Direct Simulation of a Low Momentum Round Jet in Channel Crossflow with Conjugate Heat Transfer

Abstract

Results of direct numerical simulations (DNS) of a jet in channel crossflow with conjugate heat transfer are presented. The hot laminar jet issues from a circular exit into the cold channel crossflow with a low jet-to-crossflow velocity ratio of 1/6. The steel channel wall has a finite thickness and its outer side is cooled under Robin type thermal boundary conditions for a realistic external environment, leading to a conjugate heat transfer system. The governing equations are solved by Incompact3d, an open-source code combining the high-order finite difference compact scheme and Poisson spectral solver. An internal recycling approach is used to generate the fully turbulent channel flow profile as the crossflow inflow conditions. The database is available online for open access (http://dx.doi.org/10.17632/7nx4prgjzz.3). In the fluid domain, four main flow structures are identified: 1) a large recirculation immediately downstream of the jet-exit; 2) a contour-rotating vortex pair originated from the stretching and reorientation of the injection-flow vorticity; 3) a horseshoe vortex generated as a result of the stretching of the vorticity at the jet-exit windward side; and 4) shear layer vortices coming from the lifted and shed crossflow boundary layer vorticity. Proper orthogonal decomposition and dynamic mode decomposition methods are then used to study the energy and spectrum information of structures. The results show the horseshoe vortex is related to low-frequency modes, while the shear layer vortices are connected to the high-frequency ones. In the conjugate heat transfer problem, the above coherent structures lead to a complex convective and turbulent wall heat transfer pattern around the orifice. Finally, this study evaluates the capabilities of several turbulence models in predicting this type of flow and shows how the DNS database would help test, validate and improve the turbulence models.

Date of Award	31 Dec 2017
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Dominique Laurence (Supervisor) & Imran Afgan (Supervisor)