Skip to main content Accessibility help

Inertial gravity current produced by the drainage of a cylindrical reservoir from an outer or inner edge

  • Marius Ungarish (a1), Lailai Zhu (a2) (a3) and Howard A. Stone (a2)


We consider the time-dependent flow of a fluid of density $\unicode[STIX]{x1D70C}_{1}$ in a vertical cylindrical container embedded in a fluid of density $\unicode[STIX]{x1D70C}_{2}~({<}\unicode[STIX]{x1D70C}_{1})$ whose side boundary is suddenly removed and the fluid drains freely from the edge. We show that in the inertial–buoyancy regime (large initial Reynolds number) the flow is modelled by the shallow-water equations and bears similarities to a gravity current released from a lock (the dam-break problem) driven by the reduced gravity $g^{\prime }=(1-\unicode[STIX]{x1D70C}_{2}/\unicode[STIX]{x1D70C}_{1})g$ . This formulation is amenable to an efficient finite-difference solution. Moreover, we demonstrate that similarity solutions exist, and show that the flow created by the dam break approaches the predicted self-similar behaviour when the volume ratio ${\mathcal{V}}(t)/{\mathcal{V}}(0)\approx 1/2$ where $t$ is time elapsed from the dam break. We considered two cases of drainage: (i) outward from the outer boundary in a full-radius reservoir; and (ii) inward from the inner radius in an annular-shaped reservoir. For the first case the similarity solution is expressed analytically, while the second case is more complicated and requires a numerical solution. In both cases ${\mathcal{V}}(t)/{\mathcal{V}}(0)$ decays like $t^{-2}$ , but the details are different. The similarity solutions admit an adjustable virtual-origin constant, which we determine by matching with the finite-difference solution. The analysis is valid for both Boussinesq and non-Boussinesq systems, and a wide range of geometric parameters (inner and outer radii, and height). The importance of the neglected viscous terms increases with time, and eventually the inertial–buoyancy model becomes invalid. An estimate for this occurrence is also provided. The predictions of the model are compared to results of direct numerical simulations; there is good agreement for the position of the interface and for the averaged radial velocity, and excellent agreement for ${\mathcal{V}}(t)/{\mathcal{V}}(0)$ . A box model is used for estimating the effect of a partial (over a sector) dam break. This study is an extension of the work for a rectangular reservoir of Momen et al. (J. Fluid Mech., vol. 827, 2017, pp. 640–663). We demonstrate that there are some similarities, but also significant differences, between the rectangular and the cylindrical reservoirs concerning the velocity, shape of the interface and rate of drainage, which are of interest in applications. The overall conclusion is that this simple model captures very well the flow field under consideration.


Corresponding author

Email address for correspondence:


Hide All
Acton, J. M., Huppert, H. E. & Worster, M. G. 2002 Two-dimensional viscous gravity currents flowing over a deep porous medium. J. Fluid Mech. 440, 359380.
Anderson, D. A., Tannehill, J. C. & Pletcher, R. M. 1984 Computational Fluid Mechanics and Heat Transfer. Hemisphere.
Benjamin, T. B. 1968 Gravity currents and related phenomena. J. Fluid Mech. 31, 209248.
Bonometti, T., Ungarish, M. & Balachandar, S. 2011 A numerical investigation of constant-volume non-Boussinesq gravity currents in deep ambient. J. Fluid Mech. 673, 574602.
Childs, H. et al. 2012 VisIt: an end-user tool for visualizing and analyzing very large data. In High Performance Visualization–Enabling Extreme-Scale Scientific Insight, pp. 357372. Chapman & Hall/CRC.
Hogg, A. J., Baldock, T. E. & Pritchard, D. 2011 Overtopping a truncated planar beach. J. Fluid Mech. 666, 521553.
Martin, J. C. & Moyce, W. J. 1952a An experimental study of the collapse of fluid columns on a rigid horizontal plane, in a medium of lower, but comparable, density. Phil. Trans. R. Soc. Lond. A 244, 325334.
Martin, J. C. & Moyce, W. J. 1952b An experimental study of the collapse of liquid columns on a rigid horizontal plane. Phil. Trans. R. Soc. Lond. A 244, 312324.
Momen, M., Zheng, Z., Bou-Zeid, E. & Stone, H. A. 2017 Inertial gravity currents produced by fluid drainage from an edge. J. Fluid Mech. 827, 640663.
Monismith, S. G., Mcdonald, N. R. & Imberger, J. 1993 Axisymmetric selective withdrawal in a rotating stratified fluid. J. Fluid Mech. 249, 287305.
Morton, K. W. & Mayers, D. F. 1994 Numerical Solutions of Partial Differential Equations. Cambridge University Press.
Ray, W. J. & Raba, G. W. 1991 Behavior of welded steel water-storage tank. J. Struct. Engng 117, 6179.
Roenby, J., Bredmose, H. & Jasak, H. 2016 A computational method for sharp interface advection. R. Soc. Open Sci. 3 (11), 160405.
Simpson, J. E. 1997 Gravity Currents in the Environment and the Laboratory. Cambridge University Press.
Ungarish, M. 2009 An Introduction to Gravity Currents and Intrusions. Chapman & Hall/CRC Press.
Ungarish, M. 2010 The propagation of high-Reynolds-number non-Boussinesq gravity currents in axisymmetric geometry. J. Fluid Mech. 643, 267277.
Ungarish, M. & Huppert, H. E. 2000 High Reynolds number gravity currents over a porous boundary: shallow-water solutions and box-model approximations. J. Fluid Mech. 418, 123.
Weller, H. G., Tabor, G., Jasak, H. & Fureby, C. 1998 A tensorial approach to computational continuum mechanics using object-oriented techniques. Comput. Phys. 12 (6), 620631.
Zgheib, N., Bonometti, T. & Balachandar, S. 2015 Dynamics of non-circular finite-release gravity currents. J. Fluid Mech. 783, 344378.
MathJax is a JavaScript display engine for mathematics. For more information see

JFM classification

Related content

Powered by UNSILO

Inertial gravity current produced by the drainage of a cylindrical reservoir from an outer or inner edge

  • Marius Ungarish (a1), Lailai Zhu (a2) (a3) and Howard A. Stone (a2)


Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed.