Colloquium: Rémi Allandrieu (FPP)
11 december 2018 09:30 - Locatie: Instruction Room 5, The Fellowship, Kluyverweg 5, 2629 HS, Delft.
Experimental study of fluid-structure interactions using a soap film channel
Fluid structure interactions are present everywhere, and are of crucial importance in the development of new concepts in the aerospace industry. Their study through numerical approaches such as CFD are of great interest for this sector, and understanding how to control them is key to the evolution of the industry. In this work, we present an experimental method to generate quasi-2D flows and study fluid structure interactions thanks to the use of a soap film channel and a high-speed video camera. A simple test case is presented, introducing the fluid structure instability affecting a structure composed of a rigid cylinder with a flexible silk filament attached to its rear end. A complete description of the soap filmchannel is given. Explanations on how to build the apparatus, the required lighting and recording conditions as well as the fluid to be used are also provided. A set of relevant post-processing tools including a vortex shedding frequency analysis method and a position tacking algorithm are given to extract critical information from the experimental video recordings. A theoretical model was constructed in order to predict the effect of the fluid structure instability on our test case and compare its results to the experimental data obtained with the soap film. The model proved itself to be particularly relevant for the static analysis of the phenomenon, and predicts the stationary instability affecting the average position of the filament with precision. Unlike existing models, it takes into account the action of the stiffness of the filament and provides a better description of the phenomenon. However, its simplistic construction and the strong hypothesis prevent the dynamic results to be efficiently used. We highlight the advantages and the relevance of the soap film channel to obtain experimental data in quasi-2D conditions, and provide guidance for future researchers to use this technique.