Khamsin Virtual Racecar Challenge 2015

[MadMatt]

Just for the sake of it I ran one simulation in unsteady state, and results are almost identical with steady state mode (+3% in unsteady state).

The same for me if I use a quite big timestep. Wich timestep value have you set?

I've used 0.01s (with 20 steps per time iteration) which I agree is not very small but then you have to think if the end result is worth reducing that more!

[julien.decharentenay]

Hi, I'm postprocessing the simultion done with OCCFD even if the html report has not been generated. Is there a way to extract total DF and DRAG (and COP) from Paraview?

You can extract the DF and DRAG from ParaView - you may find some information at http://www.cfd-online.com/Forums/openfo ... ssure.html.

However, there is slightly easier way: openFoam generates a file called "forces.dat" that stores the DF and DRAG as well as the moments relative to either the front wheels axis (or the (0, 0, 0) if the wheels are not recognised). The file should be in a sub-directory of the directory where analysis.foam is located.

[CAEdevice]

Hi, I'm postprocessing the simultion done with OCCFD even if the html report has not been generated. Is there a way to extract total DF and DRAG (and COP) from Paraview?

You can extract the DF and DRAG from ParaView - you may find some information at http://www.cfd-online.com/Forums/openfo ... ssure.html.

However, there is slightly easier way: openFoam generates a file called "forces.dat" that stores the DF and DRAG as well as the moments relative to either the front wheels axis (or the (0, 0, 0) if the wheels are not recognised). The file should be in a sub-directory of the directory where analysis.foam is located.

Thank you Julien! This will help very much!

[julien.decharentenay]

@CAEdevice Can you run the command below?

- Open a command prompt
- go to the directory where the images are stored
- type the command

Code: Select all

convert frontalArea1.bmp -format %c histogram:info:

You may have to specify the full path to convert as shown below alongside the output in my case:

Code: Select all

"c:\ImageMagick-6.8.6-Q16\convert.exe" frontalArea1.bmp -format %c histogram:info:
    507029: (  0,  0,  0) #000000 black
  12292971: (255,255,255) #FFFFFF white
convert.exe: length and filesize do not match `frontalArea1.bmp' @ warning/bmp.c
/ReadBMPImage/820.

Can you let me know which folder is ImageMagick installed in your case?

[cdsavage]

Now that a few of you are into the design process, what does everybody think about the size of the upper cockpit template? In my opinion it's fairly realistically sized, and I would be reluctant to make big changes this late in the process, but I am interested to hear if you think we should make any changes. It was based on the 2014 LMP1 template, but reduced in size and height.

[CAEdevice]

@Julien: I let you know about ImageMagick as soon as be in office next Monday

@Chris: the cockpit is realistic in my opinion: no need to do any change.
I would appreciate a smaller diffuser and bigger cooling inlets (but with shorter external templates), and the substitution of the side impact template with a mininum section rule, but it is only a personal opinion. The whole rulebook looks ok.

[julien.decharentenay]

I managed to get back into rendering - not best in class, but I am using free tools and wanted to get it up and running for #SimulationFriday.



Twitter message is at: https://twitter.com/FormerlyHibou/statu ... 48/photo/1 - please favorite and re-tweet.

[cdsavage]

@Chris: the cockpit is realistic in my opinion: no need to do any change.
I would appreciate a smaller diffuser and bigger cooling inlets (but with shorter external templates), and the substitution of the side impact template with a mininum section rule, but it is only a personal opinion. The whole rulebook looks ok.

The diffuser has been reduced in size already, so I'm reluctant to do this again and force another design change on everybody.

Given the freedom we have on cooling inlet positioning, I think cross-section area rules would be quite difficult to do without significantly cutting down on the design options in the mid-section of the car. For the the cooling inlets themselves, would a slightly increased minimum area require everybody to make major changes, or could this be done without too much disruption?

[MadMatt]

I think the rulebook is ok now, I don't mind increasing cooling inlet size but I don't see a reason for that.

Julien, nice rendering, did you have time to look into the mesh issue I mentioned?

The cockpit template reduction Chris was mentioning is regarding the flow from the engine intake scoop to the powertrain template:



I was suggesting that we could cut a bit the cockpit template at the rear to make it flush with the powertrain template, at least up to a 200mm distance from the symmetry plane in the X axis.

Happy Sunday KVRC peps!

[CAEdevice]

@Chris: I agree it's a bit late for any change. The only thing I'd like to change at any time (but I'm influenced from my design) is the side impact template rule. An alternative minimum external section rule would allow more design ideas, or at least the possibility to move up (50-100mm) the lower profile too. Also slightly shorter profiles would be ok.

[julien.decharentenay]

Julien, nice rendering, did you have time to look into the mesh issue I mentioned?

Just starting - Chris prepared a model for me that I can use for testing (a derivation of the one shown on the left in the rendering). I should be able to get back to you on this mid-next week. My computational resources (and time) are much less than what Matteo has...

[CAEdevice]

Julien, nice rendering, did you have time to look into the mesh issue I mentioned?

Just starting - Chris prepared a model for me that I can use for testing (a derivation of the one shown on the left in the rendering). I should be able to get back to you on this mid-next week. My computational resources (and time) are much less than what Matteo has...

Wich Matteo?
About CPU power if I can help you (and KVRC) somehow, just let me know.
About time... I should ask my little children and my rushed and angry customers

Seriously: is there any number about the test car?

[MadMatt]

I am happy to help by running simulations or else if I can, I still have some spare time at the moment, starting a new job next week, but always happy to help. I already made my design, I won't have time to do a new one so no problem, just let me know Julien, or anybody else that need simulations to be run!

[julien.decharentenay]

Thanks a lot for the offer. I will keep with my good old amazon aws - it is working fine so far (although not very fast).

About time... I should ask my little children and my rushed and angry customers

Sorry - my comment was that you had more computational power than I do. Simulations typically takes me a lot longer (in clock time). I am in no position to make a comment on your time...

Seriously: is there any number about the test car?

Can you clarify? Are you referring to a benchmark car? I know I mentioned something like this before, but haven't followed through. Mainly because it looks like you guys are able to use OCCFD. Let me (and Chris) know if a reference car is needed with accompanying CAD model and results and we will sort something out.

[cdsavage]

I mentioned a while ago that I would be happy to release a copy of the test car, I had forgotten about this. This car was built to the old diffuser dimensions though, so it won't be particularly useful for making comparisons based on CFD results.