Sumith YD | Search arXiv

arXiv2015

Fast Geometric Fit Algorithm for Sphere Using Exact Solution

Sphere fitting is a common problem in almost all science and engineering disciplines. Most of methods available are iterative in behavior. This involves fitting of the parameters in a least square sense or in a geometric sense. Here we extend the methods of Thomas Chan and Landau who fitted the 2D data using circle. This work closely resemble their work in redefining the error estimate and solving the sphere fitting problem exactly. The solutions for center and radius of the sphere can be found exactly and the equations can be hard coded for high performance. We have also shown some comparison with other popular methods and how this method behaves.

cs.CV↗

arXiv2015

A new algorithm for contact angle estimation in molecular dynamics simulations

It is important to study contact angle of a liquid on a solid surface to understand its wetting properties, capillarity and surface interaction energy. While performing transient molecular dynamics (MD) simulations it requires calculating the time evolution of contact angle. This is a tedious effort to do manually or with image processing algorithms. In this work we propose a new algorithm to estimate contact angle from MD simulations directly and in a computationally efficient way. This algorithm segregates the droplet molecules from the vapor molecules using Mahalanobis distance (MND) technique. Then the density is smeared onto a 2D grid using 4th order B-spline interpolation function. The vapor liquid interface data is estimated from the grid using density filtering. With the interface data a circle is fitted using Landau method. The equation of this circle is solved for obtaining the contact angle. This procedure is repeated by rotating the droplet about the vertical axis. We have applied this algorithm to a number of studies (different potentials and thermostat methods) which involves the MD simulation of water.

cond-mat.soft↗