TECHNICAL PAPERS: Natural and Mixed Convection

Marangoni Instability in a Finite Container-Transition Between Short and Long Wavelengths Modes

[+] Author and Article Information
L. Czechowski, J. M. Floryan

Department of Mechanical and Materials Engineering, The University of Western Ontario, London, Ontario, N6A 5B7, Canada

J. Heat Transfer 123(1), 96-104 (Sep 27, 2000) (9 pages) doi:10.1115/1.1339005 History: Received October 25, 1999; Revised September 27, 2000
Copyright © 2001 by ASME
Your Session has timed out. Please sign back in to continue.


Bénard,  H., 1900, “Les Tourbillons Cellulaires dans une Nappe Liquide,” Rev. Gen. Sci. Pures Appl., 11, pp. 1261–1271; 11, pp. 1309–1328.
Block,  M. J., 1956, “Surface Tension as the Cause of Benard Cells and Surface Deformation in a Liquid Film,” Nature (London), 178, pp. 650–651.
Pearson,  J. R. A., 1958, “On Convection Cells Induced by Surface Tension,” J. Fluid Mech., 4, pp. 489–500.
Scriven,  L. E., and Sterling,  C. V., 1964, “On Cellular Convection Induced by Surface Tension Gradients: Effect of Mean Surface Tension and Viscosity,” J. Fluid Mech., 19, pp. 321–340.
Smith,  M. K., 1966, “On Convective Instability Induced by Surface Tension Gradients,” J. Fluid Mech., 24, pp. 401–414.
Davis,  S. H., and Homsy,  G. M., 1980, “Energy Stability Theory for Free Surface Problems: Buoyancy-Thermocapillary Layers,” J. Fluid Mech., 98, pp. 527–553.
Davis, S. H., 1983, “Rupture of Thin Liquid Films,” in Waves on Fluid Interfaces, R. E. Meyer, ed., Academic Press, New York, pp. 291–302.
Davis,  S. H., 1987, “Thermocapillary Instabilities,” Annu. Rev. Fluid Mech., 19, pp. 403–435.
Goussis,  D. A., and Kelly,  R. E., 1990, “On the Thermocapillary Instabilities in a Liquid Layer Heated from Below,” Int. J. Heat Mass Transf., 33, pp. 2239–2245.
Krishnammoorthy,  S., Ramaswamy,  B., and Joo,  S. W., 1995, “Spontaneous Rupture of Thin Liquid Films due to Thermocapillary: A Full Scale Numerical Simulation,” Phys. Fluids, 7, pp. 2291–2293.
VanHook,  S. J., Schatz,  M. F., Swift,  J. B., McCormick,  W. D., and Swinney,  H. L., 1997, “Long-Wavelength Surface-Tension-Driven Benard Convection: Experiment and Theory,” J. Fluid Mech., 345, pp. 45–78.
Winters,  K. H., Plesser,  Th., and Cliffe,  K. A., 1988, “The Onset of Convection in a Container due to Surface Tension and Buoyancy,” Physica D, 29, pp. 387–401.
Dijkstra,  H. A., 1992, “On the Structure of Cellular Solution in Rayleigh-Benard-Marangoni Flows in Small-Aspect-Ration Containers,” J. Fluid Mech., 243, pp. 73–102.
Dijkstra,  H. A., 1995, “Surface Tension Driven Cellular Patterns in Three-Dimensional Boxes—Linear Stability,” Microgravity Sci. Technol., 7, pp. 307–312.
Dijkstra,  H. A., 1995, “Surface Tension Driven Cellular Patterns in Three-Dimensional Boxes—A Bifurcation Study,” Microgravity Sci. Technol., 7, pp. 307–312.
Dauby,  P. D., and Lebon,  G., 1996, “Marangoni-Benard Instability in Rigid Rectangular Containers,” J. Fluid Mech., 329, pp. 25–64.
Perez-Garcia,  C., Echebarria,  B., and Bestehorn,  M., 1998, “Thermal Properties in Surface-Tension-Driven Convection,” Phys. Rev. E, 57, pp. 475–481.
Hamed,  M., and Floryan,  J. M., 2000, “Marangoni Convection: Part 1—Cavity with Differentially Heated Side Walls,” J. Fluid Mech., 405, pp. 79–110.
Hamed,  M., and Floryan,  J. M., 1998, “Numerical Simulation of Unsteady Nonisothermal Capillary Interfaces,” J. Comput. Phys., 145, pp. 110–140.
Chen,  C., and Floryan,  J. M., 1994, “Numerical Simulation of Non-isothermal Capillary Interfaces,” J. Comput. Phys., 111, pp. 183–193.
Koschmieder,  E. L., and Prahl,  S. A., 1990, “Surface-Tension-Driven Benard Convection in Small Containers,” J. Fluid Mech., 215, pp. 571–583.
Floryan,  J. M., and Chen,  C., 1994, “Thermocapillary Convection and Existence of Continuous Layer in the Absence of Gravity,” J. Fluid Mech., 277, pp. 303–329.


Grahic Jump Location
Sketch of the flow configuration and the coordinate system used in the present analysis
Grahic Jump Location
Evolution of perturbations for a subcritical (Ma=100 — lower line) and supercritical (Ma=300 — upper line) Marangoni numbers for Ca=1.5, Bi=2, Pr=1 and L=3. The critical value of the Marangoni number is Macr=210. The maximum value of the vorticity |ω|max is used as a measure of the magnitude of the perturbation. The classical, short wavelength disturbance pattern dominates, leading to the formation of a saturated supercritical steady convection.
Grahic Jump Location
The critical Marangoni number Macr as a function of the cavity length L for different values of the Capillary number Ca and for Bi=2. The half-wavelengths of disturbances leading to the unconditional instability, according to the linear stability analysis of an infinite layer (8), are marked on the horizontal axis. Transition of flow patterns from one cell to two cells, and then to three, four, five, etc., cells are marked by points A and B, C and D, E and F, G and H, I and J, respectively, for Ca=0, and by points a and b, c and d, e and f, g and h, respectively, for Ca=1. The four and five cells S-pattern and the 3 cells L-pattern for Ca=1 are shown in Fig. 4. The thick lines indicate critical conditions for the L-pattern of response for Ca=1.
Grahic Jump Location
Flow patterns associated with the most unstable disturbances in the case of a deformable interface for Ca=1, Ma=Macr, Bi=2. The interface deformation is ignored as too small for plotting purposes. The S-pattern of response is illustrated for cavity lengths L=6.9 and 7.0 (see Figs. 4(gh), while the L-pattern is illustrated for cavity length L=7.6 (see Fig. 4(i)). The corresponding flow conditions are marked by points g,h, and i, in Fig. 3, respectively. The forms of the interface for L=7.0 and 7.6 are shown in Fig. 5. Flow normalization condition used in all cases — |ψ|max=1. The system evolution corresponds to phase 2 in Fig. 2.
Grahic Jump Location
Deformations of the interface associated with the most unstable disturbances for Ma=Macr, Ca=1, Bi=2. The S and L-patterns are illustrated for cavity lengths L=7 and 7.6, respectively. The corresponding flow fields are shown in Fig. 4. The magnitude of the deformation shown resulted from the use of the same normalization condition |ψ|max=1 in all cases. Note that the L-pattern gives rise to the deformation that is by an order of magnitude bigger than the deformation associated with the S-pattern. The system evolution corresponds to phase 2 in Fig. 2.
Grahic Jump Location
The critical length Lo for the unconditional instability as a function of the Capillary number Ca and the Biot number Bi
Grahic Jump Location
Comparison of the critical Marangoni numbers Macr for a finite container and for an infinite layer with a deformable interface for Bi=2 and for the ratio Ca/Ma=0.005 kept constant. In the case of a finite layer, L denotes the length of the container, while in the case of an infinite layer L denotes half of the wavelength of the most unstable disturbance. See text for a discussion.
Grahic Jump Location
Comparison of the stream function and temperature disturbance fields for the most unstable (linear) disturbances for Ca=1, Bi=2 for the S and L-patterns. Figures 8A, B display stream function and temperature, respectively, for L=7 (S-pattern) and Ma=Macr=215. Figures 8C, D display stream function and temperature, respectively, for L=17 (L-pattern), Ma=10 and Pr=1. The critical Marangoni number for these conditions is Macr=0. Data for presentation were taken from the exponential growth period described by the linear theory. Normalization condition used in all cases — |ψ|max=1.
Grahic Jump Location
Surface pressure distribution for the L-pattern response for Ca=1.0, Ma=150, Bi=2, Pr=1 and L=16 at t=200, 400, 593. Calculations had to be terminated at t=594. Normalization condition used — |ψmax|instantaneous=1. Slight asymmetry in the pressure distribution is due to the growth of asymmetric disturbances present in the initial conditions (see text for a discussion).
Grahic Jump Location
Maxima of the interface deformation |h−1|max, vorticity |ω|max and stream function |ψ|max as a function of time for the L-pattern response for the same conditions as in Fig. 9. A characteristic acceleration of growth of perturbations as a result of nonlinear effects can be easily observed. This behavior is qualitatively different from the one observed in the case of the S-pattern, where a nonlinear saturation and a steady final state are observed (see Fig. 2).




Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In