CFD Modelling as an Integrated Part of Multi-Level Simulation of Process Plants –
Semantic Modelling Approach
Marek Gayer
1
, Juha Kortelainen and Tommi Karhela
VTT Technical Research Centre of Finland
Vuorimiehentie 3, Espoo, P.O. Box 1000, 02044, Finland,
www.vtt.fi
Tel. +358 20 722 6246, Fax. +358 20 722 7052
ext-marek.gayer@vtt.fi
,
juha.kortelainen@vtt.fi
,
tommi.karhela@vtt.fi
Keywords: process simulation, CFD modelling, multi-level
simulation, semantic modelling, plant equipment modelling
Abstract
In this paper, we present our computational fluid
dynamics (CFD) modelling and simulation environment,
which is designed to be suitable for integration to a large-
scale system level simulation tools for industry process
simulation of plants, such as Apros 6 software.
We discuss about the coupling possibilities of these
simulations and further concentrate on the description of our
pre- processing, fluid solver integration, and post-processing
within the open-source Simantics integration platform. We
present the semantic data model and ontologies for
describing simulation models and their relations. We discuss
about suitable open-source software component candidates,
realisation of geometry, mesh, case configuration, boundary
and initial conditions, solvers, and visualization parts.
Finally, we present our proposal implementation based
on Eclipse Rich Client Platform.
1
In addition, because we
rely on open-source software tools, our proposals could be
especially interesting for developments in the area of
general process simulators, which might be eventually
extended with more detailed 3D simulation models.
1. INTRODUCTION
1.1. Large-scale dynamic simulation of process plants
and simulation information management
Dynamic-process simulation tools are used for example
in the nuclear energy sector for planning, operator support
and training, operation-state analysis, automation design and
testing, safety analysis, and verifications in various stages of
the power plant lifecycle. The advantages gained using these
tools and methods can result in significant time and money
savings, and improved safety.
The 3D flow simulation of an entire process plant is not
possible for performance and complexity reasons. Instead,
in dynamic-process simulation, we use 1D system codes for
1
This work was carried out during the tenure of an ERCIM "Alain
Bensoussan" Fellowship Programme.
simulating the process of the plant eventually with co-
operation or coupling with 3D simulation of the most
important plant components, connected at inflow and
outflow boundaries of the computational fluid dynamics
(CFD) model. In large-scale dynamic-process simulation,
there are also own models for the control and electrical
systems.
CFD is used also for design and safety analysis. There
is a clear need for combined analysis so that certain process
components are computed using CFD and the rest of the
process using large-scale process simulation.
In process simulation software systems, the simulation
modelling flow is usually expressed using graphical
diagrams. These environments also allow visualizing
simulation data in diagrams by using monitors, animations,
trends and various other, rather simple visualisation
methods. Fig.1 depicts the environment of our software
Apros 6, which is a 1D industry process plant simulator of
plants.
Figure 1. Apros 6 process simulator, integrated in Simantics
Workbench.
The studied problem is industrially relevant because the
need for multi-scale simulation tools is growing. It is
necessary to combine models of different levels of detail.
Usually there are specific simulation tools for each detail
level. By using combined approach, it is possible to make a
set of multi-domain, multi-physics, and multi-scale models
staying consistent even though the user is modelling from
one perspective at a time.
Process simulators include Information management
databases, which stores all the necessary information for
performing the simulation. However, it is usually not
possible to automatically co-operate with those databases
and transfer corresponding data for the use in simulators
outside the process simulations, such as existing 3D CFD
tools. This significantly complicates the co-use of the
simulators and it may be necessary to create separate models
with different data storage mechanisms to be used in
external simulation software applications, outside the
process simulator. It would be a huge task to combine all
these into one solver.
1.2. Coupling of CFD models with 1D process models
In coupling 3D CFD models with 1D process models,
proper mapping of the mass and heat flow variables between
the models is essential. At the inflow and outflow
boundaries, the flow variables of the 3D models have to be
reduced to 1D flow variables. In general, reduction a 3D
flow variable to a 1D flow variable can be done in a
straightforward manner by proper integration of the 3D flow
field over the coupling surface and by proper matching of
the 3D and 1D meshes. On the other hand, mapping the 1D
flow variable to a 3D flow field is somewhat problematic
because additional information is needed in order to
generate the 3D spatial distribution from the 1D flow
variable. In addition, local circumferential flow features at
the mapping interface may cause problems for the numerical
solving.
Similarly, as in coupling two different 3D codes, the
numerical stability is an important issue because the
coupling method is in practice often explicit. Therefore,
numerical instabilities may rise if the coupling between the
models is strong.
1.3. Our effort
We develop a platform for integration and coupling of
process simulation models (1D) and more detailed CFD
models (3D). For these purposes, it is necessary to establish
not only optimal and intuitive data transfer and interfaces
paradigms, but also pre-processing, post-processing, and
computational solver integration to the environment. The
latter are of primary concern and are of major topic of this
paper. For this purpose, our efforts also deal with selecting
and using optimal open-source tools to reach this goal. See
Fig. 2 for an overview of 1D process / 3D CFD modelling
connection. On the picture, it is indicated that upon selecting
one of the component from the 1D process simulator (such
as e.g. tank or pipe), we can change the modelling
perspective in the software for of such a component to the
full 3D CDF model, where we can make inspections,
modifications and simulations.
Figure 2. Illustrative schema of the link between 1D
process simulator (Apros 6) and the 3D CFD model
perspective.
2. RELATED WORK
The theoretical foundations of software platform
Simantics, which we selected as implementation
environment for process simulation and the CFD coupling,
are in detail described in [1], [2], [3] and [4].
Several related methods and concepts have been
demonstrated and documented separately. Using ontology-
based 2D vector graphics, based on scalable vector graphics
(SVG) was proposed in [5], a proposal of 3D models based
on ontologies in large-scale process simulation can be found
in [6], a marker based learning environment for detection of
equipment in plants was described in [7]. The next
generation of the large-scale process simulation software,
developed at VTT and widely used in industry, Apros
2
[8],
[9], is based on Simantics. In addition, the next version of
BALAS
3
will be integrated to Simantics. BALAS is a steady
state simulation package for pulp and paper processes
developed at VTT Technical Research Centre of Finland,
and several paper mills, engineering companies, and
equipment manufacturers currently use it.
There has been some effort in integration between
large-scale process simulation and CFD, such as a prototype
level co-simulation solutions e.g. by FLUENT and Aspen
Plus using CAPE-OPEN standard. This is a standard for
communication between computational components in
Process Modelling Environments (PME), suitable especially
for sequentially based process simulators [10]. However, for
the equation-based process simulators (such as Apros), the
standard was found to be unsuitable for this kind of
coupling [11], [12].
f CAPE OPEN has been used in implementation of
Integration Toolkit for Aspen Plus and Fluent [13], [14],
where in the bi-directional coupling, information of flow
rate, temperature, pressure, and species components can be
exchanged. The commercial applications have included
chemical reactor, fuel cell system, coal-fired power plant,
and natural gas combined cycle plant [15], [16].
Recently, component-based integration platform
CHEOPS was implemented for chemical process modelling
and simulation by using CORBA [17]. The simulation tools
tested included Fluent, Aspen Plus, gPROMS and Parsival.
As an application, dynamics of crystallisation was studied
with multi-scale coupled simulation by using Fluent and
Parsival [18].
Ontologies and semantic representations are recently
used in wide area of engineering and industry applications.
One of recent field is it’s utilization in Building Information
Modelling [19]. Also in Building Information Modelling,
there is the desire of connecting simulations of different
levels (CFD perspective of situations in the rooms vs.
situation in whole building is a task to solve).
2.1. Open-source software tools and components
Our approach utilises as far as possible open-source
software tools for the prototyping and delivery so that the
major parts of the software proposal could be public and
could be freely utilised also by others.
We have reviewed and practically experimented with
several most advanced open-source CFD solvers and
components for pre- and post-processing. The most
appealing components candidates for CFD field according
to our observations were: SALOME [20], OpenFOAM [21],
Code_Saturne [22], Gmsh [23], NETGEN [24], TetGen
[25], snappyHexMesh [21] and recently also Discretizer
2
Apros software website:
http://www.apros.fi
3
BALAS software website:
http://balas.vtt.fi
[26]. More detail reasoning can be found in [27], [28]. We
also use open-source visualization and geometry processing
libraries VTK
4
and OpenCascade
5
. The concrete selections
and usage of the components in our proposal is covered in
the next chapter.
3. OUR SOLUTION PROPOSAL
3.1. Semantic data model concept
The development of the Semantic Web
6
and its
technologies has increased activity and interest on semantic
data management and its applications. Technologies, such as
the Resource Description Framework (RDF) [29], the Web
Ontology Language (OWL) [30], SPARQL Protocol and
RDF Query Language (SPARQL) [31], and Semantic Web
Rule Language (SWRL) [32], introduce a set of basic
technologies to describe the data, to model ontologies, to
describe semantic database queries, and to model rules and
complex restrictions into ontologies respectively. The
overall concept these technologies introduce has inspired
domains outside the Web technology to consider applying
these methods for data management.
The fundamental principle of the Semantic Web, to
include or map the meaning of the data to the data itself, is
very attractive also from the system modelling data-
management point of view. In the Semantic Web, one
relatively simple data model is capable of describing
practically all kind of data and knowledge, and data
semantics allow computer-based reasoning on data, which,
on the other hand, increases the usability and value of the
data.
The Semantic Web project aims to developing
technologies for improving the usability of the data and
knowledge in the Web. Due to the fact that no-one can
know what information can be found from the Internet, the
basic assumption of the knowledge world is permissive; the
Semantic Web is based on the open world assumption
(OWA): if something is not specified in the data model it
still can be correct, it is just undefined. This apparently
simple assumption in the Semantic Web makes it difficult to
apply the Semantic Web technologies for e.g. managing of
modelling data of system simulation. This is due to the
closed and well-defined nature of modelling domains. For
setting restrictions and rules for modelling data, to e.g.
enable automatic model validation, the use of OWA would
make the domain ontology development a demanding task.
In system modelling, it is common that large amount of
data is managed during the modelling process. An example
of this is finite volume method, in which the modelled
domain, e.g. the geometry of a pipe, is divided into a set of
control volumes that fill the whole domain; this is called
4
Visualization Toolkit (VTK):
http://www.vtk.org
5
Open CASCADE website:
http://www.opencascade.org/
6
Semantic Web:
http://www.w3.org/standards/semanticweb/
discretisation. The description of the discretised geometry,
the mesh, contains usually large number of data, and for its
representation, tables and vectors are the natural choices. In
OWL, these data structures are missing, which also
decreases its attractiveness for system modelling data
representation.
3.2. Selected Approach
The selected approach in our work is based on the use
of semantic data model and ontologies for describing the
modelling data and its relations. In the semantic data model,
all data is described using simple data structures, triples,
which consist of a subject, a predicate, and an object; triples
are also called statements [29]. With this simple data model,
it is possible to describe complex data and its relations. In
addition, the data model is very flexible and extensible. The
data is described using ontologies, a kind of semantic
vocabularies, which define concepts in different domain
areas. Ontology in a semantic data model can be seen as a
class library with a specified hierarchy and properties.
Numerical simulation in general and system simulation
especially are suitable for ontology-based modelling, due to
their hierarchical and well-defined nature. The development
of a domain modelling ontology is relatively straightforward
procedure when the domain concepts are well known.
The advantage in using semantic data model related to
present common methods is that all the data in the
modelling database is described using the same simple data
model. This model allows mapping of data from one domain
to another so that all individual parts of data are captured
just once. By using ontology mapping mechanisms, these
data can form a network of data that describes, e.g. in our
case, a complete process plant model. In addition to the data
description mechanisms, the application of semantic
ontologies enables the use of computer-based reasoning to
the modelling data. Inclusion of e.g. domain modelling
constraints and rules into the ontologies enables automatic
model validation to some extent. And as the modelling
constraints and rules are described also using the same data
model mechanisms, they can be stored together with the
modelling data. In traditional system modelling and
simulation tools, the model validation information has been
a feature of the software tool, not the modelling data.
3.3. Simantics Environment
There exist different kinds of solutions for modelling
and simulation integration. Simantics platform
8
has a unique
approach that combines semantic information modelling
(ontologies) and simulation. Simantics has its own ontology
description language called Layer zero. Layer zero has
similarities with Web Ontology Language (OWL) but it has
been specially designed for the description of engineering
8
Simantics platform website:
https://www.simantics.org/simantics
and system simulation ontologies, where the user is not just
classifying the existing world but designing new products
and production processes. These domains also consist of
more complicated information structures and data types than
some more traditional modelling targets.
In Simantics, plant information can be described and
stored using a semantic knowledge database. In this
environment, integration between different domain models
can be effectively modelled and simulation tools can be
configured based on existing plant design data, where
semantic modelling in large-scale process simulation is used
[33].
From the technological point of view, the platform
applies the server-client architecture, in which the server
consists of a semantic graph database, i.e., a triplestore
(Simantics Core), and the client (Simantics Workbench) and
its user interface framework are based on Eclipse
9
plug-in
architecture.
3.4. Advantages of using CFD and ontology approach
Using of ontologies and semantic approach does not
help us in obtaining better performance when computing
CFD cases using numerical solvers. The reason for this is
that these solvers are already implemented and does not
count with ontological representations and cannot be easily
modified to directly support semantic approach internally.
Furthermore, due to performance reasons, data structures for
codes inside these solvers are already chosen by its
developers, and are suitable for underlying numerical
methods, such as finite element or finite volume method.
The ontology representations provide advantages in
other situations, such as easier integration and cooperation
in different stages of CFD modelling. CFD modelling is
used in process industry e.g. by equipment vendors during
product development phase or by engineering offices during
trouble shooting and safety analyses. Most of the
engineering information during these stages is located in
different plant design or information systems (such as
CAD). CFD modelling process can then benefit on better
integration into the design systems. Ontology approach
provides us means for mapping information between the
models that are different in nature. It also catalyses
communication between engineers of different disciplines.
Same goes also with linking simulators of different solving
level of detail, such as in our case 1D process simulators
and CFD solvers.
The CFD solvers differ in performance, requirement of
the mesh quality and stability. While some solvers would
fail to converge during simulated task due to mesh does not
conforming quality requirements, such as conformance to
Delaunay criterions or ortogonality of the mesh, others
9
Eclipse platform website:
http://www.eclipse.org/
would still be able to continue, although with considerable
performance and precisions penalties.
Another advantage of ontologies description and
semantic approach is in possible obtaining of software and
data abstraction, suitable for using of various types of
meshing and other tools used in CFD modelling.
3.5. Modules and components for plant simulation and
3D CFD coupling
The CFD simulation process consists of the following
three phases: pre-processing, solving, and post-processing.
The pre-processing phase includes the definition of the flow
domain geometry, the domain meshing, definition of
boundary and initial conditions, as well as the parameters
for the simulation. In the solving phase, the flow simulation
is numerically solved. The post-processing phase can
contain computation of additional dependent variables and
visualization of the simulation results. In the following
sections, the grouping of the software modules and
components used in our approach are introduced and
discussed. See Fig. 3 for an overview of the modelling.
Geometry
Boundary
conditions
Mesh
Case
configuration
Input
data,
files,
etc.
PRE-PROCESSING
OpenFOAM
Other
solvers
SOLVING
Internal
Visualization
External
visualization
POST-PRO.
Permanent
storage
Results
Figure 3. A schema of CFD modelling workflow and
components used in our approach.
3.5.1. Geometry
For importing existing geometry e.g. from a CAD
system we use Open CASCADE software library. It allows
the geometry to be imported into the application in STEP,
IGES, BREP format. The geometry can be edited in editors
based on Open CASCADE, like SALOME. The imported
geometry is then used in the subsequent modelling steps.
Other formats, like STL are currently not directly supported,
however in such particular cases, such as when format is
simple, it is possible to implement easy conversion. In such
case, it would be even possible for some simulation cases, to
import the geometry model from modelling tools like
Blender, Maya or Lightwave.
The Open CASCADE also provides solid geometry
modelling features, which allow the imported geometry to
be modified or to help to build an interactive geometry
editor. This would be useful for e.g. simplifying the
imported geometry for meshing. In the present version of
the environment, this feature is however missing.
3.5.2. Meshing
Obtaining a quality mesh suitable for CFD simulations
often remains a rather difficult task, namely if we request
generating these meshes by using open-source tools. Finite
volume method, which remains the most standard method in
the field of CFD simulations, brings certain requirements of
quality for the mesh and failing to meet those conditions can
result in significant performance overhead, precision
problems, or can even make simulation impossible due to
instability and divergence of the solution.
When CFD is applied on industrial computation and
complex geometries, the pre-processing phase and
especially meshing becomes critical. For this reason,
automatic meshing is an important requirement that would
allow to make the meshing easy, intuitive, and fast, and
would allow eventually changing of the modelled
component shape fast. There are several algorithms
available for automatic tetrahedral meshing of arbitrary
geometry. These algorithms are robust and fast, especially
compared to manual meshing. Currently, we use NETGEN
meshing package, which we use as a custom command line
tool created from the NETGEN libraries and integrated and
launched from within our CFD environment. NETGEN
provides an automatic tetrahedron mesh of a fair quality,
which is sufficient to use with OpenFOAM. We found
hexahedron meshes to be more performance optimal for
CFD simulations, however there are not available open-
source meshing routines for fully automatic generation of
meshes on arbitrary structures. A possible integration of
hexahedral meshers, like snappyHexMesh or Discretizer,
remains for subjects of future work.
There are several strategies for managing mesh data in
a semantic database. One is to use a fully semantic data
model, in which all mesh details, i.e. nodes, cells, sub-
meshes, and boundary patches, are described in the data
model semantically. This approach would allow great
flexibility in how the data is used, but the amount of the
triple data and the limitations the size of memory and
efficiency prevent it. Another approach is to semantically
describe only the necessary features of the mesh, namely
data tables that define the details, and boundary and initial
condition data. This approach allows still good flexibility
and requires only fraction of the resources compared to the
fully semantic approach. In the present system, only
necessary mesh data is managed semantically.
3.5.3. Solving
Currently, there are not many eventualities for open-
source or free CFD analysis tools (especially for non United
States residents, where several CFD packages are offered
for free). OpenFOAM contains a C++ library, capable of
numerical solution of partial differential equations. With
this library, different solvers (including also those provided
with the OpenFOAM distribution) can be built to solve
various classes of problems in fluid dynamics and also other
fields. We are using OpenFOAM for numerical solution of
our CFD problems.
3.5.4. Case configuration and boundary and
initial conditions
Our CFD environment is integrated into the Simantics
database, which allows using ontologies to describe types of
case configuration. In the user interface, we can define
parameters for meshing, visualization methods, solver
launching configurations and parameters.
OpenFOAM based solvers are configured by using its
dictionaries, which are plain text configuration files,
containing information’s related to solving model, such as
algorithm control, numerical schemes and numerical
solution. However, in our proposal, instead of writing
corresponding dictionaries manually, we can also use
automatic transformation to generate dictionaries from our
ontological representation. The settings are in this way also
available through the user interface, by using the graph
explorer and properties editor components, provided by
using Simantics API. This way, we can configure basic
parameters of OpenFOAM toolbox inside of our user
interface.
Boundary and initial conditions for the simulation cases
are part of the case definition for the OpenFOAM solver
dictionaries, and therefore we configure them in the similar
way.
3.5.5. Visualization
The flow field visualization in our environment is
developed on VTK, The Visualization Toolkit [34, 35]. It
includes high-level library routines for several visualization
techniques, such as cut planes, iso-surfaces and streamlines.
All visual components can be mapped with a field variable,
such as flow velocity, pressure, or temperature. The use of
this library remarkably decreases the effort for the
implementation of visualization.
Beside of these integrated visualization possibilities,
users can use several open-source visualization tools, such
as possibilities with the generated data sets.
4. IMPLEMENTATION
The implementation of our proposal is realised within
the Rich Client Platform (RCP) of Eclipse and is using
Simantics. The RCP environment provides suitable
application framework for appropriate user interface and
also for integration and similar look and feel graphical user
interface of Apros, BALAS, or of some general CFD based
applications, such as ANSYS or SALOME. This
implementation allows us to use pre-processing and post-
processing capabilities for solved simulation cases. These
possibilities include visualization of the geometry (Fig. 4),
automatic generation of tetrahedron based mesh using
NETGEN algorithm and visualization of results of the
simulation using surface plots (Fig. 5), 3D cut plot
visualization (Fig. 6) and streamlines inside the studied
objects (Fig. 7).
5. FUTURE WORK
We would like to propose an ontology-based interface,
which would allow connecting process simulators with
various family of additional, both open-source and
commercial CFD solvers, in a universal way.
Our semantic approach method between large-scale
process simulation models (1D) and more detailed CFD
models (2D/3D) will allow integration and coupling
interfaces, communication, and synchronisation between the
solvers. For this purpose, we will propose optimal data
transfer and interfaces between 1D and 3D to allow control
of simulation flow.
6. CONCLUSION
We have presented our pre-processing, post-processing
and fluid solver environment for CFD simulations, which is
a major part of a large-scale system level simulation
integration of 1D process simulation of plants and 3D CFD
modelling and simulation of selected plant components.
The base of our software proposal relies on using
ontologies approach and semantic description, which is
especially suitable for numerical simulation in general. We
described advantages by using this semantic modelling and
briefly describe our semantic integration environment
Simantics, in which our proposal is realized.
Our CFD modelling environment proposal consists of
software allowing geometry, meshing, case configuration,
boundary and initial conditions, solving and visualization
parts. The solving of simulation cases is realized by using
an integrated OpenFOAM CFD code.
We have also briefly presented our proposal
implementation in Eclipse Rich Client Platform and the
Simantics integration platform.
Because we largely rely on open-source software
components, our proposal could be interesting also in other,
general 1D process simulations, where might be necessary
modelling connection and coupling with more detailed 3D
simulation models.
7. ACKNOWLEDGMENT
This project received financial support from Tekes (the
Finnish Funding Agency for Technology and Innovation),
and from Finnish industry. We would like to also thank
Veli-Pekka Kuutti for his contribution of software
development for this project, particularly for
implementation of visualization methods.
REFERENCES
[1]
T. Lehtonen, “Ontology-based diagram methods in
process modelling and simulation,” Master’s thesis,
Helsinki University of Technology, 2007.
[2]
A. Villberg, “Design challenges of an ontology-
based modelling and simulation environment,” Master’s
thesis, Helsinki University of Technology, 2007.
[3]
M.
Luukkainen, “Use of 3D graphics for
configuration and visualization of large scale process
simulation: ontology-based approach,” Master’s thesis,
Helsinki University of Technology, 2007.
[4]
T. Kalajainen, “An access control model in a
semantic data structure: Case process modelling of a
bleaching line,” Master’s thesis, Helsinki University of
Technology, 2007.
[5]
T. Lehtonen and T. Karhela, “Ontology approach
for building and visualising process simulation models
using 2D vector graphics,” in SIMS Proceedings of the 47th
Conference on Simulation and Modeling. Finnish Society of
Automation, SIMS - Scandinavian Simulation Society,
2006, pp. 141–146.
[6] M.
Luukkainen and T.
Karhela, “Ontology
approach for co-use of 3D plant modelling and large scale
process simulation,” in The 48th Scandinavian Conference
on Simulation and Modeling (SIMS 2007). Linköping
University Electronic Press, 2008, pp. 166–172.
[7]
S. Prammanee, M. Luukkainen, T. Seuranen, and
T. Karhela, “A marker-based mobile learning environment
for a process plant,” in IADIS International Conference
Mobile Learning. IADIS, 2009.
[8]
E. Silvennoinen,
K. Juslin,
M. Hanninen,
O.
Tiihonen, J.
Kurki, and K.
Porkholm, The APROS
software for process simulation and model development.
VTT publications, 1989.
[9]
A. Niemenmaa,
J. Lappalainen,
I. Laukkanen,
S. Tuuri, and K. Juslin, “A multi-purpose tool for dynamic
simulation of paper mills,” Simulation Practise and Theory,
vol. 6, no. 3, pp. 297–304, 1998.
[10]
The GCO Consortium, “CAPE-OPEN standards,”
2003.
[11] M.
Friman,
“Modularisation issues in dynamic
equation-based process simulation solvers - comparison of
two different approaches,” Master’s thesis, Helsinki
University of Technology, Department of Chemical
Technology, 2003.
[12]
E. Karlsson, “A process simulation model as part
of a process plant life-cycle information environment,”
Master’s thesis, Helsinki University of Technology,
Department of Chemical Technology, 2004.
[13]
S. E. Zitney, “Multiscale modeling and simulation
of advanced power generation systems,” in DOE/NSF
EPSCoR 2005 Conference, Dearborn, USA, 2005.
[14] M.
Osawe,
“Fluent CAPE-OPEN COM/CORBA
bridge and CO-compliant unit operation,” in 2nd Annual
U.S. CAPE-OPEN meeting, Morgantown VW, 2005.
[15] M.
Syamlal, S.
Zitney, and M.
Osawe. (2005)
Using CAPE-OPEN interfaces to integrate process
simulation and CFD. [Online]. Available:
http://www.colan.org/CO%20Update/COUpdate07_Fluent_
Article.html
[16]
D. Sloan, “Power plant analysis using CFD and
process simulation,” in 2nd Annual U.S. CAPE-OPEN
meeting, 2005.
[17] G.
Schopfer, A.
Yang, L.
von Wedel, and
W. Marquardt, “Cheops: A tool-integration platform for
chemical process modelling and simulation,” Int. J. Softw.
Tools Technol. Transf., vol. 6, no. 3, pp. 186–202, 2004.
[18]
V. Kulikov, H. Briesen, and W. Marquardt, “Scale
integration for the coupled simulation of crystallization and
fluid dynamics,” Chemical Engineering Research and
Design, vol. 83, no. 6, pp. 706 – 717, 2005, 7th World
Congress of Chemical Engineering.
[19] P.
Siltanen, S.
Varjes, M.
Ylikerälä, and A.
S.
Kazi, “Ifc and pmo for estimating building environmental
effects,” in Proceedings of the 14th International
Conference on Concurrent Enterprising, 23-25 June,
Lisbon, Portugal, 2008, pp. 515–522.
[20] A.
Ribes and C.
Caremoli, “Salome platform
component model for numerical simulation,” in COMPSAC
’07: Proceedings of the 31st Annual International Computer
Software and Applications Conference. Washington, DC,
USA: IEEE Computer Society, 2007, pp. 553–564.
[21] OpenCFD
Limited,
OpenFOAM - The Open
Source CFD Toolbox - User Guide, 2009,
http://www.openfoam.com/docs/
.
[22] F.
Archambeau, N.
Mechioua, and M.
Sakiz,
“Code_saturne: a finite volume code for the computation of
turbulent incompressible flows- industrial applications,”
International Journal on Finite Volumes, vol. 1, pp. 1–62,
2004.
[23]
C. Geuzaine and J.-F. cois Remacle, “Gmsh: a
three-dimensional finite element mesh generator with built-
in pre- and post-processing facilities,” International Journal
for Numerical Methods in Engineering, vol. 79, no. 11, pp.
1309–1331, 2009.
[24] J.
Schöberl, “NETGEN: An advancing front
2D/3D-mesh generator based on abstract rules,” Computing
and Visualization in Science, vol. 1, no. 1, pp. 41–52, 1997.
[25] H.
Si,
A Quality Tetrahedral Mesh Generator and
Three-Dimensional Delaunay Triangulator, 2006. [Online].
Available:
http://tetgen.berlios.de
[26]
B. Bergqvist. (2010) Discretizer, a free mesh
program for CFD. [Online]. Available:
http://www.discretizer.org/
[27]
J. Kortelainen, “Meshing tools for open source
CFD - a practical point of view,” VTT, Espoo, Finland,
Tech. Rep., 2009,
http://www.csc.fi/english/pages/lscfd/Documents/MeshingT
oolsForOpenSourceCFD.pdf
.
[28]
M. Gayer and G. Iannaccone, “A software platform
for nanoscale device simulation and visualization,” in IEEE
International Conference on Advances in Computational
Tools for Engineering Applications. Zouk Mosbeh,
Lebanon: Notre Dame University, Lebanon, 2009, pp. 432–
437.
http://www.gayer.ws/en/publications/gayer09actea/
[29]
F. Manola and E. Miller, “RDF Primer,” The
World Wide Web Consortium,
http://www.w3.org/TR/2004/REC-rdf-primer-20040210/
,
W3C Recommendation, Feb. 2004.
[30] S.
Bechhofer, F.
van Harmelen, J.
Hendler,
I. Horrocks, D. McGuinness, P. Patel-Schneijder, and L. A.
Stein, “OWL Web Ontology Language Reference,” World
Wide Web Consortium (W3C), Recommendation, 2004, see
http://www.w3.org/TR/owl-ref/
.
[31]
E. Prud’hommeaux and A. Seaborne, “SPARQL
Query Language for RDF,” The World Wide Web
Consortium, W3C Recommendation, Jan. 2008. [Online].
Available:
http://www.w3.org/TR/2008/REC-rdf-sparql-
query-20080115/
[32]
I. Horrocks, P. Patel-Schneider, H. Boley, S. Tabet,
B. Grosof, and M. Dean, “Swrl: A semantic web rule
language combining owl and ruleml,” World Wide Web
Consortium, W3C Member Submission, May 2004.
[Online]. Available:
http://www.w3.org/Submission/SWRL/
[33] A.
Villberg, T.
Lehtonen, T.
Karhela, and
K. Kondelin, “Applying semantic modelling techniques in
large scale process simulation,” in ALSIS ’06 Proceedings of
the 1st IFAC Workshop on Applications of Large Scale
Industrial Systems. Suomen Automaatioseura, 2006.
[34]
W. J. Schroeder, K. M. Martin, L. S. Avila, and
C. C. Law, The VTK User’s Guide. Kitware, Inc., 2000,
http://www.kitware.com
.
[35]
W. Schroeder, K. M. Martin, and W. E. Lorensen,
The visualization toolkit: an object-oriented approach to 3D
graphics, 4th edition. Kitware, Inc., 2006.
BIOGRAPHY
Dr. Marek Gayer is a research fellow of the Computer
Simulation Technology team, which is a part of the System
Research knowledge center of Technical Research Centre of
Finland (VTT).
He obtained his Ph.D. degree in “Information Sciense
and Computer Engineering” from Czech Technical
University in Prague in 2006 after defending his thesis
“Real-time Visualization Techniques for Modelling of
Combustion and Fluids”. Prior to joining VTT, he was a
research fellow of The National Research Council (CNR) at
University of Pisa, Italy and Norwegian University of
Science and Technology (NTNU) at Trondheim, Norway.
His current research interests include computer
modelling and simulation (such as those based on finite
element and finite volume method solvers), computational
fluid dynamics, information technologies, open source
software integration, software development and software
platforms for research simulation, modelling and
visualization projects.
MSc. Juha Kortelainen earned his MSc. degree in
mechanical engineering from Helsinki University of
Technology in 1995. He has been working as a research
scientist at Helsinki University of Technology and the last
eight years as a research scientist and senior research
scientist at VTT Technical Research Centre of Finland. The
main research focus of MSc. Kortelainen has been the
application of multibody system dynamics for working
machines and computational fluid dynamics of internal
combustion engines. His current research interest includes
application of semantic data model for modelling data
management and integration.
Prof. Tommi Karhela
is a Research Professor and leader
of the Computer Simulation Technology research team at
the System Research knowledge center of VTT. Karhela
joined VTT 1996 and received his Ph.D. (Tech) degree at
Helsinki University of Technology in 2002. Doctoral thesis
was titled “A Software Architecture for Configuration and
Usage of Process Simulation Models”. Prior to joining VTT,
Karhela has worked short periods at CERN Geneva,
Technology Center of Sandvik-Tamrock and Theoretical
Research Center of Physics at Helsinki University.
Karhela’s current research interests include process
simulation, semantic data modelling and software
architectures.
Figure 4. Visualization of geometry of a plant tank.
Figure 5. Mesh and surface plot of temperatures in the tank.
Figure 6. 3D cut plot visualization of pressures inside a
pipe.
Figure 7. Stream lines visualization in a pipe.
