Lotus E20 VD

[godlameroso]

The flow is exactly what's creating the pressure differentials to begin with; no airflow no downforce. Airflow + Surface interaction = Pressure differentials. It's not just about downforce to the max it's downforce to the max, while under certain conditions. For all we know, at 220kph the Mercedes is making the most efficient downforce out of every car out there, thing is there aren't a lot of turns in F1 that you take at such speed. It could be that the Lotus is making more efficient downforce than the Merceds at 170kph, and there are a lot more turns at that speed, hence the Lotus is faster overall.

[olefud]

different way to skin the cat - I suppose
in your Newtonian way of thinking, less upward momentum (less wake height) would require less energy, which has to come from somewhere.
The end result would be the same, just the way to illustrate it is different - IMHO

Sorry, I wasn’t clear. I see no problem between Newton and Bernoulli; the latter creates the delta P and the former reacts thereto.
My inquiry was with the language,

“That for the oncoming air the cross section of the car including his (its) wake could look something like this.” relating to the illustration.

My question concerned including the wake profile with the car cross section as a drag factor, i.e. as if the wake was moving through the oncoming air. But, as originally metioned, perhaps I'm reading more than was intended.

[gato azul]

it's o.k. I think I know what you mean, want to say.
Maybe it was not the best way to phrase it, but I think their is little doubt, that it has a effect on the air volume in front of the car.
(control volume - how much air will be affected by the movement of the object through a fluid volume, as more of the volume is set into motion (and as more intense this motion will be), as more energy will be required to do so)
Sorry with English not being my native language, I may can't explain it in an more understandable way - Sorry

[gato azul]

Rear Wing Tests
The lift and drag coefficients, measured in the wind tunnel, for the rear wing in free-stream are shown in Figure 8 below.
The CFD predictions are provided on the same graph for comparison.
The wind tunnel results showed a trailing edge separation on the rear most flap beginning at 38degrees.
By 40 degrees the underside of this flap was fully separated, resulting in a plateau in the CL curve.
The flow on the underside of the main plane remained attached until 48 degrees, beyond which a complete leading edge separation was observed.
The considerable difference between the stall angles predicted by CFD and measured in the wind tunnel was most likely due to the extremely small aspect ratio of the wind tunnel tested wing (1.72).
Again, the wind tunnel drag is believed to be higher than predicted from CFD due to induced drag.

the plateau in Cl around 38° flap angle correlates with an plateau in Cd as well, post stall (>48° flap angle) Cd reduces as well - which disagrees with the conclusions drawn based on the "thin airfoil" stall theory.
Based on this data, it seems beneficial to operate a F1 style multi-element rear wing in a post "stall" or "semi stall" regime to achieve a reduction in Cd (together with a reduction in Cl)

[n smikle]

That wing profile is similar to an F1 wing though? Some wing profiles are more draggy after stalling. And another thing too... "Stalling" the F1 wing if that is indeed what was happening with F-duct in 2010, doesn't involve a change in angle of attack, so there is no rotation of the profile relative to the air stream, so it is hard to surely say that graph relates to an F1 wing.

On my test model of the E20 rear wing I finally have a semi-decent setup that I have tested with the ducts and without the ducts. It seems the ducts are there to make down-force. The model with the Ducts on the pylon had about 13 to 20% more down-force but only about 6% more drag. So it is basically a blown wing. This is just simple CFD though. I can refine it a bit more and post some pictures later.

[gato azul]

it's closer to an F1 wing, then some of the single airfoil CFD's and drag data shown earlier in this thread and others, as that it comprises of an relative flat (in terms of AoA) main plane and flap(s) (or tab) at very high AoA.
The wing shown has more then one flap, it's closer to an older style F1 rear wing, but the 38° flap data point shows the direct relation between Cl and Cd in a condition where the rearmost flap is "stalled". "stalling" the flap does not cause the lift of the whole wing to increase.

My point is, that it is the induced drag, which has a larger effect on the overall drag then the pressure drag in this cases.
The increase or reduce in drag is directly related to the loading of the wing (downforce produced per square unit wing area), while the form of the whole thing remains more or less the same.
The flap is used to manipulate the "loading" / pressure distribution on the main plane, even on a F3 front wing, it can be shown, that the flap element is operating past it's own max Cl value, but that due to the effect it has on the downforce generated by the main plane/wing (keep in mind that this one has a much larger area) it is still beneficial in terms of overall downforce produced.

A extreme case to illustrate this point, would be to look at lift and drag data of an rotating cylinder (Flettner Rotor), the geometry (area exposed to the free stream) of the lift producing object does not change, compared to an airfoil at changing
angles of attack.

n smikle:
Thanks for your answers, but I fear that only integrating the surface pressure distributions over the area of the wing will not give accurate values for overall drag.
It's not that you make anything wrong, rather then a limitation of the method your software uses.
The effects of induced drag will not be properly represented by only integrating surface pressures.
Are your values "time averaged" values, or are you (your software) assuming a steady flow condition at all times?
Effects like vortex shedding etc, and the flow in the "dead zone"/separation areas are highly unsteady in nature, so can be only represented as a time averaged value in terms of their contributions for a single Cl or Cd value.
Nevertheless, keep up the good work, I like to see the evolution of your models, just keep in mind that every tool has it's limitations.
How large is the "control volume" you chose in your model? Do you look at the pressure distribution in the wake too? If you/or your software chooses the control volume (air volume affected by the wing) too small, you will get discontinuity at the boundaries of your control volume.
In a wind tunnel with a limited test section area, you would need to measure the pressure distribution along the walls as well, and account for them in your overall model, to better predict total drag.

[hardingfv32]

The model with the Ducts on the pylon had about 13 to 20% more down-force but only about 6% more drag. So it is basically a blown wing.

Assuming your data is correct:

1) What is the basic step strategy for this system? With the system off, will the RW wing be at optimum down-force or below optimum?

2) Could the wing be in a stall condition when the system is off and the blowing helps with reattachment, eliminating the stall?

3) Would it be correct to assume the system is off (not blowing) on the straights?

Brian

[n smikle]

Thanks for your answers, but I fear that only integrating the surface pressure distributions over the area of the wing will not give accurate values for overall drag.

This is the simplest way I know to get the resultant forces over the wing, so I just work with it.

It's not that you make anything wrong, rather then a limitation of the method your software uses.
The effects of induced drag will not be properly represented by only integrating surface pressures.
Are your values "time averaged" values, or are you (your software) assuming a steady flow condition at all times?

I take it as a top down approach and take the wing as a simple free-body. I take it as a scientific truth that he force felt by the wing is only the differences in air pressures across it, whether it is induced drag, from drag etc.. the way the force is developed I am not really focusing on that and it will going too deep.

Effects like vortex shedding etc, and the flow in the "dead zone"/separation areas are highly unsteady in nature, so can be only represented as a time averaged value in terms of their contributions for a single Cl or Cd value.

I can do a transient analysis, but every time point will have its own set of data so that would be gigabytes of data. For example I could accelerate the car up to 200km/hr then save the results of say every second and and review the data, but for my intentions I don't think I will learn anything more than a steady state test. There might be some significant transient effects that Lotus have to deal with in real life to get real world tenths of a seconds, but I think a simple CFD demonstration should be good enough to get a basic Idea of what Lotus is doing. I think of it as a straight line test at constant speed.

Nevertheless, keep up the good work, I like to see the evolution of your models, just keep in mind that every tool has it's limitations.
How large is the "control volume" you chose in your model? Do you look at the pressure distribution in the wake too? If you/or your software chooses the control volume (air volume affected by the wing) too small, you will get discontinuity at the boundaries of your control volume.
In a wind tunnel with a limited test section area, you would need to measure the pressure distribution along the walls as well, and account for them in your overall model, to better predict total drag.

gato azul
43

I set the control volume a few meters ahead and behind the car.. not much above and below though. It just takes more calculation time if I make it bigger the results will be different but I don't think the story will change that much.

[amc]

Great data Gato Azul. I think that it's close enough to a modern F1 rear wing to give us a decent picture of what happens when one stalls - of course, the fact it uses angle of attack skews things a bit, but since nobody has yet published a paper on: "The effects of barely understood rear wing drag reduction mechanisms on the downforce and drag produced by the Lotus E20 Renault Formula 1 car" (thesis title anyone?) I think it's good enough.

I said before that I didn't think the wing was stalled - that VD just increased static pressure behind the wing. This would appear to contradict the stall theory where we think that the boundary layer separates below the wing, reducing the lift produced and therefore induced drag. However, what if the boundary layer has been separated (the wing is stalled) but there is still a boundary layer in place...

Because the wing's AoA does not change, the only thing that can deflect the boundary layer is something else taking the place of the boundary layer, right? (Probably not - open to criticism...) If the air from the pylon slits becomes the new boundary layer, then the previous one will be deflected upwards less, but what does this mean? Less static pressure behind the wing (greater volume of air behind it), and less upwards deflection because the air from ‘exit 2’ is in the way. Once again, Bernoulli and Newton explain the same thing in different ways.

Regardless of the scientist you prefer, in order to get the required 10% drag reduction from the rear wing, its lift needs to be reduced by about 5% overall (I sense you were getting at that Gato). In the small area that VD seems to affect, the 25% drag reduction (quoted earlier, for a 280mm width of the wing) means that lift needs to be reduced by just over 13%. This assumes that all of the drag is induced, which it is not, maybe only 80% or so is, but the figures are good enough for now. Reducing lift by 13% doesn’t seem to be enough for a true stall.

If I have seen further than others it is by standing on their shoulders... so apologies if it hurt too much.

[bhallg2k]

Make the "plenum" a resonance chamber so that the output of the wing jet ("Exit 2") is pulsed, thereby amplifying its effect on the wing for any given vehicle airspeed in which the system is in operation.

http://ftp.rta.nato.int/public//PubFull ... PSF-05.pdf

[hardingfv32]

I said before that I didn't think the wing was stalled - that VD just increased static pressure behind the wing. This would appear to contradict the stall theory where we think that the boundary layer separates below the wing, reducing the lift produced and therefore induced drag. However, what if the boundary layer has been separated (the wing is stalled) but there is still a boundary layer in place...

So assuming the system is not for stalling:

1) What is the basic step strategy for this system? With the system off, the RW wing be stalled and switching the system on eliminates the stall?

2) Could the system be forming a vortex that reattaches the flow to the bottom of the wing?

3) Would it be correct to assume the system is off (not blowing) on the straights?

Brian

[N12ck]

just had a thought, if these slots are not for stalling, then why take the air from the roll hoop, why not just make a classic 15cm slot in the middle of the wing to make more downforce (keep flow attached) or just take air from above the wing (high pressure and blow it in the middle 15cm)

I cant work out why you would make all this ducting if all you are after is effectively trying to keep air more attached

[bhallg2k]

Don't confuse the fact that the wing is not being stalled with the idea that the system is in place to keep air flow attached.

[N12ck]

Don't confuse the fact that the wing is not being stalled with the idea that the system is in place to keep air flow attached.

That is what I am saying, read my previous post....

[bhallg2k]

That was mostly for Brian, who apparently missed everything in the post he quoted except for that which he quoted.