A high quality interpolation method for colocated polyhedral/polygonal control volume methods

Abstract

A high quality and efficient interpolation method for polyhedral/polygonal control volume simulation data is presented. The proposed method utilizes a non-ambiguous and efficient mesh decomposition technique. A pseudo-Laplacian is used to solve an optimization problem to approximate the variation between discrete data points in a linear fashion. The interpolation method guarantees continuous interpolation data throughout the control volume mesh topology and faithfully reproduces the input control volume data. The interpolation connectivity is structured to mimic the interpolation methods utilized by the control volume discretization. The method only requires the geometry of the input data to perform interpolations. This allows key interpolation data to be calculated once and stored for efficient interpolations. The benefits of the proposed algorithm are highlighted by an interpolation test case which demonstrates the benefits of the current method compared to a popular interpolation method currently used in industry. Since the proposed method is designed to augment an existing mesh data structure it can be used to update existing control volume software.