arXiv ScienceSearch

arXiv subjects

D Duque

Publications and source records attributed to D Duque.

2 recordsLinked to original sources

A geometric formulation of the Shepard renormalization factor

The correct treatment of boundary conditions is a key step in the development of the SPH method. The SPH community has to face several challenges in this regard - in particular, a primordial aspect for any boundary formulation is to ensure the consistency of the operators in presence of boundaries and free surfaces. A new implementation is proposed, based on the existing numerical boundary integrals formulation. A new kernel expression is developed to compute the Shepard renormalization factor at the boundary purely as a function of the geometry. In order to evaluate this factor, the resulting expression is split into numerical and analytical parts, which allows accurately computing the Shepard factor. The new expression is satisfactorily tested for different planar geometries, showing that problems featuring free surfaces and boundaries are solved. The methodology is also extended to 3-D geometries without great increase in computational cost.

physics.flu-dyn

Exact results and mean field approximation for a model of molecular aggregation

We present a simple one-dimensional model with molecular interactions favouring the formation of clusters with a defined optimal size. Increasing the density, at low temperature, the system goes from a nearly-ideal gas of independent molecules to a system with most of the molecules in optimal clusters, in a way that resembles the formation of micelles in a dilution of amphiphilic molecules, at the critical micellar concentration. Our model is simple enough to have an exact solution, but it contains some basic features of more realistic descriptions of amphiphilic systems: molecular excluded volume and molecular attractions which are saturated at the optimal cluster. The comparison between the exact results and the mean field density functional approximation suggests new approaches to study the more complex and realistic models of micelle formation; in particular it addresses the long-standing controversy surrounding separation of internal degrees of freedom in the formulation of cluster association phenomena.

cond-mat.soft