Correlation between numerical simulation and limited data experimental technique for estimation of nitrogen flowing in LMMHD loop

• Improved modeling of mercury–nitrogen flow by advanced version of Fluent 14.
• A strategy is presented for obtaining a correlation relation between Fluent and computerized tomography method.
• Improved version of modeling is applied considering the factors such as: (a) calibration of CT algorithm using a-priori information; (b) same discretization grid is used for Fluent and CT; (c) fine tuning parameters for Fluent are tested.

Predictions by numerical simulation with estimations by Computerized Tomographic (CT) results for two component two phase flow system are compared. Motivation is to improve/deduce a strategy for a possible correlation between simulation and experimental results by employing appropriate mathematical and physical modeling. Three-dimensional simulation is performed using ANSYS Fluent 14.0®. Flow regimes considered in this work vary from churn-turbulent to bubbly flow. Changes in bubble diameter with pressure are considered. Grid independent adaptively-regularized entropy maximization CT algorithm is employed. It facilitates reconstruction with high grid resolution. A-priori information is used for calibration of this CT algorithm with real life measurements. Beer–Lambert law is used to interpret the measured gamma ray counts instead of preferred conventional method of chord average projection data. Mixture model shows negligible mismatch between two approaches for low (less than 10%) nitrogen holdup and Euler–Euler model beyond this range. Correlation relation is obtained and used for matching profile of 40 LPM with real time measurements.

Figure optionsDownload as PowerPoint slide