Kelvin-Helmholtz Instability

William H. Cabot, Lawrence Livermore National Lab

Andrew W. Cook, Lawrence Livermore National Lab

Kyle J. Caspersen, Lawrence Livermore National Lab

James N. Glosli, Lawrence Livermore National Lab

Liam D. Krauss, Lawrence Livermore National Lab

Paul L. Miller, Lawrence Livermore National Lab

David F. Richards ( Lawrence Livermore National Lab

Robert E. Rudd, Lawrence Livermore National Lab

Frederick H. Streitz, Lawrence Livermore National Lab

These images show results of a computer simulation of two fluid layers flowing across each other in opposite directions. The swirling waves and vortices form due to the Kelvin-Helmholtz instability, which is named after two of the scientists that first studied them nearly 150 years ago.

Kelvin-Helmholtz Instability

These images are details from a 9 billon atom molecular dynamics simulation of the Kelvin-Helmholtz instability in molten aluminum (blue) and copper (red). The height of the instabilities is about 0.5 microns.

Our simulation tracks the individual motions of over 9 billion atoms of liquid copper (red) and aluminum (purple).  The copper layer flows to the left and aluminum to the right. The resulting wave-like structures are beautifully intricate, decorated with secondary instabilities and complex mixing phenomena.

You may have seen similar waves in cloud formations or in giant storms in the atmospheres of Jupiter or Saturn. These waves in this simulation however are extremely small---less than 1 micron tall or about 1/100th of the thickness of a human hair.

Computing the motions of 9 billion atoms requires a large amount of computer time. In this case, about a week on a supercomputer with over 200,000 CPUs. The same calculation on a single desktop PC would take thousands of years to complete if it would fit in memory.

A video of this research is available at:

This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.

These images have not been previously published. However closely related images, and a description of the corresponding simulation can be found in the paper "Extending stability beyond CPU millennium: a micron scale atomistic simulation of Kelvin-Helmholtz instability" by J. N. Glosli, D. F. Richards, K. J. Caspersen, R. E. Rudd,J. A. Gunnels, and F. H. Streitz. Published in SC '07 Proceedings of the 2007 ACM/IEEE conference on Supercomputing, pages 1-11, 2007. Available at