Two-dimensional simulations of the 11 January 1972 Boulder, Colorado, windstorm, obtained from 11 diverse nonhydrostatic models, are intercompared with special emphasis on the turbulent breakdown of topographically forced gravity waves, as part of the preparation for the Mesoscale Alpine Programme field phase. The sounding used to initialize the models is more representative of the actual lower stratosphere than those applied in previous simulations. Upper-level breaking is predicted by all models in comparable horizontal locations and vertical layers, which suggests that gravity wave breaking may be quite predictable in some circumstances. Characteristics of the breaking include the following: pronounced turbulence in the 13-16-km and 18-20-km layers positioned beneath a critical level near 21-km, a well-defined upstream tilt with height, and enhancement of upper-level breaking superpositioned above the low-level hydraulic jump. Sensitivity experiments indicate that the structure of the wave breaking was impacted by the numerical dissipation, numerical representation of the horizontal advection, and lateral boundary conditions. Small vertical wavelength variations in the shear and stability above 10 km contributed to significant changes in the structures associated with wave breaking. Simulation of this case is ideal for testing and evaluation of mesoscale numerical models and numerical algorithms because of the complex wave-breaking response.

We study the onset of sidebranching of growth cells in directional solidification of impure succinonitrile. Care is taken to obtain uniform cell spacing over large distances in order to use this variable as a true control parameter. Two sidebranching modes referring to different crystalline orientations are observed and their physical equivalence is shown. The onset of sidebranching is identified according to an order parameter and scanned with respect to pulling velocity, thermal gradient, and cell spacing. Its evolution with the control parameters surprisingly reveals that increasing thermal gradient at otherwise fixed velocity and cell spacing enhances sidebranching. These results show the need for improving the experimental characterization and the theoretical description of sidebranching in directional solidification.

An oscillatory instability is experimentally evidenced in directional solidification of nominally pure succinonitrile in a unity Péclet number regime. It involves long-time oscillations of both cell width and cell tip position in a nearly phase quadrature and at a wavelength equal to twice the cell spacing. The instability displays subcritical features and two instability domains, one at large cell spacing and the other at small cell spacing. Close to its upper stability boundaries, it succeeds in saturating onto a limit cycle; farther inside the unstable domains, it mediates a transition from cells to a different branch of solution, the doublets. [S0031-9007(97)

In directional solidification, as the solidification velocity increases, the growth direction of cells or dendrites rotates from the direction of the thermal gradient to that of a preferred cristalline orientation. Meanwhile, their morphology varies with important implications for microsegregation. Here, we experimentally document the growth directions of these microstructures in a succinonitrile alloy in the whole accessible range of directions, velocities, and spacings. For this, we use a thin sample made of a single crystal on which the direction of the thermal gradient can be changed. This allows a fine monitoring of the misorientation angle between thermal gradient and preferred crystalline orientation. Data analysis shows evidence of an internal symmetry which traces back to a scale invariance of growth directions with respect to a Péclet number. This enables the identification of the relationship between growth directions and relevant variables, in fair agreement with experiment. Noticeable variations of growth directions with misorientation angles are evidenced and linked to a single parameter.