Khamsin Virtual Racecar challenge 2013 (CFD model racing)

Post here information about your own engineering projects, including but not limited to building your own car or designing a virtual car through CAD.
User avatar
RicME85
52
Joined: 09 Feb 2012, 13:11
Location: Derby

Re: F1 model CFD Challenge (Khamsin Virtual Racecar challeng

Post

I exported to Solidworks as it has an inbuilt export for that and then saved and opened in SU

User avatar
RicME85
52
Joined: 09 Feb 2012, 13:11
Location: Derby

Re: F1 model CFD Challenge (Khamsin Virtual Racecar challeng

Post

Decided to collect all recent useful info into one post for ease of use :)
Hi there. Nick, great job for following this up and keeping the conversation flowing. If you want to contact me, try info@hibouscientificsoftware.com.au or post on http://www.cfd-online.com with [KVRC] I am checking it more frequently than F1Technical.

Re wind tunnel model: please download from: https://dl.dropbox.com/u/40499338/KVRC2 ... 180mph.skp. It has annotations to try to explain how to set things up. Nick, can you supply the wheel/suspension for them to run the simulations?

For the 2012 test run, I used the following workflow:
- Set the car up in a stand-alone model (i.e group the body and the wheels in separate groups, i.e 5 groups) and then group the car in a single group;
- Move the car so that the front wheels are aligned with the red "x" axis with the car facing downwards (in a top view). The lowest point of the wheel is set to 5mm below the ground plane (z=0);
- Open the root model and copy and paste the wind tunnel around the car.
The above workflow requires that the car boundary conditions are setup again, but it is easier than trying to copy and paste the car geometry. The model provided should allow the way described above or to copy and paste the car geometry.

If you want to run preliminary simulation, you can neglect the tyre rotation and just copy and paste the whole car into the body group.

Wind tunnel only:
180mph scenario: https://dl.dropbox.com/u/40499338/root% ... 180mph.skp
100mph scenario: https://dl.dropbox.com/u/40499338/root% ... 100mph.skp
100mph scenario with 3degree yaw: https://dl.dropbox.com/u/40499338/root% ... %20yaw.skp
To run a test to check the configuration, it will be easier to just run a test. Use: https://dl.dropbox.com/u/40499338/Khams ... mplete.skp for example. Load the skp file in SketchUp, activate Khamsin tool and right-click -> Analysis -> go (local).

With the wind-tunnel:
- Open the skp file;
- Paste or import the car - with the provided wind-tunnel skp file, paste the car into the car-body group so that the imported car is part of the car-body group in the wind-tunnel skp file;
- If you are not able to paste the car into the car-body group (but do ensure it is imported as a group), you will need to do the following:

+ Khamsin tool -> right-click -> Define -> Volume
+ left button press (i.e press but do not release) on the group in which your car is imported, this should display a local menu;
+ move the mouse to the entry "Volume 0" - it should get highlighted in red, and release the left button;
+ Do the same and select "Close"
+ right-click -> Define -> Boundary Conditions
+ double left click on the the group in which your car is imported, this should display the boundary condition panel;
+ Change the name to "car" and the mesh size to "0.01"
+ Select "Close" by left button press over the group, etc.

- Then you can run the analysis: right-click -> Analysis -> go (local)
1. swak4Foam: it would be required for the wheels - to account for the rotation of the wheels. If you do not include the wheels, you can just untick the box.

How to install it? I am using the blueCFD distribution, which includes swak4Foam. blueCFD-SingleCore is a free subset of blueCFD, which does not include swak4Foam. I am unsure how it can be added manually... You can purchase blueCFD from blueCAPE (check price with them, but believe it is around Euro70). I will drop them an email to see if they might be willing to add swak4Foam to their blueCFD-SingleCore for the KVRC.

There are a number of alternatives:
- build a port of openFoam and swak4Foam for windows on your own - probably time consuming and frustrating experience, not for the faint hearted;
- install openFoam and swak4Foam on a Linux computer on the same network as your windows box and use right-click->analysis->remote->linux - it is probably the easiest way, but require two computers on the same network;
- use Amazon EC2 (right-click->analysis->remote->Amazon EC2). If you want to do this, let me know as I need your amazon user id to allow access to the image. It is reasonably experimental, and Amazon EC2 will charge you per CPU hour (not expensive though).

2. Individual components: Just place it the same way as you did with the car. You probably want to get the correct height above the ground to correctly account for ground effects, otherwise the position in the domain should not matter much. Just be aware that a component that is optimised by itself may behave differently when included with the car. To optimise run time, you can look at a number of aspects: do you need to run for 1000 iterations? Can you coarsen the mesh resolution?
For post-processing:
- Check the analysis output: right-click>Analysis>Output. Scroll down to the near the end and you should have the forces (in newton) and moment (in newton.meter to the 0,0,0 point). Both have two components: pressure induced forces/moment and viscous (molecular and turbulent) induced forces/moment.
- ParaView: right-click->Analysis>Paraview. In paraview:
+ Change time from 0 to the latest one (click on the up arrow);
+ Click on the apply button (on the left side of the screen);
+ Change the variable (may be called Solid Color) and may have to change the representation to "Surface" (it is a bit on the right of the variable in the toolbar) - on the left of the variable drop down, there are three buttons: rescale, edit scale and show legend (from right to left);
+ Just under the variable and representation in the toolbar, there are a few filters than can be applied: the section one is a good one.
+ Also, right-click (in the pipeline browser) on open and select the "xxxx.foam" file
+ In the object inspector: untick the internal mesh and tick the name containing car or car-body, then click apply. It should only show the car body. Check the pressure on the surface to see area of high and low pressure.

If all fails, have a check at the documentation...

cdsavage
cdsavage
19
Joined: 25 Apr 2010, 13:28

Re: F1 model CFD Challenge (Khamsin Virtual Racecar challeng

Post

Any update on the wheels and suspension?

Markus
Markus
0
Joined: 03 Apr 2012, 19:35

Re: F1 model CFD Challenge (Khamsin Virtual Racecar challeng

Post

Nick I don´t want to be a whiny bitch but I really would like to know when the new pack will be sent out.

astracrazy
astracrazy
31
Joined: 04 Mar 2009, 16:04

Re: F1 model CFD Challenge (Khamsin Virtual Racecar challeng

Post

ye, i know your busy nick. but i need the new pack to check that my model is legal, i've made the changes my self but need to check. + i need wheels and suspension

Markus
Markus
0
Joined: 03 Apr 2012, 19:35

Re: F1 model CFD Challenge (Khamsin Virtual Racecar challeng

Post

astracrazy wrote:ye, i know your busy nick. but i need the new pack to check that my model is legal, i've made the changes my self but need to check. + i need wheels and suspension
He tweeted me yesterday and said that he have had some computer probs lately but that the pack should be sent out on Tuesday.


julien.decharentenay
julien.decharentenay
10
Joined: 02 Jun 2012, 12:31

Re: F1 model CFD Challenge (Khamsin Virtual Racecar challeng

Post

We have put in place a web based submission portal - very rudimentary but should work. We emailed the links and login details - but had an error message when emailing. Can you confirm you received the details? Thanks. Feel free to check if it works.

User avatar
N12ck
11
Joined: 19 Dec 2010, 19:10

Re: F1 model CFD Challenge (Khamsin Virtual Racecar challeng

Post

Hi,

Recently (as Markus knows) my computer has been getting fixed, so I have been unable to view the forums, so all these messages I have not been able to see, I do have a smartphone, so if you have any queries in the future, send an email via the website, I can reply much faster then in case anything happens with computers.

Updates:
Suspension sent out
Wheels sent out
Corrected Tub sent out

Thanks for your patience
Nick
Budding F1 Engineer

Markus
Markus
0
Joined: 03 Apr 2012, 19:35

Re: F1 model CFD Challenge (Khamsin Virtual Racecar challeng

Post

julien.decharentenay wrote:We have put in place a web based submission portal - very rudimentary but should work. We emailed the links and login details - but had an error message when emailing. Can you confirm you received the details? Thanks. Feel free to check if it works.
I haven't received anything.

MisterX
MisterX
0
Joined: 04 Mar 2013, 10:18

Re: F1 model CFD Challenge (Khamsin Virtual Racecar challeng

Post

julien.decharentenay wrote:We have put in place a web based submission portal - very rudimentary but should work. We emailed the links and login details - but had an error message when emailing. Can you confirm you received the details? Thanks. Feel free to check if it works.

Thanks I saw that I listed as a team (italica corse), but I have not received any email regarding the links.

cdsavage
cdsavage
19
Joined: 25 Apr 2010, 13:28

Re: F1 model CFD Challenge (Khamsin Virtual Racecar challeng

Post

Theres still something out by 50mm in the new tub. The most obvious problem is that it doesnt meet 3.11.1 - the outboard parts of the front wing should be between 450mm and 1000mm forward of the front wheel center line, in the file they are only 400mm to 950mm. Im not 100% sure but I think the problem would be fixed if the tub, front wheels, and front part of the floor were moved 50mm rearwards (leaving the front wing and everything behind the rear of the tub where it is).

The wheelbase of the supplied tub is 3450mm, I thought it was meant to be 3400mm. If you moved the tub back by 50mm that would make it 3400mm.

Minor unrelated problem: The box for the rear wing endplate goes from 355mm to 365mm in the file, where it should be twice as wide, ie 355mm to 375mm.

astracrazy
astracrazy
31
Joined: 04 Mar 2009, 16:04

Re: F1 model CFD Challenge (Khamsin Virtual Racecar challeng

Post

ye i noticed the front wing prob, i moved it 100mm back as i guessed it was measured the same as the floor. Going to d.load new tub and check it matches with my model

i hadn't noticed the rear wing problem

trouble is a lot of people have prob made there car now so anythink like moving the tub back is gonna have a major effect. Perhaps we should decide that altho there are one or two issues, we should go by nicks tub guidelines for this season i.e the lastest tub sent out is final measurements to go by?

julien.decharentenay
julien.decharentenay
10
Joined: 02 Jun 2012, 12:31

Re: F1 model CFD Challenge (Khamsin Virtual Racecar challeng

Post

Emails have been sent - hopefully.

MisterX
MisterX
0
Joined: 04 Mar 2013, 10:18

Re: F1 model CFD Challenge (Khamsin Virtual Racecar challeng

Post

I received email but when I try to access I get this error:

"Your login details are not correct. Please try again."

As for the tub, I can download once I log in, or this is only the upload page of my model?

thanks

Team:SC Italica Corse