Flow field reconstruction method and application for turbine blade cascade based on experimental data assimilation
-
摘要:
为提高平面叶栅试验效率，基于集合卡尔曼滤波原理，采用拉丁超立方抽样组装生成shear stress transport（SST）湍流模型参数、计算数据与试验数据的集合矩阵，通过K-nearest neighbors（KNN）算法对不同数据间的分辨率进行误差对齐，编写了一套适用于涡轮平面叶栅的数据同化程序，对某宽攻角涡轮平面叶栅多种工况下的流场进行了数据同化研究。通过植入3种策略的试验数据，重新适配了SST湍流模型的系数，获得了与试验数据高度吻合的同化数据。同化结果表明，在不同工况下，3种同化策略均可使数值计算结果与试验数据的相对误差降低64.7%以上。故在试验中可适当采信同化数据，调整试验方案，起到提高试验效率，降低试验成本的作用。
Abstract:To improve the efficiency of linear cascade row experiment, this study employs the principles of ensemble Kalman filtering. A Latin hypercube sampling method is used to generate a set matrix of SST turbulence model parameters, computational data, and experimental data. The KNN algorithm is applied to align the resolution errors between different datasets. A data assimilation program suitable for linear cascade rows is developed. The program is used for data assimilation analysis of the flow field under various operating conditions of a linear cascade row with a wide range of attack angles.Experimental data from three strategies are incorporated to re-calibrate the coefficients of the SST turbulence model. The assimilation results show a high degree of agreement with the experimental data. The findings indicate that, under different operating conditions, all three assimilation strategies reduce the relative error between numerical calculations and experimental data by more than 64.7%. Therefore, assimilated data can be appropriately trusted in experiments to adjust experimental plans, which helps improve experimental efficiency and reduce costs.
-
图 2 叶栅几何构型及测点分布
Figure 2. Blade cascade geometrical configuration and measurement point distribution
图 5 不同工况下各参数计算结果
Figure 5. Calculation results of various parameters under different operating conditions
图 6 各参数计算结果与试验数据的相对偏差
Figure 6. Relative deviation between calculation results of various parameters and experimental data
图 7 各同化策略下不同工况的参数变化率
Figure 7. Parameter variation rates for different assimilation strategies under different operating conditions
图 9 同化前后前缘流线图
Figure 9. Streamlines at the leading edge before and after data assimilation
表 1 选取的SST模型常数、功能及默认值
Table 1. Selected SST model constants, functions, and default values
常数 功能 默认值 $ {\beta }^{*} $ 改变模型在强剪切流动中的控制能力 0.09 $ \alpha $ 控制湍流耗散的计算 0.31 $ {\alpha }_{\omega } $ 调节湍流频率方程的扩散项 0.55317 $ {\beta }_{\omega } $ 调整ω方程的精度与稳定性 0.075 $ {\alpha }_{\varepsilon } $ 调节湍流耗散方程的扩散项 0.44036 $ {\beta }_{\varepsilon } $ 调整ε方程的精度与稳定性 0.0828 
下载: 导出CSV
表 2 叶栅主要设计参数
Table 2. Main design parameters of the blade cascade
参数 数值 稠度t/c 0.67 来流马赫数Ma 0.6，0.9 进口构造角/（°） 63 叶片安装角/（°） 26 轴向速度密度比 1.15
下载: 导出CSV
表 3 数据同化选取的工况
Table 3. Selected operating conditions for data assimilation
攻角/（°） 马赫数 0 0.6 0 0.9 20 0.6 20 0.9 −35 0.6 −35 0.9
下载: 导出CSV
表 4 高密度同化策略下不同工况的适配参数及参数变化率
Table 4. Fitted parameters and their variation rates for different operating conditions under high-density assimilation strategy
参数 工况1 工况2 工况3 工况4 工况5 工况6 默认值 $ {\beta }^{*} $ 0.100731 0.10972 0.12493 0.12156 0.13632 0.167592 0.09 $ {\beta }^{*} $变化率/% 11.923% 21.914% 38.813% 35.066% 51.466% 86.213% $ \alpha $ 0.335735 0.35995 0.38224 0.42106 0.452807 0.500319 0.31 $ \alpha $变化率/% 8.302% 16.114% 23.305% 35.826% 46.067% 61.393% $ {\alpha }_{\omega } $ 0.709245 0.75024 0.88999 0.85829 1.038703 1.159642 0.55317 $ {\alpha }_{\omega } $变化率/% 28.215% 35.626% 60.891% 55.159% 87.774% 109.637% $ {\beta }_{\omega } $ 0.084066 0.06014 0.04511 0.051605 0.034543 0.053899 0.075 $ {\beta }_{\omega } $变化率/% 12.088% −19.807% −39.852% −31.193% −53.943% −28.135% $ {\alpha }_{\varepsilon } $ 0.527756 0.59999 0.70019 0.796681 0.991422 0.960342 0.44036 $ {\alpha }_{\varepsilon } $变化率/% 19.848% 36.252% 59.007% 80.918% 125.142% 118.084% $ {\beta }_{\varepsilon } $ 0.104842 0.07016 0.09225 0.052325 0.071246 0.11187 0.0828 $ {\beta }_{\varepsilon } $变化率/% 26.620% −15.261% 11.421% −36.806% −13.955% 35.108%
下载: 导出CSV
-
[1] 张庆典, 马宏伟, 杨益, 等. 平面叶栅气动试验研究进展与展望[J]. 力学学报, 2022, 54(7): 1755-1777. Zhang Q D, Ma H W, Yang Y, et al. Progress and prospect of aerodynamic experimental research on linear cascade[J]. Chinese Journal of Theoretical and Applied Mechanics, 2022, 54(7): 1755-1777. (in Chinese doi: 10.6052/0459-1879-21-684Zhang Q D, Ma H W, Yang Y, et al. Progress and prospect of aerodynamic experimental research on linear cascade[J]. Chinese Journal of Theoretical and Applied Mechanics, 2022, 54(7): 1755-1777. (in Chinese) doi: 10.6052/0459-1879-21-684 [2] Liang T, Liu B. Spence S. Effect of boundary layer suction on the corner separation in a highly loaded axial compressor cascade[J]. Journal of Turbomachinery, 2021, 143(6): 061002. doi: 10.1115/1.4050148 [3] Qian J Y, Wei L, Zhang M, et al. Flow rate analysis of compressible superheated steam through pressure reducing valves[J]. Energy, 2017, 135: 650-658. doi: 10.1016/j.energy.2017.06.170 [4] 凌代军, 代秋林, 朱榕川, 等. 叶栅试验技术综述[J]. 实验流体力学, 2021, 35(3): 30-38. Ling D J, Dai Q L, Zhu R C, et al. Review of the cascade experimental technology[J]. Journal of Experiments in Fluid Mechanics, 2021, 35(3): 30-38. (in Chinese doi: 10.11729/syltlx20200102Ling D J, Dai Q L, Zhu R C, et al. Review of the cascade experimental technology[J]. Journal of Experiments in Fluid Mechanics, 2021, 35(3): 30-38. (in Chinese) doi: 10.11729/syltlx20200102 [5] 张伟伟, 朱林阳, 刘溢浪, 等. 机器学习在湍流模型构建中的应用进展[J]. 空气动力学学报, 2019, 37(3): 444-454. Zhang W W, Zhu L Y, Liu Y L, et al. Progresses in the application of machine learning in turbulence modeling[J]. Acta Aerodynamica Sinica, 2019, 37(3): 444-454. (in ChineseZhang W W, Zhu L Y, Liu Y L, et al. Progresses in the application of machine learning in turbulence modeling[J]. Acta Aerodynamica Sinica, 2019, 37(3): 444-454. (in Chinese) [6] Evensen G. Sequential data assimilation with a nonlinear quasi-geostrophic model using Monte Carlo methods to forecast error statistics[J]. Journal of Geophysical Research: Oceans, 1994, 99(C5): 10143-10162. doi: 10.1029/94JC00572 [7] 朱佳楠, 郭建广, 倪彬彬, 等. 地球电子外辐射带的数据同化建模与分析[J]. 地球物理学报, 2021, 64(5): 1496-1507. Zhu J N, Guo J G, Ni B B, et al. Multi-dimensional Data Assimilation and Analyses of Earth’s Outer Electron Radiation Belt[J]. Chinese Journal of Geophysics, 2021, 64(5): 1496-1507. (in Chinese doi: 10.6038/cjg2021O0489Zhu J N, Guo J G, Ni B B, et al. Multi-dimensional Data Assimilation and Analyses of Earth’s Outer Electron Radiation Belt[J]. Chinese Journal of Geophysics, 2021, 64(5): 1496-1507. (in Chinese) doi: 10.6038/cjg2021O0489 [8] Canchumuni S W A, Emerick A A, Pacheco M A C. History matching geological facies models based on ensemble smoother and deep generative models[J]. Journal of Petroleum Science and Engineering, 2019, 177: 941-958. doi: 10.1016/j.petrol.2019.02.037 [9] Wang G Y, Pan Y L. Phase-resolved Ocean Wave Forecast with Ensemble-based Data Assimilation[J]. Journal of Fluid Mechanics, 2021, 918: A19. doi: 10.1017/jfm.2021.340 [10] 沈淳, 李健兵, 高航, 等. 基于数据同化的飞机尾流行为预测[J]. 雷达学报, 2021, 10(4): 632-645. Shen C, Li J B, Gao H, et al. Aircraft wake vortex behavior prediction based on data assimilation[J]. Journal of Radars, 2021, 10(4): 632-645. (in Chinese doi: 10.12000/JR21007Shen C, Li J B, Gao H, et al. Aircraft wake vortex behavior prediction based on data assimilation[J]. Journal of Radars, 2021, 10(4): 632-645. (in Chinese) doi: 10.12000/JR21007 [11] He C X, Liu Y Z. Time-resolved reconstruction of turbulent flows using linear stochastic estimation and sequential data assimilation[J]. Physics of Fluids, 2020, 32(7): 075106. doi: 10.1063/5.0014249 [12] Yang M C, Xiao Z X. Parameter Uncertainty Quantification for a Four-equation Transition Model using a Data Assimilation Approach[J]. Renewable Energy, 2020, 158: 215-226. doi: 10.1016/j.renene.2020.05.139 [13] 张鑫磊, 刘毅, 何国威. 集合卡尔曼方法及其在湍流建模中的应用[J]. 气动研究与试验, 2023, 1(1): 34-44. Zhang X L, Liu Y, He G W. Ensemble Kalman method and its applications in turbulence modelling[J]. Aerodynamic Research & Experiment, 2023, 1(1): 34-44. (in Chinese doi: 10.20118/j.issn2097-258x.2023.01.005Zhang X L, Liu Y, He G W. Ensemble Kalman method and its applications in turbulence modelling[J]. Aerodynamic Research & Experiment, 2023, 1(1): 34-44. (in Chinese) doi: 10.20118/j.issn2097-258x.2023.01.005 [14] He C X, Liu Y Z, Gan L. A data assimilation model for turbulent flows using continuous adjoint formulation[J]. PHYSICS OF FLUIDS, 2018, 30(10): 105108. doi: 10.1063/1.5048727 [15] Zhang X L, Xiao H, He G W. Assessment of regularized ensemble Kalman method for inversion of turbulence quantity fields[J]. AIAA Journal, 2021: 1-11. [16] 何创新, 邓志文, 刘应征. 湍流数据同化技术及应用[J]. 航空学报, 2021, 42(4): 161-178. He C X, Deng Z W, Liu Y Z. Turbulent Flow Data Assimilation and Its Applications[J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(4): 161-178. (in ChineseHe C X, Deng Z W, Liu Y Z. Turbulent Flow Data Assimilation and Its Applications[J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(4): 161-178. (in Chinese) [17] Evensen G, Leeuwen, P J. Assimilation of Geophysical Data Using the Ensemble Kalman Filter[J]. Monthly Weather Review, 1996, 124(5): 847-860. [18] Deng Z W, He C X, Wen X, et al. Recovering turbulent flow field from local quantity measurement: turbulence modeling using ensemble-Kalman-filter-based data assimilation[J]. Journal of Visualization, 2018, 21(6): 1043-1063. doi: 10.1007/s12650-018-0508-0 [19] Mons V, Chassaing J C, Gomez T, et al. Reconstruction of unsteady viscous flows using data assimilation schemes[J]. Journal of Computational Physics, 2016, 316: 255-280. doi: 10.1016/j.jcp.2016.04.022 [20] Kato H, Ishiko K, Yoshizawa A. Optimization of parameter values in the turbulence model aided by data assimilation[J]. AIAA Journal, 2016, 54(5): 1512-1523. doi: 10.2514/1.J054109 [21] 刘锬韬, 高丽敏, 蔡明, 等. 二维扩压叶栅流场的数据同化研究[J]. 工程热物理学报, 2022, 43(12): 3211-3218. Liu T T, Gao L M, Cai M, et al. Research of Data Assimilation on Two-dimensional Compressor Cascade Flow Field[J]. Journal of Engineering Thermophysics, 2022, 43(12): 3211-3218. (in ChineseLiu T T, Gao L M, Cai M, et al. Research of Data Assimilation on Two-dimensional Compressor Cascade Flow Field[J]. Journal of Engineering Thermophysics, 2022, 43(12): 3211-3218. (in Chinese) [22] 房培勋, 何创新, 徐嗣华, 等. 基于实验数据同化的湍流模型常数标定: 含滤网蒸汽阀门通流特性数值预测[J]. 空气动力学学报, 2021, 39(2): 12-22. Fang P X, He C X, Xu S H, et al. Calibration of turbulence model constants using measurement data assimilation: prediction of steam valve flow characteristics with filter[J]. Acta Aerodynamica Sinica, 2021, 39(2): 12-22. (in Chinese doi: 10.7638/kqdlxxb-2020.0183Fang P X, He C X, Xu S H, et al. Calibration of turbulence model constants using measurement data assimilation: prediction of steam valve flow characteristics with filter[J]. Acta Aerodynamica Sinica, 2021, 39(2): 12-22. (in Chinese) doi: 10.7638/kqdlxxb-2020.0183 [23] 刘锬韬, 李瑞宇, 高丽敏, 等. 基于数据同化的试验数据驱动的叶栅流场预测[J]. 航空学报, 2023, 44(14): 107-122. Liu T T, Li R Y, Gao L M, et al. Experimental data driven cascade flow field prediction based on data assimilation[J]. Acta Aeronautica et Astronautica Sinica, 2023, 44(14): 107-122. (in Chinese doi: 10.7527/S1000-6893.2023.28201Liu T T, Li R Y, Gao L M, et al. Experimental data driven cascade flow field prediction based on data assimilation[J]. Acta Aeronautica et Astronautica Sinica, 2023, 44(14): 107-122. (in Chinese) doi: 10.7527/S1000-6893.2023.28201 [24] Wilcox D C. Turbulence Modeling for CFD (3rd Edition) [M]. La Canada, US: DCW Industries, Inc, 2006. [25] Menter F R. Two-equation eddy-viscosity turbulence models for engineering applications[J]. AIAA Journal, 1994, 32(8): 1598-1605. doi: 10.2514/3.12149 -
点击查看大图
计量
- 文章访问数: 50
- HTML浏览量: 146
- PDF量: 4
- 被引次数: 0