F1-02 downforce

[NaN]

There has beens some discussion about tire limits: http://vdrift.net/Forum/viewtopic.php?t=1645

I did some testing with the F1-02 "vdrift -debug" and noticed that the rear wheels are going to exceed an upper limit of 10kN -> 1000kg.

Of course one can argue that F1 tires are going to have stiffer sidewalls than usual passenger car tires. We need to check the force curves in the pacejka editor to get the load limit. I'll add a max load parameter to the tire config.

Now to the downforce:
speed: 320km/h
front tire load: 6.292N 33%
rear tire load: 11.620N 67%
front lift: -9527N 33%
rear lift: -19055N 67%
car weight: 738kg
tire load: 3650kg
tire load factor: 5

The distribution looks ok: 33/67. But the downforce numbers look a bit too big. What I've been able to pick up online is about 1700kg half of our values.

[NaN]

Looks like the champ cars are generating up to 5703 lbs / 2586 kg. With car min weight 1741 lbs / 789,7 kg this is 3376 kg tire load. Looks like our values are not that bad if F1 is comparable to Champ Car regarding downforce. F1 seem to have faster lap times in direct comparison.

For reference: http://www.mulsannescorner.com/data.html

[NaN]

I've been checking the weight distribution. The default seems to be 44/56. This is of course not the aerodynamic distribution. Given our current one we will get an under-steering car with higher speeds. No idea if this is true for real F1. We are low on front grip with 33/67, maybe should go towards 40/60 at least.

Btw the current weight distribution is 1834/(1834+1772) = 51/49.

[NaN]

I've adjusted the aerodynamics to have a tire load of 44/56.

[NaN]

Looks like front wing and chassis mass have been applied twice. Fix gives 44/56 weight distribution.

[NaN]

Adjusting the lift to drag ratio to somewhat more realistic value(wing efficiency=0.8 ->L/D = 3.6) gives a top speed of about 230km/h. Looks like a high downforce setup. The transmission needs to be adjusted.

Anyone interested in creating two F1-02 sets F1-02H.car and F1-02L.car?

Edit: current L/D=5.5

[cologne]

http://vdrift.net/Forum/viewtopic.php?p=...ight=#3894

[cologne]

Adjusting the lift to drag ratio to somewhat more realistic value(wing efficiency=0.8 ->L/D = 3.6) gives a top speed of about 230km/h. Looks like a high downforce setup. The transmission needs to be adjusted.

Where can i find your modifications?

[NaN]

Set efficiency to 0.8.

Still looking for a realistic medium setting. This one has total L/D=3.6, wing L/D=10

Code:

[wing.center]
position = 0.00, 0.00, -0.40
frontal-area = 1.0
drag-coefficient = 0.6

[wing.front]
frontal-area = 0.20
drag-coefficient = 0.01        # most wing drag induced(efficiency)
surface-area = 1.25            # tire load front/rear 44/56
lift-coefficient = -1.5
efficiency = 0.9            # L/D = 1/(1-efficiency)
texture = body00.png, body-misc1.png
mesh = wing_front.joe
position = 0.0, 2.25, -0.44
mass = 20

[wing.rear]
frontal-area = 0.3
drag-coefficient = 0.01
surface-area = 1.75            # tire load front/rear 44/56
lift-coefficient = -1.5
efficiency = 0.9
texture = body00.png, body-misc1.png
mesh = wing_rear.joe
position = 0.0, -1.84, -0.1
mass = 20

Maximum L/D should be somewhere between 0-12 -> efficiency 0-0.92. The efficiency has to be adjusted with the lift coefficient. At Cl=1.5 it should be much worse than what I have. I think about L/D=5 -> efficiency=0.8. But the F1 have multi-element wings which might differ.


To get the body drag right. I've tried to replicate Penske-Reynard-Honda 01 super-speedway configuration(almost no wing): 1000 lbs. @ 230 mph, with 950 lbs. of drag.

[NaN]

As you see we have some redundancy in the parameters: lift-coeficient(Cl), drag-coefficient(Cd), efficiency = 1 - D/L.

Atm the drag is the sum of drag-coefficient and efficiency contributions.

Correction: Our Cd is Cd0, the offset at zero lift.

[NaN]

Another correction. For some reason the efficiency equation is:
drag = lift_coefficient * lift * (1.0 - lift_efficiency); -> eff = 1 - (D/L)/Cl

Means my above efficiency calculations are wrong.

[NaN]

OK, I think I got. Assuming elliptical lift distribution:

Lift:
L = 0.5*rho*S*v*v*Cl

Drag induced:
Di = 0.5*rho*S*v*v*Cdi with Cdi = Cl*Cl*k and k = 1/(pi*AR*e)
Di = L * Cl * k

I our case:
k = (1-efficiency)

Di/L = Cl*(1-efficiency)
efficiency = 1-Di/L*1/Cl

[NaN]

This means efficiency doesn't have to be adjusted with changing lift coefficient. But getting the L/D is somewhat of a pain. I guess the best method is to run vdrift -debug. And look at the lift numbers.