Sunday 4 March 2012

README first

News - From June 2010, I started to spend money on Google AdWords to maximise my effort to share experiences by writing this blog.

Welcome to my blog on CFD, numerical modelling and parallel programming. Though relevant things are included, the CFD part of this blog focuses on the packages SALOME, Code_Saturne and ParaView etc. The complete content can be directed by the Labels list on the right side.


All the posts here are kept updated. Following or subscribing a blog is a good idea if it really helps. Up to March 2012, as reported by Google, there are 90 friends who are following or subscribing my posts using Google.

The most popular posts are summarised as follows

1. Installation of SALOME on Ubuntu

The latest SALOME version is 6.3.1. However I haven't tried this version; I am still using the 5.1.x.

For 5.1.x the latest version is 5.1.6. The complete instruction, "Installation of SALOME 5.1.3 on Ubuntu 10.04 (64 bit)", should help when install SALOME 5.1.x onto Ubuntu; at least according to feedbacks, this works up to 5.1.4 onto Ubuntu 10.04 (32/64 bit). The tips also suit Ubuntu 9.10, if all packages are updated in the system.

For Ubuntu 9.10, an older post is "Installation of SALOME 5.1.2 on (K)Ubuntu 9.10 64 bit". The notes might be slightly outdated.

If preferring Ubuntu 9.04, it is also easy to install SALOME by only following "Installation of SALOME 5.1.1 on Ubuntu 9.04". The post actually helps up to the latest version 5.1.3.

For those who still use the old 4.* versions of SALOME, "Polishing GUI of older SALOME versions" is recommended to read.

2. SALOME tutorials

"CFD tutorial: laminar flow along a 2D channel - Part I" is a brief tutorial on using SALOME to do the geometry modelling and meshing.

More examples are still on the way.

3. Installation of Code_Saturne on Ubuntu

The latest Code_Saturne 2.1.0 is much easier to compile and install than ever. An instruction on this version is on the way. By the way I also successfully ported this version onto Windows 7.

Since version 2.0 Code_Saturne started to adopt the typical source build and install procedure on Linux. By following "Installation of Code_Saturne 2.0-rc1 on Ubuntu 10.04 (64 bit)" I am confident it would be an easy job for you. In the meantime, if you are also a SALOME user, I strongly recommend you to read "Compile Code_Saturne with SALOME binary libraries" in order to reuse the binary libraries already shipped with SALOME.

Two slightly outdated but helpful posts are "Installation of Code_Saturne 2.0.0 on Ubuntu 9.04" and "Installation of Code_Saturne 2.0-rc1 on Ubuntu 9.10".

You can also try to test installing Code_Saturne with apt-get. Please read "apt-get, the ideal way to install the software onto Ubuntu".

The instruction on installation of the last stable version, 1.3.3, was written in "Installation of Code_Saturne 1.3.3 on Ubuntu 9.04". On the other hand, for the version 1.4.0 the installation procedure is quite similar to that of 1.3.3, except 1.4.0 turned to use PyQt4 and Qt4. Please check "Installation of Code_Saturne 1.4.0 on Ubuntu 9.04" to see how to use (Py)Qt4 instead of the python-tk and tix.

4. Code_Saturne tutorials

"CFD tutorial: laminar flow along a 2D channel - Part II" gives a brief tutorial on how to apply Code_Saturne to solve a CFD problem.

More examples are still on the way.

5. Installation and usage of ParaView

Installation of ParaView is relatively easier and please read "Installation of ParaView 3.6.1 on Ubuntu 9.04". The usage of ParaView is briefly illustrated by an example in the article "Post-processing with ParaView".

For the latest version of ParaView, please refer to its official site.

6. GPGPU acceleration with CUDA techniques

A basic idea of using NVIDIA CUDA and the template library thrust is shown in "A simple test on the efficiency of thrust". An example to use CUDA practically is "Apply CUDA to solve a 1D heat transfer problem".

Last but not least, any comments are sincerely welcome and your support is always my motivation. The visitors here are from all over the world. See the visits map overlay below, which is provided by Google Analytics for the half-a-year period from January 2010.