W06 Influenced CFD Study

[chuckdanny]

Thanks to eyalynf1 i've read this post

The high angle of attack is fine at least for the wing's performance. Considering your mounting points are close together near the center, it makes sense structurally to have your maximum downforce in the center of the wing. It also makes sense 3D lift distribution wise as you reduce your induced drag at the ends of the wing, reducing the negative interactions of the flow with the tire. And yes, the inward moving vortex is high strength and useful to the undercut and batwing.

When you have high number of cambered elements, the end camber looks extremely steep, but the slots help refresh the boundary layer to each succeeding element as well as allow off the surface pressure recovery. You have a little separation, but you are still increasing the downforce with additional AOA, not even close to stalled

Here's an example of an extremely old 7 element design.
http://www.boatdesign.net/forums/attach ... age236.gif

Look familiar?

Nowadays, we are likely well into the 5.0 or even 6.0 C_l range.

And amazingly i'm right in the range 5.42 C_l.
Donwforce 130 kg
drag 24 kg
It seems small though (but there is no nose) and i don't expect to be anywhere close to the real thing, still surprisingly close C_l.

[bhall II]

Bhall, this venturi twister is highly optimised, they have a parametric model as opposed to mine (i would have to be paid for that ) and maybe you can get both caracteristics right, a quickly bursting vortex still producing very low pressure. This thing convert venturi in ground effect plus pressure over the wing in streamwise rotational speed.

I'm not sure what you mean by "convert venturi in ground effect," but is it possible that the portion of your statement I've underlined more or less agrees with my quoted statement? I'm not asking to prove a point; I just want to know if perhaps we're saying the same thing in different ways.

...the idea [with the turning vane] is to increase the dynamic pressure of air flow over the wing in order to create a stronger vortex when it merges with air flow from under the wing.

By the way, what do you mean by "worm"? "Sens"?

Here's a link to a free CAD rendering of the W06's first wing: https://grabcad.com/library/mercedes-am ... 6-hybrid-1



It includes the nosecone, under-chassis turning vanes, and updated camera pods. Maybe you can run it for comparison. If not, maybe you can incorporate the nosecone into your model. (If you search "f1" on that site, there are multiple renderings of an entire car.)

Also, if you want to send your CAD files to me, I can probably get a real-life aerospace engineer to run your model through his CFD software.

EDIT: It might not be a bad idea to avoid CFD entirely until you've gone as far as you're willing to go with the CAD model. That will reduce opportunities for confirmation bias.

[chuckdanny]

The mp4 30 is very detailed, nice job. I can do the nose its just that i don't have enough cell budget.
The y250 steer the flow from over the long nose that goes under into the S-duct.
The worm is the I branch on the article from gordon mccabe.

Well by convert "venturi in ground effect" i mean that the whole now squared nested arches thing is made of a convergent, a throat and a divergent (is it wrong?) and that part of the accelerated air flow will be converted into the core of the vortex.
The high pressure feeding slots further enhancing the twist and i didn't take this into account but you're right the turning vane vortices further increase the dynamic pressure over the wing that convert into almost stagnation pressure building and then pressure release into the slots that's what you mean? The gap of the slots is precisely dimensioned for this purpose i suppose.

[bhall II]

Well by convert "venturi in ground effect" i mean that the whole now squared nested arches thing is made of a convergent, a throat and a divergent (is it wrong?) and that part of the accelerated air flow will be converted into the core of the vortex.
The high pressure feeding slots further enhancing the twist and i didn't take this into account but you're right the turning vane vortices further increase the dynamic pressure over the wing that convert into almost stagnation pressure building and then pressure release into the slots that's what you mean? The gap of the slots is precisely dimensioned for this purpose i suppose.

I don't want to jump the gun here. But, it looks like we may now be in complete agreement!



Yes, every part of the wing not devoted to the Y250 vortex is aimed at consolidating flow with the main end plate vortex in order to strengthen it.

Once you add yaw to the equation, I think the intent of Mercedes' update becomes clear, and you can see how the "arches" are the latest in a long line of solutions implemented to promote consistency.

EDIT: I may have been too hasty earlier.

...stagnation pressure building and then pressure release into the slots that's what you mean? The gap of the slots is precisely dimensioned for this purpose i suppose.

Even though all air flow over and under the wing is directed toward the "arches," that feature should not be considered a plenum or accumulator of sorts. The "arches" form an end plate that redirects normal span-wise flow just like every other end plate. The difference is scope.


Yellow: air flow under the wing
Red: air flow over the wing
Striped: vortex formed from air flow under and over the wing
Purple: source of Y250
Everything: very, very generalized

When the tread of a tire is oriented directly behind, and perpendicular to, an end plate ("arches"), vortical separation is located at a point that's relatively low and forward, because the high-pressure area in front of the wheel, i.e. stagnant air, pushes the point of peak suction forward, thus lower. In other words, the vortex is in a weakened state. (But, it's important to realize the vortex is always there.)

When the wheels are steered, which removes the blockage caused by the stagnant air, the vortex is strengthened as the point of peak suction moves aft. That delays separation, which moves the separation point further back, thus higher. Whether the "tail" of the vortex follows an inboard or outboard path is determined by the direction the wheels are steered.

I don't know that I can ever illustrate it better than this.

[chuckdanny]

From what i understand as of now, the vortex state, its nature is highly determined by the geometry it arise from (quit obvious) and the tire downstream play its part of course.
What i mean by nature/state is whether it is a strong/weak vorticity, quickly bursting or not, thin until it burst or of a conical shape, etc... Part of this caracteristics maybe linked and defined by further geometry downstream, confinement etc...
WHat i saw on the wing without a wheel behind is that arches and gutter vortex are different, the gutter one narrower better defined and extending as long as there are cells available to pull it downstream.
While the arches one is different, it burst quickly. Hence i suppose that the geometry those 2 arise from is determining and i guess the quickly diverging part of the twister is the main reason.
Furthermore i wonder if the very nature of a venturi prevent this suction peak from shifting backward/forward up to a point that is without obstructing too much the outlet.
THe wheel is not turning much but for very slow corners. Yaw angle is 5° at max i heard would this change the vortex state?
I'm not saying you're wrong i'm skeptical (ok let's say i'm more diplomat ). For the moment, but of course it can be totally false, the arche vortex take all the space in the divergent part of the arches and i think that to go from a vortex that burst at the tire patch loosing its energy hence stagnation pressure hence droping drag from a vortex that developp downstream is narrower, become stronger etc... is a big ask nonwithstanding the change on the flow structure. But maybe you have some other sources ?
All this happening continuously that is especially under breaking whatever the wheel angle is, as soon as 1° for example.
What i call the gutter vortex would follow the same kind of behavior i suppose? And for me it make no sense because the study about wheel/wing interaction shows that a vortex going along the inside bottom flank of the tire participate to the treatment of the tire trailing vortices hence reduce drag, which behavior is highly needed at high speed, and change the inside spilling under the floor trailing vortex which is highly needed especially under braking.
But i recognize that the pincer may generate a vortex on its own.
the inside flow would create so much less drag than to let the twin repealing contra-rotating trailing vortices (thats a lot of ing thing) develop. And more clean air inside captured by the front wing may mean less total pressure drop at the back so consistency, higher downforce at the back from the efficient diffuser etc...
So... that's my maybe biased opinion but trust me i will not hide or attempt to steer my model if i see such behavior, quit the opposite i would be amazed and happy to show it.
But i'm not that naiv, maybe something is missing in my model or too raw a geometry or just that at this resolution it is just impossible to get those very subtle behaviors.
But whatever, your contribution has been tremendous, thank you!

[bhall II]

Oh, boy.

THe wheel is not turning much but for very slow corners. Yaw angle is 5° at max i heard would this change the vortex state?
I'm not saying you're wrong i'm skeptical (ok let's say i'm more diplomat ).

Have we really gone this whole time without it being patently clear that addressing the above issue is the central tenant to my theory? Here's a refresher, if you like.

Things to note as you move forward:

Try to keep it in mind that you're not actually testing this car...



...you're testing this "car," and the data will reflect it.



Context is everything...



And small changes can have large effects...


Result of 20mm tire deformation

Lastly, are you sure you can't think of a reason for how/why these might interact?

[chuckdanny]

nice one!

i'm sure if they could get rid of this ugly nose...


Or would they? In a way the nose is a big strake keeping the flow straight might it get in a bad mood and even down to counteract an upwash together with the wishbones fairings?

Yeah right, the nose is unavoidable for a test in yaw.
Quit a windy day on those sauber picture no? A storm?

[bhall II]

Quit a windy day on those sauber picture no? A storm?

I think that's probably as close as we'll ever get to an actual cornering simulation from a team. I'd love to see more, but I imagine other teams would, too.

As an experiment, try to find a way to radically alter just the "yellow" vortex. How does that then affect the "green" vortex? Maybe you'll see why I've previously referred to the "yellow" one as a "pilot light."

Also, the inner portion of the footplate noted with a white arrow should be level along the bottom, not angled.



You're definitely going to need to add something to your model, because wheel wake doesn't appear to be representative. The solution could be as easy as correcting the dimensions and orientation of the wheels or maybe adding "pincers." But, you'll likely need to separate them as a chassis would, since the "stagnation zones" just ahead of the sidepods play a significant role.



As trivial as it may seem, it wouldn't hurt to shore this up, too.



And if you haven't already read through it, I highly recommend this study of the aerodynamics of an inverted wing and a rotating wheel in ground effect.

[chuckdanny]

I've modified wheel, endplate, shored up gurney hole, corrected the angled gutter and with a buggy cad that's quit a pain i must say.
Thanks for the study, its hard to sleep sometimes

Tell me if you see something wrong before i CFDfy it!



I don't understand your question regarding vortices interaction do you want for example to artificially modify it with some turning winglet or while turning wheels?

Edit :
Ok, your idea is that the gutter vortex reinforce on outside wheel when turning so that it attract the arches vortex to the outside while at the opposite, the inside wheel, the gutter vortex loses strength together with the inside part of the wheel opening the door reducing adverse pressure etc... That's attractive

[bhall II]

I don't know if that's my idea or not, because I'm not exactly sure what you mean.

Also, I don't know if it's just me being nitpicky or not, but these wings only superficially resemble one another.



It seems we might be heading in the wrong direction. (Or maybe I've just turned this into some sort of referendum that it's not.)

Carry on.

[chuckdanny]

Of course! its an improvement!
I thought this wing was under british influence, a bit too exotic, rounded everywhere etc...
Squares, corners work well also in aerodynamics, this is more german and it has the property to be better align with orthogonal hexahedral mesh

[RicME85]

Also, I don't know if it's just me being nitpicky or not, but these wings only superficially resemble one another.

I agree, there are still a few issues with the shape although it is a good representation, it cannot be easy producing this.

The winglet that is attached to the endplate needs a bit more of a curve to represent the real thing IMO.

chuck - have you not been able to look at the model of the wing and nose that is available to download? Also, have you thought about speaking to Julien from KVRC? He might be interested in the CFD side of things, could help with his product etc.

[chuckdanny]

Its not that far off, its bhall that is trickering with perspective



I didn't merge the multybody and applied anti-interference fillets

The grabifyoucan thing is not available, its just a sample like for parfum.

[RicME85]

That looks better, the first image it looked lit it was just a diagonal line

[chuckdanny]

Yes, it must be related to tessellation precision which is not at the max quality plus zoom level etc...
My cad is completly bugged, i cannot add guide to modify loft/sweep upstream in history, it creates them downstream so that i can't use them, connectors defining seed and lofting sequence are changing for no reason what a mess!