Density and temperature reconstruction of a flame-induced distorted flow field based on background-oriented schlieren(BOS) technique

An experimental system based on the background-oriented schlieren（BOS） technique is built to reconstruct the density and temperature distribution of a flame-induced distorted flow field which has a density gradient. The cross-correlation algorithm with sub-pixel accuracy is introduced and used to calculate the background-element displacement of a disturbed image and a fourth-order difference scheme is also developed to solve the Poisson equation. An experiment for a disturbed flow field caused by a burning candle is performed to validate the built BOS system and the results indicate that density and temperature distribution of the disturbed flow field can be reconstructed accurately. A notable conclusion is that in order to make the reconstructed results have a satisfactory accuracy, the inquiry step length should be less than the size of the interrogation window.

RECONSTRUCTION OF PART OF AN ACTUAL BLAST-WAVE FLOW FIELD TO AGREE WITH EXPERIMENTAL DATA BY USING NUMERICAL METHOD WITH HIGH IDENTIFICATION

In this paper, on the basis of experimental data of two kinds of chemical explosions, the piston-pushing model of spherical blast-waves and the second-order Godunov-type scheme of finite difference methods with high identification to discontinuity are used to the numerical reconstruction of part of an actual hemispherical blast-wave flow field by properly adjusting the moving bounary conditions of a piston. This method is simple and reliable. It is suitable to the evaluation of effects of the blast-wave flow field away from the explosion center.

Reconstruction of the Unsteady Supersonic Flow around a Spike Using the Colored Background Oriented Schlieren Technique

In this paper, the improved Background Oriented Schlieren technique called CBOS (Colored Background Oriented Schlieren) is described and used to reconstruct the density fields of three-dimensional flows. The Background Oriented Schlieren technique (BOS) allows the measurement of the light deflection caused by density gradients in a compressible flow. For this purpose the distortion of the image of a background pattern observed through the flow is used. In order to increase the performance of the conventional Background Oriented Schlieren technique, the monochromatic background is replaced by a colored dot pattern. The different colors are treated separately using suitable correlation algorithms. Therefore, the precision and the spatial resolution can be highly increased. Furthermore a special arrangement of the different colored dot patterns in the background allows astigmatism in the region with high density gradients to be overcome. For the first time an algebraic reconstruction technique (ART) is then used to reconstruct the density field of unsteady flows around a spike-tipped model from CBOS measurements. The obtained images reveal the interaction between the free-stream flow and the high-pressure region in front of the model, which leads to large-scale instabilities in the flow.

Deep learning method for super-resolution reconstruction of the spatio-temporal flow field

The high-resolution(HR)spatio-temporal flow field plays a decisive role in describing the details of the flow field.In the acquisition of the HR flow field,traditional direct numerical simulation(DNS)and other methods face a seriously high computational burden.To address this deficiency,we propose a novel multi-scale temporal path UNet(MST-UNet)model to reconstruct temporal and spatial HR flow fields from low-resolu-tion(LR)flow field data.Different from the previous super-resolution(SR)model,which only takes advantage of LR flow field data at instantaneous(SLR)or in a time-series(MTLR),MST-UNet introduces multi-scale information in both time and space.MST-UNet takes the LR data at the current frame and the predicted HR result at the previous moment as the model input to complete the spatial SR reconstruction.On this basis,a temporal model is introduced as the inbetweening model to obtain HR flow field data in space and time to complete spatio-temporal SR reconstruction.Finally,the proposed model is validated by the spatio-temporal SR task of the flow field around two-dimen-sional cylinders.Experimental results show that the outcome of the MST-UNet model in spatial SR tasks is much better than those of SLR and MTLR,which can greatly improve prediction accuracy.In addition,for the spatio-temporal SR task,the spatio-temporal HR flow field predicted by the MST-UNet model has higher accuracy either.

Dynamic mode decomposition and reconstruction of transient cavitating flows around a Clark-Y hydrofoil

The transient cavitating flow around the Clark-Y hydrofoil is numerically investigated by the dynamic mode decomposition with criterion.Based on the ranking dominant modes,frequencies of the first four modes are in good accordance with those obtained by fast Fourier transform.Furthermore,the cavitating flow field is reconstructed by the first four modes,and the dominant flow features are well captured with the reconstructed error below 12%when compared to the simulated flow field.This paper offers a reference for observing and reconstructing the flow fields,and gives a novel insight into the transient cavitating flow features.

基于LSTM-POD的汽车湍流尾迹的高时间分辨速度场重构

本文针对方背Ahmed汽车标模的湍流尾迹,建立基于长短时记忆法(long short-term memory,LSTM)和本征正交分解(proper orthogonal decomposition, POD)相结合的深度学习模型LSTM-POD。通过建立非时间分辨平面速度场POD模态系数和若干离散点的时间分辨速度信号的映射关系,实现了方背Ahmed汽车标模湍流尾迹流场的高时间分辨率重构,并对比了不同时间步长配置,即单时间步长(LSTM-Sin)和多时间步长(LSTM-Mul)对重构效果的影响。研究表明:LSTM-POD模型在时间序列重构中具有较强的学习和泛化能力。另外,LSTM-Mul考虑到了时间上的连续性和相关性,相较于LSTM-Sin,其重构出的低阶模态系数和速度场与POD的重构结果更吻合。本研究提出的深度学习模型可以缓解通过实验及高精度数值模拟获取高时间分辨率流场数据资源消耗大、计算效率低等问题。

质子交换膜燃料电池泡沫金属流场传质性能

建立泡沫金属流场的三维完整形态,通过三维数值模拟进行研究,结果表明,泡沫金属流场可以有效地降低浓度极化损失,从而提高燃料电池在大电流密度下的性能,且与增加的输出功率密度相比,相应的泵送功率损失可以忽略不计。同时,在泡沫金属流场作用下,质子交换膜燃料电池内部的氧气浓度分布比传统的直流道的燃料电池更加均匀。通过台架试验,观测了泡沫流场中的液态水的流动状态,并通过三维数值模拟,得到了与试验相似的流动现象,且与传统直流道流场对比,泡沫金属流场对燃料电池水淹现象的改善有很大的帮助。

地浸采铀工艺中电潜泵内流场数值模拟研究

电潜泵是我国地浸采铀工艺中实现浸出液有效提升的主要装备,但目前针对介质举升状态下电潜泵水力部件内部流场作用机制的研究较少。试验以叶轮-导叶内部结构为研究对象,应用CFD方法对介质流动特性开展非定常分析,研究了叶轮和叶轮出口段内部的压力脉动变化规律,并利用瞬态流场数据对叶轮截面压力流场进行动模态分解(DMD)与压力流场重构,揭示了介质举升状态下由动静干涉引起的压力脉动周期性变化规律与动静部件距离对动静干涉强度的影响规律。结果表明,叶片通过频率值与快速傅里叶变换(FFT)频率值基本一致。试验结果可对减少由压力脉动引发的泵体无效震动、优化电潜泵水力部件结构参数和提升系统效率提供一定理论参考。

基于动力学模态分解的空化水射流非定常流场演化分析

空化水射流具有流速高、湍流强等特点,其空化流动结构十分复杂。因此,准确、快速地从非定常流场的海量数据中提取关键信息,并分析空化水射流的流动特性十分必要。基于FBDCM模型对RNG模型的湍流黏度进行修正,并考虑剪切应力对空泡形成的影响,对空化模型进行修正,与试验结果进行对比;采用动力学模态分解(DMD)方法对空化流场的空泡体积分数场进行模态分解。结果发现:1阶模态为平均流场,是流场的主要结构;2阶模态与空化云周期性脱落、溃灭的动力学行为有关;3、4阶模态表现出流场的高频行为;随着进口压力的增大,空泡脱落与溃灭的演变程度越来越剧烈,空泡尺寸与空穴长度逐渐增加,而2阶模态的频率逐渐降低,使得空化射流的流动周期变长。最后,将DMD重构流场与真实流场进行对比发现,随着进口压力的增加,DMD重构流场的误差增大,不过重构流场的最大误差仍低于1%,表明DMD方法能够准确地提取空化流场的流动特征。

基于声学层析成像的温度场和流速场重构

为有效重构三维管道内液体的温度场和流速场,提出了一种超声飞渡时间计算表征,并基于此求解温度场和流速场反演模型。同时考虑温度梯度导致的声线弯曲效应,采用正四面体前向展开法,追踪超声弯曲路径;开展“华龙一号”反应堆主管道冷却剂温度场和流速场重构实验。实验结果表明,所提方法能有效重构管道冷却剂温度场和流速场,且考虑弯曲效应能进一步提高重构精度。

基于集合变换Kalman滤波的流场高效重构

湍流场的准确估计在航空航天领域具有重要意义,现有的获取手段在分辨率或者准确性方面是不足的。实验测量准确却往往测点数量有限,数值计算能获得全场数据,但精度却难以保障。数据同化方法融合了实验观测和数值模拟,是进行流场重构的有效工具。探索了基于集合变换Kalman滤波(ensemble transform Kalman filter,ETKF)的数据同化方法在空间流场重构方面的有效性,并讨论了不同迭代更新模式的重构精度和计算效率,即状态变量基于湍流模型更新的ETKF-M和基于流场数据更新的ETKF-D。以ONERA M6机翼作为数值算例,结合风洞实验翼型表面271测压孔的压力测量数据进行算法实验,结果表明ETKF方法的不同迭代模式均有效修正了湍流模型的预测,并且ETKF-D相对于ETKF-M提升了83%的计算效率。此外,选取两组不同位置的1/4实验测点进行同化实验,得到不同精度的结果,这表明重构的精度与同化测点的位置和数量密切相关。

面向粒子图像测速的光流金字塔插值优化方法

提出了一种面向粒子图像测速的光流金字塔插值优化方法,可以减小重建角度误差和均方根误差,尤其是图像边缘部分的角度误差。使用该优化算法分别对单涡流、双涡流、DNS湍流粒子场进行速度场重建,得到三种流场的均方根误差和平均角度误差比优化前至少降低了15.96%和19.87%。搭建二维PIV实验系统,在待测流场中均匀撒入100μm的PSP粒子,进行旋转实验和注水实验,分别模拟涡流场和射流场,所提出的优化算法可以获得与常见算法流形一致的速度场。

Characteristic size research of human nasal cavity and the respiratory airflow CFD analysis

To study the airflow distribution in human nasal cavity during respiration and the characteristic parameters for nasal structure, thirty three-dimensional, anatomically accurate representations of adult nasal cavity models were reconstructed based on processed tomography images collected from normal people. The airflow fields in nasal cavities were simulated using the fluid dynamics with the finite element software ANSYS. The results showed that the difference of human nasal cavity structure led to varying airflow distribution in the nasal cavities and the main airflow passed through the common nasal meatus. The nasal resistance in the regions of nasal valve and nasal vestibule accounted for more than a half of overall resistance. The characteristic model of nasal cavity was extracted based on the characteristic points and dimensions deducted from the original models. It showed that either the geometric structure or the air-flow field of the two kinds of model was similar. The characteristic dimensions were the characteristic parameters of nasal cavity that properly represented the original model in research for nasal cavity.

基于扩散模型的流场超分辨率重建方法

低分辨率的流场数据具有较少的信息量,不能充分捕捉流场中的细节演化过程.尤其对于湍流的随机脉动特征和小尺度涡旋细节特征更加难以获取,这限制了对流场演化机理进行深入研究.为了解决这一局限性,并从低分辨率流场中重建高分辨率数据,文章提出一种流场超分辨率重建的生成扩散模型FlowDiffusionNet.该模型以低分辨率流场数据输入作为约束条件,采用去噪分数匹配方法,来实现高分辨率流场数据的复现.FlowDiffusionNet在结构设计上充分考虑了流场数据的低频信息与高频空间特征,采用基于扩散过程的建模方法,用于重建高分辨率流场数据的残差.该模型结构便于实现迁移学习,可在不同程度的退化流场上应用.将该方法在多种经典流场数据集上进行测试,并与双三次插值(bicubic)、超分辨率生成对抗网络(SRGAN)、超分辨率卷积神经网络(SRCNN)等方法进行比较.结果表明,该方法在各种流场上的重建性能达到最佳水平,特别是对于包含小尺度涡结构的4倍下采样流场数据,客观评价指标SSIM达到0.999.

声门下喉气管狭窄分度的数值模拟分析

根据真实人体喉气管CT影像数据构建了声门下喉气管狭窄的气道模型,并对这些模型进行了吸气条件下的计算流体力学仿真。分析比较了不同狭窄分度模型内的流场特征、不同喉气管部段的阻力特性以及喉气管壁面的压力与切应力分布,并对这些特性与临床病患表象的关联性进行了讨论。分析表明声门下喉气管狭窄分度与狭窄导致的呼吸道阻力变化具有良好的关联性,此外,喉气管狭窄处压力及剪应力效应对病程的影响也需要关注。

基于2D-PIV的调节阀三维流场特性研究

调节阀流道内流动状态直接影响其服役寿命和系统工作稳定性,为了探究调节阀内部三维流场的分布特性,通过二维粒子图像测速(2D-PIV)装置获取三维坐标系下不同切面的流场特征,利用插值计算其余节点的流场信息,实现调节阀流场的三维重构。实验结果表明:三维重构速度场与二维速度场整体呈一致趋势,油液在调节阀节流口处由于节流效应形成对冲射流,射流汇合后在阀芯头部下游形成高速的整体喷射;阀芯顶端区域由于两侧冲击射流的合并会产生一定的油液回流现象;整体流道速度随着拍摄平面外移均呈逐步下降的趋势,上游流道速度平稳下降到一定值后保持稳定,节流口附近速度呈先升高后下降的趋势且变化幅度较大,壁面阻力以及剪切力的存在使得近壁面速度保持平稳。下游区域速度呈同心圆分布,从湍流变为有序流动。仿真结果表明,三维重构流场与理论计算结果基本一致,最大重构误差为9.5%,出现在节流口处的高速流区域。实验重构结果与仿真重构结果具有高度相似性,流动平缓区域的重构精度较高。本研究可为优化调节阀结构和改善空蚀性能提供参考,也可为微小流道三维流场测量提供借鉴。

基于时间序列多品种网络流的高铁车站站改期间行车组织优化研究

将高铁车站站场基本布局抽象为站场网络图,借助时间序列多品种网络流模式进行行车组织优化。基于行车进路规则构建生成进路数据集算法,将既有行车方案与构建的站改期间站场网络图融合,推断既有行车方案中到发线是否可用以及是否存在可用进路,据此将列车分为无需优化列车和需要优化列车两类。针对无需优化列车,构建计算公式和算法,推算到发线占用时间序列和咽喉占用时间序列。针对需要优化列车,构建算法并依据到发线和咽喉的占用时间序列,推断是否存在可用到发线以及是否同时存在可用进路,将需要优化列车分为可以优化列车和无法优化列车。通过构建可行股道推断算法以及可行股道进出站进路可用推断算法,形成行车组织优化方案。

基于卷积神经网络的三维流场重构研究

三维流场的快速分析是工程应用领域最关心的内容。为了实现对三维非定常流场的降阶重构,提出了一种基于卷积神经网络和残差连接的U型架构降阶模型,通过对当前的流场快照进行多层次的特征提取,并通过解码器将多尺度特征进行融合,将其解码为未来时刻的流场数据。基于有限体积法求解Navier-Stokes方程得到的圆球绕流全阶流场快照作为真实的物理数据,评估降阶模型对三维典型流动的学习和推理能力。研究结果表明,增量预测结果与真实流场基本吻合,最大相对误差在0.2%以内,速度比数值模拟提升两个量级。所提出的降阶模型,实现了三维非定常流场快速重构的预期目标。

基于PINN方法的不可压流场求解

将残差网络(Res Net)结构引入物理信息神经网络(PINN),提出一种基于Res Net的PINN方法(Res Net-PINN)。采用该方法对二维不可压圆柱绕流尾迹流场进行重建和预测,结果表明:Res Net-PINN能更准确地重建绕流尾迹的非定常变化规律;在圆柱绕流尾迹的短期预测方面,Res Net-PINN的预测精度和收敛速度相比全连接PINN均能提升3倍左右。Res Net能提高PINN对非定常流场的求解和预测能力,该研究可为采用机器学习方法求解更复杂流动问题提供参考。

基于格子Boltzmann和POD方法的非定常流场重建

以二维圆柱雷诺数Re=100绕流为研究对象,在格子Boltzmann方法模拟非定常流动的基础上,利用本征正交分解(proper orthogonal decomposition, POD)算法获取二维圆柱绕流周期性稳定脱落阶段POD基函数及对应的系数,用以实现非定常流场的重建,并研究不同POD模态阶数对重建效果的影响。结果表明:前5阶POD模态占总能量的99%,可以准确地重构流场,流场重构误差最大绝对值为8×10^(-4);随着模态阶数的增加,流场主要特征表达得越细致,且流场重建误差由大幅度降低,缓慢减低到趋于稳定几乎保持不变。