Experiments on very large structures in fully developed turbulent pipe flow

点击翻译按钮获取中文摘要

The present work aims at investigation of turbulent pipe flow at high Reynolds numbers and low Mach numbers. During the recent years there has been an increasing interest in observation and understanding of large scale turbulent coherent structures, which are commonly known as Large and Very Large Scale Motions (LSM and VLSM). Nevertheless, a solid definition of their nature and vivid understanding of their evolution is still incomplete. Therefore, this study will focus on clarifying the nature and origin of LSM and VLSM as well as describing and identifying them in a quantitative manner. To this end, experiments and numerical computations (together with possible partners within the SPP) will be performed and matched as closely as possible. Experiments at Cottbus Large Pipe Test Facility (CoLa-Pipe), which was used successfully along the first phase, will be conducted at bulk Reynolds numbers of 6x10E4 ≤ Re b ≤ 1x10E6 (based on pipe diameter D and bulk velocity Ub) and Mach numbers Ma < 0.23, measuring turbulent flow properties using miniaturized Hot Wire Anemometry (HWA), high speed Particle Image Velocimetry (PIV) and Shake The Box (STB) particle tracking technique (in collaboration with DLR, A. Schröder).This proposal aims at two prominent objectives. Both objectives build on our findings in the first phase of this study and they will expand our understanding of turbulent structures using the measurement techniques which provide higher spatial and temporal resolutions. Our first goal is to clarify the uncertainties concerning scaling of structural turbulence properties using miniaturized HWA and 3D high resolution profile measurements. The second goal is to quantify the kinematics and dynamics of large-scale coherent structures. The length scales, energy contents and wall-normal locations of such structures have been already determined in pre-multiplied energy spectra during the first phase of this study. In the second phase, we will extract low order subspaces of highly dimensional turbulent flow, from 2D and 3D time-resolved measurements on a moving frame of reference by applying a Characteristic DMD (in collaboration with TU Berlin, J. Sesterhenn). Optimized combinations of extracted subspaces with long life times will form reduced order models, which accommodate structures that are known to be responsible for the formation of the spectral peaks. Life times and spatio-temporal evolutions of each group of structures will be studied in absence of small-scale structures. This will allow to determine how such structures contribute to turbulence properties such as TKE budgets, Reynolds stress and viscous shear stress.