This thesis focuses at designing, optimising, simulating (using computational fluid dynamics, CFD) and testing a multiphase intensified chemical reactors for the fast release of hydrogen from liquid organic hydrogen carrier (LOHC), for its eventual use on-board of ships (with hydrogen-fuelled engines). The reactor will be designed according to the specificity and requirements of the LOHC dehydrogenation chemical reaction, i.e. a slow endothermic heterogeneous catalytic chemical reaction between the LOHC and a catalyst particulate phase, generating high volumes of gas. More specifically, the chemical reaction requires: i) an intimate contact between the liquid phase and the catalyst, ii) an efficient and fast removal of the hydrogen generated without liquid entrainment, iii) an efficient heat transfer for the endothermic catalytic reaction while minimising the thermal stresses on the LOHC, iv) a short contact time between the catalyst and the LOHC, v) processing of high flowrates of LOHC to offset the dehydrogenation slow kinetics, and finally, vi) a compensation for the effect of the ship movements on the gas-liquid interface.
Designing this ideal device represents a considerable challenge, and the perfect reactor for this task does not exist yet. However, we will make use of a current trend in chemical reaction engineering that aims at adapting the geometry of chemical reactors so that the elementary steps of a global chemical reaction leading to the desired products are favoured. As part of this thesis, we will establish the building blocks of an automated chemical reactor design procedure: The optimisation of the reactor geometry will be performed using a constrained shape optimisation strategy, from an initial parameterised geometry. The constraints for the optimisation procedure are the mass, energy and momentum balances, evaluated numerically through the use of computational fluid dynamics (CFD), using the open source code OpenFoam.
An initial parameterised geometry (chemical reactor configuration to iterate from) is required. The selected doctoral student will first review the potential reactor configurations, but the promotor preliminary proposes a generalisation of the gas-solid vortex reactor (GSVR) concept for multiphase reactor flows (thus defining a Gas Solid Liquid Vortex Reactor, i.e. a GSLVR). This type of centrifugal reactor combines several interesting characteristics. At sufficiently high rotation speed, the effect of gravity can be neglected. The presence of a low pressure zone along it centre axis also allows for a preferred gas outlet. The GSVR is also a centrifugal device, thus combining reaction and separation functions. The parameters to be optimised for this reactor configuration are the number of slots (i.e. entry point for the liquid to the zone where the catalyst is located, the reactive zone), their spacing, the height of the device, the reactive zone chamber diameter, the position of both the LOHC inlet and outlet, as well as the diameter of the exhaust (gas outlet).
The "holy grail" of numerical experiments, i.e. without experimental validations, is still far from being a realistic objective in the field of CFD. Experimental validation is required, especially in the context of simulations of turbulent reactive flows using the two fluid model (Eulerian-Eulerian approach). A setup allowing for experimental validation and demonstration will be constructed. The Particle Image Velocimetry (PIV) technique will be used to validate both the liquid flow (liquid phase seeded with tracer particles) and catalyst bed (unseeded PIV). Due to the interdisciplinarity of the proposed research, the student will acquire a comprehensive knowledge in numerous complementary fields – chemical (chemistry and catalysis), mechanical (fluid mechanics), programming (C++®, Python®, CFD, etc.) – at both theoretical, computational and experimental levels.