Keywords: hypersonic flow, wall cooling, multiscale simulations, adaptive mesh refinement.

High-fidelity direct numerical simulations (DNS) of the complex flow physics of a hypersonic boundary layer interacting with the coolant flow of a porous surface require the combination of a high-order-accurate numerical scheme and an efficient mesh adaptation methodology implemented for massively parallel distributed memory systems. Our block-structured adaptive mesh refinement (SAMR) framework AMROC provides patch-based hierachical mesh refinement in space and time for explicit finite volume methods. This enables high-resolution simulations involving multiple scales, allowing for the coupling and simultaneous solution of two communicating flow regions characterised by very different length scales, as e.g. the millimiter-scale of the outer boundary layer flow and the micrometer-scale of the flow in the interior porous layer. The overall computational efficiency of the method is achieved for difficult 3D DNS problems by the integration of the SAMR methodology with a high-order, specially hybridised, finite-volume solver, which guarantees the benefits of a 6th-order central base scheme for the accurate solution of the smooth regions, and of a 6th-order WENO scheme for the treatment of discontinuities.

In the present work we show the results reached to date in the context of multiscale simulations of hypersonic flows with coolant injection from a porous wall by using the described numerical approach. The simulations are aimed at both film and transpiration cooling applications. In particular, results are presented and discussed for the case of a flat plate with coolant injection through thin slots, and a flat plate where coolant injection is provided by the flowtranspiring from an underneath layer of distributed porosity. Different species of coolant gas are also considered. Simulations have been run in the high performance computing environment based on Cray XC30 architecture of the UK supercomputer facility ARCHER. Each compute node in the HPC facility contains two 2.7 GHz 12-core series processors. Our high-resolution 3D simulations employed 3000-7000 cores.

Due to the numerical constraints linked to the high complexity of the hypersonic flow physics and the intrinsic multiscale nature of the porous media flow, the full flowfield resolution by direct simulation enabled by parallel adaptive mesh refinement is a challenging but at the same time novel approach in this specific field. Moreover, the considered applications are of high interest and increasing importance in the hypersonic scientific community, as well as, in general, in all the engineering applications concerning wall cooling in harsh high-temperature flow environments.

purchase the full-text of this paper (price £22)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £25 +P&P)