Degree Name

Doctor of Philosophy


Department of Mechanical Engineering


An accurate estimation of the total pipeline air pressure drop is considered widely to be one of the most important aspects of pneumatic conveying system design. Since there are numerous influential parameters, such as velocity, particle properties, pipe length, diameter and material, the determination of the total pipeline air pressure drop in pipelines of different length, diameter, step and bend number is based largely on the scale-up of test rig data.

In many systems, the air-only component of pressure drop is significant (e.g. dilutephase, long-distance conveying). This thesis modifies initially the correlations for predicting pressure drop caused by an incompressible fluid flowing through bends and straight sections of pipe. These correlations are sufficientiy accurate as long as mean conditions (based on average air density) for each straight pipe and conditions at the outlet of each bend are used in the analysis.

Based on detailed mathematical and dimensional analyses, semi-empirical correlations then are set up for predicting the pressure drop due to the presence of solids in bends and straight pipes. Fly ash is used as the test material. B y using the data from any constant diameter pipeline with one bend and two long straight pipes, the exponents in the semi-empirical correlations are determined. These correlations then are used to predict the total pipeline air pressure drop in pipelines of different length, diameter and bend number. A close agreement between predicted and experimental results is obtained as long as there are no short straight pipes in the pipeline.

However, in industry, short straight pipes are in c o m m o n use. It is found that the previously 'determined' values of exponents underpredict significantly the total pipeline air pressure drop in pipelines that contain short straight pipes. Therefore, the semi-empirical correlation for bend pressure drop is modified by determining the exponents from data obtained on a simple configuration of constant diameter pipeline, which comprises two bends and three sections of pipe, two long and one short. The operating conditions of several pipeline having different lengths and diameters are predicted accurately and this demonstrates the good reliability and stability of this n e w scale-up procedure.

Finally, the results from two other test materials with significantly different particle properties (viz. pulverised coal and plastic pellets) demonstrate further the high accuracy and reliability of the scale-up procedure developed in this thesis.