Reduced FV modelling based on CFD database and experimental validation for the thermo-fluid dynamic simulation of flue gases in horizontal fire-tubes

The present work aims at proposing an optimized physical heat transfer correlation in smooth horizontal fire-tubes (applying a one-dimensional reduced Finite Volume Method - FVM), validated over a wide range of operating conditions. An experimental activity is first carried out, in which the temperature and pressure profiles of the flue gases, streaming through pipes while applying hot-water-boiler conditions, are measured. A Computational Fluid Dynamics (CFD) set-up is developed for simulating the thermo-physical behaviour of the flue gases undergoing a wide range of operating conditions (replicating the conditions that these streams experience in the most common industrial boiler applications). The acquired experimental data with a detailed uncertainty analysis are employed for validating the developed CFD results. Subsequently, the reduced one-dimensional FVM physical models for estimating the heat transfer (taking into account the convective and radiative terms) and pressure drops are implemented. The most accurate models are calibrated using the results of the fluid-dynamics simulations. In this procedure, the coefficients of the standard models are optimized, employing a Multi Objective Genetic Algorithm optimization procedure (MOGA). This work demonstrates that the CFD simulations can reproduce the experimental data accurately. Furthermore, the estimations of the FVM model based on the calibrated physical correlations are in good agreement with numerical results, but with a reduced computational cost. Results indicate an improved estimation of tubes outlet temperature with respect to standard correlation (42.4% error reduction), while the estimated outlet pressure shows only a slight increase of the error with respect to the Fang correlation.