Study on Configuration and Position of Heating Elements in Ductwork to Enhance Forced Convective Heat Transfer in HVAC System

Abstract

This study investigates the influence of the configuration and position of strip heaters in the heating chamber of a mechanical ventilation system on forced convective heat transfer occurring in the ductwork. Originated from past commissioning experience, the measured discharge air temperature at the outlet of ductwork was found not able to meet the designed value after several attempts were made, e.g. adjustment of response time of digital controllers of the strip heaters, checking on the air leakage at equipment and air ducts, other than a non-efficient method of increasing the power output of strip heaters. A full scale experiment comprised two axial supply air fans connected in series, silicon controlled rectifier, insulated air ducts and ten strip heaters housed in a heating chamber was set up for measurement of temperature and velocity distribution along the ductwork in a factory. The fans drew in the ambient air, pre-heated and maintain a constant temperature of 30 °C, and positively pressurised the heating chamber before the air was discharged through the air duct to the surroundings. The temperatures and velocities along the vertical and horizontal centre lines, namely, the y- and z-axes, at four particular Section A-A to D-D, and the total power consumption of the strip heaters were measured by hand held instrument. Moreover, the initial cost in setting up this experiment and the running cost was high. The period of measurement was limited to Spring time in order to reduce the power consumption. Pre-heating section was applied to maintain the intake temperature during the variation in weather condition, and the ductwork was well insulated to obviate conductive heat transfer from the air stream to the surrounding. Concurrently, the computational fluid dynamics (CFD) approach was also adopted to simulate the results of changes in the temperature, velocity profiles and total pressure across the bundle of strip heaters. This simulation was successfully validated by the experimental data. In this study, i) the temperature profile, ii) velocity profile and iii) pressure distribution obtained from simulation were analysed to explore the influence of the six cases of the bundle of strip heaters in the heating chamber on these parameters, named Cases A to F.A widely used commercial code, ANSYS Fluent v13, was adopted to resolve these cases. The approach was Reynolds-averaged Navier-Stokes (RANS) and the turbulence model applied was standard k-ε. Moreover, standard wall function was applied to near-wall regions. Many literatures were reviewed and reliability, upside and downside of this model were researched before the appropriate turbulence model was selected. The selected model was believed to achieve a compromise amongst the other models. With reference to the Newton’s Law of Cooling and the first law of thermodynamics, the convective heat transfer coefficient and dimensionless Reynolds number and Nusselt number were estimated from the measurements and simulated results. In addition, the Log Mean Temperature Difference (LMTD) method was adopted to calculate the Nusselt number at the aforementioned sections for verification of and comparison with that obtained via the theory of Newton’s law of cooling. Under the condition of fully turbulent and not fully developed flow in the air duct, the profile of the temperature increment was observed to differ in the six cases examined, dubbed Cases A to F, and the velocity profiles obtained also differed in both the CFD simulation and measurements. In addition, the actual total current input to the strip heaters were measured to demonstrate the relative amounts of power consumption. The velocity values close to the inner surface of the ductwork were relatively low, whilst the temperature values at the core were higher. Nusselt number was found not sensitive to the variance in Reynolds number even though an increase in Reynolds number induced a slight decrease in Nusselt number in the horizontal and vertical axes in the streamline diagram of the round air duct. No wakes or re-circulation of air flow was observed after the bundle of strip heaters due to a large aspect ratio of the length of the pipe to the diameter of the air duct. Adverse and opposite changes in velocity vectors were observed in Case C. The highest Nu value of Case B at Section D-D amongst other cases denoted most effective convective heat transfer, also demonstrating a highest temperature increment, with reference to the inlet air, but not at the lowest power consumption amongst other cases. An initial relatively high Nu values at Section A-A does not imply the highest temperature increment, because the temperature distribution also respond to the changes in velocity profiles along the pipe. The configuration of strip heaters in Case D was the most ideal case in eliminating the fluctuation in velocity vectors, so that losses in turbulence kinetic energy on shear stress when passing over the surface of the strip heaters were minimized. Therefore a competitive temperature increment other than Case B was observed. In conclusion, electrical power consumption, total pressure drop and velocity profile are the key factors that determine the enhancement of convective heat transfer in ductwork.

Date of Award	15 Jan 2016
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	W. Z. LU (Supervisor)