Khamsin Virtual Racecar challenge 2013 (CFD model racing)

[variante]

For those who like this kind of stuff, i've calculated the efficiency of the cars... nothing more than Cl.A / Cd.A.
It should be interesting especially thinking that we'll have to race on tracks like Monza, where we'll have to achieve an extremely low drag, preserving as much downforce as we can.

Variante --------1,78
STZ Racing -----1,55
MARELI ---------1,46
RicMe M. -------1,33
Cottrell M. -----1,24
WigglesGP -----1,22
Mecury M. ------1,00
Blindblom GP --0,11

BTW It's not a very significant datum because of the few info we have for the moment, and because a lot of things can change till the beginning of the actual season

[machin]

Good stuff; the more numbers the better!

[julien.decharentenay]

Love the discussions. Sorry for not joining in earlier I was catching up on sleep and family life.

Have a look at the car results (180mph) at https://picasaweb.google.com/1108457918 ... 3019525682. The results show some of the things we will try to improve:

- Where you see patchy colors: the geometry is not watertight (i.e forming a completely closed shell) and the CFD analysis is also done inside the car. This is not an issue as the pressure inside the car is around 0Pa and have little effect on the drag and downforce. We have to think how we can address this in a simple and effective manner - in the CFD world, geometry preparation is where a significant amount of time is spent;
- The mesh is based on a mesh 1cm resolution at the car. Small gaps of less than 1cm can create issues - this will be clearer on the CFD result files. This is required to allow us to get a reasonable compromise between accuracy and turn-around time;
- CPU time: I quoted earlier that some simulation took over 60hours. While this is correct, I am suspecting that there was network issues that made the simulation very sluggish.

I will try to do a bit more analysis on some of the cars to see where difference in performance comes from (time permitting) and report back with some more numbers...

[flynfrog]

How are the lap times calculated? Do they take into account the unbalance front and rear and its effect on handling? Or are you simply putting in a Raw Cl and Cd number into a simple sim?

Nick's link to my website explains it in a more detail, but essentially the program uses the COP to determine front and rear downforce split, the program then calculates vertical loads on the tyres and from there the cornering forces at each end of the car. You can see this in the results; Marelli had the second highest Cl.A but because most of it was generated at the front his lap times were well off the fastest car.

The car's mass distribution is fixed as per the real 2012 F1 rules and runs at minimum weight plus 10kg of fuel.

Again, as Nick says; the Khamsin guys test the cars at two speeds straight ahead and one speed at 3 degrees yaw. the program then iterates for other speeds and yaw angles; in effect we have a simple aero map for each car.

I'm happy to answer any further questions....?

Thats cool. I Like the simple approach to keep the sim simple to understand yet not a simple beauty contest.

[RicME85]

I assume that looking at that picture alone, for my 2013 car I need to work on getting air around the front wheels more and also work on getting airflow to attach over and around the rear of the sidepods?

[astracrazy]

i'd say so based on the red area just behind the front wing on the wheels, although i can't imagine anyone here is going to get it perfect.

Perhaps barge boards may help around your sidepods, it looks like no air is getting to your beam wing or going around the pods? Perhaps try removing those flow conditioners, based on the amount of low pressure on top of them they could be sending the air up which is why your getting nothing over your sidepods and also creating lift.

perhaps also get rid of that red bull slot, if its blocked off then its just creating drag

of course this is just guess work, i'm no expert! lol

-----

julien: might it be a good idea to lower the pressure range for the images? For example instead of 4000 to - 4000 do 1000 to -1000 so you can clearly see high a low pressure even better?

heres what i mean if i haven't explained well.

big difference
http://img855.imageshack.us/img855/9127/rw1pressure.jpg
small
http://img109.imageshack.us/img109/9742 ... essure.jpg

[RicME85]

That nose and bulkhead will be different for the 2013 car.
Originally the duct was open and it had a pathway in and around the cockpit but i blocked it off for the test along with the airbox and intakes.

Im thinking, like you, some decent barge boards will help with the coke bottle rear as there is a distinct lack of pressure around the beam wing.
If my plans come to fruition this car will be a distant memory.
Im going to go out and get some graph paper to do some design ideas in pencil first, then using the grids import the drawings into SU.

[julien.decharentenay]

It is hard to get the pressure scale right - 4000 is based on 0.5*rho*U*U at 180mph (i.e the dynamic pressure contribution). We provides VTK files (link in the PDF document that Nick has/will send) for you to have a look at the results - look around the car, do streamlines, change the scale...
Front wheels: the red area will create drag. But due to the rotation of the wheels, there is also a significant effect on downforce.

[astracrazy]

Ric: i wouldnt say your design is bad, just need to add and remove one or two things and i would guess you'd see some decent results. thats a good idea with grids.

i'm guessing in theory you are getting air around the pods actually, low pressure (blue) = high velocity or is that wrong?

[RicME85]

Could do with one of the Codemasters cars being run so we can see what the results look like to see if we are in the right sort of area with our designs.

Really need to figure out how to use Khamsin

[machin]

Ric,
That's a nice looking car, thanks for sharing!

What surprises me is that the upper surface of the main elements of the front wing aren't generating much positive pressure... are they simply not inclined enough to the air flow? That explains the Fr:Rr Downforce balance....

Is anyone else going to share their cars?!

[etsmc]

there are images of all the cars in the link provided by julien. mine (mercury motorsport) apears to be also not getting much air flow to the beam wing and the frony wings not prodicing downforce. is the front wing main plane meant to be airofoil shape or just the other ellements?

[astracrazy]

on the one i'm designing at the moment, i'm doing the underside an aero shape (curving up), seems to help

[variante]

etsmc,
astracrazy is right, and i suggest you to have a look at the way the elements of your wings overlap (most of the guys are doing this same mistake, loosing a lot of DF).
As for the beam wing, it may help to increase a bit the angle.

[CottrellGP]

Interesting.