Lintermann, Andreas; Schröder, Wolfgang Lattice–Boltzmann simulations for complex geometries on high-performance computers Journal Article In: CEAS Aeronautical Journal, 2020, ISBN: 1869-5582. @article{Lintermann2020c,
title = {Lattice–Boltzmann simulations for complex geometries on high-performance computers},
author = {Lintermann, Andreas and Schröder, Wolfgang },
url = {http://link.springer.com/10.1007/s13272-020-00450-1},
doi = {10.1007/s13272-020-00450-1},
isbn = {1869-5582},
year = {2020},
date = {2020-05-13},
journal = {CEAS Aeronautical Journal},
abstract = {Complex geometries pose multiple challenges to the field of computational fluid dynamics. Grid generation for intricate objects is often difficult and requires accurate and scalable geometrical methods to generate meshes for large-scale computations. Such simulations, furthermore, presume optimized scalability on high-performance computers to solve high-dimensional physical problems in an adequate time. Accurate boundary treatment for complex shapes is another issue and influences parallel load-balance. In addition, large serial geometries prevent efficient computations due to their increased memory footprint, which leads to reduced memory availability for computations. In this paper, a framework is presented that is able to address the aforementioned problems. Hierarchical Cartesian boundary-refined meshes for complex geometries are obtained by a massively parallel grid generator. In this process, the geometry is parallelized for efficient computation. Simulations on large-scale meshes are performed by a high-scaling lattice–Boltzmann method using the second-order accurate interpolated bounce-back boundary conditions for no-slip walls. The method employs Hilbert decompositioning for parallel distribution and is hybrid MPI/OpenMP parallelized. The parallel geometry allows to speed up the pre-processing of the solver and massively reduces the local memory footprint. The efficiency of the computational framework, the application of which to, e.g., subsonic aerodynamic problems is straightforward, is shown by simulating clearly different flow problems such as the flow in the human airways, in gas diffusion layers of fuel cells, and around an airplane landing gear configuration},
keywords = {Airway, Computational Fluid Dynamics, High performance computing, Large-Scale Simulation Data, Lattice-Boltzmann method},
pubstate = {published},
tppubtype = {article}
}
Complex geometries pose multiple challenges to the field of computational fluid dynamics. Grid generation for intricate objects is often difficult and requires accurate and scalable geometrical methods to generate meshes for large-scale computations. Such simulations, furthermore, presume optimized scalability on high-performance computers to solve high-dimensional physical problems in an adequate time. Accurate boundary treatment for complex shapes is another issue and influences parallel load-balance. In addition, large serial geometries prevent efficient computations due to their increased memory footprint, which leads to reduced memory availability for computations. In this paper, a framework is presented that is able to address the aforementioned problems. Hierarchical Cartesian boundary-refined meshes for complex geometries are obtained by a massively parallel grid generator. In this process, the geometry is parallelized for efficient computation. Simulations on large-scale meshes are performed by a high-scaling lattice–Boltzmann method using the second-order accurate interpolated bounce-back boundary conditions for no-slip walls. The method employs Hilbert decompositioning for parallel distribution and is hybrid MPI/OpenMP parallelized. The parallel geometry allows to speed up the pre-processing of the solver and massively reduces the local memory footprint. The efficiency of the computational framework, the application of which to, e.g., subsonic aerodynamic problems is straightforward, is shown by simulating clearly different flow problems such as the flow in the human airways, in gas diffusion layers of fuel cells, and around an airplane landing gear configuration | |
Göbbert, Jens Henrik; Habbinga, Sonja; Lintermann, Andreas Comprehensive Visualization of Large-Scale Simulation Data Linked to Respiratory Flow Computations on HPC Systems Online Jülich, Forschungszentrum (Ed.): Forschungszentrum Jülich 2017, visited: 24.10.2017. @online{jscaia-flow,
title = {Comprehensive Visualization of Large-Scale Simulation Data Linked to Respiratory Flow Computations on HPC Systems},
author = {Göbbert, Jens Henrik and Habbinga, Sonja and Lintermann, Andreas },
editor = {Forschungszentrum Jülich},
url = {https://www.youtube.com/watch?v=FmPvHIZSjyk, Link to Video},
year = {2017},
date = {2017-10-24},
urldate = {2017-10-24},
organization = {Forschungszentrum Jülich},
abstract = {Conditioning large-scale simulation data for comprehensive visualizations to enhance intuitive understanding of complex physical phenomena is a challenging task. This is corroborated by the fact that the massive amount of data produced by such simulations exceeds the human horizon of perception. It is therefore essential to distill the key features of such data to derive at new knowledge on an abstract level.
Furthermore, presenting scientific data to a wide public audience, especially if the scientific content is of high societal interest, i.e., as it is the case for fine dust pollution, is not only difficult from a visualization but also from an information transfer point of view. Impressive visual and contextual presentation are hence key to an effective knowledge transfer of complicated scientific data and the involved methods to arrive at such data. This is presented for highly-dense simulation data stemming from HPC simulations of inspiratory flows in the human respiratory tract. The simulations are performed on JUQUEEN using a coupled lattice-Boltzmann/Lagrange method and aim at understanding the microscopic interactions of flow and particle dynamics in highly intricate anatomically correct geometries. As such, they deliver insights on the impact of particulate matter on the human body.},
howpublished = {You-Tube Channel Forschungszentrum Jülich},
keywords = {Large-Scale Simulation Data, Respiratory Flow Computation, Visualization},
pubstate = {published},
tppubtype = {online}
}
Conditioning large-scale simulation data for comprehensive visualizations to enhance intuitive understanding of complex physical phenomena is a challenging task. This is corroborated by the fact that the massive amount of data produced by such simulations exceeds the human horizon of perception. It is therefore essential to distill the key features of such data to derive at new knowledge on an abstract level.
Furthermore, presenting scientific data to a wide public audience, especially if the scientific content is of high societal interest, i.e., as it is the case for fine dust pollution, is not only difficult from a visualization but also from an information transfer point of view. Impressive visual and contextual presentation are hence key to an effective knowledge transfer of complicated scientific data and the involved methods to arrive at such data. This is presented for highly-dense simulation data stemming from HPC simulations of inspiratory flows in the human respiratory tract. The simulations are performed on JUQUEEN using a coupled lattice-Boltzmann/Lagrange method and aim at understanding the microscopic interactions of flow and particle dynamics in highly intricate anatomically correct geometries. As such, they deliver insights on the impact of particulate matter on the human body. | |
