the particle positions are reselected; particles that meet the overlap conditions are retained. After the initial arrangement is completed, discrete element calculations are used to adjust and rearrange the particles under contact force to reduce unreasonable local overlaps generated during the initial filling process and to gradually bring the particle system to a stable state.
[0074] The number of coarse binder particles is controlled by the total number of particles determined in S3 based on the target binder mass and a preset mass tolerance. Under the condition of meeting the predetermined binder mass tolerance requirements, the rationality of particle filling is improved by reselecting pore locations and adjusting and rearranging the particles under contact force. After filling is completed and a stable state is reached, the residual porosity of the model is calculated based on the total volume of the computational domain and the volume occupied by each solid particle. In this embodiment, the residual porosity obtained after filling and reaching a stable state is approximately 2.6%, ultimately forming a CGDEM microstructure model composed of oxidant particles, metal fuel particles, and coarse binder particles.
[0075] During the aforementioned filling and particle position adjustment process, the interparticle contact employs a contact mechanics model suitable for coarse-grained systems, such as the Hertz-Mindlin model, with relevant contact parameters determined by a scaling factor. l Corrections were made to ensure that the contact mechanical response of the coarse-grained system was consistent with its equivalent characterization of the real particle system. In the final generated CGDEM mesoscopic model, the binder phase consists of approximately 1.3 × 10⁻⁶ particles. 3 Characterization of coarse-grained particle units, compared to directly using the actual particle size to establish approximately 1.6 × 10⁻⁶ particles. 5 The model uses real binder particle units, which significantly reduces the number of particles and improves computational efficiency.
[0076] In this scheme, the schematic diagram of the binder model in the solid propellant is as follows: Figure 2 As shown, a microscopic model is constructed by filling the interstitial spaces or pores of the skeleton with coarse-grained particles. The filling morphology is verified by statistically analyzing the residual porosity after filling, which supports subsequent microscopic mechanical behavior simulation, so that the particle size mapping results can be applied to a solidified model that can be used in engineering.
[0077] Model validation: The CGDEM model constructed in this embodiment is validated.
[0078] It should be noted that in this scheme, "skeleton gap space" refers to the space in the AP / Al solid skeleton that can be filled by the binder, while "residual porosity" refers to the voids that are not completely filled after the binder is applied. Statistically, the residual porosity of the model after binder application is 2.6%, which deviates from the target residual porosity of 2.8% by approximately 7.1%, showing good consistency between the two.
[0079] The Kolmogorov-Smirnov test was performed on the actual particle diameter distribution of the binder. Compared with the pore characteristic length distribution extracted in S2, the test statistic showed a higher accuracy. D = 0.038, p-value 0.97, indicating no significant difference between the two, and the distribution of the true particle diameter of the binder and the distribution of the characteristic length of the pores have good statistical consistency.
[0080] In terms of computational efficiency, this embodiment introduces CGDEM coarsening equivalence based on the actual particle size matching the pore scale, significantly reducing the number of binder particle units actually involved in the calculation. This embodiment demonstrates that this scheme can significantly improve modeling efficiency while maintaining the physical fidelity of the model, and is suitable for large-scale simulation of the microstructure of solid propellants.
[0081] S5. Based on the actual particle diameter distribution of the binder determined in S2, calculate the volumetric surface area of the binder phase, determine the thermal feedback parameters of the combustion surface, and combine with the burning rate model to predict the burning rate and pressure index of the solid propellant.
[0082] The specific application of the CGDEM model constructed according to this scheme in the prediction of solid propellant combustion performance.
[0083] Using the CGDEM mesoscopic model constructed in the above steps (AP mass fraction 68%, Al mass fraction 18%, HTPB binder mass fraction 14%), the actual particle size distribution data of the binder, determined by the pore feature length, was extracted from the model. Specifically, the equivalent diameter of the actual binder particles corresponding to each coarse-grained particle unit was statistically analyzed to obtain the particle size distribution function. P ( This distribution is consistent with the characteristic length distribution of gaps or pores in the skeleton within the model; in the limited sample statistics of this embodiment, it is mainly concentrated around 15 μm, and the average Sottle diameter is about 13.8 μm.
[0084] Based on the above particle size distribution, the volumetric surface area of the binder phase is calculated. For the spherical particle system, the volumetric surface area is... Calculate using the following formula:
[0085] in, The Sauter mean diameter is defined as follows:
[0086] The Sottle mean diameter characterizes the equivalent interface size per unit volume of binder and is a key bridging parameter connecting the micro-particle size distribution and macro-combustion thermal feedback. In this embodiment, the calculated... ≈ 13.8 μm, corresponding to a specific surface area. S sp ≈ 4.73 × 10 5 m 2 / m 3 .
[0087] The specific surface area calculated above S sp These parameters are substituted into the empirical model of solid propellant burning rate.
[0088] This embodiment uses a modified Arrhenius-type burning rate formula:
[0089] in, Linear burning rate (mm / s). This refers to the combustion chamber pressure (MPa). n The pressure index is a dimensionless parameter. The activation energy for the thermal decomposition of the binder (J / mol). R It is the universal gas constant (8.314 J / (mol·K)). The combustion surface temperature (K) is the temperature of the combustion surface. a As a pre-exponential factor, in Using mm / s, P When using MPa, its unit is mm·s. -1 ·MPa -n .
[0090] In this model, the combustion surface temperature It is regulated by solid-phase thermal feedback, and the thermal feedback coefficient is related to the volumetric surface area of the binder / pore interface. S sp There is a positive correlation, specifically:
[0091] in, The reference temperature (K) is used. β This is a proportionality constant used to characterize the contribution of volumetric surface area to surface temperature. Because... S sp The unit is m2 / m 3 , i.e. m 1 ,therefore β The unit is K·m. The physical meaning of this relationship is that the smaller the binder particle size and the larger the specific surface area, the higher the heat conduction efficiency of the binder / pore interface, the stronger the heat feedback received by the combustion surface, the higher the surface temperature, and thus promote an increase in the combustion rate.
[0092] For the pre-exponential factor in the above model a proportionality coefficient β and pressure index n The parameters, such as the pre-exponential factor, can be determined based on existing experimental data, publicly available literature data, or validated empirical parameter ranges for the propellant system under study, and can be calibrated in conjunction with macroscopic responses such as burning rate and combustion surface temperature. a The proportionality coefficient can be determined by inversion based on the burning rate data of the corresponding propellant system under reference operating conditions. β This is used to characterize the effect of changes in the volumetric surface area of the binder / pore interface on the combustion surface temperature, and can be calibrated by fitting the temperature or combustion rate response under different pore structure conditions; pressure index n The parameters can be obtained by fitting data on combustion rates under different pressure conditions. For parameters lacking direct experimental data, a reasonable range of values can first be determined based on literature reports of similar materials or similar formulation systems. Then, the parameters can be corrected by parameter sensitivity analysis and comparison of model results with existing macroscopic experimental results, so that the obtained parameters can reasonably characterize the corresponding physical processes.
[0093] Furthermore, in this embodiment, the pressure index n The calibration is performed based on the combustion rate response under different pressure conditions, and the changes are used to characterize the effect of changes in pore structure on combustion pressure sensitivity.
[0094] Specific surface area is a key intermediate quantity connecting microscopic particle size distribution and macroscopic combustion thermal feedback: the smaller the particle size and the larger the specific surface area, the higher the thermal conductivity of the binder / pore interface, the stronger the thermal feedback received by the combustion surface, the higher the surface temperature, and the greater the combustion rate. Through a quantitative chain of "specific surface area → thermal feedback parameters → combustion rate model," the prediction results can change with the pore structure, solving the problem of existing technologies being unable to distinguish the differences in combustion rate under different pore structures and having poor prediction accuracy due to fixed particle size.
[0095] To verify the ability of this scheme to identify different pore structures, while keeping the propellant formulation unchanged, three CGDEM models with different variances of pore feature length distribution were constructed by adjusting the packing density of AP and Al particles in step S1 (variances were 0.08 μm and 0.08 μm respectively). 2 0.104 μm2 (Benchmark), 0.14 μm 2 The actual particle size distribution of the binder was extracted from the three models, the specific surface area was calculated, and the results were substituted into the burning rate model for prediction.
[0096] The prediction results are shown in Table 1:
[0097] As shown in Table 1, under the pore structure control method set in this embodiment, as the variance of the pore characteristic length distribution increases, the dispersion of the actual particle size distribution of the binder increases accordingly, the Sottle mean diameter decreases, the volumetric surface area increases, and the predicted burning rate and pressure index both show a monotonically increasing trend. The physical mechanism of this trend is that, under the control method of this embodiment, the more dispersed the pore distribution, the higher the proportion of small-scale pores, which increases the proportion of small-diameter particles in the actual binder particles, resulting in an increased volumetric surface area, enhanced thermal feedback on the combustion surface, and consequently, an increase in burning rate and pressure index.
[0098] In contrast, a traditional DEM modeling method was used (the actual particle size of the binder was taken as 10 μm, which was used as a fixed particle size based on engineering experience; at the same time, the total mass of the binder was kept consistent with the present scheme and did not change with the characteristic length of the pores) to model and predict the burning rate of the same propellant formulation under the same computational domain size and particle filling conditions. The predicted burning rate at 7 MPa pressure was 7.52 mm / s, which is significantly different from the baseline burning rate of 7.61 mm / s predicted by the present scheme.
[0099] A comparative diagram of the traditional DEM model and the CGDEM model mapped to the pore size in this scheme is shown below. Figure 3 As shown, Figure 3 (a) is a schematic diagram of the binder particles generated using a fixed particle size in a traditional DEM model; Figure 3 In (b), the actual particle diameter of the binder in the CGDEM model of this scheme is based on the characteristic length of the pores between adjacent particles. A schematic diagram illustrating adaptive determination and further improved computational efficiency through coarse-grained equivalence.
[0100] Furthermore, traditional methods, due to the fixed true particle size of the binder, cannot distinguish the influence of three different pore structures (low variance, baseline, and high variance) on the burning rate. The predicted results remain unchanged among the three groups (all 7.52 mm / s). However, this scheme can effectively reflect the difference in burning rate caused by changes in pore structure (from 7.20 mm / s to 8.24 mm / s, a change of approximately 14.4%). This indicates that by establishing a quantitative mapping relationship between the true particle diameter of the binder and the characteristic length of the pores, this scheme enables the model to identify different pore structures. At the same time, by using coarse-grained equivalence, computational efficiency can be improved while maintaining this physical mapping relationship, which is difficult to achieve with existing methods.
[0101] This embodiment demonstrates that the CGDEM model established in this scheme can effectively characterize the internal pore structure of the propellant through the actual particle size distribution of the binder, thereby affecting the prediction results of burning rate and pressure index. The modeling method provided in this scheme establishes a quantitative correlation between microstructural features (pore distribution) and macroscopic combustion performance (burning rate, pressure index), which can effectively support the optimized design of propellant formulations and rapid evaluation of combustion performance. Furthermore, the numerical simulation methods and empirical models used in this embodiment are conventional techniques in the field. The core contribution of this scheme lies in providing these methods and models with input parameters reflecting the actual particle size distribution of the binder that can reflect the pore structure characteristics, and further improving computational efficiency through coarse-graining equivalence, thereby enhancing the accuracy, discriminative power, and engineering computability of combustion performance prediction.
[0102] Correspondingly, solid propellant combustion performance prediction systems that consider pore structure characteristics include: The solid skeleton generation module is configured to generate and fill oxidant particles and / or metal fuel particles in the computational domain to form a solid particle skeleton structure. The pore identification module is configured to: identify pores in the solid particle skeleton structure, extract the interstitial space between skeleton particles or the equivalent feature length of the pores, and obtain the distribution of the equivalent feature length of the pores. The particle size mapping module is configured to: determine the true particle diameter distribution of the binder phase based on the equivalent characteristic length distribution, so that the true particle diameter of each binder is matched with the equivalent characteristic length of the corresponding pore within a preset tolerance range; The combustion performance prediction module is configured to: calculate the volumetric surface area of the binder phase based on the actual particle diameter distribution, determine the thermal feedback parameters of the combustion surface, and predict the burning rate and pressure index of the solid propellant in conjunction with the burning rate model.
[0103] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A method for predicting the combustion performance of solid propellants considering pore structure characteristics, characterized in that, Includes the following steps: Oxidant particles and / or metal fuel particles are generated and filled in the computational domain to form a solid particle skeleton structure. Pore identification is performed on the solid particle skeleton structure, and the equivalent feature length of the interstitial space or pores between skeleton particles is extracted to obtain the distribution of the equivalent feature length of pores. Based on the equivalent characteristic length distribution, the true particle diameter distribution of the binder phase is determined so that the true particle diameter of each binder is matched with the equivalent characteristic length of the corresponding pore within a preset tolerance range. Based on the actual particle diameter distribution, the volumetric surface area of the binder phase is calculated to determine the thermal feedback parameters of the combustion surface. Furthermore, combined with the burning rate model, the burning rate and pressure index of the solid propellant are predicted.
2. The method for predicting the combustion performance of solid propellants considering pore structure characteristics as described in claim 1, characterized in that, Extract the equivalent characteristic length of the interstitial space or pore between the skeletal particles. Specifically, calculate the equivalent diameter, hydraulic diameter or maximum inscribed sphere diameter of each interstitial space or pore, and use the equivalent diameter, hydraulic diameter or maximum inscribed sphere diameter as the equivalent characteristic length.
3. The method for predicting the combustion performance of solid propellants considering pore structure characteristics as described in claim 1, characterized in that, The actual diameter of each binder particle and the equivalent characteristic length of the corresponding pore are matched within a preset tolerance range. Specifically, the equivalent diameter of the actual binder particle and the equivalent characteristic length of the pore satisfy a preset tolerance relationship, as shown in the following formula: ; in, This represents the equivalent diameter of the actual binder particles. The equivalent characteristic length of the pores, with a preset tolerance. ε This represents the upper limit of the relative deviation between the equivalent diameter and the equivalent feature length.
4. The method for predicting the combustion performance of solid propellants considering pore structure characteristics as described in claim 1, characterized in that, The true particle diameter distribution of the binder phase is consistent with the equivalent characteristic length distribution of the pores. Specifically, the particle size distribution function of the true binder particles is consistent with the characteristic length distribution function of the pores. Consistency includes that the expected value and variance of the two are the same, or that the Kolmogorov-Smirnov test statistic of the cumulative distribution function of the two is less than a preset threshold.
5. The method for predicting the combustion performance of solid propellants considering pore structure characteristics as described in claim 1, characterized in that, After determining the actual particle diameter of the binder, the actual binder particles are equivalently mapped to coarse-grained particle elements using a scaling factor to improve computational efficiency. The equivalent calculated particle size of the coarse-grained particle elements is related to the actual particle diameter of the binder and satisfies the following equation: ; in, The equivalent calculated particle size for coarse-grained particle units. Equivalent diameter of the actual binder particles; scaling factor λ The value of is determined based on the computational accuracy requirements, computational efficiency requirements, particle number control targets, and numerical stability conditions.
6. The method for predicting the combustion performance of solid propellants considering pore structure characteristics as described in claim 5, characterized in that, Based on the principle of mass conservation, assuming that the actual binder particles and the coarse-grained particle units have the same equivalent density and are both represented by a spherical equivalent, the mass of the coarse-grained particle unit is scaled according to a scaling factor, as shown in the following equation: ; in, The quality of coarse-grained particle units, For the actual mass of the binder particles, λ This is the scaling factor.
7. The method for predicting the combustion performance of solid propellants considering pore structure characteristics as described in claim 5, characterized in that, The time step of the numerical model using the coarse-grained discrete element method is determined based on the scaled particle mass, contact stiffness, and stability criterion. Under the adopted contact parameter scaling strategy, the time step is amplified or corrected according to the scaling factor, and checked according to the Rayleigh time step or the contact vibration period stability criterion, as shown in the following equation: ; in, The time step of the numerical model using the coarse-grained discrete element method. The time step is the numerical simulation time using the real particle-scale discrete element method. λ This is the scaling factor.
8. The method for predicting the combustion performance of solid propellants considering pore structure characteristics as described in claim 5, characterized in that, The number of coarse-grained binder particles generated is determined based on the target binder mass and a preset mass tolerance. The cumulative mass of each coarse-grained binder particle unit matches the target binder mass calculated from the actual mass fraction of binder in the solid propellant within the preset mass tolerance range, and satisfies the following: ; in, This represents the total number of coarse-grained particle units in the binder. The target binder mass is determined based on the actual mass fraction of binder in the solid propellant. For the density of the adhesive, Let be the equivalent particle size of the i-th coarse-grained binder particle unit. This refers to the allowable relative deviation between the actual total mass of the coarse-grained binder particles and the target binder mass.
9. The method for predicting the combustion performance of solid propellants considering pore structure characteristics as described in claim 1, characterized in that, The particle size distribution of the actual binder particles is used to characterize the thermal conduction boundary conditions and mass transport boundary conditions during the combustion process of solid propellants, so as to simulate the changes in the burning rate and pressure index of solid propellants under different actual particle size distributions.
10. A solid propellant combustion performance prediction system considering pore structure characteristics, characterized in that, include: The solid skeleton generation module is configured to generate and fill oxidant particles and / or metal fuel particles in the computational domain to form a solid particle skeleton structure. The pore identification module is configured to: identify pores in the solid particle skeleton structure, extract the interstitial space between skeleton particles or the equivalent feature length of the pores, and obtain the distribution of the equivalent feature length of the pores. The particle size mapping module is configured to: determine the true particle diameter distribution of the binder phase based on the equivalent characteristic length distribution, so that the true particle diameter of each binder is matched with the equivalent characteristic length of the corresponding pore within a preset tolerance range; The combustion performance prediction module is configured to: calculate the volumetric surface area of the binder phase based on the actual particle diameter distribution, determine the thermal feedback parameters of the combustion surface, and predict the burning rate and pressure index of the solid propellant in conjunction with the burning rate model.
