A numerical scheme with good spectral properties is important for the simulation of compressible flows with various of length scales for fine flow scales resolving. The MDAD-HY scheme (Li et al., 2022) using a discontinuity detector and scale sensor achieves the minimized dispersion and adaptive dissipation property. However, the discontinuity detector is devised based on the ratio of the 1st-order and 2nd-order derivatives on two sides of the interface introducing excessive numerical cost. To address this issue, an efficient hybrid WENO scheme with minimized dispersion and adaptive dissipation properties is proposed in this work. Based on the characteristic-decomposition approach, the numerical flux of the present hybrid scheme is achieved by switching between the linear MDAD scheme and the MDAD-WENO scheme according to a new efficient non-dimensional discontinuity detector. The linear flux is reconstructed in a component-wise method to decrease the characteristic-projection operations. To further improve the spectral property of the present scheme, an adaptive parameter controlling the contribution of the optimal linear scheme according to the discontinuity indicator is introduced. Several benchmark test cases involving broadband of length scales and discontinuities are adopted to verify the efficiency and the high-resolution capability of the present scheme.

With the increasing use of deep neural networks as surrogate models for accelerating computational simulations in mechanics, the application of artificial intelligence in computational fluid dynamics has seen renewed interest in recent years. However, the application of deep neural networks for flow simulations has mainly concentrated on relatively simple cases of incompressible flows. The strongly discontinuous structures that appear in compressible flows dominated by convection, such as shock waves, introduce significant challenges when approximating the nonlinear solutions or governing equations. In this work, we propose a novel physics-constrained, flow-field-message-informed (FFMI) graph neural network for spatiotemporal flow simulations of compressible flows involving strong discontinuities. To enhance the nonlinear approximation capability of strong discontinuities, a shock detector method is leveraged to extract the local flow-field messages. These messages are embedded into the graph representation to resolve the discontinuous solutions accurately. A new FFMI sample-and-aggregate-based message-passing layer, which aggregates the edge-weighted attributes with node features on different hop layers, is then developed to diffuse and process the flow-field messages. Furthermore, an end-to-end paradigm is established within the encoder–decoder framework to transform the extracted information from the flow field into latent knowledge about the underlying fluid mechanics. Finally, a variety of one- and two-dimensional cases involving strong shock waves are considered to demonstrate the effectiveness and generalizability of the proposed FFMI graph neural network.

In recent years, the applications of graph convolutional networks (GCNs) in hyperspectral image (HSI) classification have attracted much attention. However, hyperspectral classification faces problems such as complex noise effects, spectral variability, labelled training sample deficiency, and high spectral mixing between materials. Furthermore, the available GCN-based methods are computationally complex and cannot automatically adjust aggregate paths. To mitigate these issues, we propose a novel multiadaptive receptive field-based graph neural framework (MARP) for HSI classification. In our method, an adaptive receptive path aggregation (ARP) mechanism is proposed to suppress the impact of noise nodes on classification and automatically explore an adaptive receptive field, where a graph attention (GAT) neural network is introduced to learn the importance of different-sized neighbourhoods and a long short-term memory (LSTM) method is adopted to update the nodes and preserve the local convolutional features of the nodes. To address the problem that ARP may fall into a local optimum, we design a multiscale receptive mechanism. Extensive experimental results obtained on four public HSI datasets demonstrate that the proposed MARP method can mitigate oversmoothing and reduce computational complexity while achieving competitive performance when compared to several state-of-the-art methods.

The bypass dual throat nozzle (BDTN) is capable of achieving thrust vectoring by introducing secondary flow from upstream through the bypass channel. In order to investigate the aerodynamic vectoring characteristics of the axisymmetric BDTN, numerical simulations are performed to analyze the internal flow field of the nozzle at different nozzle pressure ratios (ratio of total inlet pressure to ambient pressure, NPR) in three dimensions. The results show that the influence of the bypass secondary flow causes a significant asymmetry in the parameter distribution of the flow field of the axisymmetric nozzle, and the flow also undergoes lateral expansion in the axial direction. As the total inlet pressure increases, the internal flow velocity and temperature are more stable, while the pressure and density gradually increase. Compared with the 2D configuration, the lateral expansion within the nozzle cavity reduces the degree of asymmetric difference in the internal flow field structure, resulting in a vector deflection effect. The thrust vector angle decreases as the NPR increases, and the thrust coefficient increases slightly and then decreases.

A numerical scheme with good dissipation and dispersion properties is essential for simulations of complex compressible flows in resolving small-scale structures and capturing discontinuities. Although a sufficient numerical dissipation benefits the shock-wave capturing, the fine scales in smooth regions are dissipated either. In this work, a framework to construct arbitrarily high-order spatio-temporal optimized finite difference schemes with adaptive dispersion and critical-adaptive dissipation is proposed, where the dispersion and dissipation properties are characterized by two free parameters respectively. The proposed scheme can automatically adjust these two free parameters to control the dispersion and dissipation according to the local flow-field properties quantified by the scale sensor. As the first step to optimize the spectral properties of the fully discrete scheme, the total dispersion and dissipation errors induced by spatio-temporal discretization are studied. Then, the dispersion property is optimized by a new integrated error function devised for fully discrete scheme. To improve the precision of prediction, the scale sensor employed to quantify the local scaled wavenumber of flow fields is optimized in wavenumber space. Furthermore, a modified dispersion-dissipation condition is developed to characterize the relationship between the total dispersion and dissipation errors. Building upon the optimized scale sensor and modified dispersion-dissipation condition, the critical-adaptive dissipation parameter is obtained, which keeps the numerical dissipation as low as possible to resolve more small-scale structures in the low-wavenumber region and introduces sufficient dissipation to capture strong discontinuities successfully. Moreover, the adaptive dispersion parameter related to the local wavenumber and Courant number is obtained to reduce the total dispersion error significantly. Meanwhile, the critical-adaptive dissipation surface and adaptive dispersion surface are constructed for different Courant number utilized in actual flow simulations, which improves the spectral properties of the proposed scheme. Finally, some benchmark cases involving strong discontinuities and multi-scale structures are employed to verify the attractive performance of the proposed scheme.

Due to prior knowledge deficiency, large spectral variability, and high dimension of hyperspectral image (HSI), HSI clustering is extremally a fundamental but challenging task. Deep clustering methods have achieved remarkable success and have attracted increasing attention in unsupervised HSI classification (HSIC). However, the poor robustness, adaptability, and feature presentation limit their practical applications to complex large-scale HSI datasets. Thus, this article introduces a novel self-supervised locality preserving low-pass graph convolutional embedding method (L2GCC) for large-scale hyperspectral image clustering. Specifically, a spectral–spatial transformation HSI preprocessing mechanism is introduced to learn superpixel-level spectral–spatial features from HSI and reduce the number of graph nodes for subsequent network processing. In addition, locality preserving low-pass graph convolutional embedding autoencoder is proposed, in which the low-pass graph convolution and layerwise graph attention are designed to extract the smoother features and preserve layerwise locality features, respectively. Finally, we develop a self-training strategy, in which a self-training clustering objective employs soft labels to supervise the clustering process and obtain appropriate hidden representations for node clustering. L2GCC is an end-to-end training network, which is jointly optimized by graph reconstruction loss and self-training clustering loss. On Indian Pines, Salinas, and University of Houston 2013 datasets, the clustering accuracy overall accuracies (OAs) of the proposed L2GCC are 73.51%, 83.15%, and 64.12%, respectively.

In this paper, based on our previous optimal compact schemes with minimized dispersion and controllable dissipation (OC–WENO schemes) (Sun et al. in Sci China-Phys Mech Astron 57:971–982, 2014), a spatio-temporal optimized, hybird compact-WENO scheme with minimized dispersion and critical-adaptive dissipation is developed for solving compressible flows. Firstly, the spectral properties of the fourth-order OC–WENO scheme is researched within the spatio-temporal discrete framework. In conjunction with total dispersion error of the fully scheme, an integrated error function is designed to optimize the dispersion property. Secondly, the scale sensor leveraged to quantify the local scaled wavenumber is optimized in the wavenumber space to improve the accuracy of estimating. Moreover, a dispersion-dissipation condition, controlling the relative proportion of dispersion and dissipation errors, is developed for the fully discrete scheme. Thirdly, by exploiting the optimized scale sensor and the dispersion-dissipation condition, the critical-adaptive dissipation surface is constructed to achieve the adaptive dissipation property relating to the local characteristics of the flow fields and different Courant number. To have the shock-capturing capability, the proposed compact scheme is blended with fifth-order WENO scheme to form the MDADFC–WENO scheme. Finally, a set of benchmark test cases is employed to validate the good performance of the proposed scheme.

Aiming at the problem of silo design, the silo ejector model and ejector function were proposed. The properties of the static pressure matching function were studied. The critical conditions were analyzed when the engine total pressure, the diameter of the silo, and the outlet pressure of the mixing chamber changed. It is proved that the curve that shows the inlet pressure of the mixing chamber changes along with the outlet pressure of the mixing chamber is a straight line parallel to the horizontal axis when the outlet pressure of the mixing chamber is smaller than the stagnation critical point, the curve that shows the inlet pressure of the mixing chamber changes along with the total pressure of the nozzle develops along the optimal pressure function curve when the total pressure of the nozzle is greater than the stagnation critical point, and the curve that shows the inlet pressure of the mixing chamber changes along with the diameter of the mixing chamber develops along the optimal pressure function curve when the diameter of the mixing chamber is smaller than the stagnation critical point. The characteristic curves of the silo ejector and the ideal ejector were compared when imbalance or wall friction were of concern, respectively, which show that the degree of imbalance has little effect on the calculation results of the silo ejector and the wall friction has a significant effect, especially near the stagnation point. The results provide important guidance for the design of silos and ejectors. Finally, the reliability of the ejector function method is verified by comparing the experimental data with the theoretical results.

A small-scale plasma ablation facility was employed to test the C/C-SiC composite material for investigating the thermal performance and ablation characteristics under two heat flux conditions, 3593.54 kW.m- 2 and 5644.86 kW.m-2. The morphology of post-test specimens was analyzed with the ablation rates calculated. The average mass ablation rates of two group specimens were 0.01735 and 0.10620 g.s-1 respectively with average linear ablation rate of 0.00680 and 0.09407 mm.s-s1. Specimen surface could be divided into three regions with typical layered structure characteristics. For the stagnation point ablation test, the structural deformation in the ablation surface area featured in vertical layering and lateral regionality, forming an ablation pit near the stagnation point. In the center region, sublimation occured primarily, accompanied by a serious jet scouring of the molten liquid phase, as well as a small amount of oxidation reaction; Jet erosion with thermal sublimation was the main factor for the mass loss in the transitional region; Thermochemical reactions were mainly carried out in the marginal region. The SiO2 generated from the thermochemical reaction of the material filled the interspace well and prevented the thermochemical reaction from penetrating deeper through the crack. The protective layer in the molten state with high viscosity reduced the damage of the high-speed jet impact material.

针对考虑多因素综合影响的引射器优化问题,基于引射器函数法,深入研究了静压协调函数的数学特性并分析了其数学曲线上的奇异点.在此基础上,通过程序设计和基准推进归纳法得到了多分支工作特性曲线,分析了不同分支下解的特性.另外结合工作特性曲线重点研究了混合室背压、主次流总压比和混合不均匀度对引射性能的影响,提出了静压特性曲面、临界曲线的设计概念.通过试验数据对比分析验证了引射器函数法的可靠性.结论表明:靠近临界曲线工况时引射器设计性能较好;混合不均匀度对引射性能影响重大,不均匀度为1.5时对比理想状态,引射系数最大误差达到32.73％;考虑壁面摩擦设计时需对摩擦因数公式模型进行修正.研究结果为引射器优化设计提供了重要指导.

随着俄罗斯匕首、锆石等高超声速武器飞速发展,非对称作战逐渐成为未来新型作战样式以及影响各国全球战略威慑的重要因素.主要介绍了近期美俄高超声速武器研究和试验情况,分析了各型号高超声速助推滑翔导弹、高超声速吸气式导弹和高超声速飞机的作战性能、战略用途以及发展联系.在此基础上,论述了未来高超声速飞行器发展过程中超燃冲压发动机与组合循环动力系统、热防护结构设计与高温材料应用和制导与控制等关键技术面临的难题,介绍了相关技术的研究进展并作出展望.