Invited talks
7th International Berlin Workshop (IBW7)
"Transport Phenomena with Moving Boundaries and More"
Download full program [pdf]
Keynote lectures will be presented by:
Prof. Dr. Dieter Bothe
"Mass Transfer in Pure and Contaminated Fluid Systems - Continuum Thermodynamics and Detailed Numerical Simulation"
Institute of Mathematical Modeling and Analysis, TU Darmstadt, Germany
http://www.mma.tu-darmstadt.de
Prof. Dr. Amparo Galindo
"Theoretical and Computational Development for Next Generation Thermodynamic Modelling of Complex Fluids"
Department of Chemical Engineering, Imperial College, London, Great Britain
http://www.imperial.ac.uk
Prof. Dr. Johannes Khinast
"From Multiphase DNS to Large-scale Bioreactor Models"
Institute for Process and Particle Engineering, University of Technology Graz, Austria
http://portal.tugraz.at/
Prof. Dr. Daniele Marchisio
"Beyond Simple Masstransfer Models for Polydisperse Systems with Fluidic Interfaces: A Population Dynamics Approach"
Institute of Applied Science and Technology, Politechnico di Torino, Italy
http://www.disat.polito.it
Prof. Dr.-Ing. Jadran Vrabec
"Molecular Simulation of Fluid Phase Boundaries"
Institute of Thermodynamics and Energy Technology,
University of Paderborn, Germany
http://www.uni-paderborn.de/
Mass Transfer in Pure and Contaminated Fluid Systems - Continuum Thermodynamics and Detailed Numerical Simulation
Dieter Bothe
Institute of Mathematical Modeling and Analysis
TU Darmstadt
Email: bothe@csi.tu-darmstadt.de
http://www.mma.tu-darmstadt.de
Abstract
Reactive mass transfer from rising gas bubbles to the ambient liquid is the basis for many chemical processes of industrial importance. Besides experimental investigations, the necessary intensification requires numerical simulations based on mathematical modeling. Our approach is based on continuum mechanical sharp-interface balances of mass, momentum and species mass. It employs the Volume of Fluid method to solve the two-phase Navier-Stokes equations with capillary interface, complemented by convection-diffusion-reaction equations for the involved chemical components. The method employs two scalar variables for the concentration field of each transfer component, avoiding artificial mass transfer and enabling the use of separate one-sided concentration limits at the interface. To capture thin concentration boundary layers at the bubble surface, we employ a subgrid-scale model for the concentration profile as well as further computational techniques.
In real world applications, impurities and additives are present in the liquid phase which often leads to partially immobilized fluid interfaces, significantly changing the bubble hydrodynamics and mass transfer rates. We report on a novel approach to incorporate such effects into numerical simulations, using a modified momentum transmission condition at the interface. Finally, an outlook on recent extensions toward local volume effects and reactive mass transfer subgrid-scale modeling is given.
close window X
Theoretical and Computational Development for Next Generation Thermodynamic Modelling of Complex Fluids
Amparo Galindo
Department of Chemical Engineering
Imperial College, London
Email: a.galindo@imperial.ac.uk
http://www.imperial.ac.uk
Abstract
The fundamental phase diagram of a pure substance exhibiting gas, liquid and solid phases is reasonably well understood. As the pioneering work of van der Waals showed, the fluid phase behaviour of "simple" fluids can be understood in terms of the balance of spherical repulsive and attractive forces, and even the fluid-solid transition of such systems can be explained in terms of the freezing of a hard-sphere system. A challenge, however, arises when trying to describe, and even predict, accurately the properties of a given substance of mixture of substances. Modern equations of state, such as SAFT (statistical associating fluid theory), which are based on detailed molecular models have greatly enhanced the capability of analytical methods and provide a tool that can be used to study complex fluids. I will discuss some of our recent advances in this area; especially the recasting of the free energy expression into a group contribution method and the advantage of incorporation of generalised Lennard-Jonesium potentials. With this framework in place, the analytical approach (the equation of state) can be used to develop so-called force-fields for use in computer simulation. The importance of taking a molecular perspective towards complex systems is highlighted as the design of novel materials is guided by a theoretical understanding.
close window X
From Multiphase DNS to Large-scale Bioreactor Models
Johannes Khinast
Institute for Process and Particle Engineering
University of Technology Graz
Email: khinast@tugraz.at
http://portal.tugraz.at/
Abstract
Large-scale bio reactors are used in a wide variety of process industries,
ranging from waste-water treatment to the production of novel bio-pharmaceuticals.
The associated (bio-)reactor sizes can range anywhere from small ml-lab systems to
large industrial fermenters with several hundreds of meter cubes of fermentation broth.
In such large-scale systems thermal events become important as well, in addition to
mixing and mass transfer. Clearly, the processes in large-scale bio reactors are complex
and occur on many scale, yet have so far escaped a straightforward multi-scale analysis.
Processes on the molecular scale involve the bio-molecular expression process of the
fermentation products and the metabolism of the involved micro-organisms or cell systems.
On the micro-scale mass (and heat) transfer effects of bubbles and solid particles have
to be considered, as well as small-scale turbulence that interacts with any surface in
the system and is responsible for micro-mixing. Meso-scale effects include the dynamics
of bubble plumes, stirrers, heat exchangers and other internals. Macro effects describe
the macro-mixing in the reactor, i.e., the mixing fluid elements and the exchange between
different reactor zones.In this talk an introduction to bioreactor design is given,
followed by an overview of various approaches to the modeling of such multiphase systems
are presented, ranging from the direct numerical simulation of bubble swarms (DNS) and the
associated mass transfer in Newtonian and non-Newtonian fluids, meso-scale models developed
via model reduction to large-scale Euler-Lagrangian or Euler-Euler models. Finally, novel
computational methods based on GPU computing and Lattice-Boltzmann Models (LBM) are presented
that allow the analysis of industrial-scale systems.
close window X
Beyond Simple Masstransfer Models for Polydisperse Systems with Fluidic Interfaces: A Population Dynamics Approach
Daniele Marchisio
Institute of Applied Science and Technology
Politechnico di Torino
Email: daniele.marchisio@polito.it
http://www.disat.polito.it
Abstract
In this work the lim itations of the standard approaches to describe mass transfer in fluid-fluid systems
are critically discussed. Emphasis is placed on polydisperse systems, such as gas-liquid bubbly flows, liquid-gas
sprays and liquid-liquid emulsions and dispersions. In the simplest possible approach the fluid-fluid disperse
system is supposed to be spatially homogeneous and monodisperse: namely all the elements of the disperse
phase have the same properties. This very simple, albeit inaccurate, approach is the one typically employed for
design, scale-up and optimization of unit operations and of the relative equipment. The inadequacy of the model
depends on the competition between four phenomena: spatial mixing (induced by convection and turbulent
diffusion), coalescence, breakage and mass transfer (often triggered on enhanced by chemical reactions). When
spatial mixing dominates the system can be considered homogeneous, whereas when coalescence and
breakage dominate the only polydispersity that one has to account for is size. On the contrary when mass
transfer dominates the process, very often both spatial inhomogeneities and polydispersity with respect to size,
composition and velocity (of the elements of the disperse phase) must be considered. This can be efficiently
done by describing the dynamics of the entire population of elements of the disperse phase, by solving the socalled
generalized population balance equation. This latter equation is in turn generally tackled by using
quadrature-based moment methods, which can be easily implemented in computational fluid dynamics codes. In
this work the governing equations will be discussed and made dimensionless by using characteristic time-scales.
These time-scales will be used to define different regimes where the spatial homogeneous monodisperse model
can be used and where the other models (of increasing complexity) have to be used. Eventually some examples
taken from our current work on gas-liquid stirred tanks and bubble columns are also discussed.
close window X
Molecular Simulation of Fluid Phase Boundaries
Jadran Vrabec
Institute of Thermodynamics and Energy Technology
University of Paderborn
Email: jadran.vrabec@upb.de
http://www.uni-paderborn.de/
Abstract
Molecular modeling and simulation with molecular dynamics (MD) or Monte Carlo (MC) techniques allows for tackling
numerous problems in engineering science that were inaccessible to numerical modeling before. On the basis of
physically sound and quantitatively validated models of the molecular interactions, relying on quantum chemistry
data and experimental measurements, technically relevant fluid systems may be analyzed with molecular simulation.
The field is rapidly evolving, driven by methodological progress and also by the ongoing exponential growth of computing power.
The author presents an introduction into molecular modelling and simulation and its application to the thermophysical
properties of fluids, for which accurate results may be achieved. The focus lies on the properties of inhomogeneous
systems that exhibit a phase boundary. Surface structure and its features are discussed also for mixtures. Dynamic
processes, like nucleation at the onset of droplet formation and evaporation, are addressed.
close window X
