Aerospace Science and Technology最新文献_第4页

Flexible composite phase change material with highly latent heat enables superior temperature uniformity for high-power batteries 具有高潜热的柔性复合相变材料为大功率电池提供了优越的温度均匀性

IF 5.8 1区工程技术 Q1 ENGINEERING, AEROSPACE

Aerospace Science and Technology

Pub Date : 2026-06-01 Epub Date: 2026-02-03 DOI: 10.1016/j.ast.2026.111811

Yuxiang Wang , Wei Wang , Yawei Xu , Yong Shuai

Extreme thermal control technology based on phase change heat storage has important scientific value and practical significance for ultra-high heat flux density thermal control serving aerospace, high-power electronic equipment, and semiconductor integrated circuits. This study addresses the critical limitations of traditional phase change materials (PCM), including low thermal conductivity and latent heat, leakage, and rigidity, by developing a novel flexible composite PCM (CPCM) for advanced thermal management. The CPCM was fabricated via a hybrid hot-press method using paraffin (PA) as the phase change matrix, olefin block copolymer (OBC) and styrene-ethylene-butylene-styrene (SEBS) as encapsulation supports, and expanded graphite (EG) as a thermal conductive filler. The optimized CPCM with a PA/OBC/SEBS ratio of 16:2:2 and containing 5 wt % EG exhibited a leakage rate of 4.75 % after 70 thermal cycles. Its mechanical flexibility was significantly enhanced, with the maximum strain increasing from 36 % to 122 %, while the thermal conductivity improved by 66.7 % to 0.60 W/(m·K), all while maintaining a high latent heat of up to 194 J/g. Numerical simulations of a battery pack consisting of 16 cells further demonstrate that, with PCM-based cooling, the average battery temperature can be effectively regulated and stabilized in the vicinity of the PCM melting temperature (35 °C), even under the 4C discharge condition. Specifically, the average temperature was maintained at 35.13 °C, which is 7.7 °C lower than that achieved using natural convection. These results demonstrate superior temperature uniformity and thermal management performance, highlighting the strong potential of the proposed CPCM for applications in high-power electronics and extreme operating environments.

基于相变蓄热的极端热控制技术对于服务于航空航天、大功率电子设备、半导体集成电路等领域的超高热流密度热控制具有重要的科学价值和现实意义。本研究通过开发一种用于高级热管理的新型柔性复合相变材料（CPCM），解决了传统相变材料（PCM）的关键局限性，包括低导热性和潜热、泄漏和刚性。以石蜡（PA）为相变基体，烯烃嵌段共聚物（OBC）和苯乙烯-乙烯-丁烯-苯乙烯（SEBS）为包封载体，膨胀石墨（EG）为导热填料，采用复合热压法制备CPCM。优化后的CPCM， PA/OBC/SEBS比为16:2:2，含5 wt % EG，经过70次热循环后，泄漏率为4.75 %。其机械柔韧性显著增强，最大应变从36 %增加到122 %，导热系数提高了66.7 %，达到0.60 W/(m·K)，同时保持了高达194 J/g的高潜热。对由16个电池组成的电池组的数值模拟进一步表明，即使在4C放电条件下，基于PCM的冷却可以有效地调节和稳定电池的平均温度在PCM熔化温度（35 °C）附近。具体而言，平均温度维持在35.13 °C，比自然对流低7.7 °C。这些结果证明了优越的温度均匀性和热管理性能，突出了所提出的CPCM在高功率电子和极端操作环境中的应用潜力。

{"title":"Flexible composite phase change material with highly latent heat enables superior temperature uniformity for high-power batteries","authors":"Yuxiang Wang , Wei Wang , Yawei Xu , Yong Shuai","doi":"10.1016/j.ast.2026.111811","DOIUrl":"10.1016/j.ast.2026.111811","url":null,"abstract":"<div><div>Extreme thermal control technology based on phase change heat storage has important scientific value and practical significance for ultra-high heat flux density thermal control serving aerospace, high-power electronic equipment, and semiconductor integrated circuits. This study addresses the critical limitations of traditional phase change materials (PCM), including low thermal conductivity and latent heat, leakage, and rigidity, by developing a novel flexible composite PCM (CPCM) for advanced thermal management. The CPCM was fabricated via a hybrid hot-press method using paraffin (PA) as the phase change matrix, olefin block copolymer (OBC) and styrene-ethylene-butylene-styrene (SEBS) as encapsulation supports, and expanded graphite (EG) as a thermal conductive filler. The optimized CPCM with a PA/OBC/SEBS ratio of 16:2:2 and containing 5 wt % EG exhibited a leakage rate of 4.75 % after 70 thermal cycles. Its mechanical flexibility was significantly enhanced, with the maximum strain increasing from 36 % to 122 %, while the thermal conductivity improved by 66.7 % to 0.60 W/(m·K), all while maintaining a high latent heat of up to 194 J/g. Numerical simulations of a battery pack consisting of 16 cells further demonstrate that, with PCM-based cooling, the average battery temperature can be effectively regulated and stabilized in the vicinity of the PCM melting temperature (35 °C), even under the 4C discharge condition. Specifically, the average temperature was maintained at 35.13 °C, which is 7.7 °C lower than that achieved using natural convection. These results demonstrate superior temperature uniformity and thermal management performance, highlighting the strong potential of the proposed CPCM for applications in high-power electronics and extreme operating environments.</div></div>","PeriodicalId":50955,"journal":{"name":"Aerospace Science and Technology","volume":"173 ","pages":"Article 111811"},"PeriodicalIF":5.8,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146109799","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Framework of typical defects detection method for solid rocket motor based on interpretable knowledge-driven three-dimensional modeling and X-ray digital twin simulation 基于可解释知识驱动三维建模和x射线数字孪生仿真的固体火箭发动机典型缺陷检测方法框架

IF 5.8 1区工程技术 Q1 ENGINEERING, AEROSPACE

Aerospace Science and Technology

Pub Date : 2026-06-01 Epub Date: 2026-01-24 DOI: 10.1016/j.ast.2026.111761

Liangliang Li , Peng Wang , Ruohai Di , Chao Xu , Mengyu Sun , Zhigang Lü

Solid rocket motors play a crucial role in the aerospace field, however, detecting internal defects in these motors has long been a technical challenge. This paper proposes a framework for detecting typical defects in solid rocket motors, based on interpretable knowledge-driven three-dimensional modeling and X-ray digital twin simulation. Initially, three-dimensional models of typical defects-porosity, cracks, inclusions, and debonding - are constructed using interpretable knowledge. These models are abstracted into ellipsoid geometries, fractal structures, irregular three-dimensional shapes, and locally irregular surfaces, respectively. An interaction model based on force fields is introduced to simulate defect distribution. Subsequently, a digital imaging model for typical defects is established, and X-ray digital simulation imaging is achieved using digital imaging principles. Additionally, the first simulated defect dataset for solid rocket motors is constructed. Lastly, a defect detection model with a self-supervised auxiliary branch is proposed, featuring an attention convolution residual decoding module, an advanced dynamic kernel updating module, and an self-supervised auxiliary branch module, enabling quantitative detection of various defects. Experimental results indicate that the model’s Dice coefficient is 0.7836, its precision is 0.8404, recall is 0.7557, and the Jaccard is 0.6718-significantly outperforming existing advanced methods. This research offers a novel and effective approach for detecting defects in solid rocket motors, holding significant theoretical and practical value.

固体火箭发动机在航空航天领域发挥着至关重要的作用，然而，检测这些发动机的内部缺陷一直是一个技术难题。提出了一种基于可解释知识驱动三维建模和x射线数字孪生仿真的固体火箭发动机典型缺陷检测框架。首先，使用可解释的知识构建典型缺陷的三维模型——孔隙、裂纹、夹杂物和脱粘。这些模型分别抽象为椭球几何、分形结构、不规则三维形状和局部不规则曲面。引入了一种基于力场的相互作用模型来模拟缺陷分布。随后，建立了典型缺陷的数字成像模型，利用数字成像原理实现了x射线数字模拟成像。建立了首个固体火箭发动机缺陷仿真数据集。最后，提出了一种具有自监督辅助分支的缺陷检测模型，该模型具有注意卷积残差解码模块、高级动态核更新模块和自监督辅助分支模块，能够定量检测各种缺陷。实验结果表明，该模型的Dice系数为0.7836，准确率为0.8404，召回率为0.7557，Jaccard为0.6718，显著优于现有的先进方法。该研究为固体火箭发动机缺陷检测提供了一种新颖有效的方法，具有重要的理论和实用价值。

{"title":"Framework of typical defects detection method for solid rocket motor based on interpretable knowledge-driven three-dimensional modeling and X-ray digital twin simulation","authors":"Liangliang Li , Peng Wang , Ruohai Di , Chao Xu , Mengyu Sun , Zhigang Lü","doi":"10.1016/j.ast.2026.111761","DOIUrl":"10.1016/j.ast.2026.111761","url":null,"abstract":"<div><div>Solid rocket motors play a crucial role in the aerospace field, however, detecting internal defects in these motors has long been a technical challenge. This paper proposes a framework for detecting typical defects in solid rocket motors, based on interpretable knowledge-driven three-dimensional modeling and X-ray digital twin simulation. Initially, three-dimensional models of typical defects-porosity, cracks, inclusions, and debonding - are constructed using interpretable knowledge. These models are abstracted into ellipsoid geometries, fractal structures, irregular three-dimensional shapes, and locally irregular surfaces, respectively. An interaction model based on force fields is introduced to simulate defect distribution. Subsequently, a digital imaging model for typical defects is established, and X-ray digital simulation imaging is achieved using digital imaging principles. Additionally, the first simulated defect dataset for solid rocket motors is constructed. Lastly, a defect detection model with a self-supervised auxiliary branch is proposed, featuring an attention convolution residual decoding module, an advanced dynamic kernel updating module, and an self-supervised auxiliary branch module, enabling quantitative detection of various defects. Experimental results indicate that the model’s Dice coefficient is 0.7836, its precision is 0.8404, recall is 0.7557, and the Jaccard is 0.6718-significantly outperforming existing advanced methods. This research offers a novel and effective approach for detecting defects in solid rocket motors, holding significant theoretical and practical value.</div></div>","PeriodicalId":50955,"journal":{"name":"Aerospace Science and Technology","volume":"173 ","pages":"Article 111761"},"PeriodicalIF":5.8,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146047963","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Complex nonlinear dynamic behaviors and dual-parameter multi-pulse chaotic vibrations of GPR rotating pretwisted airfoil blade for aero-engine, theory and experiment 航空发动机GPR旋转预扭翼型叶片复杂非线性动力学行为及双参数多脉冲混沌振动，理论与实验

IF 5.8 1区工程技术 Q1 ENGINEERING, AEROSPACE

Aerospace Science and Technology

Pub Date : 2026-06-01 Epub Date: 2026-01-25 DOI: 10.1016/j.ast.2026.111759

Y.Z. Lian , W. Zhang , Y.F. Zhang

Aero-engine blades exhibit large-amplitude nonlinear vibrations during continuous operation, which leads to fatigue damage accumulation of the aero-engine blades. Considering geometric large-deformation nonlinearities and modal coupling terms, Lagrange’s method is employed to establish the two-degree-of-freedom nonlinear coupled reduced-order model. In this paper, the complicated nonlinear vibration behaviors of the graphene platelets reinforced rotating pretwisted airfoil blade under combined the in-plane and transverse loads are explored. Incorporating the graphene as the reinforcement into the composite materials of the aero-engine blades can significantly enhance the specific strength (strength/density) and specific stiffness (stiffness/density) of the material. The excitation frequency of the aero-engine blades primarily falls near the natural frequencies of the first bending mode and the third torsional mode, which satisfy the 1:2 internal resonance condition. Therefore, the two modes are selected as the dominant modes, and the modal coordinates as the new degrees of freedom. Based on energy-phase method, it can be found that energy dissipation in the system leads to the disappearance of multi-pulse homoclinic orbits, which can readily induce vibration for the system. According to the amplitude-frequency response curves of the system, the first-order mode is excited under in-plane load, while the second-order mode is excited under transverse load. The parameter regions of Shilnikov-type dual-parameter multi-pulse chaotic motions offer direct theoretical guidance for the safety design of engineering blades. The validity of the proposed analytical method is verified against the Runge-Kutta numerical benchmark, yielding a maximum discrepancy of <0.5% in the steady-state response. Quantitative analysis reveals that the 1:2 internal resonance induces a saturation phenomenon in the first mode (amplitude locked at ∼9.1), while triggering a 130% surge in the second mode's amplitude (jumping from 4.9 to 11.5) as the excitation varies.

航空发动机叶片在连续工作过程中会产生较大的非线性振动，导致叶片疲劳损伤的累积。考虑几何大变形非线性和模态耦合项，采用拉格朗日方法建立了两自由度非线性耦合降阶模型。研究了石墨烯片增强预扭旋转翼型叶片在面内载荷和横向载荷联合作用下的复杂非线性振动行为。将石墨烯作为增强剂加入到航空发动机叶片复合材料中，可以显著提高材料的比强度（强度/密度）和比刚度（刚度/密度）。航空发动机叶片的激励频率主要落在第一弯曲模态和第三扭转模态的固有频率附近，满足1:2的内共振条件。因此，选择这两个模态作为主导模态，并将模态坐标作为新的自由度。基于能量相法，可以发现系统中的能量耗散导致多脉冲同斜轨道的消失，容易引起系统的振动。从系统的幅频响应曲线可以看出，一阶模态在面内载荷作用下被激发，二阶模态在横向载荷作用下被激发。希尼科夫型双参数多脉冲混沌运动的参数区为工程叶片的安全设计提供了直接的理论指导。通过龙格-库塔数值基准验证了该分析方法的有效性，稳态响应的最大偏差小于0.5%。定量分析表明，1:2的内部共振在第一模态中引起饱和现象（振幅锁定在~ 9.1），而随着激励的变化，在第二模态的振幅中触发130%的激增（从4.9跳到11.5）。

{"title":"Complex nonlinear dynamic behaviors and dual-parameter multi-pulse chaotic vibrations of GPR rotating pretwisted airfoil blade for aero-engine, theory and experiment","authors":"Y.Z. Lian , W. Zhang , Y.F. Zhang","doi":"10.1016/j.ast.2026.111759","DOIUrl":"10.1016/j.ast.2026.111759","url":null,"abstract":"<div><div>Aero-engine blades exhibit large-amplitude nonlinear vibrations during continuous operation, which leads to fatigue damage accumulation of the aero-engine blades. Considering geometric large-deformation nonlinearities and modal coupling terms, Lagrange’s method is employed to establish the two-degree-of-freedom nonlinear coupled reduced-order model. In this paper, the complicated nonlinear vibration behaviors of the graphene platelets reinforced rotating pretwisted airfoil blade under combined the in-plane and transverse loads are explored. Incorporating the graphene as the reinforcement into the composite materials of the aero-engine blades can significantly enhance the specific strength (strength/density) and specific stiffness (stiffness/density) of the material. The excitation frequency of the aero-engine blades primarily falls near the natural frequencies of the first bending mode and the third torsional mode, which satisfy the 1:2 internal resonance condition. Therefore, the two modes are selected as the dominant modes, and the modal coordinates as the new degrees of freedom. Based on energy-phase method, it can be found that energy dissipation in the system leads to the disappearance of multi-pulse homoclinic orbits, which can readily induce vibration for the system. According to the amplitude-frequency response curves of the system, the first-order mode is excited under in-plane load, while the second-order mode is excited under transverse load. The parameter regions of Shilnikov-type dual-parameter multi-pulse chaotic motions offer direct theoretical guidance for the safety design of engineering blades. The validity of the proposed analytical method is verified against the Runge-Kutta numerical benchmark, yielding a maximum discrepancy of <0.5% in the steady-state response. Quantitative analysis reveals that the 1:2 internal resonance induces a saturation phenomenon in the first mode (amplitude locked at ∼9.1), while triggering a 130% surge in the second mode's amplitude (jumping from 4.9 to 11.5) as the excitation varies.</div></div>","PeriodicalId":50955,"journal":{"name":"Aerospace Science and Technology","volume":"173 ","pages":"Article 111759"},"PeriodicalIF":5.8,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146047958","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Geometric skill learning paradigm for cellular space robots: Achieving cross-task and cross-configuration generalization 元胞空间机器人几何技能学习范式：实现跨任务和跨构形泛化

IF 5.8 1区工程技术 Q1 ENGINEERING, AEROSPACE

Aerospace Science and Technology

Pub Date : 2026-06-01 Epub Date: 2026-01-27 DOI: 10.1016/j.ast.2026.111783

Xiaomeng Liu, Dexiao An, Yu Chen, Guoliang Tang, Xin Yuan, Haiyu Gu, Bindi You

Cellular Space Robots (CSRs) function as self-reconfigurable systems with variable topologies and possess great potential for spatial non-cooperative target acquisition. However, task uncertainty and large workspaces pose significant challenges to control strategy generalization. Traditional reinforcement learning often designs policies for specific configurations, which can cause limited transferability across configurations and tasks. To address this issue, an innovative geometric skill learning paradigm is proposed, which construct a configuration-independent capability space using spinor theory, decouple the robot’s physical capabilities from its configurations, and decompose tasks into atomic skill primitives. The Bellman operator of the geometric skill is formulated using optimal transport theory, and its contraction mapping property is rigorously proven to ensure the algorithmic convergence and uniqueness of the optimal policy. On this basis, a task-independent performance bound theorem is derived and further extended to configuration and task generalization theory. The quadratic generalization error of this method is substantially lower than that of traditional algorithms under cross-configuration and zero-sample transfer conditions. The proposed method enabled universal control with only a few training samples. The simulation results indicate that it is effectively applied to 7-DOF and 5-DOF CSRs, achieving capture success rates of 96.3 % and 93.6 %, and significantly exceeds baseline algorithms in zero-shot generalization across tasks and configurations. This study establishes a complete framework from geometric distribution learning to generalization theory, providing a theoretical foundation for self-reconfigurable robotic applications and promoting the transition of robot learning from behavioral imitation to capability acquisition with substantial application potential.

元胞空间机器人作为具有可变拓扑结构的自重构系统，在空间非合作目标获取方面具有很大的潜力。然而，任务的不确定性和大的工作空间给控制策略的泛化带来了巨大的挑战。传统的强化学习通常为特定的配置设计策略，这可能导致配置和任务之间的可转移性有限。为了解决这一问题，提出了一种创新的几何技能学习范式，该范式利用旋量理论构建构型无关的能力空间，将机器人的物理能力与其构型解耦，并将任务分解为原子技能原语。利用最优传输理论构造了几何技巧的Bellman算子，并严格证明了其收缩映射性质，保证了算法的收敛性和唯一性。在此基础上，导出了与任务无关的性能界定理，并进一步推广到组态和任务泛化理论。在交叉配置和零样本转移条件下，该方法的二次泛化误差大大低于传统算法。该方法只需要少量的训练样本就能实现通用控制。仿真结果表明，该方法能够有效地应用于7自由度和5自由度CSRs，捕获成功率分别为96.3%和93.6%，在跨任务和配置的零射击泛化方面明显优于基线算法。本研究建立了从几何分布学习到泛化理论的完整框架，为机器人的自重构应用提供了理论基础，促进了机器人学习从行为模仿向能力获取的转变，具有较大的应用潜力。

{"title":"Geometric skill learning paradigm for cellular space robots: Achieving cross-task and cross-configuration generalization","authors":"Xiaomeng Liu, Dexiao An, Yu Chen, Guoliang Tang, Xin Yuan, Haiyu Gu, Bindi You","doi":"10.1016/j.ast.2026.111783","DOIUrl":"10.1016/j.ast.2026.111783","url":null,"abstract":"<div><div>Cellular Space Robots (CSRs) function as self-reconfigurable systems with variable topologies and possess great potential for spatial non-cooperative target acquisition. However, task uncertainty and large workspaces pose significant challenges to control strategy generalization. Traditional reinforcement learning often designs policies for specific configurations, which can cause limited transferability across configurations and tasks. To address this issue, an innovative geometric skill learning paradigm is proposed, which construct a configuration-independent capability space using spinor theory, decouple the robot’s physical capabilities from its configurations, and decompose tasks into atomic skill primitives. The Bellman operator of the geometric skill is formulated using optimal transport theory, and its contraction mapping property is rigorously proven to ensure the algorithmic convergence and uniqueness of the optimal policy. On this basis, a task-independent performance bound theorem is derived and further extended to configuration and task generalization theory. The quadratic generalization error of this method is substantially lower than that of traditional algorithms under cross-configuration and zero-sample transfer conditions. The proposed method enabled universal control with only a few training samples. The simulation results indicate that it is effectively applied to 7-DOF and 5-DOF CSRs, achieving capture success rates of 96.3 % and 93.6 %, and significantly exceeds baseline algorithms in zero-shot generalization across tasks and configurations. This study establishes a complete framework from geometric distribution learning to generalization theory, providing a theoretical foundation for self-reconfigurable robotic applications and promoting the transition of robot learning from behavioral imitation to capability acquisition with substantial application potential.</div></div>","PeriodicalId":50955,"journal":{"name":"Aerospace Science and Technology","volume":"173 ","pages":"Article 111783"},"PeriodicalIF":5.8,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146072612","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Effects of helium injection on thrust performance of a solid-gas hybrid rocket motor under two-phase non-equilibrium flow conditions 两相非平衡流动条件下氦气注入对固气混合火箭发动机推力性能的影响

IF 5.8 1区工程技术 Q1 ENGINEERING, AEROSPACE

Aerospace Science and Technology

Pub Date : 2026-06-01 Epub Date: 2026-01-29 DOI: 10.1016/j.ast.2026.111797

Chengke Li, Zenan Yang, Ge Wang, Haiwei Yang

This study aims to elucidate the coupled mechanisms of helium injection and two-phase non-equilibrium flow in solid-gas hybrid rocket motors (SGHRM), and to systematically evaluate thrust performance. A multiphase Euler-Lagrange model was adopted, incorporating aluminum particle combustion to simulate gas-particle interactions and accurately capture the coupled flow-combustion behavior within the motor. The influence of helium injection mass flow rate and aluminum particles size distribution is systematically analyzed. Results indicate that, under two-phase non-equilibrium flow, the maximum specific impulse gain reaches 9.63 % at a helium mass flow rate of 1.5 kg/s, which is lower than the equilibrium-flow optimum of 2.0 kg/s due to small gas-phase mass under the non-equilibrium flow. Helium reduces the molecular weight of the gaseous combustion products, increases the gas-phase exit velocity, and in turn accelerates the particle-phase through aerodynamic drag, thus enhancing the motor specific impulse. Additionally, small aluminum particles with high specific surface area experience stronger aerodynamic drag and contribute more effectively to reducing velocity and temperature lag losses. Consequently, helium injection results in a higher specific impulse gain in cases with smaller aluminum particles. Although helium injection intensifies particle temperature lag as a result of inefficient interphase heat conduction, the associated particle heating locally enhances gas-phase velocity in high-helium regions near the axis. Specifically, at equal mixing ratio, the gas-phase velocity under non-equilibrium flow conditions exceeds that under equilibrium flow conditions. In conclusion, helium injection remains significantly increasing the motor thrust performance under two-phase flow conditions, particularly with finer aluminum particles, thereby advocating for the use of fine aluminum powder in propellant formulations of SGHRMs to achieve optimal performance.

本研究旨在阐明固气混合火箭发动机（SGHRM）中氦气注入与两相非平衡流动的耦合机理，并对推力性能进行系统评价。采用多相欧拉-拉格朗日模型，结合铝颗粒燃烧模拟气-颗粒相互作用，准确捕捉电机内部流动-燃烧耦合行为。系统分析了注氦质量流量对铝颗粒粒度分布的影响。结果表明，在两相非平衡流条件下，当氦气质量流量为1.5 kg/s时，比冲增益最大可达9.63%，但由于非平衡流条件下气相质量较小，比冲增益低于平衡流最佳值2.0 kg/s。氦气降低了气体燃烧产物的分子量，增加了气相出口速度，进而通过气动阻力加速颗粒相，从而提高了电机比冲。此外，具有高比表面积的小铝颗粒经历更强的气动阻力，有助于更有效地减少速度和温度滞后损失。因此，在铝颗粒较小的情况下，氦注入导致更高的比脉冲增益。注氦虽然由于相间热传导效率低下而加剧了粒子的温度滞后，但在靠近轴的高氦区域，粒子的局部加热提高了气相速度。具体而言，在混合比相同的情况下，非平衡流动条件下的气相速度大于平衡流动条件下的气相速度。综上所述，在两相流条件下，特别是在铝颗粒更细的情况下，氦气注入仍能显著提高发动机的推力性能，因此建议在sghrm推进剂配方中使用细铝粉以达到最佳性能。

{"title":"Effects of helium injection on thrust performance of a solid-gas hybrid rocket motor under two-phase non-equilibrium flow conditions","authors":"Chengke Li, Zenan Yang, Ge Wang, Haiwei Yang","doi":"10.1016/j.ast.2026.111797","DOIUrl":"10.1016/j.ast.2026.111797","url":null,"abstract":"<div><div>This study aims to elucidate the coupled mechanisms of helium injection and two-phase non-equilibrium flow in solid-gas hybrid rocket motors (SGHRM), and to systematically evaluate thrust performance. A multiphase Euler-Lagrange model was adopted, incorporating aluminum particle combustion to simulate gas-particle interactions and accurately capture the coupled flow-combustion behavior within the motor. The influence of helium injection mass flow rate and aluminum particles size distribution is systematically analyzed. Results indicate that, under two-phase non-equilibrium flow, the maximum specific impulse gain reaches 9.63 % at a helium mass flow rate of 1.5 kg/s, which is lower than the equilibrium-flow optimum of 2.0 kg/s due to small gas-phase mass under the non-equilibrium flow. Helium reduces the molecular weight of the gaseous combustion products, increases the gas-phase exit velocity, and in turn accelerates the particle-phase through aerodynamic drag, thus enhancing the motor specific impulse. Additionally, small aluminum particles with high specific surface area experience stronger aerodynamic drag and contribute more effectively to reducing velocity and temperature lag losses. Consequently, helium injection results in a higher specific impulse gain in cases with smaller aluminum particles. Although helium injection intensifies particle temperature lag as a result of inefficient interphase heat conduction, the associated particle heating locally enhances gas-phase velocity in high-helium regions near the axis. Specifically, at equal mixing ratio, the gas-phase velocity under non-equilibrium flow conditions exceeds that under equilibrium flow conditions. In conclusion, helium injection remains significantly increasing the motor thrust performance under two-phase flow conditions, particularly with finer aluminum particles, thereby advocating for the use of fine aluminum powder in propellant formulations of SGHRMs to achieve optimal performance.</div></div>","PeriodicalId":50955,"journal":{"name":"Aerospace Science and Technology","volume":"173 ","pages":"Article 111797"},"PeriodicalIF":5.8,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146072601","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Adjusting the hovering position of the tethered multicopter to counteract stronger winds 调整系绳多旋翼机的悬停位置以对抗强风

IF 5.8 1区工程技术 Q1 ENGINEERING, AEROSPACE

Aerospace Science and Technology

Pub Date : 2026-06-01 Epub Date: 2026-01-29 DOI: 10.1016/j.ast.2026.111793

Haoyu Wei, Hanli Chen, Quan Quan

Tethered multicopters are aerial robots working in cooperation with cables. When the tethered multicopter hovers at a fixed point in strong winds, the cable will exert a force in the same direction as the aerodynamic resistance on the multicopter due to its bend, which is adverse for hovering and may lead to a crash. However, if the requirements for the positional accuracy of the tethered multicopter are relaxed, the cable can help counteract the aerodynamic resistance, thereby enabling the multicopter to counteract stronger winds. In this paper, the dynamic model of the tethered multicopter with the slack cable is first established. Then, the wind-rejection performance of the tethered multicopter is analyzed under different wind conditions and hovering positions. It is shown that the tethered multicopter that adjusts its hovering position within a specified range can counteract a stronger wind. Moreover, an impedance controller is designed for hovering, in which the position is used as the input, and the desired rotor thrust is obtained from the preset impedance model. Simulations and real-world experiments verify that the above control strategy effectively enables the tethered multicopter to counteract stronger winds. Especially in real-world experiments, the tethered multicopter can hover in winds up to 15 m/s.

系绳多旋翼机是一种空中机器人，通过电缆协同工作。当系绳多旋翼机在强风中悬停在固定点时，由于缆绳的弯曲，会对多旋翼机施加与气动阻力方向相同的力，不利于悬停，可能导致坠毁。然而，如果对系绳多旋翼机定位精度的要求放宽，缆绳可以帮助抵消空气动力学阻力，从而使多旋翼机能够抵御更强的风。本文首先建立了带松弛索的系留多旋翼机的动力学模型。然后，分析了系留多旋翼在不同风况和悬停位置下的抗风性能。结果表明，在一定范围内调整悬停位置的系留多旋翼机可以抵消较强的风力。设计了以悬停位置为输入的悬停阻抗控制器，由预先设定的阻抗模型得到所需的旋翼推力。仿真和实际实验验证了上述控制策略能有效地使系留多旋翼机抵抗更强的风。特别是在现实世界的实验中，系绳多旋翼机可以在风速高达15 米/秒的情况下悬停。

{"title":"Adjusting the hovering position of the tethered multicopter to counteract stronger winds","authors":"Haoyu Wei, Hanli Chen, Quan Quan","doi":"10.1016/j.ast.2026.111793","DOIUrl":"10.1016/j.ast.2026.111793","url":null,"abstract":"<div><div>Tethered multicopters are aerial robots working in cooperation with cables. When the tethered multicopter hovers at a fixed point in strong winds, the cable will exert a force in the same direction as the aerodynamic resistance on the multicopter due to its bend, which is adverse for hovering and may lead to a crash. However, if the requirements for the positional accuracy of the tethered multicopter are relaxed, the cable can help counteract the aerodynamic resistance, thereby enabling the multicopter to counteract stronger winds. In this paper, the dynamic model of the tethered multicopter with the slack cable is first established. Then, the wind-rejection performance of the tethered multicopter is analyzed under different wind conditions and hovering positions. It is shown that the tethered multicopter that adjusts its hovering position within a specified range can counteract a stronger wind. Moreover, an impedance controller is designed for hovering, in which the position is used as the input, and the desired rotor thrust is obtained from the preset impedance model. Simulations and real-world experiments verify that the above control strategy effectively enables the tethered multicopter to counteract stronger winds. Especially in real-world experiments, the tethered multicopter can hover in winds up to 15 m/s.</div></div>","PeriodicalId":50955,"journal":{"name":"Aerospace Science and Technology","volume":"173 ","pages":"Article 111793"},"PeriodicalIF":5.8,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146072591","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Flow field created by a plasma actuator in quiescent air under the influence of a single sessile droplet 静空气中等离子体致动器在单个无底液滴作用下产生的流场

IF 5.8 1区工程技术 Q1 ENGINEERING, AEROSPACE

Aerospace Science and Technology

Pub Date : 2026-06-01 Epub Date: 2026-01-31 DOI: 10.1016/j.ast.2026.111824

Chang Li , Xin Zhang , Zhihong Zhou

The anti/de-icing technology using a DBD (dielectric barrier discharge) plasma actuator has attracted extensive attention. However, the underlying anti/de-icing mechanism of such plasma actuators remains insufficiently addressed, and the interaction between droplets and the plasma actuator has not been investigated in detail. To address the aforementioned research gaps, the flow field induced by a symmetric DBD plasma actuator in the presence of a single sessile droplet was studied using a high-speed particle image velocimetry (PIV) system. The results indicate that the droplet reduces the discharge intensity of the plasma actuator, leading to a decrease in the wall-jet velocity induced by the actuator during the droplet's stable stage. Nevertheless, the instability of the separated shear layer is enhanced, generating a series of roll-up vortices. Additionally, a train of coherent structures is formed near the wall during the droplet distortion stage. Notably, the broken droplet acts as a "virtual electrode" and alters the actuator's discharge position. Finally, in the water film stage, the broken droplet evolves into a water film-this not only shifts the discharge position but also transforms the discharge mode from steady (in the dry state) to unsteady within the water film.

采用介质阻挡放电（DBD）等离子体作动器的防除冰技术引起了广泛的关注。然而，这种等离子体致动器的潜在防/除冰机制仍然没有得到充分的研究，液滴与等离子体致动器之间的相互作用也没有得到详细的研究。为了解决上述研究空白，使用高速粒子图像测速（PIV）系统研究了对称DBD等离子体致动器在单个无底液滴存在下诱导的流场。结果表明，液滴降低了等离子体致动器的放电强度，导致致动器在液滴稳定阶段引起的壁面射流速度降低。然而，分离剪切层的不稳定性增强，产生一系列卷升涡。此外，在液滴变形阶段，在壁面附近形成一列相干结构。值得注意的是，破碎的液滴充当“虚拟电极”并改变执行器的放电位置。最后，在水膜阶段，破碎的液滴演变成水膜，这不仅改变了排放位置，而且将水膜内的排放模式从稳态（干燥状态）转变为非稳态。

{"title":"Flow field created by a plasma actuator in quiescent air under the influence of a single sessile droplet","authors":"Chang Li , Xin Zhang , Zhihong Zhou","doi":"10.1016/j.ast.2026.111824","DOIUrl":"10.1016/j.ast.2026.111824","url":null,"abstract":"<div><div>The anti/de-icing technology using a DBD (dielectric barrier discharge) plasma actuator has attracted extensive attention. However, the underlying anti/de-icing mechanism of such plasma actuators remains insufficiently addressed, and the interaction between droplets and the plasma actuator has not been investigated in detail. To address the aforementioned research gaps, the flow field induced by a symmetric DBD plasma actuator in the presence of a single sessile droplet was studied using a high-speed particle image velocimetry (PIV) system. The results indicate that the droplet reduces the discharge intensity of the plasma actuator, leading to a decrease in the wall-jet velocity induced by the actuator during the droplet's stable stage. Nevertheless, the instability of the separated shear layer is enhanced, generating a series of roll-up vortices. Additionally, a train of coherent structures is formed near the wall during the droplet distortion stage. Notably, the broken droplet acts as a \"virtual electrode\" and alters the actuator's discharge position. Finally, in the water film stage, the broken droplet evolves into a water film-this not only shifts the discharge position but also transforms the discharge mode from steady (in the dry state) to unsteady within the water film.</div></div>","PeriodicalId":50955,"journal":{"name":"Aerospace Science and Technology","volume":"173 ","pages":"Article 111824"},"PeriodicalIF":5.8,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146089652","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Theoretical study on the primary combustion of the boron-based fuel-rich propellant for solid rocket scramjet 固体火箭超燃冲压发动机含硼富燃料推进剂一次燃烧的理论研究

IF 5.8 1区工程技术 Q1 ENGINEERING, AEROSPACE

Aerospace Science and Technology

Pub Date : 2026-06-01 Epub Date: 2026-01-30 DOI: 10.1016/j.ast.2026.111813

Chaolong Li, Lei Bao, Guobin Zhang, Zhengtao Guo, Yi An, Xianzhong Gao, Likun Ma

The study on the primary combustion of boron-based fuel-rich propellants is conducive to bridging the gap between primary combustion and secondary combustion in solid rocket scramjet and providing important information for the prediction and organization of secondary combustion processes. In this work, a theoretical investigation on the primary combustion process of the boron-based fuel-rich propellant based on thermochemical equilibrium was presented. By analyzing the mass and energy properties of the primary combustion products, it can be concluded that the high performance of the solid rocket scramjet requires that the energetic condensed products must achieve sufficient heat release in the secondary combustion process since the mass and energy fractions of the condensed combustible products are above 70%. The effects of combustion pressure and fuel-rich degree of the solid propellant on the primary combustion process were studied accordingly. The results show that the increase of pressure can improve the ignition and combustion characteristics of the primary combustion products by raising its temperature and changing the compositions. And the possible mechanism from theoretical analysis was further proposed. The increase of fuel-rich degree has both advantages and disadvantages to the overall performance of the solid rocket scramjet. The heat value of solid propellant and the existing combustion organization technology should be considered comprehensively in the design of fuel-rich degree. If the upper limit of combustion organization technology is reached, the fuel-rich degree of the solid propellant should be designed as high as possible to make scramjet carry more chemical energy, thus improving the overall performance of the solid rocket scramjet.

研究硼基富燃料推进剂的一次燃烧，有助于弥补固体火箭超燃冲压发动机一次燃烧和二次燃烧之间的差距，为二次燃烧过程的预测和组织提供重要信息。本文从热化学平衡的角度对硼基富燃料推进剂的一次燃烧过程进行了理论研究。通过对一次燃烧产物的质量和能量特性的分析，可以得出固体火箭超燃冲压发动机的高性能要求高能凝聚产物在二次燃烧过程中必须实现充分的放热，因为凝聚可燃产物的质量和能量分数在70%以上。研究了固体推进剂的燃烧压力和富油度对一次燃烧过程的影响。结果表明，压力的增加可以通过提高一次燃烧产物的温度和改变其成分来改善其点火和燃烧特性。并从理论分析上进一步提出了可能的机理。富燃料度的提高对固体火箭超燃冲压发动机的整体性能既有有利的一面，也有不利的一面。在设计富燃料度时，应综合考虑固体推进剂的热值和现有的燃烧组织技术。如果达到燃烧组织技术的上限，则应设计尽可能高的固体推进剂富燃料度，使超燃冲压发动机携带更多的化学能，从而提高固体火箭超燃冲压发动机的整体性能。

{"title":"Theoretical study on the primary combustion of the boron-based fuel-rich propellant for solid rocket scramjet","authors":"Chaolong Li, Lei Bao, Guobin Zhang, Zhengtao Guo, Yi An, Xianzhong Gao, Likun Ma","doi":"10.1016/j.ast.2026.111813","DOIUrl":"10.1016/j.ast.2026.111813","url":null,"abstract":"<div><div>The study on the primary combustion of boron-based fuel-rich propellants is conducive to bridging the gap between primary combustion and secondary combustion in solid rocket scramjet and providing important information for the prediction and organization of secondary combustion processes. In this work, a theoretical investigation on the primary combustion process of the boron-based fuel-rich propellant based on thermochemical equilibrium was presented. By analyzing the mass and energy properties of the primary combustion products, it can be concluded that the high performance of the solid rocket scramjet requires that the energetic condensed products must achieve sufficient heat release in the secondary combustion process since the mass and energy fractions of the condensed combustible products are above 70%. The effects of combustion pressure and fuel-rich degree of the solid propellant on the primary combustion process were studied accordingly. The results show that the increase of pressure can improve the ignition and combustion characteristics of the primary combustion products by raising its temperature and changing the compositions. And the possible mechanism from theoretical analysis was further proposed. The increase of fuel-rich degree has both advantages and disadvantages to the overall performance of the solid rocket scramjet. The heat value of solid propellant and the existing combustion organization technology should be considered comprehensively in the design of fuel-rich degree. If the upper limit of combustion organization technology is reached, the fuel-rich degree of the solid propellant should be designed as high as possible to make scramjet carry more chemical energy, thus improving the overall performance of the solid rocket scramjet.</div></div>","PeriodicalId":50955,"journal":{"name":"Aerospace Science and Technology","volume":"173 ","pages":"Article 111813"},"PeriodicalIF":5.8,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146089658","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Design optimization methods for a compressor rotor under inlet distortion 进口畸变下压气机转子的设计优化方法

IF 5.8 1区工程技术 Q1 ENGINEERING, AEROSPACE

Aerospace Science and Technology

Pub Date : 2026-06-01 Epub Date: 2026-01-30 DOI: 10.1016/j.ast.2026.111809

Hefang Deng , Songan Zhang , Honglin He , Mingmin Zhu , Xiaoqing Qiang , Jinfang Teng

Axial compressors are usually inevitably required to operate under inlet distortions, but the accurate prediction of the interaction between rotors and inlet distortions requires full-annulus unsteady simulations, which are computationally expensive and impractical for rapid iterative and optimization. To address the challenge of high computational cost in compressor rotor design optimization under complex distorted inflows, a design optimization method based on transfer learning is first developed in this study. Specifically, extensive single-passage steady simulation results are utilized as the source domain for pre-training the neural network, whereas a limited number of full-annulus unsteady simulations serve as the target domain for fine-tuning. Through this strategy, a high-accuracy surrogate model is established with significantly lower computational expense, enabling efficient design optimization of the compressor rotor under complex distortions. This study also finds a strong linear correlation between rotor performance under the complex distortion and radial distortion, based on which a low-cost linear correlation method is further proposed. By leveraging the strong linear correlation, the overall performance variations under the complex distortion can be effectively estimated based on the results from the radial distortion, thereby significantly reducing the computational cost associated with full-annulus unsteady simulations during the design process. The methods and findings presented in this study can provide valuable insights for the optimization of compressor rotors under inlet distortions.

轴向压气机通常不可避免地需要在进气道畸变下运行，但要准确预测转子与进气道畸变之间的相互作用，需要进行全环空非定常模拟，这不仅计算成本高，而且难以实现快速迭代和优化。针对复杂畸变流条件下压气机转子优化设计计算量大的问题，提出了一种基于迁移学习的设计优化方法。具体而言，利用大量的单通道稳态模拟结果作为预训练神经网络的源域，而利用有限数量的全环非定常模拟作为微调的目标域。通过该策略，建立了高精度的代理模型，大大降低了计算成本，实现了复杂畸变条件下压缩机转子的高效设计优化。研究还发现复合畸变下转子性能与径向畸变之间存在较强的线性相关关系，在此基础上提出了一种低成本的线性相关方法。利用强线性相关性，可以根据径向畸变的结果有效地估计复杂畸变下的整体性能变化，从而显著降低设计过程中全环空非定常模拟的计算成本。本文提出的方法和研究结果可以为进气道畸变下压气机转子的优化提供有价值的见解。

{"title":"Design optimization methods for a compressor rotor under inlet distortion","authors":"Hefang Deng , Songan Zhang , Honglin He , Mingmin Zhu , Xiaoqing Qiang , Jinfang Teng","doi":"10.1016/j.ast.2026.111809","DOIUrl":"10.1016/j.ast.2026.111809","url":null,"abstract":"<div><div>Axial compressors are usually inevitably required to operate under inlet distortions, but the accurate prediction of the interaction between rotors and inlet distortions requires full-annulus unsteady simulations, which are computationally expensive and impractical for rapid iterative and optimization. To address the challenge of high computational cost in compressor rotor design optimization under complex distorted inflows, a design optimization method based on transfer learning is first developed in this study. Specifically, extensive single-passage steady simulation results are utilized as the source domain for pre-training the neural network, whereas a limited number of full-annulus unsteady simulations serve as the target domain for fine-tuning. Through this strategy, a high-accuracy surrogate model is established with significantly lower computational expense, enabling efficient design optimization of the compressor rotor under complex distortions. This study also finds a strong linear correlation between rotor performance under the complex distortion and radial distortion, based on which a low-cost linear correlation method is further proposed. By leveraging the strong linear correlation, the overall performance variations under the complex distortion can be effectively estimated based on the results from the radial distortion, thereby significantly reducing the computational cost associated with full-annulus unsteady simulations during the design process. The methods and findings presented in this study can provide valuable insights for the optimization of compressor rotors under inlet distortions.</div></div>","PeriodicalId":50955,"journal":{"name":"Aerospace Science and Technology","volume":"173 ","pages":"Article 111809"},"PeriodicalIF":5.8,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146089752","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Transonic aerodynamic predictions with sparse edge-augmented transformers and graph attention networks: A comparative study 稀疏边缘增强变压器与图注意网络跨声速气动预测的比较研究

IF 5.8 1区工程技术 Q1 ENGINEERING, AEROSPACE

Aerospace Science and Technology

Pub Date : 2026-06-01 Epub Date: 2026-01-27 DOI: 10.1016/j.ast.2026.111762

Saad Hussain , Yang Pei , Muhammad Hammad Hassan , Yuxue Ge

Accurate prediction of aerodynamic behavior in transonic flows is critical for aerospace design, particularly in capturing complex flow phenomena dominated by shockwaves. Although Computational Fluid Dynamics (CFD) provides high-fidelity solutions, its computational cost limits its scalability. This study explores graph-based neural models as computationally efficient surrogates for CFD, using NASA’s Common Research Model (CRM). Two custom architectures, i.e., AeroFormer, a Transformer model with flow-adaptive sparse attention, and MeshGAT, a GNN with dynamic edge updates, were trained on CFD datasets validated against wind tunnel data. Sparse attention enabled AeroFormer to handle high-resolution meshes ( ∼ 166k nodes) on a single RTX 3060 GPU while preserving global flow context. The training datasets included 40, 90 and 180 cases, covering coarse to medium resolution meshes, with training in Mach (0.6-0.95) and angle of attack (-6° to 18°). The models were evaluated on their ability to predict surface pressure distributions, aerodynamic coefficients (C_L and C_D), and shock capturing capability. Under standard test conditions, AeroFormer outperformed MeshGAT, achieving an R² score of 0.976, and less than 6% average error in aerodynamic coefficients. The performance gap becomes more pronounced under edge-case scenarios, highlighting AeroFormer’s superior generalization. Despite these strengths, both models show limitations near Mach 1, where drag divergence and shock-induced separation dominate. This study highlights the efficacy of graph-based learning models for transonic aerodynamic predictions, demonstrating significant reductions in computational cost while maintaining high accuracy.

跨声速流动中气动特性的准确预测对于航空航天设计至关重要，特别是在捕捉由激波主导的复杂流动现象方面。尽管计算流体动力学（CFD）提供了高保真的解决方案，但其计算成本限制了其可扩展性。本研究利用NASA的通用研究模型（CRM），探索基于图的神经模型作为计算效率高的CFD替代品。在针对风洞数据验证的CFD数据集上训练了两种定制架构，即具有流动自适应稀疏关注的Transformer模型AeroFormer和具有动态边缘更新的GNN MeshGAT。稀疏注意力使AeroFormer能够在单个RTX 3060 GPU上处理高分辨率网格（ ~ 166k节点），同时保留全局流上下文。训练数据集包括40例、90例和180例，涵盖粗到中分辨率网格，训练速度为马赫（0.6-0.95），攻角为攻角（-6°至18°）。评估了这些模型预测表面压力分布、气动系数（CL和CD）以及冲击捕获能力的能力。在标准测试条件下，AeroFormer优于MeshGAT， R2得分为0.976，气动系数平均误差小于6%。在极端情况下，性能差距变得更加明显，突出了AeroFormer优越的通用性。尽管有这些优势，但两种模型在接近1马赫时都显示出局限性，此时阻力散度和激波分离占主导地位。该研究强调了基于图的学习模型在跨声速气动预测中的有效性，在保持高精度的同时显著降低了计算成本。

{"title":"Transonic aerodynamic predictions with sparse edge-augmented transformers and graph attention networks: A comparative study","authors":"Saad Hussain , Yang Pei , Muhammad Hammad Hassan , Yuxue Ge","doi":"10.1016/j.ast.2026.111762","DOIUrl":"10.1016/j.ast.2026.111762","url":null,"abstract":"<div><div>Accurate prediction of aerodynamic behavior in transonic flows is critical for aerospace design, particularly in capturing complex flow phenomena dominated by shockwaves. Although Computational Fluid Dynamics (CFD) provides high-fidelity solutions, its computational cost limits its scalability. This study explores graph-based neural models as computationally efficient surrogates for CFD, using NASA’s Common Research Model (CRM). Two custom architectures, i.e., AeroFormer, a Transformer model with flow-adaptive sparse attention, and MeshGAT, a GNN with dynamic edge updates, were trained on CFD datasets validated against wind tunnel data. Sparse attention enabled AeroFormer to handle high-resolution meshes ( ∼ 166k nodes) on a single RTX 3060 GPU while preserving global flow context. The training datasets included 40, 90 and 180 cases, covering coarse to medium resolution meshes, with training in Mach (0.6-0.95) and angle of attack (-6° to 18°). The models were evaluated on their ability to predict surface pressure distributions, aerodynamic coefficients (<em>C<sub>L</sub></em> and <em>C<sub>D</sub></em>), and shock capturing capability. Under standard test conditions, AeroFormer outperformed MeshGAT, achieving an <em>R</em><sup>2</sup> score of 0.976, and less than 6% average error in aerodynamic coefficients. The performance gap becomes more pronounced under edge-case scenarios, highlighting AeroFormer’s superior generalization. Despite these strengths, both models show limitations near Mach 1, where drag divergence and shock-induced separation dominate. This study highlights the efficacy of graph-based learning models for transonic aerodynamic predictions, demonstrating significant reductions in computational cost while maintaining high accuracy.</div></div>","PeriodicalId":50955,"journal":{"name":"Aerospace Science and Technology","volume":"173 ","pages":"Article 111762"},"PeriodicalIF":5.8,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146174580","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0