[KVRC] Khamsin Virtual Racecar Challenge 2016

[RicME85]

...

Julien, where are you at with AWS?

[etsmc]

@CAEdevice Nice car and I love the rendering details. Well done.

On my quest to get a parallel setup to work for someone else than me (and giving you some more tools), I have done the following two things:

- OCCFD v1.1 patch 008 - I added a parallel test option to the KVRC Testing - so that you don't have to wait a while to test if the parallel simulation is working. I also added a Light Single Core version. Download link http://www.khamsinvirtualracecarchallen ... .1-008.zip

- i made a small test case to run in parallel. Download it from https://drive.google.com/file/d/0B7FzSB ... sp=sharing and extract it. Open a command prompt and try running in parallel using the command
mpiexec -n 2 simpleFoam.exe -parallel

You can try variations: calling mpiexec with the path, or trying within cygwin and outside of cygwin. For my purpose it is important that it works outside of cygwin. In cygwin, use the command "which" to find where the executable is located (for example which mpiexec) and the command "where" in the windows command prompt.

It would be great if you can feedback what works and doesn't work.

I have just started a Light Single core test and will update on how it goes.

[RicME85]

So this is the car that I will most likely be submitting for the test race.

[etsmc]

Well, the light single core test failed on simulation 0 with divergence detected so decided to try the local single core test and that has also failed with divergence detected.
this is the same STL file that I was using previously with the multicore test and that got further into the simulation 1

[RicME85]

I too tried a Light Single Core run and got a failure early doors.

One-click CFD Calculator for Vehicle
Version 1.1 (008)
Copyright (C) 2014-2016 Julien de Charentenay, PhD
Distributed by Khamsin Virtual Racecar Challenge (aka KVRC) - http://www.khamsinvirtualracecarchallenge.com
Simulation starting time 2016-03-16 00:04:24 +0000
===========================================================================================
openfoam distribution: C:/OpenFOAM/cygwin64/opt/OpenFOAM
Tool pvbatch: C:/Program Files/ParaView 4.4.0/bin/pvbatch.EXE
Tool meshlabserver: C:/Program Files/VCG/MeshLab/meshlabserver.EXE
Tool convert: C:\Program Files\ImageMagick-6.9.3-Q16/convert.EXE
Tool gnuplot: C:\Program Files\gnuplot\bin/gnuplot.EXE
Tool mpiexec.exe: C:\Program Files\Microsoft MPI\Bin\mpiexec.exe or C:/Program Files/ParaView 4.4.0/bin/mpiexec.exe
Starting simulation for file:
C:/Users/ricme/Desktop/Test 1/OCCFD/body.stl
C:/Users/ricme/Desktop/Test 1/OCCFD/cooling-exhaust.stl
C:/Users/ricme/Desktop/Test 1/OCCFD/cooling-intake.stl
C:/Users/ricme/Desktop/Test 1/OCCFD/engine-exhaust.stl
C:/Users/ricme/Desktop/Test 1/OCCFD/engine-intake.stl
C:/Users/ricme/Desktop/Test 1/OCCFD/wheel_front_left.stl
C:/Users/ricme/Desktop/Test 1/OCCFD/wheel_front_right.stl
C:/Users/ricme/Desktop/Test 1/OCCFD/wheel_rear_left.stl
C:/Users/ricme/Desktop/Test 1/OCCFD/wheel_rear_right.stl
C:/Users/ricme/Desktop/Test 1/OCCFD/wing-front.stl
C:/Users/ricme/Desktop/Test 1/OCCFD/wing-rear.stl
Velocity: 44.7
Ground level: -0.0445
Using light simulation parameters
Simulating full car
Force simulation in sequential mode (ie on only 1 processor even if more are available).
Check file and solid names for unsupported characters.
Supported characters are letters, numbers, "-", "_".
File and solid names checked. No issue found.
Assessing physical bounds of the geometry
Physical bounds: [-1.00596,-0.934684,-0.05] to [1.00596,4.001127,0.965341]
Simulation starting time 2016-03-16 00:04:43 +0000
==================== Running simulation 0 ================================
Generate background mesh using blockMesh...
Extract geometry features using surfaceFeatureExtract...
Generate computational mesh using snappyHexMesh...
Running analysis using simpleFoam...
Divergence has been identified. Sorry we were not able to find a solution to your analysis
Error occured with message:
Divergence detected. Analysis aborted....
Backtrace: C:/Users/ricme/AppData/Local/Temp/ocrBBB2.tmp/src/mainOneClickCFDVehicle.rb:974:in `block in simulate'
C:/Users/ricme/AppData/Local/Temp/ocrBBB2.tmp/src/Shared/FoamHandler.rb:295:in `call'
C:/Users/ricme/AppData/Local/Temp/ocrBBB2.tmp/src/Shared/FoamHandler.rb:295:in `block (2 levels) in solve'
C:/Users/ricme/AppData/Local/Temp/ocrBBB2.tmp/src/Shared/RunCommands.rb:140:in `run'
C:/Users/ricme/AppData/Local/Temp/ocrBBB2.tmp/src/Shared/FoamHandler.rb:293:in `block in solve'
C:/Users/ricme/AppData/Local/Temp/ocrBBB2.tmp/src/Shared/FoamHandler.rb:88:in `each'
C:/Users/ricme/AppData/Local/Temp/ocrBBB2.tmp/src/Shared/FoamHandler.rb:88:in `solve'
C:/Users/ricme/AppData/Local/Temp/ocrBBB2.tmp/src/mainOneClickCFDVehicle.rb:944:in `simulate'
C:/Users/ricme/AppData/Local/Temp/ocrBBB2.tmp/src/mainOneClickCFDVehicle.rb:238:in `start'
C:/Users/ricme/AppData/Local/Temp/ocrBBB2.tmp/src/mainOneClickCFDVehicle.rb:1722:in `<main>'
Press enter to close the window

[etsmc]

Looks very clean and a great start for development.
the ridges over the front wheel pod, are they to help combat lift that seems to be generated from the top of the pods?

the colour scheme makes it look like what Ferrari would produce. i like it

So this is the car that I will most likely be submitting for the test race.

http://s21.postimg.org/d2vza6nh3/car.jpg

[RicME85]

Its not a ridge, its just poor modelling with a not so good render.
I had surgery 3 weeks ago and am at the hospital in a few hours for assessment to ahead of another surgery (this one is pretty major, will be having my elbow replacement removed and a new one fitted) so rushes to get a car complete that I can try and refine if my arm lets me use my PC post surgery.
I took the rear half of the front wheel pod that came in the intro class pack and cut it in a few places to get what I wanted from it (I don't like the open side nor the shape of the top where it has a large opening but I do like the trailing edge and the general curves that fit with the supplied sidepod. I kept half of the rounded section in the wheel well to have a smooth transition for the wheel wake to flow out of, the supplied version is rounded to the outside of the car and also the inside so was a bit of a cut and shut job. Will try and tidy it before submission if I get chance though.
The only area I have anything trick is in the diffuser where I have a double deflection diffuser with vortex generators between the strakes.

[etsmc]

so, I am not the only one to have saved some time by pillaging and pilfering from the supplied files

I noticed last year that most of the competitors that shared information and even myself that there always seems to be a small amount of lift above the wheel pods. I was wondering if some sort of ridge or I guess small gurney flap may help reduce this lift. hence the question.

good luck with the surgery by the way hope all goes well and a speedy recovery.

[RicME85]

Thanks, Im going to need it!

Yes there was definitely some lift in that region. Last year I was planning on investigating some sort of opening at the top like LMP cars have to see what effect this would have.

I always take something from the supplied file, last year it was the nose/cockpit section, this year its more substantial obviously. After last season I sat and went through lots of images of LMP cars and looked at Mulsanne Corner for inspiration. I had an outline in my head of what I wanted my car to be like, the sidepods that are supplied are not far off what I wanted to do myself so if I get time I will try and rework them to get the look I wanted along with possibly designing a new cockpit and nose before moving on to the wheel pods. Im hoping what I have done at the back of the car will give a decent amount of downforce so that it is just a case of working with AoA on the front.

[julien.decharentenay]

...

Julien, where are you at with AWS?

It should be working out of the box. I rebuilt the Amazon image from scratch using a public one, and made mine public as well. It defaults to openFoam v3.0, but can be made to run to v2.3 if required.

Sorry about the divergence on the light version. Can you send me a copy of the log file?

[RicME85]

OK, will try it later, just wanted to make sure it was working and not having the same issues as the Windows version so that I am not using money or failed runs. You mentioned looking into a cheaper tier, did you get anywhere with that?

Will try and send the log later, didn't sleep last night due to pain then had an 8.10am pre-op assessment at the hospital where I was told I would be having another procedure before the elbow replacement...it's like they don't realise I'm trying my best to produce a fast race car!

[CAEdevice]

I finally was able to complete simulation with OCCFD (single core).

I received strange results about the inlet/outlet surface areas (opt1):

Code: Select all

Cooling intake, Area: 0.103m2 - Not compliant !!!
Cooling exhaust, Area: 0.076m2 - Not compliant !!!

Why are they not compliant? The rulebook says that there values are ok, am I wrong?

Hi, anyone of the staff could help me with this report? I have the suspect that OCCFD uses the old rules, with larger inlet/outlet surfaces.

[julien.decharentenay]

I finally was able to complete simulation with OCCFD (single core).

I received strange results about the inlet/outlet surface areas (opt1):

Code: Select all

Cooling intake, Area: 0.103m2 - Not compliant !!!
Cooling exhaust, Area: 0.076m2 - Not compliant !!!

Why are they not compliant? The rulebook says that there values are ok, am I wrong?

Hi, anyone of the staff could help me with this report? I have the suspect that OCCFD uses the old rules, with larger inlet/outlet surfaces.

Sorry, for this. OCCFD uses the old rules (0.12m2 for intake and 0.08 for the exhaust) - so ignore the message. The compliance is manually checked by Chris and his expertise is the only one that count...

[CAEdevice]

Finally... a high dwownforce rear wing.



PS: Is Mantium still hiding?

[MadMatt]

Love the car, guys, RicME good job, I can really see a big level improvement compared to beginning of last year, kudos. Should help others see that with work and dedication you move forward! Can't wait to see the other designs!