Predicting the airside performance of fin-and-tube heat exchangers with plain and herringbone fin geometries: A simulation-based approach compared to established correlations

Abstract

Fin-and-tube heat exchangers are widely utilized to control air temperature and humidity in heating, ventilation, and air conditioning systems. In this thesis, a simulation-based approach is employed to study air-side performance characteristics of fin-and-tube heat exchangers with plain and herringbone fin geometry. Computational fluid dynamics simulations are performed using the OpenFOAM 12 software package.

The simulation domain is confined to a region between two adjacent fins and two tube rows parallel to the air flow. Simulations of the plain fin geometry are conducted with six circular tubes and one fin, applying the symmetry condition on the midplane between adjacent fins. Simulations of the herringbone fin geometry are performed with two circular tubes and two fins.

A novel pre-processing workflow based on open-source software was developed as part of this thesis. Alternative pre-processing workflows and numerical methods are evaluated and compared. Conjugate heat transfer between air and the fin-and-tube heat exchanger is modeled by solving the steady-state Reynolds-averaged Navier–Stokes equations using the PIMPLE pressure-velocity coupling algorithm and the k-ω SST turbulence model.

The Colburn j-factor and Fanning f-factor are derived from the simulation results and compared against established empirical correlations. The dependence of the j-factor and f-factor on the Reynolds number is analyzed. The open-source simulation workflow developed in this thesis provides a foundation for future optimization studies of fin-and-tube heat exchanger geometries.

Significant discrepancies, reaching up to 60 %, were observed between the established jfactor and f-factor correlations for plain and herringbone fin geometries. Since none of the existing correlations accurately predict the simulated j-factor for the herringbone fin geometry, a novel jfactor correlation has been proposed.

The choice of numerical scheme for discretizing the divergence terms of the velocity and energy fields influences the simulation results. Using the upwind scheme for discretizing the velocity divergence term tends to oversimplify the flow field, smoothing out sharp gradients. Additionally, employing the upwind scheme for discretizing energy terms leads to an overestimation of heat transfer.