Keynote and Invited Talks
Keynote and Invited Talks
High-Fidelity Simulation and Machine-Learning for Heat Transfer Problems
Professor Richard Sandberg
Department of Mechanical Engineering, University of Melbourne
CFD predictions are becoming increasingly important in the design of gas turbines because correlation-based methods are unable to further improve efficiency and laboratory experiments with the required fidelity are prohibitively expensive.
This presentation will show how physical insight relevant to designers can be extracted from high-fidelity simulations enabled by the latest HPC systems. It will also discuss how the HPC-generated data can be used to develop better predictive models using a machine-learning approach that is based on gene-expression programming. The case studies presented include effects of inflow turbulence characteristics on heat transfer to a high-pressure turbine blade, turbulent Prandtl number variations for wall jets and cooling efficiency of jets in cross flow. It will be shown that the machine-learnt models outperform traditional models both for the cases they were trained on and for cases not seen before.
Richard is Chair of Computational Mechanics in the Department of Mechanical Engineering at the University of Melbourne. His main interest is in high-fidelity simulation of turbulent flows and the associated noise generation in order to gain physical understanding of flow, noise and heat transfer mechanisms. He also uses the data to help assess and improve low-order models that can be employed in an industrial context, in particular by pursuing novel machine-learning approaches.
He received his PhD in 2004 in Aerospace Engineering at the University of Arizona and prior to joining the University of Melbourne, he was a Professor of Fluid Dynamics and Aeroacoustics in the Aerodynamics and Flight Mechanics research group at the University of Southampton and headed the UK Turbulence Consortium, coordinating the work packages for compressible flows and flow visualisations and databases. He was awarded a veski innovation fellowship in July 2015 entitled: “Impacting Industry by enabling a step-change in simulation fidelity for flow and noise problems” and has been granted an Australian Research Council Future Fellowship for 2020-2023.
When Surfaces Matter: Using Nano-scaled Surface Features to Control Macro-scale Transport
Professor Gary Rosengarten
School of Mechanical and Automotive Engineering, RMIT University
Typical heat transfer correlations do not consider the effect of surface properties for processes such as convection or radiation. However, even nano-scaled features can have a massive effect on transport properties. In this presentation I will show some of our work demonstrating how micro- and nano-scale engineering can be used to control and enhance both heat and mass transport.
Professor Gary Rosengarten is head of the Laboratory for Innovative Fluid Thermal Systems (LIFTS) in the School of Engineering at RMIT University and leader of the Energy Cluster at RMIT. Prior to joining RMIT University in 2012, he spent 6 years at UNSW running the heat transfer group, and being head of the thermal fluids research area. He also has 2 years experience in consulting for sustainable building design. He has first class honours degrees in Mechanical Engineering and in Physics from Monash University, and a PhD in Mechanical Engineering from the University of NSW. He won the inaugural American Society of Mechanical Engineers (ASME) Solar Energy Division Graduate Student award in 2000. In the last 6 years he has been awarded over $6 million in funding from ARENA for various renewable energy project. He has approximately 150 refereed papers in fields ranging from Solar Energy to Biotechnology.
High-order Multi-component Lattice Boltzmann Method and its Capabilities for Non-ideal Fluid Mixtures
Associate Professor Emilie Sauret
School of Mechanical, Medical & Process Engineering, Queensland University of Technology
Multi-component and multiphase fluid flows are of great interest for a large scientific and industrial community. For example, Organic Rankine Cycles (ORC) for waste heat recovery use mixtures of refrigerants and tune their composition to provide a new degree of freedom for the design of adapted ORC systems while viscoelastic polymer solutions in microfluidic applications open new avenues for increasing mixing at the microscale. However, those mixtures can exhibit complex non-ideal behaviour due to nonequilibrium statistical fluctuations. Computationally predicting this complex behaviour then becomes a challenge. High-order Lattice Boltzmann (LB) models with pseudopotential interactions set a compelling case as a numerical method for solving the dynamics of those non-ideal fluid mixtures and predicting essential transport properties of nonequilibrium flows.
In this context, this talk will focus on the development of an innovative high-order multi-component LB model to solve complex physical fluid phenomena. By expanding the LB model to higher orders, it is possible to solve for non-ideal and nonequilibrium flow quantities and fully recover the Navier-Stokes equations. High-order LB pseudopotential model capabilities to simulate non-ideal fluid mixtures are firstly explored by simulating a binary non-ideal fluid mixture undergoing phase-change for both the case of miscible and immiscible components. To further explore the capabilities of this method, the properties of a real refrigerant mixture, namely R245fa and R134a, which exhibits zeotropic behaviour, is considered. Finally, the approach is extended to model single-phase viscoelastic polymer flows using the four-roll mill benchmark at high Weissenberg numbers. The presented numerical approach serves as a preliminary step towards an efficient, robust and realistic numerical method for simulating complex non-ideal fluid flow mixtures that can pave the way to new technologies for bioengineering and energy applications.
Dr. Emilie Sauret is currently Associate Professor in the School of Mechanical, Medical & Process Engineering, Queensland University of Technology (QUT), and an elected council member of the Australasian Fluid Mechanics Society. She received her PhD degree in Turbulence Modelling from the University Pierre & Marie Curie, Paris, France in 2004. Prior to her postdoctoral position (2009-2012) at the University of Queensland, she spent 5 years in the automotive and oil and gas industry both in France and in Australia. In 2013, she was awarded an ARC-DECRA and joined QUT where she teaches in the Mechanical Engineering degree. Dr. Sauret has extensive interdisciplinary research experience in computational fluid dynamics, applied mathematics and applied physics. Her current research focusses on the development of advanced computational techniques to accurately simulate complex non-ideal fluid flow behaviours that are critical for the rational design and robust optimisation of engineering applications, in particular in the field of energy and biomedical engineering. She has produced over 80 publications, attracted over $8M in research funding and established collaborations across the globe. In 2019, an Endeavour Leadership Fellowship supported her visiting position at MIT and in 2020, Dr. Sauret was awarded an ARC-Future Fellowship to uncover fundamental microscale physics, pioneering research on computational microfluidics.
Turbulent Vertical Natural Convection Boundary Layers – Insights Gained from DNS up to Grd = 1.8 x 108
Dr Nicholas Williamson
School of Aerospace, Mechanical & Mechatronic Engineering, The University of Sydney
The natural convection boundary layers (NCBLs) that form adjacent to a vertical heated surface occur widely in engineering problems and also in the natural environment. Some of these applications occur at very large scales and at Grashof numbers (Gr) which significantly exceed the conditions obtained by laboratory experiments or DNS to date. This talk will focus on the mechanics of these large scale turbulent flows and recent observations made using DNS of temporally evolving parallel NCBL flow up to Grd = 1.8×108 at Pr = 0.71 (Grx ~ 2.6 x 1011). At high Gr, this flow is comprised of a turbulent boundary layer region that is coupled with a turbulent outer plume. The DNS show that the boundary layer exists within a constant forcing layer and suggest that direct contributions of buoyancy to this forcing remain significant until Grd > 109. Within this range, the direct action of buoyancy leads to modification of the turbulent log-law of the wall for the mean velocity profile, the form of which will be discussed. The outer bulk plume-like region is shown to attain self-similar Gr-independent behaviour after Grd = 107 and the surface heat transfer and shear are directly related to the top-hat scales which characterise the plume-like flow. The heat transfer at Grd = 1.8 x 108 is shown to attain characteristics of the ultimate flow regime, with Nu ~ Gr0.37 in comparison with Nu ~ Gr0.33 seen in laboratory studies at lower Gr. The high Gr scaling is analogous to that identified for Rayleigh Bernard (Grossman and Lohse, 2000; 2011). The transition to and characteristics of this regime will be discussed.
Dr Nicholas Williamson is a Senior Lecturer in the School of Aerospace, Mechanical and Mechatronic Engineering at the University of Sydney. His research focuses on buoyant turbulent flows, in particular natural convection boundary layers and free shear flows such as negatively buoyant jets. He also works collaboratively in riverine science towards understanding the physical mechanisms which cause algal blooms and how stable thermal stratification aids bloom formation.