Hi, Sorry for the late responses. I have (am still) trying my best to get through the issues. I have updated OCCFD to have a better recognition of the openFoam distribution being used. There are teething issues with the cfd support distribution - sorry to have them so close to the first race.
The main issue at this stage seems to be:
1)
Parallel running: The new patch (007)
http://www.khamsinvirtualracecarchallen ... .1-007.zip has been updated to identify the openFoam distribution based on path location and using a logic based on the identified openFoam distribution.
If using openFoam from cfdsupport: Please install HPC MS-MPI from
https://www.microsoft.com/en-us/downloa ... x?id=36045 with the default installation directory (ie C:\Program Files\Microsoft HPC Pack 2012 ). Ideally you might want to remove openMPI prior to installing MS-MPI.
As an alternative, there is now an option for forcing OCCFD to run on single processor... So at least there is a way around that does not involve cygwin, etc.
2)
Meshing: The mesh is generated using snappyHexMesh. In some cases (usually non watertight geometry), the inside of the car may be meshed. When this happens, it is likely to have the following symptoms: divergences (usually the inside mesh will have a worst mesh quality than the outside and more prone to corners doing strange things) and wrong results (the inside mesh will have some level of pressure that will act on the shell to affect drag and downforce). There are two markers: the mesh size that should be around 4,5 millions cells normally and goes up to around 6 millions cells with the inside mesh and the force distributions that look strange. When this happens, the easier solution is to use a tool to close the holes (I personally use netfabb, but other tools can be used - see mesh cleaning steps in this blog post
http://blog.othermachine.co/convert-stl-to-step)
Last but not least, please note that the Very Stable option has been moved to be the default option.
