Anton P. Markesteijn

Concurrent multiscale modelling of atomistic and hydrodynamic processes in liquids

Fluctuations of liquids at the scales where the hydrodynamic and atomistic descriptions overlap are considered. The importance of these fluctuations for atomistic motions is discussed and examples of their accurate modelling with a multi-space–time-scale fluctuating hydrodynamics scheme are provided. To resolve microscopic details of liquid systems, including biomolecular solutions, together with macroscopic fluctuations in space–time, a novel hybrid atomistic–fluctuating hydrodynamics approach is introduced. For a smooth transition between the atomistic and continuum representations, an analogy with two-phase hydrodynamics is used that leads to a strict preservation of macroscopic mass and momentum conservation laws. Examples of numerical implementation of the new hybrid approach for the multiscale simulation of liquid argon in equilibrium conditions are provided.

Time asynchronous relative dimension in space method for multi-scale problems in fluid dynamics

A novel computational method is presented for solving fluid dynamics equations in the multi-scale framework when the system size is an important parameter of the governing equations. The method (TARDIS) is based on a concurrent transformation of the governing equations in space and time and solving the transformed equations on a uniform Cartesian grid with the corresponding causality conditions at the grid interfaces. For implementation in the framework of TARDIS, the second-order CABARET scheme of Karabasov and Goloviznin 1] is selected for it provides a good combination of numerical accuracy, computational efficiency and simplicity of realisation. Numerical examples are first provided for several isothermal gas dynamics test problems and then for modelling of molecular fluctuations inside a microscopic flow channel and ultrasound wave propagation through a nano-scale region of molecular fluctuations.

CABARET GPU Solver for Fast-Turn-Around Flow and Noise Calculations

A GPU implementation of an aero-acoustic solver based on the low-dissipative, low-dispersive CABARET scheme is demonstrated. The implementation makes industrial relevant (10-50 million cells) LES studies of jet noise modeling possible in reasonable time (several days) using just a workstation computer, therefore avoiding the need for supercomputing facilities for these cases. Besides the solving of the turbulence, the post-processing relevant to the applications in mind can be done on the fly on the GPU as well.

Empiricism-free noise calculation from LES solution based on Goldstein generalized acoustic analogy: volume noise sources and meanflow effects

The hybrid CAA method with propagation equations corresponded to the Goldstein generalized acoustic analogy and noise sources extracted from time-resolved LES solution is considered. In the current implementation, which includes sound propagation meanflow effects, noise source information was extracted directly from LES without any assumptions of statistic properties such as fourth-order velocity correlations, which makes this tool applicable for investigation of jet mixing noise without using any intermediate assumptions. Far-field noise and noise source statistics have been calculated for two subsonic single-stream jets. One jet is isothermal and the other is heated corresponding to temperature ratio ( ); they have the same acoustic Mach number ( ) and correspond to jet conditions of model-scale tests recently conducted at the QinetiQ Noise Test Facility. For unsteady flow calculation, Monotonically Integrated Large Eddy Simulation (MILES) is used based on the high-resolution CABARET scheme. Flow field solutions and far-field noise predictions are compared with the experimental data available as well as with the results of the Ffowcs Williams Hawkings solution based on the same set of LES data.

Acoustic Wave Focusing by Non-uniform Mean Flow in a Rectangular Duct with Viscous Walls

The high-frequency acoustic wave focusing by a viscous flow profile and reflective duct walls encasing the flow is considered. For a range of the non-uniform meanflow and duct parameters, the problem of image sources generation upstream of the physical source in a rectangular duct is investigated computationally and analytically. This is done first by solving the Navier-Stokes equations at high Reynolds number in the Eulerian framework, then by using a Lagrangian particle method based on the high-frequency approximation, and finally, with a simple analytical plug-flow-profile model based on geometrical acoustics. It is shown that both the meanflow-refraction and wall-reflection play a key role in the focusing of acoustic waves. The analogy with geometrical acoustics is further emphasized by comparing the Navier-Stokes simulation results with the experimental data from the literature on acoustic wave focusing by a lens made of a negatively refractive material.