[MVRC] Mantium Virtual Racecar Challenge 2016

[Alonso Fan]

Hi guys,

together with Alex from HP-Racing we found and solved another small Windows related bug. The runCase.sh and the one to check the stl files will not work under Windows. As a workaround for now you can just manually execute every line.
I have already fixed it. As soon as Alex can confirm that the whole case runs I will upload the hopefully last update to get every thing working.

So far nobody has sent any log files. At least the few Linux users should have had some success by now. It would be very helpful if you could send the log files, just so I can have a look and be sure that everything is working as intended.

where would you like them sent?

[CAEdevice]

Thank you (and thanks to HP team).

I am not an expert with Linux, but I am not enthusiastic about the Windows porting of OF+, so I was wondering, considering that we have to do with a VirtualBox machine in any case, if it would not be simpler to provide a Linux virtual machine (based one of the light Ubuntu releases) with all the Mantium WFlow software pre inatalled (and other tools: Paraview, FreeCad, STL utilities). I mean something like the old CAElinux project.

Anyway, tomorrow I will test Wflow on Win again (@LDVH: is the bug you found related to the messages that I reported above?)

Finally: would it be possible to have a list of the commands to run manually? I guess they are included into runCase.sh.

[Ft5fTL]

Well i managed to run the test case but not the Mantiumwflow.

local_OF_location=C:\Program Files\ESI\OpenFOAM\1606\Windows\
local_OF_vers=v1606+
OS=Windows

Isnt this right?

[LVDH]

Hi guys,

together with Alex from HP-Racing we found and solved another small Windows related bug. The runCase.sh and the one to check the stl files will not work under Windows. As a workaround for now you can just manually execute every line.
I have already fixed it. As soon as Alex can confirm that the whole case runs I will upload the hopefully last update to get every thing working.

So far nobody has sent any log files. At least the few Linux users should have had some success by now. It would be very helpful if you could send the log files, just so I can have a look and be sure that everything is working as intended.

where would you like them sent?

When you start MantiumWFlow there is some nice ascii art, under that you find an email address.

Thank you (and thanks to HP team).

I am not an expert with Linux, but I am not enthusiastic about the Windows porting of OF+, so I was wondering, considering that we have to do with a VirtualBox machine in any case, if it would not be simpler to provide a Linux virtual machine (based one of the light Ubuntu releases) with all the Mantium WFlow software pre inatalled (and other tools: Paraview, FreeCad, STL utilities). I mean something like the old CAElinux project.

Actually this is a cool idea. Right now it looks as if the issues with the Windows OF+ version are manageable.
Also I see that issues are that people do not know how to get from folder to folder using commands.

Anyway, tomorrow I will test Wflow on Win again (@LDVH: is the bug you found related to the messages that I reported above?)

No, it is more like the runCase.sh file is not executable. This is technically not what is going on but that is how it appears to the users. So in case you want to get running before I release the newest update you can just have a look at the file and run it manually command by command.

Finally: would it be possible to have a list of the commands to run manually? I guess they are included into runCase.sh.

Well as soon as the case is written you get two executable files. Just run them both. And once more, it is written in the readme file that ships with MantiumWFlow.

Well i managed to run the test case but not the Mantiumwflow.

local_OF_location=C:\Program Files\ESI\OpenFOAM\1606\Windows\
local_OF_vers=v1606+
OS=Windows

Isnt this right?

Looks good. Where is the problem? You have to open a command window in Windows, go to the location of MantiumWFlow and type:

Code: Select all

MantiumWFlow_MVRCed_win.exe <path_to_your_case>

[Ft5fTL]

Getting this.

[SR71]

you joined a virtual challenge that didnt know when to say enough is enough....

we could have continued to grow, but we wanted "more cooling"....

Why are you soo salty about the cooling dude? Race cars needs cooling, and the competition is about race cars.

I agree with Ft5fTL. Cooling simulation is more important than (for example) wheels rotation in a race car.

Is you car complete? I remember some stunning renderings, but I am convinced that my car (not the fastest one), despite being compliant with cooling requirements, is faster than yours without cooling devices.

All I'm saying is that you guys crossed the acceptable simulation based threshold for a free competition without even a blink.

When was the last race?

When is the next? I'll prepare my 2018 car given this timeline.

This thing should be simple and results should be quick.

[etsmc]

just been to the website to get the latest MantuimWFlow and there is no download button? I am logged in but can't access what I need.
also, could someone tell me what way I need to rotate my car from the old OCCFD to MantiumWFlow?
say I was the driver sitting in the car looking forward doi turn it left/right and by how much? if I have read it correctly I believe it to be 90 degrees to the left

[CAEdevice]

Also I see that issues are that people do not know how to get from folder to folder using commands.

I am among them who have never used a Unix command before (and only a few time Dos or Linux) but the most frequent issue is that I am not sure about the starting directory (es. Sometimes the OF shell creates a "username" folder, sometimes not). Anyway, in my opinion, the weak point is OF+ porting, not WFlow.

[LVDH]

Getting this.
http://i.imgur.com/S8hHHbb.jpg
http://i.imgur.com/KfBaeoe.jpg?1

That was the odd bug from yesterday. You can now download MantiumWFlow for Windows again. It will produce a new runCase.sh file. It looks absolutely the same but works.
How I deacitaved the download button is another question but that is back up again.
I really hope this now concludes the teething issues.

Also I see that issues are that people do not know how to get from folder to folder using commands.

I am among them who have never used a Unix command before (and only a few time Dos or Linux) but the most frequent issue is that I am not sure about the starting directory (es. Sometimes the OF shell creates a "username" folder, sometimes not). Anyway, in my opinion, the weak point is OF+ porting, not WFlow.

I agree Windows is not the best platform for CFD. However most participants have entered Windows as their standard OS so we needed a good solution. I asked people to try out OF+ for Windows ahead of time. There was only little response.
But OF+ does offer some cool features. It should produce the same results as OF does on Linux. Also I want to use the OF+ post-processing utilities soon:
http://openfoam.com/version-v3.0+/post- ... Extraction
This will mean you get a cool report like the full edition of MantiumWFlow creates but do not have to install a single additional piece of software. CAE and CFD never work super smoothly and always need a user to do some magic. But if we only need OF+ and my tool we are down to only two pieces of software which should make things easy.

[CAEdevice]

I successfully launched the simulation: the mesher worked well (multicore! =D> ), but then I received this error:

Code: Select all

No MRF models present

Creating finite volume options from "system/fvOptions"

Selecting finite volume options model type explicitPorositySource
    Source: porosity_heat_exchanger_dx
    - selecting cells using cellZone heat_exchanger_dx
    - selected 39741 cell(s) with volume 0.0187211
Porosity region porosity_heat_exchanger_dx:
    selecting model: DarcyForchheimer
    creating porous zone: heat_exchanger_dx

Starting time loop

forceCoeffs forceCoeffs:
--> FOAM Warning :
    From function Foam::labelHashSet Foam::polyBoundaryMesh::patchSet(const Foam::UList<Foam::wordRe>&, bool, bool) const
    in file meshes/polyMesh/polyBoundaryMesh/polyBoundaryMesh.C at line 782
    Cannot find any patch or group names matching "body.*"
--> FOAM Warning :
    From function Foam::labelHashSet Foam::polyBoundaryMesh::patchSet(const Foam::UList<Foam::wordRe>&, bool, bool) const
    in file meshes/polyMesh/polyBoundaryMesh/polyBoundaryMesh.C at line 782
    Cannot find any patch or group names matching "rear_wheel_SX.*"
--> FOAM Warning :
    From function Foam::labelHashSet Foam::polyBoundaryMesh::patchSet(const Foam::UList<Foam::wordRe>&, bool, bool) const
    in file meshes/polyMesh/polyBoundaryMesh/polyBoundaryMesh.C at line 782
    Cannot find any patch or group names matching "Front_wheel_SX.*"
    Including porosity effects

[2]
[2]
[2] --> FOAM FATAL ERROR:
[2] faceSource faceSource_engine_intake: patch(engine_intake):
    Unknown patch name: engine_intake. Valid patch names are:
11
(
auto_wt_side
auto_wt_sym_plane
auto_wt_inlet
auto_wt_outlet
auto_wt_floor
auto_wt_top
rear_wheel_DX
Front_wheel_DX
procBoundary2to0
procBoundary2to1
procBoundary2to3
)

[2]
[2]
[2]     From function void Foam::fieldValues::faceSource::setPatchFaces()
[2]     in file fieldValues/faceSource/faceSource.C at line 178.
[2]
FOAM parallel run exiting
[2]
[0] --------------------------------------------------------------------------
MPI_ABORT was invoked on rank 2 in communicator MPI_COMM_WORLD
with errorcode 1.

NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
You may or may not see output from other processes, depending on
exactly when Open MPI kills them.
--------------------------------------------------------------------------

[0]
[0] --> FOAM FATAL ERROR:
[0] faceSource faceSource_engine_intake: patch(engine_intake):
    Unknown patch name: engine_intake. Valid patch names are:
10
(
auto_wt_side
auto_wt_sym_plane
auto_wt_inlet
auto_wt_outlet
auto_wt_floor
auto_wt_top
rear_wheel_DX
Front_wheel_DX
procBoundary0to1
procBoundary0to2
)

[0]
[0]
[0]     From function void Foam::fieldValues::faceSource::setPatchFaces()
[0]     in file fieldValues/faceSource/faceSource.C at line 178.
[0]
FOAM parallel run exiting
[0]
[1]
[1]
[1] --> FOAM FATAL ERROR:
[1] faceSource faceSource_engine_intake: patch(engine_intake):
    Unknown patch name: engine_intake. Valid patch names are:
11
(
auto_wt_side
auto_wt_sym_plane
auto_wt_inlet
auto_wt_outlet
auto_wt_floor
auto_wt_top
rear_wheel_DX
Front_wheel_DX
procBoundary1to0
procBoundary1to2
procBoundary1to3
)

[1]
[1]
[1]     From function void Foam::fieldValues::faceSource::setPatchFaces()
[1]     in file fieldValues/faceSource/faceSource.C at line 178.
[1]
FOAM parallel run exiting
[1]
[default:00473] 2 more processes have sent help message help-mpi-api.txt / mpi-abort
[default:00473] Set MCA parameter "orte_base_help_aggregate" to 0 to see all help / error messages
[ofuser@default workingDir]$

Here you can find the log(s):
https://dl.dropboxusercontent.com/u/522 ... 60822A.zip

[LVDH]

I overlooked someone else's question. I think it affects you as well. Your car is not in the correct position. That I forgot to include that check into the MWFlow has been bothering me for a while now. Why it produces such strange errors, I do not understand but I guess it is better than wasting the time with a useless simulation.

I hope the image helps.

In the output of MFWlow, you should find something like this:

Code: Select all

('max BB dimensions: ', [-1.012876, 4.0, -1.00597, 1.00597, -0.05, 1.014646])

I guess your Bounding Box will look a bit different. I will implement a check to make sure that the cars are in a reasonable box and if not a warning is displayed.

[CAEdevice]

I overlooked someone else's question. I think it affects you as well. Your car is not in the correct position. That I forgot to include that check into the MWFlow has been bothering me for a while now. Why it produces such strange errors, I do not understand but I guess it is better than wasting the time with a useless simulation.
I will make a small screenshot to show how the car should be oriented.

Are you sure it is a position issue? The car has been modeled "over" the templates!

I can send you the STLs if you think it may be useful.

[LVDH]

I just edited the post above.
Maybe you can have a look if you find that line in your output. But from looking at the surface check logs I am quite sure.

[CAEdevice]

Code: Select all

('max BB dimensions: ', [-1.012876, 4.0, -1.00597, 1.00597, -0.05, 1.014646])

I guess your Bounding Box will look a bit different. I will implement a check to make sure that the cars are in a reasonable box and if not a warning is displayed.

You are right:

Code: Select all

Bounding Box : (-1 -1.005 -0.999986) (1 -0.039994 4)

... but I am not sure how to change it, since I used the template as reference. I will try to rotate the car somehow and I'll test it.

EDIT: It is a bug... but of my CAD software!

[LVDH]

I guess I should have a look at what Chris did, maybe he made a mistake.