基于概率场分量分解与m阶SPOD的喷嘴空化流场时空解耦试验研究
Experimental study on spatiotemporal decoupling of nozzle cavitating flow based on probability-field component decomposition and m-decomposed SPOD
摘要
针对喷嘴空化流场多尺度时空结构高度耦合、现有模态分解方法难以实现解耦且流场图像识别缺乏独立动力学交叉验证的问题,搭建了同步原位采集试验系统,对三种不同喷嘴开展对比试验.提出了基于概率场的分量分解方法,成功实现三类耦合时空结构的像素级软分离,并结合方位角傅里叶分解与m阶谱本征正交分解构建频率-模态联合分析方法.结果表明:三种喷嘴工况下流场结构分解分类清晰度均高于90%,波结构向溃灭事件分量的能量泄漏率低于1%;随着喷嘴扩张角增大,m=0轴对称模态能量占比由4.84%增至11.71%,m=1 一阶非轴对称模态能量占比由45.25%降至13.47%,壁面压力信号冲击强度与周期性增强,能量分布向低频区间集中.研究证实,喷嘴扩张方式可同步调控空化流场模态结构与壁面脉动特性,可为喷嘴优化设计与冲击特性调控提供支撑.
To address the difficulties that the multiscale spatiotemporal structures in nozzle cavitating flow are highly coupled,existing modal decomposition methods are unable to achieve effective decoupling,and image-based flow-field identification lacks independent dynamic cross-validation,a synchronous in situ measurement system was established and comparative experiments were conducted on three different nozzles.A probability-field-based component decomposition method was proposed to achieve pixel-level soft separation of three coupled spatiotemporal structures.In combination with azimuthal Fourier decomposition and m-decomposed spectral proper orthogonal decomposition,a frequency-mode joint analysis method was further developed.The results show that,under all three nozzle conditions,the classification clarity of flow-structure decomposition exceeds 90%,and the energy leakage from the wave component to the collapse-event component is lower than 1%.As the nozzle divergence angle increases,the energy fraction of the axisymmetric mode(m=0)increases from 4.84%to 11.71%,whereas that of the first-order non-axisymmetric mode(m=1)decreases from 45.25%to 13.47%.Meanwhile,the impact intensity and periodicity of the wall-pressure signal increase,and the energy distribution becomes more concentrated in the low-frequency range.The results confirm that the nozzle divergent geometry can simultaneously regulate the modal structure of cavitating flow and the wall-pressure pulsation characteristics,providing support for nozzle optimal design and impact-characteristic control.
年,卷(期):
2026, 45(20)
分类号:
TH113.1
O352
机标分类号:
TP391.41
O4
TQ323.5
在线出版日期:
2026年11月3日
基金项目:
2024QN-B043:火箭军工程大学青年基金,