Date of Award

Spring 2023

Document Type

Open Access Dissertation


Mechanical Engineering

First Advisor

Jamil Khan


This work is focused on dilute and dense phase pneumatic conveying of bulk solids. The motivation of this work is to use industrial data to scale up existing models for pressure drop in pneumatic conveying systems, determine which models best predict pressure drop in full conveying systems, and find the empirical coefficients that go with these models.

Several major research objectives were accomplished by this work. First, system pressure drop methods were evaluated in this research. Other researchers have focused on single pipe section type pressure drops. Second, the data used in this research both for the machine learning evaluations and the development of flow splitter pressure drop methods is from industrial scale systems. Previous researchers have used shorter conveying pipelines with smaller diameters that contain fewer bends. Third, pressure drop methods and gas and solids distribution methods are generated for pneumatic conveying flow splitters. This is a novel area of research which has not previously been investigated.

The first part of this research uses machine learning and industrial data from six pneumatic conveying pipelines to determine which collections of methods best models pressure drop in dilute phase pneumatic conveying systems. The results of this evolutionary modelling allows for generalizations to be made about the best families of methods for different pipe types and identifies common features shared by methods that most successfully model the data.

The second part of this work takes the best method collections identified by the evolutionary models and evaluates them with a larger dataset. This allows for a more comprehensive analysis to be performed on the method collections that were identified in the first part of this research.

The third part of this research analyzes the factors that impact pressure drop in pneumatic conveying systems to develop new methods for total system pressure drop in these systems. A nondimensional analysis of the factors contributing to pressure drop is performed to develop multiple new methods to determine system pressure drop. These new methods are then evaluated and compared to the results from the results of the evolutionary model and the larger dataset evaluations.

The fourth part of this work develops novel methods for pressure drop through pneumatic conveying flow splitters. These new methods are compared with a baseline method collection to determine which of the new methods best predicts pressure drop in pneumatic conveying systems.

Lastly, a final method for each pressure drop type is selected and the industrial scale empirical coefficients which were found as part of this research are presented. This gives a final collection of methods and new coefficients which allow for pressure drop for industrial scale dilute phase pneumatic conveying systems to be modeled.

This newly developed Gorman modelling system for dilute phase pneumatic conveying has extensive industrial applications and will provide for more effective design and optimization of pneumatic conveying systems. This research will be especially useful when used with pneumatic conveying systems that contain splits.