Use this URL to cite or link to this record in EThOS:
Title: Improved prediction methods for finned tube bundle heat exchangers in crossflow
Author: McIlwain, Stuart Russell
ISNI:       0000 0001 3624 7695
Awarding Body: University of Strathclyde
Current Institution: University of Strathclyde
Date of Award: 2003
Availability of Full Text:
Access from EThOS:
Access from Institution:
Previous methods for predicting the heat transfer and pressure drop performance of crossflow heat exchangers with finned tubes have concentrated on developing correlations. These correlations have been based on the researchers observations of what geometric parameters may affect the performance, and then dimensionless groups developed to allow a correlation to be developed. This work shows that many of these models are limited either by design or by their databases, and often are not general enough to cater for air-cooled heat exchangers as well as the generally larger scale heat recovery bundles. The most recent prediction methods have been developed as more aerodynamically based models, although these still encompass an element of empiricism to account for effects that are not readily understood. This new work develops from these physically based models. An improved method for the prediction of the pressure drop of staggered finned tube bundles is presented, based on high quality test data and the results of a CFD (Computational Fluid Dynamics) study. This is shown to perform better than previous models, and also correct a defect in the formulation of a previous method. A new prediction scheme for inline finned tube bundles is also presented. Experimental work was performed on nine inline air coolers to determine their performance characteristics and, along with open literature data, develop a reliable databank for prediction method development. The models incorporate a new approach to the pressure drop prediction using a sophisticated gap flow model, and a multiple term heat transfer model, that considers heat transfer and flow mixing between the main flow streams. This method is shown to significantly improve on previous methods. Experiments were conducted on an isothermal staggered air-cooler bundle that allowed differing wall sealing devices (corbels) to be used, or allow a bypassing lane. Flow visualization tests were performed on this bundle, and observations of the flow patterns compared with a simple two-dimensional CFD model. From the test results a new method of predicting the pressure drop performance of staggered bundles with various corbels was developed. Using the bypassing air-cooler data and new data taken from a heat recovery bundle an iterative method to predict the pressure drop when a bypassing lane is present is presented. This method is shown to be both simple and computationally cheap, and is used in conjunction with the new staggered bundle pressure drop method. The experimental inline air cooler results were used in conjunction with CFD to provide data to investigate the effect on heat transfer with an increasing number of rows through the bundle. It was found that the key factors in determining this are turbulence and the temperature difference between the tubeside and crossflow fluids, and also that the fin frequency plays a key role. A model is presented to predict the local heat transfer coefficient, which uses sub-models to express the two contributory factors. The results of this approach are shown to be very good, and promote better understanding of tube row heat transfer duty than previously developed models.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral