In the present work the complex subject of thermal convection is tackled both numerically and experimentally. In the past, this subject has attracted the interest of many researchers due to its importance in many technological processes and natural phenomena. The present thesis has been devoted to explore some important aspects still requiring attention due to the inherent (physical) complexity and the difficulties embedded in the underlying (mathematical) models. In particular, its main objective
has been an investigation into the specific routes taken by thermal convection to evolve from initial (steady and laminar) states towards (low-dimensional or high-dimensional) chaos. Towards this end, the problem has been approached in the framework of numerical (direct numerical solution of the governing non-linear equations) and experimental activities; these have been used in synergy with existing theories on the behaviour of non-linear systems (e.g. the bifurcation theory) and tools for the analysis of chaotic systems (algorithms for the evaluation of the fractal structure of attractors). Moreover, much effort has been devoted to identify the existence in the space of parameters of different branches of laminar flows that coexist and can be effectively selected by a fluid-dynamic process depending on its initial state, i.e. the so-called multiple solutions. In order to identify ’universality classes’ in such dynamics, different types of convection (driven by thermal buoyancy and/or by surface-tension gradients) and different fluids (namely, liquid metals, gases, ordinary liquids and even non-Newtonian liquids) have been considered. As another possible degree of freedom potentially affecting the considered problems, the role of thermal boundary conditions has also been explored (assuming adiabatic and conducting boundaries, or walls with finite thickness, which can effectively exchange heat with the external environment). The thesis has been structured in order to elaborate a relevant physical and mathematical context in the
first part and present extensively all the obtained results in the second part.
|Date of Award||23 Mar 2020|
- University Of Strathclyde
|Sponsors||University of Strathclyde|
|Supervisor||Marcello Lappa (Supervisor) & Monica Oliveira (Supervisor)|