Development of gas temperature imaging by Laser-Induced Fluorescence for internal cooling channels

Forced convection is a phenomenon already widely investigated in literature. However, there remain potential important improvements on its understanding and numerical prediction. The classical RANS approaches, widely used in industry for the design of cooling systems, are based on a constant turbulent Prandtl number value and isotropic turbulence. This results in a limited fidelity for the prediction of the convective heat transfer coefficient (at best 10-15% of accuracy). The present investigation, therefore, aims at supporting the development of CFD codes by providing detailed and accurate experimental data. Unlike most of the investigations available in literature, which provide the wall heat transfer and/or the velocity fields, the present research addresses the flow temperature. This temperature information allows obtaining the turbulent heat transport term required for the determination of the turbulent Prandtl number value. Laser-Induced Fluorescence (LIF) has the advantage to provide a 2D temperature distribution by a non-intrusive optical approach. Gas-phase measurements, required for a consistent heat transfer investigation, were never applied before in internal flows as in the present application. First of all, the use of a fluorescent tracer is investigated regarding its impact on the Prandtl number value of the gas. After a detailed evaluation of the technique, including the image processing, the first experiments in a static cell show encouraging results of the first two-color measurements. A novel facility was designed and built at the von Karman Institute (VKI) to extend the study to the selected application. For simplicity, the test section is a smooth high aspect ratio duct to investigate a quasi 2D internal flow, as this configuration is closer to a cooling channel. The measurements performed with the one-color methodology and using toluene as the fluorescent tracer show a good agreement between the obtained mean temperature profile and the one resulting from a traversing thermocouple probe. On the other hand, temperature fluctuations are strongly affected by the signal fluctuations recorded by the camera. The facility is prepared for possible nitrogen flow in order to eliminate the undesirable effect of oxygen quenching. Besides, the implementation of a Particle Image Velocimetry (PIV) system allows simultaneous velocity and temperature measurements with results of the turbulent heat transport term. The deep discussion and conclusions allow the author to outline possible future perspectives for further development of this technique.(FSA - Sciences de l'ingénieur) -- UCL, 201