Abstract

The performance of large mechanical draft air-cooled heat exchangers is directly related to fan performance which is influenced by atmospheric wind conditions, as well as the plant layout. It is often necessary to numerically model the entire system, including fans, under a variety of operating conditions. Full three-dimensional, numerical models of axial flow fans are computationally expensive to solve. Simplified models that accurately predict fan performance at a lesser expense are therefore required. One such simplified model is the actuator disk model (ADM). This model approximates the fan as a disk where the forces generated by the blades are calculated and translated into momentum sources. This model has been proven to give good results near and above the design flow rate of a fan, but not at low flow rates. In order to address this problem two modifications were proposed, namely the extended actuator disk model (EADM) and the reverse engineered empirical actuator disk model (REEADM). The three models are presented and evaluated in this paper using ANSYS FLUENT. The models are simulated at different flow rates representing an axial flow fan test facility. The resulting performance results and velocity fields are compared to each other and to previously simulated three dimensional numerical results, indicating the accuracy of each method. The results show that the REEADM gives the best correlation with experimental performance results at design conditions (ϕ = 0.168) while the EADM gives the best correlation at low flow rates. A comparison of the velocity profiles shows that none of the three models predict the radial velocity distribution at low flow rates correctly, however the correlation improves at flow rates above ϕ = 0.105. In general the upstream velocity profiles, where reversed flow occurs through the fan, are poorly predicted at low flow rates. At the flow rates above ϕ = 0.137 the correlation between the velocity profiles for the simplified modes and the three dimensional results is good.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call