Khamsin Virtual Racecar challenge 2013 (CFD model racing)

[machin]

.... run the same races again if other tracks are not possible.

To confirm; I have track data for a lot more tracks, so this isn't the issue; (mainly uk tracks that I've raced on myself and therefore collected data as I've driven). Some data is derived from friends cars who've driven on more "exotic" tracks like Magny Cours.... I have even generated track data from things like google earth prior to actually getting to a venue to try and figure out lap times/cornering speeds etc to give me a steer as to how fast I should be going at each point and that has been surprisingly accurate....

[variante]

I propose to get rid of the 3° yaw simulation. It takes like 1/3 of the computing time, so we would gain 1 more race.

I understand it wouldn't be fair to do so from the next race... but at least from next season.

For those who don't know: 100mph at 3° yaw simulation gives almost identical results of downforce and drag to a 100mph straightline sim.

[N12ck]

However the 3 degrees Yaw is there to create a better aeromap for the virtual stopwatch, which is why we have it,

[astracrazy]

you cant really drop any, they are all required

perhaps next season limit the amount of applicants. running something 15 times will surely use less cpu time than running it 35 times?

still i don't want to moan to much because i think this is a great thing and i hope it lasts a few seasons. i don't want nick etc thinking f**k this we don't need this s**t. just some positive feedback/ideas

[variante]

However the 3 degrees Yaw is there to create a better aeromap for the virtual stopwatch, which is why we have it

I understand, but how much would the result be compromised without yaw data? Maybe Machin could help us.
If the result doesn't differ that much then it would be about choosing between extreme accuracy and 1 more race.

[RicME85]

Surely a fair bit as the yaw will help with the downforce levels when cornering?

Seems odd to of come to the decision to allow so many teams to enter at the detriment of running a proper season of races.

I fully understand that the competition is being run thanks to Sabalcore giving CPU time so asking for anything more from them is unreasonable.

For the future of KVRC it would be good if we could explore other avenues, approaching more companies maybe or if 1 or 2 of the 3 tests were done on our own PC's and the results sent in alongside the car model for Julien and Nick to complete the rest?
I would even be happy to set something up where my media server can crunch numbers as well if it means more races. I might do a test run on one of the laptops I have access too, my local community centre has 6 or 7 laptops that dont get used at all, I could use my own HDD's so they all ran Linux and set them up in a cluster.

I dunno, just bouncing ideas around so that this doesnt die out in a fizzle.

[cdsavage]

So where do we submit the file?

[astracrazy]

Julien should of e-mailed you a few weeks ago with a link, username and password where you log in and upload it. its like a portal. you must upload it by 23:59 Friday in the way Nick e-mailed you about 2 weeks ago:

For those wondering what will be policed on legality and what wont be I will explain:

Engine cover areas will be checked (must be larger than or equal to the triangle shown in the welcome pack)
Chassis Dimensions
Intakes
Outlets
'Holes in the floor/ bodywork from underneath the car' will be checked
Wing Dimensions/ Diffuser Dimensions
Mirrors must be sensible, but the regulations for the mirrors will not be checked.
Nose Height will be checked (no higher than 625mm)

Reminders:
Suspension + Wheels must be added before the race is started
Must not send half a car, it must be setup fully
Gearbox + Engine must be added

[machin]

how much would the result be compromised without yaw data? Maybe Machin could help us.

I've not actually seen the full results from the test race (I supply the program, but the aero coefficients get inputted by Jules/Nick), but assuming that your car loses 5% downforce in the 3 degree yaw condition that translates to a 0.3second lap time loss around a lap of Monaco... of course all cars will be affected differnetly in the 3 degree yaw case; some will be affected more than others; some may even improve.

I see that no two cars in the Monaco test were within 0.3 seconds of eachother.... so potentially it wouldn't have changed the overall result in terms of placings.... but that assumes 5% change in aero coefficients at 3 degrees yaw.. what sort of numbers were you guys actually getting?

[astracrazy]

i only test my parts/car at 100mph 3deg yaw

[variante]

I see that no two cars in the Monaco test were within 0.3 seconds of eachother.... so potentially it wouldn't have changed the overall result in terms of placings.... but that assumes 5% change in aero coefficients at 3 degrees yaw.. what sort of numbers were you guys actually getting?

Drag: 1% better, downforce: 2,5% worse.

If everybody gets similar results, wouldn't be the same thing to ignore those data?

[RicME85]

I reckon that it will be a lot larger difference for some cars due to the design and build quality.

[astracrazy]

maybe a silly question, but i think i've made a big mistake with my tests. does tunnel size you use effect the results you get very much?

[variante]

I reckon that it will be a lot larger difference for some cars due to the design and build quality.

Yes, but again, we could just ignore it (maybe with default numbers for everybody), and of course telling everybody there's no need to design the cars to face the yaw scenario.

[RicME85]

whats the point in that?

KVRC wants to be what Formula Sketch wasnt, accurate.

a smaller grid with more races would of been better than the current set up.
I would understand the current amount of entrants if there was a fee to enter as obviously more entrants = more profit but its free so why not less teams and have a better championship that would bring more attention to the Khamsin brand as well as the sponsors.