Pikes Peak

[langwadt]

This setup resulted in a time of 484,95 s. The change from AWD to 2WD increased the time to 514,44 s, a massive change!

My own program also suggests a similar difference; in the range of 25 to 30 seconds difference between AWD and 2WD for an 875kg car with 875bhp...

But, sticking with the graphs above rather than post my own which would only be slightly different, 25 to 30 seconds is approximately equivalent to a weight difference of 250 to 300kg.... so if the hillclimb singleseater weighs no more than 625kg (and everything else being equal), it would be quicker according to those graphs... and, as Andy says, its likely that the aero figures for the singleseater would be better too....

can you make an estimate on how many kWh it took to do the Loeb run? and how much of it is drag?

[Smokes]

Isn't dpcars building a turbo 4wd car to attack the peaks with next year in the unlimited class
They won 2012 with n/a v8

http://dpcars.net/

[andylaurence]

Brilliant work! How did you create the track for Pikes Peak? Can you post your track as a .OLTra so we can try some other simulations. I have setups for a few cars in Optimum Lap, including my own and I'm interested to see what comes out the other end compared to this simulation of the 208. I'm stunned you managed to find a dyno readout of the 208!

[Blanchimont]

Good to see you like my work!

As requested, the track is available to download now: Pikes Peak Track

The track was created with the help of Google Maps/Google Earth to know the radius of each corner and the website GPSies.com to know how long every single corner or straight is. It took some time to measure the whole track.

OptimumLap Tracks are defined as the next picture shows, through the radius and length.



This approach unfortunately adds up all the little errors that come with each corner and in the end the complete track doesn't look exactly the same as the real track. But for the calculations this shouldn't matter that much, the correct radii should matter the most. But you're free to modify the track and improve it!

Here is the result, start at the upper right of the picture:



A "real" map, the orientation is different with the start at the bottom left:



Some stats on the track from OptimumLap:

[Blanchimont]

I remember that in OptimumLap you cannot include elevation which is a real shame because it is responsible for a lot of losses in the car acceleration.

That's correct. The tracks in OptimumLap are flat and because of that the acceleration simulated should be slightly higher than at the the real track. But on the other hand it is possible to decelerate harder when the road is rising.

The additional power for a car of 875kg going up an incline of 10% varies with the speed, as

Power [W] = mass * 9,81 N/kg * SIN ( ARCTAN (0,1) ) * velocity [m/s]

So for different speeds the numbers are:

15 m/s = 54 km/h: 12,8 kW 2,1% of peak power of 600kW
30 m/s = 108 km/h: 25,6 kW 4,2%
60 m/s = 216 km/h: 51,2 kW 8,4%

That's the one effect of the incline, the other is that the forces between the tyre and the road (traction limit) caused by gravity are also changed.

Available tyre force:

F [N] = mass * 9,81 N/kg * COS (road angle) + downforce

The reduction for a 10% incline in tyre force is: 1 - COS(ARCTAN(0,1)) = 0,005. Only half a percent, i thought it's more than that. With the calculated power loss of around 5% and the tyre force almost being constant at 10% incline, i would say the best method to simulate an ascent in OptimumLap is to further reduce the engine power by 5 percent!

[Blanchimont]

can you make an estimate on how many kWh it took to do the Loeb run? and how much of it is drag?

OptimumLap says that 133 MJ were needed for the run.

An estimation of the energy for the real run of Loeb could be: average engine power * lap time
Loeb's car had a peak power of 817 hp at the wheels which equals to 817/1,36 = 600 kW. If we assume the average engine power for the whole run is half of that, the energy is calculated as:

600 kW / 2 * 493,8 s = 148,1 MJ

If divided by an assumed engine efficiency of 35%, we would need 148,1 MJ / 0,35 = 423,2 MJ of chemical fuel energy, which is about the energy contained in 10kg or 13,3l of petrol! That would be a fuel consumption of 50kg or 66,6l per 100km.

The amount of energy that is lost by drag could be calculated accordingly as: rho/2 * cd * Area * (average speed)^2 * track length.
Loeb's average speed was 145 km/h.

Energy lost: 1,1/2 * 1 * 2 * (145/3,6)^2 * 20000 = 35,7 MJ

For a more precise calcuation you would need to integrate the drag force over the track length.

[WhiteBlue]

...If divided by an assumed engine efficiency of 35%, we would need 148,1 MJ / 0,35 = 423,2 MJ of chemical fuel energy, which is about the energy contained in 10kg or 13,3l of petrol! ..

That engine never has a brake thermal efficiency of 35%. It is a turbo engine but has none of the modern bells and whistles of the 2014 generation. We are talking 2002 to 2006 vintage with a bit of cosmetic. Quick and dirty I would say you can expect something around 30% AFAIK.
http://www.f1technical.net/forum/viewto ... 03#p440103

MadMatt said the consumption was 100L/100 km = 20L.
If we use that figure we get:
Thermal energy petrol 32.7 MJ/L. -> 654 MJ thermal energy -> 23% BTE
That looks a bit lowish. I would say the truth is probably somewhere in between those two figures.

[Blanchimont]

...If divided by an assumed engine efficiency of 35%, we would need 148,1 MJ / 0,35 = 423,2 MJ of chemical fuel energy, which is about the energy contained in 10kg or 13,3l of petrol! ..

That engine never has a brake thermal efficiency of 35%. It is a turbo engine but has none of the modern bells and whistles of the 2014 generation. We are talking 2005 or 2006 vintage and you can expect something around 30% AFAIK.

You may be right, but after all it's just an estimation. If you look at the way i estimated (=half the peak wheel power), i don't think the engine efficiency plays an important role in THIS calculation.


Two more pictures for the 208 simulations. Lateral tyre friction (which could be a way to describe the tyre and suspensions quality) is more important than longitudinal friction.




And the last one shows lap time as a function of both drag and downforce. If you know the different configurations regarding lift:drag that a car allows, the "theoretical best" can found out with the help of this diagram.




The next step could be to simulate other cars than the Peugeot to find out which car configuration is the fastest up the hill in theory. Of course there need to be some limitations in weight, power and downforce...but i hope we'll find out together in this thread!

[MadMatt]

You could use the "tyre data" that Peugeot reviled on the car:

- max acceleration: 1.3g
- max cornering: 1.5g
- max braking: 2.2g

OptimumLap does also not include the aero balance (CoP position) which is very important, especially if you have a tyre map. I think there are too many approximations there, plus it is all steady state. A software like Chassissim would be more adapted but then lack of information on the car would make us assume too many things again.

Interesting posts nonetheless Blanchimont, keep up the work. What would be nice as you said is to get other car data to compare and correlate your model. I will give you my Pikes Peak Subaru model and see what it makes with your settings, it might still be interesting!

[andylaurence]

This is getting very interesting. I've run a couple of simulations based on your version of Pikes Peak to keep things comparable. A V8 hillclimb car similar to the one I posted earlier managed 501.95. I cribbed a quick version of the GWR Raptor and ran that up the hill. Despite a normally aspirated 1600cc engine, it managed 506.31. I also simulated the new Empire that's due to run this weekend at Loton Park with a 999cc engine and it made the hill in 527.55. Admittedly, I've had to make educated guesses at some of these figures, but I used some figures from previous Aerobytes articles and graphed something reasonable from known power figures for the V8. The power/weight is known for the new Empire as it was announced over the weekend by the new owner and the ratios are also known.

Lastly, I simulated my own car, which managed 588.68. Surprisingly fast, I thought, given the relatively heavy weight (510kg including driver), poor aero (1.48 -L/D) and lack of power (180bhp or so). Compared to the similarly-engined Empire, I lose pace through an inability to take some corners flat, which the better aero package of the Empire allows. There are just a few places where I'm quicker and that's the really fast sections where my lower drag counts. Amazingly, fitting a larger turbocharged engine to my car only gets me half way to the performance of the Empire, which shows I really need to improve my aero!

I'd be happy to do some proper simulation in LapSim for better accuracy, but I need to find the time to do it properly and as much data as possible for the cars.

[P.S.]

Lastly, I simulated my own car, which managed 588.68. Surprisingly fast, I thought, given the relatively heavy weight (510kg including driver), poor aero (1.48 -L/D) and lack of power (180bhp or so).

As it was said, Optimum Lap don´t calculates the losses at climbing. For exampel your car with its pretty low power/weight ratio (in comparison to the slightly modified 208) will accelerate much slower as this tool predicts.

My thoughts are:
with 1,44 km difference in hight and 20 km track length we have: sin^(-1)(1,44/20)=4,129[°] average uphill.

So with a weight of 510 kg it generate an aditional constant drag of 360 [N].
And this is roughly the same drag as the aero produces at 135 kph.

So maybe not the best comparison, but I think your car would be far of 588 seconds.

I already was thinking how to modify the data in OptimumLap to get an approach, but I don´t have an idea.

[andylaurence]

Yes, you're right. It's also worth pointing out that OptimumLap also believes I will complete Crystal Palace in 31.16 seconds, yet my record is 34.32 seconds. Assuming a linear error, then that bumps 588 seconds up to 647 (10:47). Still pretty quick and I'd be happy with that.

[MadMatt]

Below is the only picture I could find of the floor of that Peugeot 208 T16 (open in new tab for bigger resolution). They say it is responsible for most of the downforce on the car, what do you guys think?


http://imageshack.us/a/img7/7536/siqo.jpg

Doesn't look that impressive. Their value of 10[kN]+ is irrelevant because they don't mention the speed at what this would be achieved.

[andylaurence]

Doesn't look that impressive.

Seriously? It's wide, tall and long. I'd say it's quite a monster and bigger than anything you'd find in a regulated series.

[MadMatt]

Doesn't look that impressive.

Seriously? It's wide, tall and long. I'd say it's quite a monster and bigger than anything you'd find in a regulated series.

Wide I agree, long I agree, but it is not really tall, and I was surprised about this. I like how it expends outward tho. Don't get me wrong, it is a wonderful machine, but I was wondering why they didn't make it higher. The suspension components are not on the way at this point.


http://s.tf1.fr/mmdia/i/01/3/la-peugeot ... o_2038.jpg

Maybe I'm having in mind those group C machines and their Venturi tunnels?