My PhD work involved simulation and design of a novel membrane reactor coupling dehydrogenation of ethylbenzene with hydrogenation of nitrobenzene. Membrane reactors are generally catalytic reactors in which in-situ separation and heat transfer take place. Hydrogen and oxygen are most commonly removed components depending on the type of the permeation membranes installed. Permeation of hydrogen from the reacting chamber assists greatly in shifting the equilibrium conversions and yields of thermodynamically limited reactions like dehydrogenation of ethylebenzene to styrene. Modelling and simulation this type of reactors are very crucial to proof the concept, evaluate the effect of operating and design variables then design an experimental set-up.
Optimizing the performance of chemical reactors leads to
designing efficient reactors in which some measures of performance are
maximized/minimized while not violating certain physical or chemical
constraints.
Optimization can be looked at from many prospectives. One way is to optimize
catalytic chemical reactors so that certain performance measures are
maximized within certain constraints considering operating and design
parameters as optimization variables. Normally, the solution set is a
single-value vector containing the optimal set of the design and the
operating parameters.
Another way is to obtain the control signal maximizing/ minimizing certain
objectives while not violating any physical constraints (Optimal Control).
The solution set here is a profile rather than a single-value vector.
It happens sometimes that certain performance measures show conflicting
behavior making the process of optimizing the systems falls into the area of
multiobjective optimization.
Computational mathematics:
The mathematical models describing the performance of catalytic reactors are highly nonlinear due to the presence of the chemical reaction rates. Their numerical solutions require strong background in computational mathematics and applying sometimes "tricks" for choosing good initial guesses. Different mathematical techniques are sometimes applied and combined so that efficient algorithms are obtained.
Ideas for MS.c and PhD students:
Different membrane reactor configurations were
proposed in last two decays. However, optimal designs have been
addressed for a few of them. In this work, rigorous reactor models will
be derived and then the effects of design and operating parameters are
evaluated. These types of problems can fit to be master works.
Interested students are required to have an excellent background in
reaction engineering, optimization, and using Matlab, C++ and Fortran.
Fluidized bed reactors are novel alternatives
for reactions suffering from intraparticle mass and heat transfer
limitations due to the small size of catalyst particles which are
usually in micrometers. However, these rectors are less efficient than
fixed bed reactors due to the transportation limitation between phases.
Here, it is required to break these limitations using the concept of
optimal control. Optimal transportation profiles are intended to be
obtained and implemented. This project can be completed by PhD students
having good background in reaction engineering and programming skills in
C++ and Matlab.
Currently, I'm working on breaking the thermodynamic limitation of dehydrogenation of propane in a novel membrane reactor which can be operated cocurrently, countercuttently, cocurrent-countercurrently. If you are interested, you are free to join!!!!