[MVRC] Mantium Virtual Racecar Challenge 2017

[rjsa]

Very good looking car. I hope you did not forget the air intake.

I have uploaded a new version of the software. As I hope it now concludes the teething issues it is time to think about new features. The top three on my list are, streamline visualizations, forces per part and adding AWS to the HPC tab, you might have noticed. As rsja is trying to use MantiumFlow on Windows to create a Linux case, this feature would be at least useful to him but probably also to others and that is what that tab is made for.
Does anyone know if we can run a poll on this forum? If someone has a good recommendation for a Wordpress plugin we can also do it on the Challenge page and decide there what comes next.

The bottom blue block below the page appears when comosing. In thread creation mode it has a 'Poll' tab besides the 'Options' one.

On the intake you only have to be above 0 or am I wrong? Try using a negative value. Machin is currently updating the software, it will soon as work for the intro class cars.
How did you solve the buffeting issue? Was it the car or the software? I have not had time to look into it yet.

OK. So intake above 0 and exhaust below 0. Is this bound to change and the 4.5 value be enforced?

I think it was the car: replaced my crazy made up intake channel with a proper NACA duct and voilà.

[rjsa]

Are we supposed to use MWFlow_PP_lin or MantiumWFlow_ppCase_linux?

And what are the dependencies?

ubuntu@ip-172-31-53-116:~/OpenFoam/run/MantiumWFlow_MVRC_linux$ ./MantiumWFlow_ppCase_linux
Traceback (most recent call last):
File "/home/andre/Documents/Projects/MWFlow/mantiumWFlow_MVRC/MantiumWFlow_ppCase_lin.py", line 1, in <module>
ImportError: No module named matplotlib.pyplot
ubuntu@ip-172-31-53-116:~/OpenFoam/run/MantiumWFlow_MVRC_linux$ ./MWFlow_PP_lin
Traceback (most recent call last):
File "/home/andre/Documents/Projects/MWFlow/mantiumWFlow_PyInstallertests/MWFlow_PP_m.py", line 24, in <module>
ImportError: No module named vtk

[LVDH]

Are we supposed to use MWFlow_PP_lin or MantiumWFlow_ppCase_linux?

And what are the dependencies?

ubuntu@ip-172-31-53-116:~/OpenFoam/run/MantiumWFlow_MVRC_linux$ ./MantiumWFlow_ppCase_linux
Traceback (most recent call last):
File "/home/andre/Documents/Projects/MWFlow/mantiumWFlow_MVRC/MantiumWFlow_ppCase_lin.py", line 1, in <module>
ImportError: No module named matplotlib.pyplot
ubuntu@ip-172-31-53-116:~/OpenFoam/run/MantiumWFlow_MVRC_linux$ ./MWFlow_PP_lin
Traceback (most recent call last):
File "/home/andre/Documents/Projects/MWFlow/mantiumWFlow_PyInstallertests/MWFlow_PP_m.py", line 24, in <module>
ImportError: No module named vtk

MantiumWFlow_ppCase_linux is from last year, you can stop using it.
Unfortunately MWFlow_PP_lin does not yet support off-screen-rendering. So you currently still have to pack up the post-processing folder and the solver.log and return to your computer to post-process your case.
The error messages are because you are using the old version. The new one is packaged so you do not need to install anything (not Linux style as you can see when comparing the file sizes).

[Alonso Fan]

Here goes nothing:

http://i.imgur.com/JYlhvzE.jpg
http://i.imgur.com/BaOk6KJ.jpg

Still on development obviously but this is the base im working on atm.

That's neat!

[Ft5fTL]

Very good looking car. I hope you did not forget the air intake.

Same concept as last year.

Here goes nothing:

http://i.imgur.com/JYlhvzE.jpg
http://i.imgur.com/BaOk6KJ.jpg

Still on development obviously but this is the base im working on atm.

That's neat!

Thanks. Sadly i couldnt quite manage to add some sponsorships for render like yours.

[rjsa]

In Linux that line is necessary. Are you creating it in Windows with the ¨hack¨ I suggested a while ago?

Here's what I get editing the default_settings.ini.

local_OS=Linux
numCPUs=6

###### OF Settings ##########
local_OF_vers=v1606+
local_OF_location=/opt/OpenFoam

#!/bin/bash


. /opt/OpenFoam\OpenFOAM-v1606+\etc\bashrc
nCPUs=12
CASE="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
#rm -rf proc*

blockMesh -case $CASE
decomposePar -force -case $CASE
mpirun -np $nCPUs snappyHexMesh -case $CASE -parallel -overwrite 2>&1 | tee $CASE/mesh.log
mpirun -np $nCPUs renumberMesh -case $CASE -constant -overwrite -parallel 2>&1 | tee $CASE/renumMesh.log
mpirun -np $nCPUs checkMesh -case $CASE -parallel -constant 2>&1 | tee $CASE/checkmesh.log

for procDir in $CASE/proc*; do
mkdir $procDir/0
cp $CASE/0/* $procDir/0
done

mpirun -np $nCPUs potentialFoam -case $CASE -noFunctionObjects -writep -initialiseUBCs -parallel 2>&1 | tee $CASE/potentialFoam.log
mpirun -np $nCPUs simpleFoam -case $CASE -parallel 2>&1 | tee $CASE/solver.log

mpirun -np $nCPUs execFlowFunctionObjects -case $CASE -dict system/controlDict2 -latestTime -parallel 2>&1 | tee $CASE/flowFunctionObjects.log

mpirun -np $nCPUs sample -case $CASE -latestTime -noFunctionObjects -parallel 2>&1 | tee $CASE/sample.log
touch $CASE/case.foam

So I'm editing the .sh file and fixing the line like this:

. /opt/OpenFOAM/OpenFOAM-v1606+/etc/bashrc

It isn't a problem when I'm on the iMac/OSX because I run the cases with scripts started interactively from within a docker session.

Became a headache in AWS when I want to run cases from the host prompt using "docker exec" to be able to shut the system (and the $$$ charging) down after the processing is done.

[etsmc]

Well after trying to develop last season's car during the offseason I came to the conclusion of needing to move to a more simple layout.
So the last week I have spent an hour a day and created the new MercuryMotorsport MM17-02
Still a little development to go official CFD testing should be starting soon.

[Ft5fTL]

Doesnt that front fender design add some unnecessary drag?

[etsmc]

surprisingly they seem to produce very minimal lift across the top of the fender but down the side near the front wheel arch they do create drag so this is one of the areas I am developing some more. Current Cd is only 0.7 so if I can trim some drag from the front and the downforce is at an acceptable level I am hopeful.

Again the aim of this new car is to keep it simple and work up from there.

My goal for this season is to move up the standing to 5th, 4th would be nice but I think the same contenders will be fighting for the top 4 spots again.

[Ft5fTL]

Well there is no surprise on the Cd side. I'm adding winglets litteraly everywhere atm and getting around 1.15-1.20 Cd.

[LVDH]

It has gotten very quiet here, is nobody working or is the software doing its job and nobody has to complain?
BTW 90% of the images can now be produced in off-screen mode. I am only missing the plots.

[rjsa]

I'm taking a beating trying to bring the COP back into target. And HX flow also quite stubborn.

And, of course, having the simulation times double did cramp my style

[etsmc]

Can someone please confirm the ride height used for CFD testing. I am guessing something around 40mm-50mm as the reference plane to the lowest point of the tyre is ~58mm so there has to be some squat in the tyre. Looked through the rules but couldn't see anything to state what was being used tor testing.

[LVDH]

The floor is at Z=-0.0455mm. Does this help?

I'm taking a beating trying to bring the COP back into target. And HX flow also quite stubborn.

And, of course, having the simulation times double did cramp my style

I am running experiments on your car as it is not very stable. It is possible that there will be modifications with the numeric settings. These might slightly increase speed. Mostly the run times are larger because of the generally more accurate simulations we are running this year.

[rjsa]

The floor is at Z=-0.0455mm. Does this help?

I'm taking a beating trying to bring the COP back into target. And HX flow also quite stubborn.

And, of course, having the simulation times double did cramp my style

I am running experiments on your car as it is not very stable. It is possible that there will be modifications with the numeric settings. These might slightly increase speed. Mostly the run times are larger because of the generally more accurate simulations we are running this year.

It's behaving way better with a new intake duct.