City Research Online

Abstract

This paper presents an approach both to receiving multiple sensor data from a flow in real time and to analyzing these data in order to characterize the flow condition and, if necessary, control the flow. In order to obtain the data, an optical micro-pillar array acting as distributed wall-shear sensor was developed and interrogated optically with an LDM (long distance microscope). Together, the micro-pillar array and the LDM form a channeling optics, which allows magnified imaging of larger numbers of individual pillars simultaneously. The sensor was tested in a turbulent wall shear stress field under varying conditions (Reynolds number). A frame rate of 3000 fps was used since the higher the temporal resolution is, the more specific flow control strategies might be applied later in realistic application. However, the temporal high resolution would lead to a vast amount of data, which is difficult to analyze in real time. Therefore, a fast image processing algorithm is developed, which detects the tip deflections of the pillars and vectorizes the wall-shear stress field online. The extracted data fields are then broken down into equidistant and overlapping windows in order to guarantee fast POD (proper orthogonal decomposition) modes calculation. The POD is applied to each of these windows and the extracted modes are compared, summarized and collected in a library. Finally, this library is again applied to the flow but under different conditions in order to identify the state of the current flow in real time.

Publication Type:	Conference or Workshop Item (Paper)
Subjects:	T Technology > TJ Mechanical engineering and machinery T Technology > TL Motor vehicles. Aeronautics. Astronautics
Departments:	School of Science & Technology > Engineering

Export

Downloads