Interpreting proper orthogonal decomposition modes extracted from partial cavity oscillation

被引:1
|
作者
Yin, Tingyun [1 ]
Pavesi, Giorgio [2 ,3 ]
机构
[1] Univ Cagliari, Dipartimento Ingn Meccan Chim & Materiali, Via Marengo 2, I-09123 Cagliari, Sardegna, Italy
[2] Univ Padua, Turbomachinery & Energy Syst TES Grp, Via Venezia 1, I-35131 Padua, Veneto, Italy
[3] Univ Padua, Dipartimento Ingn Industriale, Via Venezia 1, I-35131 Padua, Veneto, Italy
关键词
LARGE-EDDY SIMULATION; BUBBLY SHOCK PROPAGATION; CLOUD CAVITATION; NUMERICAL-SIMULATION; SHEET; TRANSITION; FLOW; MECHANISM; DYNAMICS;
D O I
10.1063/5.0244165
中图分类号
O3 [力学];
学科分类号
08 ; 0801 ;
摘要
This study employs the two-dimensional proper orthogonal decomposition approach to analyze the pressure, vapor fraction, and streamwise velocity flowfields of partial cavity oscillation. The interrelations among mode, energy ratio, temporal coefficient, and flowfield reconstruction are thoroughly examined, thereby augmenting comprehension of the cavitating flow mechanism and bubble dynamics. It is found that the first modes of the pressure, vapor fraction, and streamwise velocity flowfields contain 56.31%, 36.37%, and 31.81% energy, respectively; the decrease in energy ratio results in the variation of its temporal coefficient close to sinusoidal configurations. Moreover, the temporal coefficient of the first mode varies closely related to the flowfield-relevant variable. The first modes of the pressure, vapor fraction, and streamwise velocity flowfields are significantly different, but all have two highlighted structures closely related to the self-variable system. The strong nonlinearity and high dimensionality of the cavitation flowfield render precise reconstruction using a limited number of modes exceedingly challenging. The data approximate the original snapshot more closely when the flow field is reconstructed with a greater number of modes. Although the location with a relatively high root mean square reconstruction error is significantly different when the first nine modes are used for flowfield reconstruction, its order of magnitude is less than the self-variable system, and the order discrepancy is fixed, equal to 1. (C) 2025 Author(s).
引用
收藏
页数:16
相关论文
共 20 条