Subaru Impreza Pikes Peak

[Alonso Fan]

ok, wesley123, you win.

[machin]

Certainly Peugeot seem to think the "up and over" approach is the way to go:-

[MadMatt]

My suggestions:
1. Start your rear diffuser farther forward
2. Add a front diffuser vented behind the front wheels
3. Find a better method of attaching rear wing assembly( maybe two centered swanneck mounts and to attachments to the endplates. Also maybe move whole assembly just slightly forward so the mounting isn't so stressed and difficult to work with.
4.On the front endplate wings, are they three stacked wings or a traditional three element wing? I'd go with the latter.
5. Ignore anyone talking about drag. I'm sure you've noticed that other Unlimited cars for Pikes Peak ignore drag. If they think there's too much drag, they just add power.
6. Start working on fine geometry refinement rather than macro changes. There's a huge amount of DF in detail refinement.

I would definitely add a front diffuser.

Good points there!

1. I cannot really start it before the rear wheel axle because I do not know how big is the transmission, where is the diff, where stops the chassis, so I cannot take risks there.
2. I do not know if I got you there, but I have added diffusers underneath the front splitter. I will show pictures of this later this week. But I am interested in something I could do behind the wheels to help the flow to exit there, considering I cannot modify the chassis.
3. I would expect the wing assembly to be moved backward a bit in order to give the lower beam a better flow and help to interact more with the diffuser, but this will require CFD simulations (I have planned to do that).
4. They are 3 stacked wings, and I found them to be quite unproductive, especially the 2 upper profiles which could be predicted. I will indeed try with a single element or with a typical assembly (as you mentioned).
5. This is correct, however, I have a L/D ratio objective of -1.5:1 which I see no problem to obtain.
6. Again, I agree there, but before refinement, I have to seek for the best wing position/AoA for example, or diffuser geometry, devices to be added/removed/modified. I will probably won't be able to do small adjustments in the time scale that I have, so I try to focus on trying devices (gurney flaps, vortex generators, small wings, etc.) rather than changing angle of wings by 0.1°. But you are right in saying detail makes a big difference. Its just that I have first to find a kind of "optimum" with the large modifications.

[machin]

Looking forward to seeing some aero figures from the V1 kit. Will you be checking Front:Rear downforce split? As you may've seen from the Khamin Challenge; getting high downforce is one thing... but getting it in the correct split front:rear (and hence a good lap time as a result) is something else....

[MadMatt]

Looking forward to seeing some aero figures from the V1 kit. Will you be checking Front:Rear downforce split? As you may've seen from the Khamin Challenge; getting high downforce is one thing... but getting it in the correct split front:rear (and hence a good lap time as a result) is something else....

Yes I will assess aerodynamic balance as well. The objective is 25% at the front (after looking at what Roger Clark Motorsport wants to achieve with their Time Attack Impreza which is the same model, detail in Racecar Engineering June and July 2013). I know 25% is not really what you might call balanced, but considering the amount of power and the repartition rear biased, a CoP more at the back will increase driver confidence and help to put power down as well! One of the thing I will try is add rake and see the effect which is supposed to be quite big!

[Alonso Fan]

ok, ok, up and over is the way to go

happy now?



(ignore my drag comments from earlier)

[machin]

The objective is 25% at the front (after looking at what Roger Clark Motorsport wants to achieve with their Time Attack Impreza which is the same model, detail in Racecar Engineering June and July 2013).

I must admit I'm skeptical how they arrived at that figure... if you look at Peugeot's 208 T16 it looks more aero balanced than 25:75 and we've seen how quick that is up Pikes Peak from the practice day (easily fastest). Even if you assume that the Impreza is more rear biased than the Peugeot I'm still skeptical; an F1 car ("100% rear biased"!) has a much more even Downforce split front to rear, and in power to weight terms is higher than the RCM impreza.... I would love to know what lap time program they used to come to the 25% conclusion (if they used any program at all) because my program, and the real life examples above, do not concur....

[MadMatt]

The objective is 25% at the front (after looking at what Roger Clark Motorsport wants to achieve with their Time Attack Impreza which is the same model, detail in Racecar Engineering June and July 2013).

I must admit I'm skeptical how they arrived at that figure... if you look at Peugeot's 208 T16 it looks more aero balanced than 25:75 and we've seen how quick that is up Pikes Peak from the practice day (easily fastest). Even if you assume that the Impreza is more rear biased than the Peugeot I'm still skeptical; an F1 car ("100% rear biased"!) has a much more even Downforce split front to rear, and in power to weight terms is higher than the RCM impreza.... I would love to know what lap time program they used to come to the 25% conclusion (if they used any program at all) because my program, and the real life examples above, do not concur....

I am also surprised, but here is what they wrote in the mag:

Chassis dynamics analysis had determined that the car would naturally oversteer when under power and that it would need a significant proportion of its total downforce to be generated at the rear. So the car's aerodynamic package was devised with this in mind using CFD and CAD model prepared from scan data. [...] However, it was not too far from the balance target of 75-80 per cent rear.

I will try to have it more balanced and still meet my objectives.

[variante]

I've never really followed Pikes Peak, but is it possible that the 25/75 balance is due to very different vehicle dynamics compared to ordinary cars? After all, those cars behave more like rally cars than "ordinary" racecars, so we see a lot of drifting and unusual trajectories.
So, Machin, does your program take into account those kind of dynamics too? (...honestly i fail to imagine how to translate "drifting" into simple algorithms)

[Blanchimont]

In my opinion the more a car drifts (rally car on dirt track) the more downforce should be on the rear of the car.

But today's Pikes Peak is similar to an asphalt rally where cars only drift in very narrow corners. Corners of higher speeds are taken without drifting during asphalt rallys. And in such higher speed corners the aero balance should be in the same region as the front to rear weight ratio, i would say.

[machin]

The RCM impreza referred to above (the one with 25:75 aero split) runs on "conventional" UK tarmac circuits like Croft, Snetterton etc. granted it has more power than a typical race car, but its power:weight is no greater than a powerful single seater, and its 4-wheel drive...

[Blanchimont]

My idea behind aero balance and weigth balance being similar is the following:

When a car corners the total force of all tyres has to be m*v²/r. If we split the total car mass into m_front and m_rear the force at the front wheels is m_front*v²/r and m_rear*v²/r at the rear. Additional downforce pushes the tyres down and increases tyre force and therefore cornering speed. In an extreme case where all the downforce is placed at the rear tyres, the rear could corner faster than the front, or vice versa. The car then would understeer or oversteer. If downforce is split between the front and the rear according to the weigth distribution this problem shouldn't occur. Of course this is related to cornering and if the car is very powerful and has problems to grip the driven wheels properly, it could be better to shift aero balance to the rear of front.

From the videos i googled it seems that the RCM Impreza has problems with getting the power down on the road!

[machin]

I pretty much concur with what you are saying.. of course its slightly more complicated since the driven wheels always need to use up some of their grip "capacity" to overcome the car's resistance (aero drag, rolling resistance, road inclination) even when the cornering speed is constant. But at slow to medium speeds these resistances are small compared to the cornering forces, even with these big draggy wings. Assuming that the drag effect is negligible, what you have described above is essentially what happens mid-corner when the car's speed is neither increasing, nor decreasing (and therefore there is little motive force being applied other than to overcome the small amount of resistances I mention above), there are other affects such as load transfer, front:rear roll stiffness distribution, tyre sizes etc of course, but lets assume you've done your home work and mechanically balanced all those effects (I never think its a good idea to use aero as a "crutch" to solve a mechanical issue). By moving the balance back from the front to the rear you are effectively trading apex speed for more rear tyre grip coming out of the corner... Is there much benefit to this for a 4 wheel drive car? My program suggests not, and I'm glad to see that this conclusion is borne out visually by the vehicles of those competitors right at the top of the motorsport tree, such as the Peugeot 208 T16 (see image above) and the Toyota-Radical, below, both of which also have 4wd and lots of power:weight:-

[MadMatt]

I tend to agree with Blanchimont's arguments, but machin has also good points there. From my personal knowledge, I would also balance the aero depending on the weight distribution of the car, and why not maybe shifting a bit the CoP towards the back for more stability. But 75% at the back sounds like a lot anyway. It would be great to have RCM's data and how they came up with this conclusion.

For my project, I will concentrate on reaching at least 25% of the aero balance at the front (as a start), which is not that easy considering that the rear diffuser is not starting early as you can see on some cars. We will see what comes up but 25% is more than reasonable to achieve and still remain above the pure downforce values that I have fixed.

[Pierce89]

Certainly Peugeot seem to think the "up and over" approach is the way to go:-

http://images.thecarconnection.com/lrg/ ... 4994_l.jpg

The side fences on the front splitter are quite popular at Pike's Peak because down force matters much more than drag.