THE EFFECTS OF FLOW MALDISTRIBUTION ON THE THERMAL PERFORMANCE DEGRADATION OF FIN-TUBE HEAT EXCHANGERS

Maldistribution of flow in a heat exchanger has an adverse effect on its thermal
and hydraulic performance. Not only does the heat duty reduce but the fluid pressure
drop across the exchanger increases too. The characteristics of a maldistribution
profile are described by its four statistical moments of probability density function,
viz. mean, standard deviation, skew and kurtosis. A novel mathematical analysis
technique has been developed to demonstrate the influence of these statistical
moments on the heat transfer and pressure drop performance of an exchanger.
The analysis has shown that both the mean and standard deviation have the
highest degradation effect on the heat exchanger performance while subsequent
higher moments have declining effects until the fourth moment kurtosis, which has no
significant effect. A discretized numerical method was then used on the fin-tube heat
exchanger coil to calculate the magnitudes of thermal degradation as the statistical
moments of the air inlet velocity distribution and geometrical parameters of the
exchanger are systematically changed. The results show that the degradation is not
only dependent on the moments but also on the exchanger NTU, ratio of external to
internal heat transfer coefficients, R, and the number of tube rows in the coil.
Consequently, new correlation equations have been developed to predict the
magnitude of deterioration from a known air velocity maldistribution profile, for a
given heat exchanger geometry.
An experimental test rig was fabricated to validate the correlation equations. The
same experimental data were then used to validate the Computational Fluid Dynamics
(CFD) model of the fin-tube heat exchanger. With the same modelling technique,
simulations with various exchanger geometry and layout designs can be performed to
extract the statistical moments of the maldistribution and predict the heat exchanger
performance. By doing so, the design could be optimized to find the lowest possible
degradation effects. Maldistribution with low standard deviation and high positive
skew is seen to give low thermal performance deteriorations.