Document Type


Date of Award


Degree Name

Doctor of Philosophy in Chemical Engineering - (Ph.D.)


Chemical and Materials Engineering

First Advisor

Ecevit Atalay Bilgili

Second Advisor

Rajesh N. Dave

Third Advisor

Lisa Axe

Fourth Advisor

Roman S. Voronov

Fifth Advisor

R. Bertrum Diemer


Cement milling is known to be inefficient and energy-intensive. Thus, even small improvements in cement milling's performance could significantly reduce operation costs. This dissertation aims to develop a simulation tool for dry milling and generate a fundamental process understanding, which enables process optimization. To this end, a true unsteady-state simulator (TUSSIM) for continuous dry milling is developed and applied to model various processes: (a) open circuit continuous mills, (b) closed-circuit continuous mills, and (c) vertical roller mills. TUSSIM is based on the solution of the cell-based population balance model (PBM) for continuous milling, which consists of a set of differential algebraic equations (DAEs). Moreover, air classifier parameters and ball size distribution for the closed-circuit operation are tailored to maximize production capacity while achieving desirable cement product qualities. Discrete element method (DEM) and PBM are coupled to simulate lab-scale batch milling of cement clinker to gain fundamental understanding of the roles of ball size and material (steel vs. alumina).

First, dynamic simulations are performed to investigate the impact of ball mill operation parameters on the full-scale open-circuit ball milling of cement clinker without an external air classifier. Parameters for the simulation are taken from the literature. Simulation results suggest that a single-compartment mill produces the desired cement size, but it requires pre-milled fresh feed. Depending on the ball sizes used, a two compartment mill produces cement sizes similar to those produced by a three-compartment mill. A uniform mass of balls achieves an 8% higher specific surface area (SSA) compared to a uniform number of balls. The classifying liners have negligibly finer cement products compared to a uniform mass distribution.

TUSSIM is also incorporated with a variable Tromp curve model for classification to simulate full-scale closed-circuit ball milling with an air classifier. The simulation results suggest that a faster rotor speed or lower air flow rate leads to a finer cement product and increases the dust load of the classifier feed. Integrating air classifiers into open-circuit ball milling increases the production rate by 15% or cement SSA by 13%. Operation failure due to overloading of the entire circuit is detected when dust load is too high. Process optimization with a global optimizer?DAE solver is performed to identify either the air classifier's parameters or the ball size distributions that yield desirable cement quality while maximizing production rate. Optimization results show that the production rate can be increased by 7% compared to the baseline process. Unlike open circuits, a two-compartment mill produces a finer cement product than a three-compartment mill. Optimal ball mixtures are identified in a two-compartment mill, suggesting a 14% increase in production rate at a desirable cement quality.

A global optimizer-based back-calculation method, based on PBM, is used to determine the breakage kinetics parameters of cement clinker in a lab-scale ball mill loaded with steel or alumina balls of three single ball sizes and their mixtures. The motion of the balls in the mill is simulated via the DEM. The results show that steel balls achieve faster breakage of clinker than alumina balls, which is explained by the higher total?mean energy dissipation rates of the steel balls. The particle size distribution (PSD) becomes finer as smaller balls are used. The ball mixture is the most effective overall. Significant energy can be saved if steel balls are replaced with alumina balls, but the slower breakage with the alumina balls needs to be accounted for.

Finally, steady-state cement PSD obtained from a full-scale vertical roller mill is fitted with TUSSIM. The fitted results show good agreement compared to the experimental PSD. Overall, this dissertation has provided a novel process simulator, TUSSIM, and many fundamental insights into the continuous milling of cement clinker and its optimization.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.