Hello,
Professor Heinz Preisig has attended a workshop in Stuttgart and
here follows his report-
http://www.chemeng.ntnu.no/person/person.php?id=preisig
---
3rd International Workshop on
Modeling, Simulation
and
Design in Process Engineering
February 27 - 28, 2003
University of Stuttgart, Germany
http://www.sfb412.uni-stuttgart.de
Overall impression:
The workshop was extremely
well visited with 150+ people from universities but also industry. The
arrangement of having reputed lecturers in a chosen field and
presentations from the program was well balanced and it was not only PhD
students to present their material, but it was also the masters that gave
a survey or indulged in a detail on occasions. The atmosphere was open
and constructive.
Written Material:
And now to the
workshop itself with comments from Heinz:
Thursday, February 27,
2003
Modeling in Process
Engineering
09.15
Modeling and Simulation in a
Specialty Chemicals Company
U. Plöcker, Degussa AG,
Hanau
Interesting discussion on the structure of the chemical business with
a distinction of molecule suppliers, materials and problem solvers as
structural components in
today’s business. Emphasis was placed on the need for rigorous models
mainly also for design purposes but also operations. Flowsheeting is seen
as a commodity.
10.00
Molecular Simulation in Process
Design
R. Krishna, University of
Amsterdam, NL
Molecules in zeolites was the theme. Krishna showed that simple
arguments, based on geometry and entropy considerations provide a
relatively easy insight into the separation behaviour of zeolites.
Overall an interesting talk with a little too much emphasis on appealing
to the public thus too much repetition on the cost of more information.
11.15
Molecular Simulation of
Thermodynamic Properties of Real Fluids
J. Vrabec, J. Stoll, H. Hasse,
University of Stuttgart
Computational thermodynamics for liquid mixtures computing effective
potentials. Basis is a polar, two-centre model and Lennard-Jones
potentials incorporating shape dependent quantities such as energy
associated with size and elongation, as well as polarity dependent
quantities such as the dipole and quadrupole. Models describe crit temp,
sat density and heat of vaporization mostly in the form of exponentials.
Viral coefficients and prediction of surface tension as well as self
diffusion coefficients, latter based on Green-Kubo model. Peng Robertson
models are seen as fine in some cases but not in others. The results from
molecular data are better than empirical equations of state such as
Peng-Robertson. Ternary mixtures are OK taking into account binary
interactions only. Three and more molecule interactions are not of
importance.
11.45
Nonlinear Wave Propagation in
Combined Reaction/Separation Processes
A. Kienle, Max Planck Institute,
Magdeburg, S. Grüner, University of Stuttgart Bio Process
Engineering
12.15
Systems Biology: The Reincarnation
of Systems Theory Applied to Biology?
O. Wolkenhauer, University of
Manchester, UK
Current state of the art is to accumulate a lot of data and then use
data mining techniques, thus correlation techniques, to extract models.
It is believed that this will rapidly change in the near future as more
mechanistic views are coming to bear. Hybrid behaviour is likely to play
an important role. There is also quite a bit of redundancy in biological
systems.
14.30
Dynamic Behaviour of Microbial
Populations in Stirred Bioreactors Simulated with EULER-LAGRANGE
METHODS
M. Reuss, A. Lapin, University
of Stuttgart
Demonstration of impressive combination of computational fluid
mechanics with biological reaction kinetics. Marvelous pictures and
movies showing spread and growth of “bugs” in the tanks with different
stirrer configurations.
Numerical
Simulation
15.00
Consistent Modeling of Multiphase
Flows
S. Subramaniam, Rutgers
University, Piscataway, NJ, USA
A discussion on different approaches to CFD including
Eulerian-Lagragian approach. Detailed technical discussion. One of the
main things in my point of view was the clear understanding that
assumptions such as incompressibility leads to index problems and
requires or allows for state reduction, however view one
prefers.
15.45
Numerical Solution and Parameter
Estimation of Particulate Processes
M. Wulkow, CiT,
Rastede
Detailed technical discussion on the numerics and their intrinsic
properties. The mix of stochastic with mixing poses indeed difficult
problems. On the other hand, the approaches are quite down to earth as
one defines largely a network of class transitions and an initial
distribution.
17.00
A Novel Reverse-Flow Reactor Concept
for the utothermal Coupling of Endothermic and Exothermic eactions
B. Glöckler, G. Kolios, G.
Eigenberger, University of Stuttgart
The discussion centred on how combine storage of energy with a
endothermic reaction with the energy carried in through an exothermic
reaction the two reaction schemes being alternately initiated. Obviously,
at least I thought so, is the heat distribution and the locality of where
the reactions consequently take place of dominant importance. I side feed
turns out, not too surprisingly, deliver a more uniform heat distribution
during the “heating” cycle, which yields a higher efficiency when
switching to the endothermic productive reaction.
17.30
Simulation of Acoustic
Fluid-Structure Interaction by Hybrid Boundary Element - Finite Element
Formulations
L. Gaul, University of
Stuttgart
This is not my speciality, but I did ask for papers after this talk,
because I felt that there is something in it that one should look at
carefully. He maps the 3D behaviour onto the 2D boundary. I think this
could be very interesting when thinking control of distributed systems.
That was the reason I asked for the literature.
18.00
Reduced Kinetic Models for CVD
Processes
J. Bauer, U. Maas, University of
Stuttgart
This man has been making his reputation on studying time-scale-based
model reduction for large chemical reaction systems as they are common in
incineration processes. I think we are doing quite the same in our
software, though in distinction with the extensive quantities and not in
the concentration space. Interesting are the procedures on how they
decide on the time-scale splits.
Friday, February 28,
2003
Process Analysis and
Design
08.30
Low-Dimensional Models for
Homogeneous and Catalytic Reactors
V. Balakotaiah, University of
Houston, TX, USA
His work is suggesting to use finely-granular models to start with and
then find more simple models through averaging procedures. This would
need a careful study in order to judge its true contents. The theory
being used is Liapunov-Schmidt.
Population Balance Modeling:
Granulation
M. J. Hounslow, University of
Sheffield, UK
I felt this was pretty much straightforward, though certainly relevant
as chemical engineers do not so much lean towards population balance
models just yet. The subject is gaining on interest, though.
10.00
Pressure Swing and Simulated Moving
Bed Technologies for Adsorption Processes
A. Rodrigues, University of
Porto, Portugal
I did not get much out of this story. Lots of experimental experience
though.
11.15
Chemical Process Synthesis
Fundamentals and Tools
M. F. Doherty, University of
California, Santa Barbara, CA,USA
As usual, Mike gave a very pleasant, competent talk. Interesting how
Fritz Horn is coming to the surface again. Indeed his work has not yet
been digested, though he is dead for about 30 years.
Process
Control
12.00
Analysis and Control of Nonlinear
Process Systems based on Physical Insights
K. Hangos, Hungarian Academy of
Sciences, Budapest, Hungary
Erik Ydstie left his marks on her. She is trying to view processes
through the glasses of Hamilton and Lagrange. The two representations
were discussed. I did not quite like it. It was, for example not
mentioned that the two are linked through Legendre
transformations.
14.15
Neurobiology and Industrial Process
Control: Development and Implementation of a High Performance Sensor
System for an Industrial Polymer Reactor
B. A. Ogunnaike, University of
Delaware, Newark, DE, USA
Tunde talked about his work in DuPont, as slowly sufficient time has
passed that makes it possible to do so. The idea was to look at
neurological systems and their control behaviour. The example taken was
the heart and the control of pressure and flow. The key observation is
that there are two controllers active, one that acts quickly but then
passes control on to a slower one as time progresses. The quick response
yields non-sustainable conditions, which is the reason that the first
controller is overruled with time by a second control action that yields
more sustainable conditions but does not deliver the same high level of
energy, thus performance is lower. The control loops also use a multitude
of sensor information of quite different quality. For example the fast
part is not very accurate, but then it is fast. This work will be
published. Certainly something to look out for.
15.00
Control of Periodical Fixed-bed
Processes
M. Bitzer, M. Zeitz, University
of Stuttgart
I did not pay much attention to this one, as I have been reviewing the
paper recently. It is a detailed analysis on how to control periodic
fixed-bed reactors with an application to PSA.
16.00
Nonlinear Model Reduction and
Control of Complex Distributed Processes
P. D. Christofides, University
of California, Los Angeles, CA, USA
Model predictive control in distributed systems. He uses a combination
of nonlinear model reduction techniques and geometric as well as Lyapunov
methods to come up with a good control scheme. Performance was
demonstrated on several examples. The low order of the models is rather
relative. Complexity is high.
16.45
How much Nonlinearity is too much
Nonlinearity?
T. Schweickhardt, F. Allgöwer,
University of Stuttgart
The core of this presentation was the discussion of a new measure for
nonlinearity. Elements in the measure include process dynamics, operating
point and operating region, and controller design procedure and
performance specs. The new measure is based on the optimal control law.
The computational procedure does not require the computation of the
optimal controller, which initially is a surprise, at least it was to
me.
17.15
Closing Remarks by the Speaker of
the SFB 412
H. Hasse, University of
Stuttgart
This archive was generated by hypermail 2.1.3 : Fri Sep 05 2003 - 16:42:09 CEST