F1technical.net

I'm about to get a new pc, so perhaps the simulation could be done on it. Would 6GB suffice? Given of course I could somehow get the software.

I don't know if it's worth it, maybe it's far off already no? i've got 8gb

turbof1 wrote:I'm about to get a new pc, so perhaps the simulation could be done on it. Would 6GB suffice? Given of course I could somehow get the software.

Memory is dirt cheap these days, and 6GB hardly cuts it at all these days. I'd suggest going with 16GB(or possibly even more)

chuckdanny wrote:I will not be able to do the asymetrical model, it would need to double the cell budget

I don't know what software you use, but there are solvers for cornering flow available.

Absent that, I wonder if we can still identify trends like this...



I'm 99.99999999999999999999999999999999999999% sure that the entire purpose of the widened "arches" is to reduce the amount of steering angle necessary to unblock the end plate. And if it's not feasible to simulate an angled front wheel, maybe just moving it over, or otherwise manipulating its dimensions, might do the trick. It won't be definitive by any stretch, but still.

The underwing strakes, however, are another story altogether. Without a study of transient events, I don't think we can even point to a trend.

Most folks seem to think these are vortex generators. I don't.



I think they're slot-gap separators that also function as a fulcrum about which the flap pivots under load. Like this...



If that slot closes, the wing will "stall," and that means the flap won't pivot. (It would also likely cause instability on corner entry.)

The strakes are only ajoined to the main plane in order to accommodate the flap as it moves down...



That wasn't always the case...



Those strakes can certainly help "connect" flow from the wing to the "pincers" to be directed away from the front edge of the floor. Like a more regulated version of this...



It's also possible they increase the strength of the end plate vortex by reducing the dynamic pressure of the flow under the wing that merges with air flow over the wing.

In a left turn, for instance, flow will shift to the right. The strakes will selectively exclude air flow from under the wing, thereby increasing the pressure differential between flow under the wing and flow over the wing, which will strengthen the vortex.



Speculative? Absolutely.

EDIT: For future reference, in case I revert to another way of naming things, what you call the "gutter," I call the footplate; the "pincer" is a brake duct turning vane; and I consider the "v-section" to be the beginning of the end plate.

Ben the Almighty wrote:Most folks seem to think these are vortex generators. I don't.

Nope they are not. I've come across those things multiple times. They are actually the underbody strakes completely stretching inbetween the slots on top of an element.

I think these are structural supports along being slotgap seperators. Perhaps they are vital in flexing the wing.

EDIT: Perhaps... . Most likely. As you said to stop the elements from closing the slots. The second and thirth elements are connected by these 3 underbody strakes. The upper flaps however are connected to the mainplane elements by the flap adjuster structure. It might suggest a different flexing is going on there.

bhall II wrote: It's puzzling to see so much air flow going inside the wheel given a wheel placement that more closely mimics a 2009-2013 no-doubt-about-it outwash wing. That's a head-scratcher. (Or maybe I just need to wash my hair.)

I think there is no doubt that the wing is designed to flow air outwards. I'm minded to remember the video of the butterfly going through the upper turning vanes on the Lotus. The acceleration across the front, and outboard of, the front tyre was very marked. That isn't going to be an isolated flow structure. I note that this sort of flow isn't obviously identified in this study which suggests that we're missing some detail that matters.

I also think there is no doubt that some flow will go inwards (it has to because most of the wing's flow is inside of the wheel). That's why we see complex "brake duct" geometries inside of the wheel - they're trying to trap/direct/tidy up some flow structure that is coming off the front wing. Perhaps one of the flow structures seen on this work is similar to the real thing in that regard.

I also think the turning vanes below the forward chassis will have an effect on the flow behind the wheel. This will affect the flow in front of the wheel.

One thing this piece of work has done is reiterate how complex the front wing's aero is; also, how easily small variations might have a large effect on the overall flow field around the rest of the car. I congratulate chuckdanny; I just wish we could give him access to some industrial-level IT kit so the model could be developed so much further.

That's ridiculous, i don't know what the fia flap test were but it pushes mercedes to effectively put a slotgap separator between 1st and 2nd flap while they stall the wing by closing 1st flap/3rd element gap.

I don't know what software you use, but there are solvers for cornering flow available.

I thought it was still only in the research phase, interesting. While openfoam is free this module seems to be totalsim property and not available though. It must really shine with the whole car, a yaw angle should suffice for the truncated front only model

For me the footplate is composed of the gutter and the slotted part between arches and the endplate so we need to give it a specific name to make the distinction. Well i know nobody but me calls it like that but it is very similar in shape no? not in function of course.

The strakes might help the wing to recover from a 3rd element stalled state maybe also. And determining the stalled area more precisely.

What do you think of the idea that the v-section might control the positioning of the turning winglets vortices?


Interestingly on the R28 the clockwise vortex is directed on the tire tread near the flank. That's why i think with a counterclockwise vortex you should direct it toward the tire flank instead to have a similar effect on the tire wake trailing vortex.

Regarding the inflow paradox, is it not precisely because you prevent a certain amount of air from flowing behind by outwashing it that together with the endplate bending you create this inwash ?

@wesley
my motherboard is limited to 8Gb, i'm not very good at soldering.

chuckdanny wrote:That's ridiculous, i don't know what the fia flap test were but it pushes mercedes to effectively put a slotgap separator between 1st and 2nd flap while they stall the wing by closing 1st flap/3rd element gap.

[...]

The strakes might help the wing to recover from a 3rd element stalled state maybe also. And determining the stalled area more precisely.

I don't think teams are necessarily stalling the wings as much as they're just reducing induced drag along straights. It's kinda hard to get away with much more than that these days.

That said, I think you're absolutely correct about the stabilizing nature of the strakes. Even without a stall, flow patterns will change on a wing that has a dynamic AoA.

chuckdanny wrote:What do you think of the idea that the v-section might control the positioning of the turning winglets vortices?

I still don't see the "v-section" as being anything more than the boundary between the wing and the end plate.

To create a strong vortex, you've got to turn the flow so it will accelerate. In an ideal world, you'd rely solely on the cumulative camber of all wing elements to provide that turn, because it would maximize effective wingspan, i.e. create higher peak downforce.



The problem is the adverse pressure gradient that exists between the trailing edge of the wing and the area in front of the spinning wheels. It pushes the suction peak/separation point forward, and thus down, which greatly reduces the efficiency of the wing. The solution then is to keep the flow low and turn it outward instead. This allows more consistency, but it comes at the expense of a reduced effective wingspan, i.e. creates lower peak downforce.



As I said at one point to Turbo, Ferrari was criticized for having a "simple" wing on the F138. In reality, that wing was aggressive as hell, because it aimed for the ideal.

In any case, I think the Mercedes wing is a further optimization of the compromise solution. When the wheels are steered straight ahead, they form downstream blockages that hugely reduce the efficiency of the wing. Upon further reflection, though, the "paradoxical" inwash flow in this situation actually makes sense, because even though the wing isn't doing much in this configuration, it's still doing something, and it seems logical for air flow to be pulled inward as a result. (J_a_f figured that out, too.)

When steering angle changes, the outer wheel moves such that it unblocks the outer-most portion of the end plate, and vortical flow will follow the outward path (blue). The inner wheel moves such that it unblocks the inner-most portion of the end plate, and vortical flow will follow the the wing's camber (green). The update specifically enhances this effect. Or so my logic goes.



That's the reason for the angled trailing edge on the vertical end plate. It accommodates the various separation points for the end plate vortex as it shifts with steering angle changes. Or, again, so my logic goes.



(I feel like I'm getting better with that spiel every time I repeat it.)

Now I'm gonna mess with some noodles...

I think every explanation I've read for blown axles/hubs is wrong.

How does the system work ? The oversized brake duct captures the airflow and splits it into two streams. Part of the air helps cool the carbon disk, while the remainder feeds a channel running through the spinning axle and exits out of the centre of the hollow wheel-nut. As a consequence, the stream creates a low-pressure zone which pushes the wake coming off the twisted front-wing endplates further outward. This in turn reduces drag and channels the airflow towards the sidepods and crucially, towards the diffuser. Last year’s technical regulations led to narrower front wings, which meant the airflow started to decrease in intensity. This is why, in order to make up for lost ground and have the endplates less curved, aerodynamicists re-introduced the open front wheel hub design.

What makes more sense? Increasing the size of the wheels' wake, which increases the size of the downstream blockage the end plates must overcome and increases overall form drag, all with the hope of affecting an interaction that's ~3m away?



Or does it make more sense to use the pumping action of the spinning wheels to mitigate the interference caused by the wake and literally pull air flow from the wing and then push it outboard so it doesn't adversely affect the sealing vortices aimed at the front of the floor?



You be the judge.

(Actually, I shouldn't say "every explanation I've read," because that implies I read a lot of them, and I don't since they're most often wrong. So, it's entirely possible someone has already identified this.)

EDIT:

turbof1 wrote:EDIT: Perhaps... . Most likely. As you said to stop the elements from closing the slots. The second and thirth elements are connected by these 3 underbody strakes. The upper flaps however are connected to the mainplane elements by the flap adjuster structure. It might suggest a different flexing is going on there.

If they were just the leading edges of the strakes mounted to the main plane underneath, then you'd try to avoid having them interfere with the underside of the flap elements since the underside of an airfoil is where the magic happens.



For the exact same reason, if they were just slot-gap separators or any other sort of structural support, they'd most likely look like this:

I got sidetracked before and totally missed these.

Just_a_fan wrote:

bhall II wrote: It's puzzling to see so much air flow going inside the wheel given a wheel placement that more closely mimics a 2009-2013 no-doubt-about-it outwash wing. That's a head-scratcher. (Or maybe I just need to wash my hair.)

I think there is no doubt that the wing is designed to flow air outwards. I'm minded to remember the video of the butterfly going through the upper turning vanes...I note that this sort of flow isn't obviously identified in this study which suggests that we're missing some detail that matters.

Probably not a fair comparison since the Lotus had a longer wingspan...

I also think there is no doubt that some flow will go inwards (it has to because most of the wing's flow is inside of the wheel). That's why we see complex "brake duct" geometries inside of the wheel - they're trying to trap/direct/tidy up some flow structure that is coming off the front wing. Perhaps one of the flow structures seen on this work is similar to the real thing in that regard.

In my view, only Y250 flow should pass between the wheels. Everything else should be directed outboard by the contours of the wing. (That doesn't mean it happens, though.)



Still, I think I understand the inward flow. It's just the result of lower pressure behind the wing than free stream flow outside the wheels as the car's driven straight. It will change in yaw.

chuckdanny wrote:Interestingly on the R28 the clockwise vortex is directed on the tire tread near the flank. That's why i think with a counterclockwise vortex you should direct it toward the tire flank instead to have a similar effect on the tire wake trailing vortex.

Regarding the inflow paradox, is it not precisely because you prevent a certain amount of air from flowing behind by outwashing it that together with the endplate bending you create this inwash ?

I'm not sure what you mean here. Elaborate?

bhall II wrote: Probably not a fair comparison since the Lotus had a longer wingspan...

Yes, and the flow was turned even further across the tyre. The position of the Lotus turning vane is similar, in relation to the front tyre, to that on current cars.

One further odd thought that came to mind - could one run a vortex (a low pressure zone don't forget) across the tyre in order to reduce the adverse pressure gradient behind the front wing? Not sure if it would be possible, whether it would work or whether it would be worth the effort.

Just_a_fan wrote:

bhall II wrote: Probably not a fair comparison since the Lotus had a longer wingspan...

Yes, and the flow was turned even further across the tyre. The position of the Lotus turning vane is similar, in relation to the front tyre, to that on current cars.

You can sorta see where maybe a vortex has formed (top).

chuckdanny wrote:

It's tough to really judge without a spinning wheel behind it.

...could one run a vortex (a low pressure zone don't forget) across the tyre in order to reduce the adverse pressure gradient behind the front wing?

I'm not exactly sure what you have in mind here. But, I will say that if you get it right, I think the adverse pressure gradient can be beneficial.

Its negative impact on efficiency is greatest when the car is traveling in a straight line with very little need for downforce. In those instances, if the wing doesn't operate at peak efficiency, then it also doesn't create peak drag.

Like justafan says one of the purpose of this venturi twister must be to reduce adverse pressure gradient.
The point to clarify here is whether the strength of a vortex increase only by allowing it to develop further downstream or if this arches/slot mechanism of high pressure release that feeds it plus the whole car aerodynamic downstream is replacing the freely developping case.
A constraint may accelerate the flow like a leading edge does.

For my r28 bold allegation i was just observing the fact that they used to bend an endplate vortex on top of the wheel tread like those turning winglet and that the v-section canyon or riverbed could direct and bend it at a similar place. But both kind of vortices have opposite sens so it couldn't produce the same result being thrown in the same place unless...
But i'm not sure of the purpose (i've got to test this) whether it reduces the acceleration on top of the wheel hence the lift or is it changing the flow structure upstream so that it changes the tire trailing vortex downstream.

For the 2nd floor outwash inducing or participating to the inflow pattern of the 1st floor lets think of a case where there is no outwash that means that part of this flow goes inside the wheel instead of the flow of the 1st floor so you exchange one for the other in a very naiv probably way of explaining what my model is telling me maybe wrongly.

If we think of a simple classical sloted multi element wing that wouldn't be bent at it's extremities and with a span greater than the wheel track by a large margin. THe flow acceleration pattern would be altered by the tire where a symetrical wake of high static pressure would develop that create a perfect blocage that would be a barrage for an inflow pattern. The simple fact that there are tires create a confinement and then an acceleration. On my model tunnel walls are a bit too close so that there is the same effect on the outside part of the tire which should vanish or reduce with far wall hence increasing the pressure difference between outside and inside flow. That's why also i think that the part of the outwashed flow that comes from over the wing that is slowed down and build pressure participate to the pressure difference. Air flows from high to low pressure.
Those pattern are established in a transient way and we are just seing the end result.
I'm not cfd specialist but although called steady the algorithm uses virtual time step it doesn't get rid of the time completly at least in turbulent region because turbulence is unsteady by nature, that's just that those sometimes variable time step are not always physical but we need the explanation of a specialist here i'm getting out of the envelop or blackbox, i may crash.
The arches vortex could also inwash (wash in machin) the part of the wake that builds in front of the tire at the 1st floor level.
Now i'm tired of my own naiv explanations i've got to have a rest.

I wonder if vortex formation at micro time scale is not like those fireworks with orthoradial microimpulse of high pressure release that twist or turn the core.

(Wo, with all those pictures the page load in minutes, i don't like pictures anymore)

The point I was trying to make about the trajectory of the turning vane vortex is that, in reality, it's going to be strongly affected by the flow patterns associated with a spinning wheel, which we don't have. You can see that it tries to go outboard, but is diverted by something. On a more representative model, I think we'd more or less see the same result as the "organic" FloVis on the Lotus E21, if not a little weaker, because post-2013 front wings are weaker all around.



That conclusion based on the fact that the vortex would contribute to lift if it passed over the wheel as depicted, and no one wants lift on a Formula One car.

There's a significant language barrier here. I've read through everything at least a half-dozen times, and I don't know that I'm any closer to understanding it. So, bear with me if I've misinterpreted things.

chuckdanny wrote:Like justafan says one of the purpose of this venturi twister must be to reduce adverse pressure gradient...

Why? I don't see a need for it. The adverse pressure gradient is resolved when steering angle changes. Otherwise, it's favorable because cars don't need downforce on straights.

For my r28 bold allegation i was just observing the fact that they used to bend an endplate vortex on top of the wheel tread like those turning winglet and that the v-section canyon or riverbed could direct and bend it at a similar place...

I still can't see the logic behind somehow using the "v-section" for any sort of directional control. Why would you reduce the wing's effective span to create an effect that could easily be facilitated by simply adding another turning vane?



(By the way, what is R28?)

For the 2nd floor outwash inducing or participating to the inflow pattern of the 1st floor lets think of a case where there is no outwash that means that part of this flow goes inside the wheel instead of the flow of the 1st floor so you exchange one for the other in a very naiv probably way of explaining what my model is telling me maybe wrongly.

I'm totally lost here, because I don't know what you mean by "1st floor" and "2nd floor." If it's about the "paradoxical" inwash flow I mentioned, yeah, I get it now.

To echo what J_a_f said a bit earlier, your model looks great, and the CFD results are (surprisingly) compelling. It just seems like you might be reading into the results a bit more than is prudent. That's not necessarily to say you're wrong. But, I think simulated steering angle changes would make the purpose of the "arches" crystal clear; a spinning wheel would demonstrate how the turning vane vortices "connect" to the trailing edge of the end plates to increase dynamic pressure in order to strengthen end plate vortices; and incorporating the rest of the car would illustrate how allowing "outwashed flow that comes from over the wing [to slow] down and build pressure" would likely just be a source of drag.

Nearly everything aerodynamacists do these days is predicated upon, and specifically designed for, transient conditions that are a pain in the ass to model. Steady-state downforce is easy; dynamic downforce is not, but it's the whole ballgame. It's the difference between Mercedes and Marussia.

Bhall, i can not answer all now, i will try later.
The brake duct turning vane are catching a vortex, i think we agree on this. It's not coming from the strakes you said it. In order to do this it must be fully developped even when wheels are straight at least for the outer wheel of the turn because if the purpose is to keep it close to the trailing edge of the tire its because it has a tendency of going elsewhere. Then if its not there when the wheel turn then there is no way it can catch it.
The r28 is the renault f1 of 2008 from which i posted a cfd picture. You can see that they purposely (dented ramped endplate) design the endplate to direct a vortex on top of the tire.