Petascale algorithms for reactor hydrodynamics

Fischer, Paul; Lottes, James; Pointer, W. David; Siegel, A.

doi:10.1088/1742-6596/125/1/012076

Cited by 107 publications

(89 citation statements)

References 4 publications

(4 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Below, we discuss mainly three different experiments from low to large forcing. Numerical computations are performed with spectral element methods [28,29]. The geometry of the numerical setup closely reproduces the experimental one.…”

mentioning

confidence: 99%

Energy cascade in internal-wave attractors

Brouzet¹,

Ermanyuk²,

Joubaud³

et al. 2016

EPL

View full text Add to dashboard Cite

-One of the pivotal questions in the dynamics of the oceans is related to the cascade of mechanical energy in the abyss and its contribution to mixing. Here, we propose internal wave attractors in the large amplitude regime as a unique self-consistent experimental and numerical setup that models a cascade of triadic interactions transferring energy from large-scale monochromatic input to multi-scale internal wave motion. We also provide signatures of a discrete wave turbulence framework for internal waves. Finally, we show how beyond this regime, we have a clear transition to a regime of small-scale high-vorticity events which induce mixing.Introduction. -The continuous energy input to the ocean interior comes from the interaction of global tides with the bottom topography [1]. The subsequent mechanical energy cascade to small-scale internal-wave motion and mixing is a subject of active debate [2] in view of the important role played by abyssal mixing in existing models of ocean dynamics [3][4][5]. A question remains: how does energy injected through internal waves at large vertical scales [6] induce the mixing of the fluid [2]?In a stratified fluid with an initially constant buoyancy frequency N = [(−g/ρ)(dρ/dz)] 1/2 , where ρ(z) is the density distribution (ρ a reference value) over vertical coordinate z, and g the gravity acceleration, the dispersion relation of internal waves is θ = ± arcsin(Ω). The angle θ is the slope of the wave beam to the horizontal, and Ω the frequency of oscillations non-dimensionalized by N . The dispersion relation requires preservation of the slope of the internal wave beam upon reflection at a rigid boundary. In the case of a sloping boundary, this property gives a purely geometric reason for a strong variation of the width of internal wave beams (focusing or defocusing) upon reflection. Internal wave focusing provides a necessary condition for large shear and overturning, as well as shear and bottom layer instabilities at slopes [7][8][9][10].In a confined fluid domain, focusing usually prevails, leading to a concentration of wave energy on a closed loop, the internal wave attractor [11]. Attractors eventually

show abstract

mentioning

confidence: 99%

Energy cascade in internal-wave attractors

Brouzet¹,

Ermanyuk²,

Joubaud³

et al. 2016

EPL

View full text Add to dashboard Cite

show abstract

“…Likewise, we will describe P3DFFT, a 3D fast Fourier transformation library, and compare its relative performance on both networks. NEK5000 [5] is a spectral element CFD code developed at Argonne National Laboratory, which features spectral element multigrid solvers coupled to a highly scalable parallel coarse grid solver. It was recognized in 1999 with a Gordon Bell prize and is used by more than two dozen research institutions worldwide for projects including ocean current modeling, thermal hydraulics of reactor cores and spatiotemporal chaos.…”

Section: Application Resultsmentioning

confidence: 99%

Improving Resource Availability by Relaxing Network Allocation Constraints on Blue Gene/P

Desai

Buntinas

Buettner

et al. 2009

2009 International Conference on Parallel Processing

View full text Add to dashboard Cite

show abstract

“…Many simulation [6,14,17], visualization, and analysis frameworks [1,7,33,38] are data-parallel, meaning each process executes the same program on a different part of the data. There is, however, a subtle but important difference between those parallel models and ours.…”

Section: Data Parallelism and Block-structured Abstractionsmentioning

confidence: 99%