OpenFOAM version 1.2 has been released 22/Aug/2005. Major new developments include reorganisation of the pre-processor FoamX for easier customisation, several new multi-phase flow solvers and numerical schemes. Improvements in the topology handling module continues, with a number of pre-packaged dynamic mesh classes.
Announcing big news: OpenFOAM
is now public domain! In order to
satisfy the requirements of the academic CFD community for an
advanced and unified research platform, OpenFOAM is released under
Gnu Public Licence (GPL). Announced on
CFD-Online on 12/Dec/2004, I hope this will
expand the user and development base of
FOAM/OpenFOAM (still
FOAM for me!)
The core technology of OpenFOAM is a flexible set of efficient C++ modules. These are used to build an archive of: solvers, to simulate specific problems in engineering mechanics; utilities, to perform pre- and post-processing tasks ranging from simple data manipulations to visualisation and mesh processing; libraries, to create toolboxes that are accessible to the solvers/utilities, such as libraries of physical models.
For interested parties, a full list of features is given in the Features section of the OpenFOAM web site.
Over the last year, the scope of the OpenFOAM project has grown
massively and beyond my expectations. In my position as the
Technical Director of Nabla Ltd. and
chief architect of OpenFOAM,
I have driven the development effort in order to satisfy both
the needs of commercial clients and the academic community. The
new role is even more challenging and involves contributing to
the project not only by writing new code but also working with
new research groups worldwide to contribute the the project.
The new setup effectively removes the commercial pressure,
making the development work much more fun!
Also, I hope
the number of publications and University collaboration will
significantly expand in the future.
My future involvement will develop along two lines: collaboration with Universities and high-end consultancy based on the capability of OpenFOAM to handle complex physics.
Collaborative project through Wikki Ltd. on modelling of cement castables. The objective of the project is to write a OpenFOAM solver for heat, mass and energy transfer simulating the drying process of pre-fabricated blocks cast from high-aluminium content cement.
Collaboration with the Internal Combustion Engine Group, Dipartimento di Energetica. The objective of the project is to test the topological changes system and implementation and perform moving piston and valve flow simulations of the intake and exhaust stroke. With this functionality, OpenFOAM will be able to re-use KIVA-3V meshes with minimal change.
NSP-ME Project (JEP-19017-2004), financed from the CARDS 2004 Curriculum Development funds of the European Union Programs (NSP) at the St. Cyril and Methodi University in Skopje Macedonia. EU partners include Bureau Veritas, France; CAE Consulting, Belgium; Wikki Ltd, United Kingdom and University of Stuttgart, Germany.
NUSIC Project (JEP-18085-2003), financed from the CARDS 2003 UE programme Curriculum Development funds targets the establishment of higher education curricula in the field of Numerical Simulation Programs (NSP) at the University of Zagreb, Croatia. EU partners include Bureau Veritas, France; CAE Consulting, Belgium; Wikki Ltd, United Kingdom; University of Stuttgart, Germany.
As a part of the project, I will deliver two lecture courses on Numerical simulations of fluid flows for aeronautics applications and Computational Continuum Mechanics over the next two years (2005-2007). The project has been extended for an additional lecture cycle in 2007-2008.
I am proud to announce the completion of doctoral Thesis of Dr. Zeljko Tukovic. Public presentation of the work will be done at the Faculty of Mechanical Engineering and Naval Architecture, University of Zagreb, Croatia on Monday 14/Feb/2005 at noon (looks like the "Students" page is coming soon!). The Thesis is currently available only in Croatian, but an English translation is coming soon after the viva.
Ongoing collaboration with prof. Ivankovic's research group on numerical modelling in non-linear stress analysis, including contact stress simulations, non-linear materials, large deformations and solid-fluid interaction.
Over the past four years OpenFOAM has been developed along commercial lines, with the idea of providing a high-end tool for complex physics simulations and offering consultancy for customers who need help with complex modelling. Unfortunately, it turns out this approach was wrong.
The new phase of OpenFOAM development starts with the end of the commercial approach, recognising that OpenFOAM is essentially a research tool. Therefore, its home is really in the academic community and its future can only be guaranteed through University collaboration.
When I have started doing my PhD at Imperial College in 1993, I have been given an old Fortran 77 single-block structured CFD legacy code to work with. After about 3 weeks of working with the code I was having serious doubts if this is the way to go: my previous education has shown me Fortran in all its glory, together with the messy brains spaghetti programming produces. My biggest concern, however, was the fact that once I finish my PhD I may end up with lots of interesting stuff that no-one can use; the only alternative was to start with something new.
Luckily for me, the project was already under way: Mr. Henry Weller, sitting 3 desks away from me has thrown away all of his Fortran on the basis that it was not good enough to do what he wanted to do and concentrated on getting FOAM (Field Operation and Manipulation) to work. When I joined the project, FOAM has just done its first unsteady laminar flow simulation. Have in mind that in 1993 it was by no means clear if one could write a "production quality" CFD code in a language that was evolving just about as quickly as the code itself. By combination of careful programming, brain power and luck, we got where we are today.
Henry and I have hit it off spectacularly well from the start. Over the next 3 years I have had the opportunity to learn a lot of Finite Volume Numerics, discretisation, solver technology, parallelism and, most of all, numerical modelling from him. He used to impress me by quoting the precise form of various turbulence models over lunch and going in detail over the terms and what they" do". On top of the CFD education I have received, I have learned a lot about software engineering and C++, but also developed intolerance to sloppy work (sometimes slows me down but always pays in the long run).