Richard J.A.M. Stevens

Professor at the University of Twente

Wind energy · Turbulence · Environmental flows

Turbulent superstructures

Our AFiD simulations reveal that turbulent thermal superstructures survive in fully turbulent convection, even in very wide domains; see figure 1 [1]. Later work showed that temperature and vertical-velocity superstructures have comparable characteristic sizes and that this size increases with Rayleigh number [2].

The required domain size depends on the observable. In the studied cases, heat transfer and total flow strength reach their wide-domain limits near an aspect ratio of 4, whereas temperature and velocity variance profiles converge near 8 in periodic domains and near 16 in cylindrical cells [3]. Thus, large-scale organization remains important for the flow statistics even after global transport appears converged.

Temperature field snapshots showing thermal superstructures and plumes in a wide Rayleigh-Bénard convection domain
Figure 1. Snapshots of the temperature field in a horizontal domain that is 64 times as long as the domain height. The columns from left to right show successive zooms of the area indicated by the black box. The upper panels show the flow structure close to the bottom plate, where small hot thermal plumes carry away heat, while the lower panels reveal the large-scale flow patterns in the center of the domain. Full-resolution figure.

Convection in horizontally extended systems

Many natural instances of convection take place in horizontally extended systems. Rayleigh-Bénard flow, in which a fluid layer is heated from below and cooled from above, provides a controlled system for studying the resulting large-scale patterns. Hot fluid rises and cold fluid descends, linking small-scale thermal plumes to the wider flow organization.

Small-aspect-ratio domains greatly reduce the experimental or computational cost of reaching high Rayleigh numbers. This enables heat-transfer studies at lower cost, but restricts the large-scale horizontal organization that the domain can contain. Our wide-domain simulations test how these restrictions affect transport and flow statistics.

Superstructures and confinement

Computing project report

The Gauss Centre report Superstructures in Turbulent Thermal Convection (November 2019) describes the large-domain simulation campaign on SuperMUC and lists Detlef Lohse and Richard Stevens as principal investigators.

References

  1. R.J.A.M. Stevens, R. Hartmann, R. Verzicco, D. Lohse,
    How wide must Rayleigh-Bénard cells be to prevent finite aspect ratio effects in turbulent flow?
    J. Fluid Mech. 1000, A58 (2024). Read the finding
  2. D. Krug, D. Lohse, R.J.A.M. Stevens,
    Coherence of temperature and velocity superstructures in turbulent Rayleigh-Bénard flow,
    J. Fluid Mech. 887, A2 (2020). Read the finding
  3. R.J.A.M. Stevens, A. Blass, X. Zhu, R. Verzicco, D. Lohse,
    Turbulent thermal superstructures in Rayleigh-Bénard convection,
    Phys. Rev. Fluids 3, 041501(R) (2018) Read the finding