AFiD-GPU: A versatile Navier–Stokes solver for wall-bounded turbulent flows on GPU clusters

Zhu, Xiaojue; Phillips, Everett; Spandan, Vamsi; Donners, John; Ruetsch, Gregory; Romero, Joshua; Ostilla–Mónico, Rodolfo; Yang, Yantao; Lohse, Detlef; Verzicco, Roberto; Fatica, Massimiliano; Stevens, Richard J. A. M.

doi:10.1016/j.cpc.2018.03.026

Cited by 82 publications

(67 citation statements)

References 72 publications

(109 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition to the experiments, three-dimensional direct numberical simulations (DNS) of the Boussinesq equations are performed by using the in-house AFiD code [30,31]. An immersed boundary method [32] is implemented to simulate the ratchet surfaces.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Controlling Heat Transport and Flow Structures in Thermal Turbulence Using Ratchet Surfaces

et al. 2018

Self Cite

View full text Add to dashboard Cite

In this combined experimental and numerical study on thermally driven turbulence in a rectangular cell, the global heat transport and the coherent flow structures are controlled with an asymmetric ratchet-like roughness on the top and bottom plates. We show that, by means of symmetry breaking due to the presence of the ratchet structures on the conducting plates, the orientation of the Large Scale Circulation Roll (LSCR) can be locked to a preferred direction even when the cell is perfectly leveled out. By introducing a small tilt to the system, we show that the LSCR orientation can be tuned and controlled. The two different orientations of LSCR give two quite different heat transport efficiencies, indicating that heat transport is sensitive to the LSCR direction over the asymmetric roughness structure. Through a quantitative analysis of the dynamics of thermal plume emissions and the orientation of the LSCR over the asymmetric structure, we provide a physical explanation for these findings. The current work has important implications for passive and active flow control in engineering, bio-fluid dynamics, and geophysical flows.Turbulent convective flows over rough surfaces are ubiquitous in engineering and geophysical flows. Examples include convective flows in the atmosphere and in oceans, where the ground, sea-bed and ocean floor are generally not smooth. As the ability to enhance convective heat transfer is crucial in many industrial applications, numerous strategies have been proposed to efficiently enhance it. Among several approaches, introducing wall roughness is an effective way to do so. Indeed, the study of surface roughness effects in wall-bounded turbulent flows has been an area of intense research (see e.g. some recent work [1][2][3][4][5][6][7][8], the reviews [9,10], and the textbooks [11,12]). Similarly, several studies have been conducted on turbulent thermal convection over rough plates [13][14][15][16][17][18][19][20]. The vast majority of these studies with rough walls adopt some ordered and symmetrical structures, such as pyramids, squares, rectangles etc. However, the rough surfaces in engineering applications and in nature are in general not symmetric, resulting in complex interactions between the flow and the asymmetric roughness elements. Examples are wind blowing over a landscape with asymmetric slopes and ocean flows over an asymmetric sea-bed, etc. Other examples include marine animals which can actively change the asymmetric roughness for maneuverability.In this work, we aim to study the influences of ratchetlike wall structures on the flow organization and heat transfer in fully developed convective thermal turbulence. Indeed, building on the classical Feynman-Smoluchowski ratchet, in various contexts researchers have proposed ratchet-type mechanisms and devices that operate outside of thermal equilibrium [21]. Examples include the so-called 'capillary-ratchet' in zoology [22], a rotational ratchet in a granular gas [23], self-propelled Leidenfrost droplets and solids on ratchet sur...

show abstract

mentioning

confidence: 99%

“…An immersed boundary method [32] is implemented to simulate the ratchet surfaces. The code has been extensively validated and used [30,31,33]. Adequate resolutions [34] are ensured for all simulations so that the results are grid independent.…”

mentioning

confidence: 99%

Controlling Heat Transport and Flow Structures in Thermal Turbulence Using Ratchet Surfaces

et al. 2018

Self Cite

View full text Add to dashboard Cite

show abstract

“…The open source finite-difference Navier-Stokes solver AFiD [12] was written in Fortran 90 to study large-scale wall bounded turbulent flow simulations. In collaboration with NVIDIA, USA, the code was ported in its newest version to a GPU setting using an MPI and CUDA Fortran hybrid implementation optimized to run and solve large flow fields [13].…”

Section: Numerical Simulations Of Sheared Thermal Convectionmentioning

confidence: 99%

A comparative evaluation of three volume rendering libraries for the visualization of sheared thermal convection

Favre

Blass

2019

Parallel Computing

View full text Add to dashboard Cite

Oceans play a big role in the nature of our planet, about 70% of our earth is covered by water [1]. Strong currents are transporting warm water around the world making life possible, and allowing us to harvest its power producing energy. Yet, oceans also carry a much more deadly side. Floods and tsunamis can easily annihilate whole cities and destroy life in seconds. The earth's climate system is also very much linked to the currents in the ocean due to its large coverage of the earth's surface, thus, gaining scientific insights into the mechanisms and effects through simulations is of high importance. Deep ocean currents can be simulated by means of wall-bounded turbulent flow simulations. To support these very large scale numerical simulations and enable the scientists to interpret their output, we deploy an interactive visualization framework to study sheared thermal convection. The visualizations are based on volume rendering of the temperature field. To address the needs of supercomputer users with different hardware and software resources, we evaluate different volume rendering implementations supported in the ParaView [2] environment: two GPU-based solutions with Kitware's native volume mapper or NVIDIA's IndeX library, and a CPU-only Intel OSPRay-based implementation. Figure 1: Snapshot of the three-dimensional temperature field of sheared thermal convection at Ra = 4.6 × 10 6 and Re w = 6000 [3].

show abstract

“…Weak scaling analysis up to an extreme scale mesh with more than a trillion computational points9.3. Comparison with Other SoftwareAFiD-GPU[18] is an open-source incompressible Navier-Stokes solver developed with a mixed MPI-CUDA Fortran implementation. AFiD simulates wallbounded buoyancy-driven turbulent flows.…”

mentioning

confidence: 99%

PittPack: An open-source Poisson’s equation solver for extreme-scale computing with accelerators

Hasbestan

Xiao

Şenocak

2020

Computer Physics Communications

View full text Add to dashboard Cite

We present a parallel implementation of a direct solver for the Poisson's equation on extreme-scale supercomputers with accelerators. We introduce a chunkedpencil decomposition as the domain-decomposition strategy to distribute work among processing elements to achieve superior scalability at large number of accelerators. Chunked-pencil decomposition enables overlapping nodal communication and data transfer between the central processing units (CPUs) and the graphics processing units (GPUs). Second, it improves data locality by keeping neighboring elements in adjacent memory locations. Third, it allows usage of shared-memory for certain segments of the algorithm when possible, and last but not least, it enables contiguous message transfer among the nodes. Two different communication patterns are designed. The fist pattern aims to fully overlap the communication with data transfer and designed for speedup of overall turnaround time, whereas the second method concentrates on low memory usage and is more network friendly for computations at extreme scale. To ensure software portability, we interleave OpenACC with MPI in the software. The numerical solution and its formal second order of accuracy is verified using method of manufactured solutions for various combinations of boundary conditions. Weak scaling analysis is performed using up to 1.1 trillion Cartesian mesh points using 16384 GPUs on a petascale leadership class supercomputer.

show abstract

AFiD-GPU: A versatile Navier–Stokes solver for wall-bounded turbulent flows on GPU clusters

Cited by 82 publications

References 72 publications

Controlling Heat Transport and Flow Structures in Thermal Turbulence Using Ratchet Surfaces

Controlling Heat Transport and Flow Structures in Thermal Turbulence Using Ratchet Surfaces

A comparative evaluation of three volume rendering libraries for the visualization of sheared thermal convection

PittPack: An open-source Poisson’s equation solver for extreme-scale computing with accelerators

Contact Info

Product

Resources

About