Doctor of Philosophy (PhD)


Mechanical Engineering

Document Type



The present effort is organized into five chapters that encompass problems of interest to Mechanical and Chemical Engineering. The problems tackled include applying the science of nonlinear optimization to enhance the design of a packed bed reactor and the shape of a catalyst pellet. Likewise, the present work addresses the flow over a flat plate by perturbation series and asymptotic expansions to correct the similarity solution. Similarly, this thesis deals with the need to increase the accuracy of droplet vaporization models while maintaining the computational cost suitable for multiple droplets (sprays).

The work included in Chapter 1 is motivated by the imperative of improving the safety of chemical reactors and preventing the deactivation posed by temperature spikes. Topology optimization by a gradient-based approach reveals that implementing a conical region with a high void fraction (or filled inert pellets) near the entrance of a packed bed reactor (PBR) can reduce the peak temperature of the PBR and can be used to achieve a higher conversion by safely raising the temperature of the cooling jacket.

In Chapter 2, shape optimization of a catalyst pellet shows that an optimized catalyst can give 20% more conversion than a spherical pellet. Chapter 3 tackled a catalyst with a reaction limited by mass transfer, and it served as the stepping stone for Chapter 2. The work in Chapters 1-3 reveals that the present sophistication of optimization algorithms combined with today's computational capabilities allows for tackling reactor design and catalysis problems.

Chapter 4 deals with the flow over a flat plate near the leading-edge using similarity transformations and perturbation series. The perturbation series approach transforms the Navier-Stokes equations into a sequence of ODEs that extend the applicability beyond x ->0.

In Chapter 5, the modeling of sprays with short droplet lifetimes is advanced by enhancing the Euler-Lagrange approach to account for non-uniform temperature. By dividing the droplet's mass into heated and unheated, the prediction of the surface temperature, the heating rate (dT/dt), and the droplet lifetime are improved without introducing PDEs in the numerical method.



Committee Chair

Wong, Harris



Available for download on Sunday, May 20, 2029