Film cooling and heat transfer performance on the squealer tip with different degrees of cooling-hole blockage are investigated using numerical methods.The film cooling effectiveness and heat transfer coefficient distribution on the squealer tip are obtained under three blowing ratios(M=0.5,1.0,and 1.5)and five different blockage ratios(B=0,0.2,0.4,0.6,and 0.8).The results show that the cooling-hole blockage causes decrease of film cooling effect and increase of thermal load on the squealer tip.The film cooling effect and heat transfer performance on the squealer tip are deteriorated with increasing the blockage ratio.The blockage in film cooling holes significantly alters the flow structures in the squealer tip gap.With cooling-hole blockage,the lift-off effect of coolant in the squealer cavity will be aggravated,resulting in earlier leave of cooling flow from the squealer cavity.Generally,the blowing ratio is an important factor that affects the heat transfer and film cooling effect on the squealer tip under blockage conditions.Compared with the unblocked cases,the area-averaged film cooling effectiveness on the squealer tip decreases by 64.88%,and the heat transfer coefficient increases by 13.01%,at blowing ratio M=1.0 and blockage ratio B=0.8.Reducing the blowing ratio is conducive to the improvement of film cooling effectiveness on the squealer tip with small blockage ratio.The area-averaged film cooling effectiveness on the squealer tip only decreases by 6.82%at blowing ratio M=0.5 and blocking ratio B=0.4.However,in the case of small blow ratio,for example M=0.5,the film cooling effect on the squealer tip is deteriorated significantly under the condition of large blockage ratio.At blowing ratio M=0.5 and blocking ratio B=0.8,the area-averaged film cooling effectiveness on the squealer tip decreases by 82.09%compared with the unblocked case.The increase in blowing ratio can improve the film cooling effect on the squealer tip with large blockage ratio.Compared with the unblocked case,the area-averaged film cooling effectiveness on the squealer tip only decreases by 51.34%and the heat transfer coefficient increases by 11.52%at blowing ratio M=1.5 and blockage ratio B=0.8.
Liquid film cooling is broadly applied in the thermal protection of liquid rocket engine. An experimental system of liquid film cooling for a heated curved stainless-steel wall is built in this study to investigate the effects of jet mass flow rate, jet angle, and nozzle diameter on the cooling performance. The temperature distribution of the heated wall is demonstrated with infrared thermography technology and the flow spreading shape of the liquid film is captured with high-speed camera. A numerical liquid film cooling model based on the volume of fluid (VOF) coupled Level-set method is established to calculate the evaporative heat absorption and spreading shape of liquid film. The experimental study indicates that the peak film thickness in the hydraulic jump region increases with the jet mass flow rate, such as the peak film thickness increases from 265.15 to 632.25 mu m at Q = 100-600 mL.min(-1). Besides, a jet angle of 35 degree can maintain the maximum wall temperature drop of 29.4 degrees C. In addition, the simulated peak film thickness presents a minimum deviation of 5.2 % from the experimental data. The numerical model provides important reference for the study of liquid film cooling performance of the liquid rocket engine.
This article presents a fuzzy simple additive weighting multi-objective micro-genetic algorithm (FSAW-MO-MGA) for high-power microwave (HPM) sources optimization. The FSAW-MO-MGA and a single-objective genetic algorithm are independently used to optimize the same Ka-band relativistic backward-wave oscillator (RBWO) device. The use of the FSAW-MO-MGA allowed us to obtain an optimized device structure of RBWO with comprehensively improved performance. Particle-in-cell (PIC) simulation results show that the optimized RBWO can generate microwave pulses with an output power of 616.0 MW and an operating frequency of 30.33 GHz under a diode voltage of 625.6 kV, and the diode current is 6.59 kA at a guiding magnetic field of 0.8 T. Compared with the original RBWO, the output power of the optimized RBWO has been increased by 197.6%, the beam-to-microwave conversion efficiency has increased from 5.0% to 14.9%, and the beam-current transmission has increased from 70.13% to 87.72%. The reasons for these improvements in device performance are analyzed in detail.
The communication base station management system mostly adopts an empirical test method at present, which cannot simulate the test cases with a long period and reasonable occurrence time. To solve this problem, a test case modeling method based on the base station alarm log is proposed. Firstly, taking the log generation time interval of the base station as the research object, it is found that the log generation time interval obeys power-law distribution by analyzing data entirety, device type, log type, and device combination log type. Then, the least square method, maximum likelihood estimation, and maximum posterior estimation are compared and analyzed. Based on the estimation error, it is found that the least square method has the fitting of a higher degree and the smallest residual error, the value of goodness of fit is 0.96%, and mean absolute percentage error is 3.5%. Finally, the least square method is used to estimate the log time interval by power exponent, and the alarms with the value of goodness of fit greater than 0.7 are reserved to form test cases within a location. The model is applied to estimate the interval distribution of alarm logs of communication base stations in different cities throughout the country. The experimental results show that the proposed model can calculate more than 85% of log distribution law, which lays a foundation for arranging test cases close to the actual operating environment and further improves the reliability of test results.
为了提高姿轨控发动机结构环境试验的设计水平,迫切需要研究相应的环境试验响应分析方法.基于姿轨控发动机中频振动特征明显这一特点,综合考虑各组件的模态数及结构特征,为姿轨控发动机建立了适用于中频响应分析的混合有限元-统计能量分析(FE-SEA)模型.在地面试验中,由于使用振动台,输入功率无法直接采用现有理论进行计算,提出了一种基于连接界面随机振动加速度功率谱密度的振动台激励输入功率计算方法,用以实现发动机结构在振动台激振下的响应预示.基于混合有限元-统计能量分析的姿轨控发动机结构响应预示结果与振动台试验的响应测量结果比较,结果表明,混合有限元-统计能量分析方法在中频分析频段内能够准确预示结构响应,当随机子系统的模态数大于3时,对应中心频率下的结构中频响应预示误差均小于±3 dB,具有良好的环境试验响应预示精度.混合FE-SEA方法的预示结果准确性与随机子系统在分析频带内的模态数密切相关,当模态数较小时,预示结果与试验结果在较低频带的误差变得显著.该研究内容为姿轨控发动机结构的中频振动响应分析提供了有效参考.
为应对振动环境下管路结构高可靠性寿命评估问题,建立材料、载荷等底层因素随机性模型,并考虑随机性传递、综合作用,获得管路结构振动疲劳寿命可靠性分析模型.以典型材料试验结果及单自由度时域系统,证明了随机性模型的适用性及寿命模型的高效性.以液体火箭发动机典型管路结构为分析对象,评估其高可靠性寿命.分析结果表明:在试车振动载荷作用下,管路结构振动疲劳寿命符合对数正态分布,且寿命存在较大的分散性;所考虑的随机因素中,材料疲劳性能对寿命分散性的影响较大;对管路结构,在确定性寿命的基础上,应考虑至少不小于5倍的寿命缩减系数,以覆盖载荷及材料的随机性影响而获得其高可靠度振动疲劳寿命.该研究为液体火箭发动机可重复使用技术发展提供重要技术支撑.
为进一步探究阴极催化层铂载量对燃料电池低温启动过程中传质反应过程及宏观性能特性的影响机制,建立了质子交换膜燃料电池一维多相非等温非稳态低温启动模型,耦合了考虑氧气在催化层多组分中传质阻力的电化学反应动力学模型,考虑了电池多孔介质内水的输运及相变、电化学反应、电渗拖拽、热量传递等过程,研究了阴极催化层铂载量对低温启动过程中电池性能、水含量、冰的体积分数、电池温度等变化的影响.研究结果表明:过小的铂载量不利于低温启动过程的进行,但过高的铂载量又不利于电池性能的提升,0.1 mg/cm2为最优铂载量;氧气从催化层孔隙向铂表面进行输运的过程存在时间效应;随着铂载量的增加,阴极催化层中冰的最初生成时间推迟,冰的生成速率下降;随着铂载量的增加,阳极催化层和阴极催化层中膜态水含量的分布梯度变小.
针对可调谐微波器件对材料高介电可调性与低介电损耗的性能需求,采用传统固相烧结法,设计并制备了具有优异调谐性能的施受主复合掺杂(Ba0.675Sr0.325)1-0.005La0.005Ti1-xCrxO3(x=0.3％、0.5％、0.7％、1.0％)陶瓷.引入施主杂质La取代A位元素改善介电可调性,受主杂质Cr取代B位元素降低介电损耗,并改变施主/受主掺杂比,研究了施主/受主掺杂比对钛酸锶钡(BST)陶瓷微观结构与电学性能的影响规律.实验结果表明:随着受主杂质摩尔分数的增加,晶粒尺寸先增大后减小,致密度先提升后降低,即合适的施主/受主掺杂比更有利于试样的致密化;当Cr3的摩尔分数为0.7％时,试样的调谐性能最佳,其介电可调性为83.6％,介电损耗为0.004 1,优值因子可达到204,此时,试样的击穿场强达到14 kV/mm.研究结果可为材料在强外加电场下使用稳定性提供参考.
针对传统子域模型在分析电机电磁性能时无法考虑软磁材料磁导率的缺陷,提出了一种考虑软磁材料磁导率为具体值的三段式Halbach阵列永磁电机多层解析模型.根据内部激励源和媒介的不同,将永磁电机全域划分成Halbach阵列永磁体层、气隙层、定子齿尖层和电枢绕组层4类磁场求解层.利用柯西乘积和复数形式的傅里叶级数来描述磁场求解层中媒介磁导率分布情况.基于麦克斯韦方程组建立了每个磁场求解层的拉普拉斯方程或泊松方程,结合边界条件求解出每个磁场求解层的磁矢位.利用多层解析模型计算了电机的气隙磁密、空载反电动势和输出转矩等电磁特性,并与有限元模型计算结果进行了对比.计算结果表明:多层解析模型和有限元模型计算得到的气隙磁密波形吻合得很好,两者的空载反电动势峰值和输出转矩平均值的相对误差均小于1％,证明了多层解析模型的正确性.最后利用多层解析模型研究了槽开口宽度、磁极宽度和充磁夹角对电机输出转矩的影响规律,给出了输出转矩最优的设计方案.仿真结果表明:优化后的输出转矩脉动削减了75.7％,证明所提出的多层解析模型不仅计算准确而且计算速度快,在电机的初始设计和电磁性能影响规律的探究上具有明显的优势.
针对现有网络流量分类方法难以在样本稀缺场景下快速识别早期未知应用这一问题,提出一种基于序列特征与知识引导的未知网络应用早期识别方法.一方面基于Transformer-encoder框架构建流量分类模型(FAIN),该模型利用自注意力机制挖掘流量序列中数据包之间的全局依赖关系,生成具有可分辨性的流量表示向量用于分类.另一方面,为提升FAIN模型在样本稀缺场景下的适应能力,采取监督预训练与元学习相耦合的模型优化策略,赋予模型在小样本场景下快速学习流量分类任务的能力,使其满足识别早期未知网络应用的需求.在公开数据集与真实校园网流量合成的小样本数据集上进行了深入的对比实验.结果表明:所提出的流量分类模型FAIN在公开分类任务上优于现有方法,且优化后的FAIN模型在XJTU-FSTC和CSTNET数据集的5类和10类小样本分类任务上,准确率最高分别提升了16.75％、10.08％和11.57％、8.24％.该研究结果为未知网络应用的早期识别提供了有效的方法支撑.
氦氙混合物为工质的布雷顿循环具有循环效率高、系统结构紧凑、化学稳定性好等优势,适合作为空间核反应堆的能量转换系统.在深入调研空间堆氦氙布雷顿循环发展历史和国内外研究进展的基础上,对其关键技术问题和重点研究方向的相关研究进展进行了综述,发现主要研究方向包括氦氙混合物工质特性、氦氙布雷顿循环关键部件、循环性能提升、循环动态特性及控制策略等方面;而有待继续深入研究的关键技术问题包括不同比例氦氙工质的高精度物性及流动传热模型、高性能叶轮及高效紧凑式换热器设计及试验、不同功率等级下系统全工况优化、耦合反应堆的系统全局动态特性及控制策略等.分析结果可为推动空间堆氦氙布雷顿循环技术发展提供参考.
为提高粒子法研究中初场计算节点的空间分布质量,包括计算域内部计算节点分布的各向同性和计算域边界附近计算节点对理论边界的贴体性,提出了一种计算节点(即粒子)初场布置的优化方法.该方法借助网格控制求解域的边界,采用粒子迁移模型使流体粒子自适应地发展为均匀各向同性的分布状态.为了将网格的信息耦合进粒子法的粒子迁移模型中,在粒子迁移模型中引入了基于边界网格的数值积分算法.验证算例包括二维椭圆、二维转子泵、三维复杂腔体、三维球体等,验证结果表明:所提出的初场布置优化算法能够保证流体域内粒子的均匀各向同性分布,并减小边界粒子与理论边界的偏差值,降低边界粒子数密度的噪点;在液滴模拟中,优化初场布置后的圆度误差较传统初场布置减小了78.6％.
针对火箭橇试验凿削磨损现象影响因素作用机制尚不明确、临界条件难以界定的难题,基于有限元法建立了特有的三维凿削磨损模型,在热力耦合场下对该过程进行了模拟.模拟结果表明:在初始航向速度和温度条件下,竖向速度达到1.75 m·s-1以上时凿削磨损发生;在初始航向速度和竖向速度条件下,温度达到673 K以上时凿削磨损发生.初始航向速度条件下,凿削磨损率随着竖向速度和温度的上升而增大.当航向速度由1 200 m·s-1增加到1 800 m·s-1时,临界竖向速度由1.65 m·s-1上升至3.00 m·s-1,临界温度由673 K上升至873 K,临界夹角由0.079°上升至0.096°.通过降低滑靴竖向速度和温度能够有效避免凿削磨损的发生.该研究可为进一步阐明火箭橇试验中的凿削磨损机制提供理论支撑,并为避免凿削磨损发生、提高试验安全性提供思路.
为深入了解低雷诺数下增压级正交叶片叶型表面的流动特性,发展高效的附面层调控策略,以高负荷增压级正交叶片某一截面处的叶型为研究对象,通过高精度参数化方法对该叶型进行改型设计.利用数值模拟手段,研究了低Re(Re=1.0×105)下层流分离及转捩对不同叶型的响应特性,阐明了前缘形状及叶片压力梯度对典型增压级叶型附面层发展的调控机制.结果表明:前缘压力尖峰和前加载压力梯度分布均能够有效抑制低R e下分离泡发展,消弱近壁区回流强度,将设计工况点叶型损失分别降低12.4％和12.5％,同时拓宽了低损失攻角范围;前缘压力尖峰耦合前加载压力梯度分布同时继承了上述优点,能够更大限度地提升低雷诺数下叶型的气动性能.相比于原始叶型,设计工况点叶型损失降低18.9％.该研究结果为低Re下增压级叶型的气动设计和流动调控提供有益借鉴.
为了对高功率微波(HPM)在线测量系统中的核心器件耦合器进行准确标定,采用锥形渐变加圆柱的内导体结构,设计了一种S波段同轴TEM模式转圆波导TM01模式的模式转换器.从模式耦合和阻抗匹配的基本思路出发,对模式转换过程、工作带宽的扩展以及非TM01模式的抑制等方面进行了理论分析.使用CST仿真软件对模式转换器的结构参数进行了优化,对模式转换器的反射性能、传输性能、工作带宽、输出模式和脉冲响应进行了实验测量.仿真和实验结果均表明:所设计的模式转换器中几乎不含有TE11和TM01模式以上的高次模式,TM01模式纯度较高;在工作频点2.1 GHz处的反射系数小于-30 dB,传输系数大于-0.2 dB,实测模式转换效率可以达到96％以上,工作带宽为74MHz,而且脉冲响应时间很短.该模式转换器目前已成功应用于S波段在线耦合器的标定之中.该研究为HPM在线测量过程中模式转换器的设计提供了新思路.
Aiming at the problems that the low energy absorption performance of uniform lattice structures and poor stiffness performance of cross section gradient lattice structures， a multilayer size gradient FCC lattice structure was designed and its mechanical properties were investigated through theoretical， numerical， and experimental methods in this paper. Firstly， size graded face centre cubic （SG-FCC） and uniform face centre cubic （U-FCC） lattice structures were fabricated by SS316L stainless steel and selective laser melting （SLM） technology. The quasi-static compression tests were also carried out. Then according to the tests and finite element analysis results， the mechanical properties of the two lattice structures during compression were obtained. In addition， the equivalent elastic modulus analytical model of both lattice structures were proposed based on the deformation theory of beam and the stiffness set method of gradient lattice structures. Finally， the influence of different gradient factors on the compressive performance of SG-FCC lattice structures was also studied. The results showed that the mechanical properties of SG-FCC lattice were better than that of U-FCC lattice. The stiffness and specific stiffness of SG-FCC lattice were increased by 17.8% and 16.2% respectively， and the absorption energy and specific absorption energy were increased by 10.8% and 10.36% respectively. The theoretical elastic modulus model of SG-FCC lattices were in good agreement with the experimental and finite element results. The prediction error was less than 10%， which could be used to predict the stiffness of SG-FCC lattice structure. When the gradient factor was 1.5， the stiffness properties of SG-FCC lattice structure were better than those of other gradient factors. When the gradient factor was 3， the energy absorption properties of SG-FCC lattice structure were better than those of other gradient factors. This research can provide some guidance for the design and engineering application of the lightweight structures with high stiffness and large energy absorption performance.
为解决时空尺度上扑翼流动控制方程求解需要花费大量时间和计算资源的问题,基于强非线性曲线拟合能力的物理信息神经网络(PINN)深度学习方法,提出了一种混合粗糙数据驱动物理信息神经网络模型(HCDD-PINN),研究了模型对涉及非定常流动特征和动边界二维俯仰扑翼问题的训练和预测性能.通过使用相较于传统计算流体动力学方法(CFD)更为粗糙的数据驱动模型训练,将扑翼流动控制方程嵌入神经网络损失中,并施加初始条件和边界条件约束,采用一阶自适应矩优化算法(ADAM)和二阶拟牛顿法优化算法(L-BFGS-B),以前馈-反向传播方式最小化模型损失函数,从而提高模型预测控制方程数值解的准确性和可靠性.结果表明:与原始PINN模型相比,HCDD-PINN模型显著降低了流场的预测误差,能够准确地预测扑翼瞬时气动力和瞬时速度及压力场,训练时间缩短了75％.此外,训练完成的HCDD-PINN模型可以快速获得流场任意时刻的物理信息,而传统CFD方法则需要重新对流场进行计算.该研究为求解扑翼流动控制方程乃至流体非线性偏微分方程组(PDEs)提供了一种有效的替代方案.
针对电站锅炉工况下水冷壁超温现象导致的水冷壁腐蚀问题,研究了电站锅炉水冷壁T22和12Cr1MoVG在600 ℃与650 ℃超温工况下的腐蚀行为及机理.利用分析天平获取两种耐热钢的腐蚀动力学曲线;利用X射线衍射仪、扫描电子显微镜与能谱分析仪获得腐蚀产物的物相、微观形貌与元素分布,结合腐蚀产物热力学计算分析其演变规律和剥离机理.研究结果表明:深度调峰超温工况下,两种耐热钢的腐蚀动力学曲线在反应初期符合抛物线型规律,之后转变为直线型规律;600 ℃和650 ℃下,腐蚀规律转折点时间分别为25 h和4 h,说明温度能够加速耐热钢腐蚀规律转变;耐热钢T22和12Cr1MoVG表面形成的腐蚀产物主要为Fe3O4、FeCr2O4、Fe2O3和FeS;T22和12Cr1MoVG在600 ℃下的腐蚀层厚度大于650 ℃工况.耐热钢表面腐蚀产物随着反应的进行发生剥离,甚至部分发生剥落,主要原因归结于硫化物的出现.综合耐热钢腐蚀的质量变化、腐蚀层厚度和腐蚀产物微观形貌成分分布规律,T22的抗腐蚀性能优于12Cr1MoVG.
为了解决单一的相关性分析方法在分析临近空间飞行器遥测数据时存在局限性以及证据冲突问题,在相关系数分析评价与优势组合的基础上,提出了基于支持因子的证据理论融合算法.首先分别利用Pearson相关系数、Spearman相关系数与距离相关系数对遥测数据进行相关性评价分析,表明3种相关系数可优势互补.其次,建立了基于支持因子的证据理论融合算法,实现证据的冲突基本概率赋值函数分配,避免Dempster-Shafer证据理论(D-S证据理论)的一票否决和合成规则失效问题.最后,利用3种相关系数构造相关性证据并开展遥测数据相关性分析实验.结果表明:基于支持因子的证据理论融合方法能使Pearson相关系数、Spearman相关系数与距离相关系数的证据融合更加合理;在证据冲突大的情况下,相比相容系数的证据理论融合方法,相关概率计算结果准确度提高约6.55％,能更有效地处理证据冲突问题.
稠油和低成熟页岩等重质油资源潜力巨大,但其低流动性和高干酪根含量导致难以高效开采,针对现有注蒸汽等方法存在采收率低、加热效率差和热流体成本高等问题,提出了以超临界水热化学转化为核心的重质油原位转化开采新思路,基于超临界水独有的高溶解性、高扩散性和高反应性,采用油田有机废液为原料通过超临界水气化与氢氧化原理生产超临界多元热流体,并将超临界多元热流体注入地层加热储层、同时原位转化稠油和干酪根生成轻质油气,通过超临界混相驱大幅提高采收率,从而实现了高效、清洁、低成本的热流体生成、储层加热转化与油气高效采出.研究结果表明,油田有机废液碳转化率超98％,稠油岩心驱替效率超97％,单井4轮次吞吐效率超75％,低成熟页岩有机质生烃制油气有机碳转化率达60％.该技术的突破可对变革传统重质油开发方式、保障我国油气能源安全提供参考.
