Numerical pressure retrieval from velocity measurement of a turbulent tornado-like vortex

This paper presents the computational reconstruction of a three-dimensional distribution of time-averaged pressure of a large-scale tornado-like vortex based on measurement data of time-series velocities. The large-scale tornado simulation facility of Texas Tech University, known as VorTECH, was used to acquire the time-series velocities over a cross-section through the apparatus center as well as the time-series pressure on the floor. The three-dimensional pressure was numerically retrieved using a solution of the pressure Poisson equation discretized by finite element approximation while taking into consideration the effect of turbulence on the pressure in terms of the Reynolds stress. The computed pressure distribution near the floor was validated using the measured floor pressure. The effect of turbulence was found to be significant when attempting to accurately compute pressures using turbulent flow data. Furthermore, it was shown that the computer code is capable of assimilating a measured flow field into the divergence-free field by determining the stationary solution of a functional using the Lagrange multiplier. It was found that the effect of assimilation on the retrieved pressure is not significant.