W06 Front Wing Discussion

[chuckdanny]

You know absolutly nothing about methodology, do you think they put a new wing concept and bang it's spot on!
Your are trying to make an impression with your monkey and --- in every discussion, you're a child! Ridiculous.

[eyalynf1]

I have to say the non-stop assault on the CFD efforts here suck all the fun out of this.

We need to take a cue from Bernie on this; We are here to be entertained! Why so serious? If I really, really, really want to get down to business on this stuff, I'll enroll in an aeronautical engineering graduate program. I would guess that most people are satisfied to spectate on the efforts and engage in "shop talk" on the principles involved. No disparagement, discouragement, or witty, pithy put downs are necessary.

And even on the substance of the criticisms about potential accuracy, the teams themselves get things wrong all the time. They miss correlation of their CFD models to the way the real wing works all the time! The evidence is the fact that they drive around with those ridiculously unsafe pitot grids attached to the car to sort out their models out. They often bring new aero bits to the track that just don't f-ing work! We are still entertained, are we not? How is this different from one of us missing how the real wing works vs. how his CFD one does?

It is a common theme on these forums, someone puts forth extraordinary efforts to understand this subject we all love, and is rewarded with scorn from the self-appointed gatekeepers of authenticity.

I understand that certain members may be well regarded, but that is no excuse for the kinds of high-handed hazing I've read on the past few pages. I feel the moderators should be shutting this kind thing down, not sympathizing with a wink and a nod.

[chuckdanny]

Thanks eyalynf1 i appreciate, i don't understand why he is insulting me all the time, i never pretended to show the truth of the w06 wing to the world, it's an attempt, of course they are many pitfalls and basically i know all of that because as a monkey i'm equipped with a basic logic system that tells me : if they are using massive computing power, throwing millions at it how could i get the same results with my low res experiment hmm ?

He's telling me very seriously : they are no slots in your footplate, can you believe this! I've done the cad, i've got eyeballs mark I.

I'm not pretending to be an aerodynamicist because i can set up a cfd model, that's ridiculous. I know all the hard work it is to really master this science, i'm very far from it.

[RicME85]

Would be interesting to get the front wing from the W06 model in Codemasters F1 2015 when it is released. Their models are signed off by the teams for accuracy and use some CAD data. Only thing is it would be the Melbourne spec wing and will need some work on it (the models when extracted are minute, a whole car is about a foot in length)

[bhall II]

Do you really think that the best way to have the big picture of a complex interacting system is to put all at once and either it works and you don't know why or it diverge (hit or miss like Platinumzealot says)

YES! That's the whole --- point!

To analyze a system that exists in the real world is to somehow faithfully copy it all as thoroughly as possible and then see what happens. You learn nothing if you manipulate the parameters in order to fulfill some predetermined idea about what the outcome should be. That results in what's called confirmation bias.

That the complexity of the system in question is so great is precisely why I've been saying this cannot be done.

And you say I'm the one who doesn't understand appropriate methodology?

Seriously, am I being "punked" here? Is this the Twilight Zone?

How in the world can someone claim to understand the limitations of a simulation...

He's telling me very seriously : they are no slots in your footplate, can you believe this! I've done the cad, i've got eyeballs mark I.

...and yet still somehow try to draw real-world conclusions from it anyway?

Actually the suction peak is not where you think it is like on an old school wing. Ok i know the footplate is slotted and don't extend like that under the arches but...

Cognitive dissonance: psychological conflict resulting from simultaneously held incongruous beliefs and attitude.

I need a drink.

EDIT: Can I get a mod to ban me for a few days to eliminate any temptations I might have to keep this up? At this point, I think I may need some help with that!

[turbof1]

After consulting with Steven, we decided to split the discussion off into the engineering projects subforum.

The reason so is that we indeed will not be able to simulate the wing's real life counterpart to its exact flow details. However, we might learn a few patterns and a few of the solution the wing applies.

Furthermore, splitting the topic and giving the topic title a subtle different name, will protect it from ranting that "it isn't worthwhile since you can't ever simulate the real deal". This topic descended into bickering due insistance of how "wrong" the project was. It is not wrong, and honestly I really don't appreciate the smear campaign against it.

This means all together:
-Chuckdanny will continue on his fine work here: http://www.f1technical.net/forum/viewto ... 14&t=22203, in all earnest.
-Criticism is allowed, but it has to helping the project, not work against him.
-Effectively, any ranting that hurts the project will be removed!

I'll clean up this topic later on. In all honesty, there's only so much crap one can digest at a time...

[bhall II]

Furthermore, splitting the topic and giving the topic title a subtle different name, will protect it from ranting that "it isn't worthwhile since you can't ever simulate the real deal". This topic descended into bickering due insistance of how "wrong" the project was. It is not wrong, and honestly I really don't appreciate the smear campaign against it.

What you call a "smear campaign," I call a "reality check."

If the idea here on this forum is to learn/discuss the technical aspects of Formula One, then we ought to look at the actual technical aspects of Formula One. You wouldn't practice for a basketball game by taking a cooking class, would you?

The cavalier attitudes that supported this project in its original context betray a staggering inability for some folks to fully appreciate how even minute alterations can have significant results.

Without taking everything into account, nothing can be gleaned from an inadequate experiment, because aerodynamic phenomena don't always show an intuitive linear progression from A to Z; sometimes things go from A to orange, and ignoring that reality doesn't mean it just goes away.

For example, here's a snippet from a study that included research into the effects of tire deformation on downforce.

When the vehicle is cornering, accelerating, or decelerating, the tires are constantly deformed due to vertical, lateral, and longitudinal forces.

[Methodology, including acquisition of real-world track data that showed a 20mm tire deformation]

The measured front tire shape was analyzed using CFD. This analysis showed that, in comparison to a tire with no side force acting on it, approximately 5% of the vehicle’s downforce was lost when a side force of 9000 N acted on the tire.

Figure 17 shows the total pressure distribution close to the road surface with and without a side force acting on the tire. The results show that the position of the separation point on the outboard-side wall of the tire moves back significantly when a side force acts on the tire. This backwards shift of the separation point changes the circulation around the tire in the XY planes, and the tire wake which previously flowed to the outboard of the vehicle now flows under the vehicle.

The fact that this reduces the dynamic pressure underneath the vehicle, resulting in a decline in downforce, can be seen from the change in the static pressure underneath the vehicle when the tire goes from a state of no side force to one in which side force is acting.

[...]

That's a wholesale change in flow characteristics, and it was all because of a deformation that measured a scant 20mm. What do you think will happen to a guesswork model with even more variance?

The results won't even be indicative of Formula One, much less a specific car, and no one will have learned anything in the process.

I fully acknowledge that I've been an asshole here, and I apologize for the troubles. But, I defy anyone to bear witness to a flagrant display of willful ignorance and not get pissed off.

[eyalynf1]

Furthermore, splitting the topic and giving the topic title a subtle different name, will protect it from ranting that "it isn't worthwhile since you can't ever simulate the real deal". This topic descended into bickering due insistance of how "wrong" the project was. It is not wrong, and honestly I really don't appreciate the smear campaign against it.

What you call a "smear campaign," I call a "reality check."

If the idea here on this forum is to learn/discuss the technical aspects of Formula One, then we ought to look at the actual technical aspects of Formula One. You wouldn't practice for a basketball game by taking a cooking class, would you?

The cavalier attitudes that supported this project in its original context betray a staggering inability for some folks to fully appreciate how even minute alterations can have significant results.

Without taking everything into account, nothing can be gleaned from an inadequate experiment, because aerodynamic phenomena don't always show an intuitive linear progression from A to Z; sometimes things go from A to orange, and ignoring that reality doesn't mean it just goes away.

For example, here's a snippet from a study that included research into the effects of tire deformation on downforce.

When the vehicle is cornering, accelerating, or decelerating, the tires are constantly deformed due to vertical, lateral, and longitudinal forces.

[Methodology, including acquisition of real-world track data that showed a 20mm tire deformation]

The measured front tire shape was analyzed using CFD. This analysis showed that, in comparison to a tire with no side force acting on it, approximately 5% of the vehicle’s downforce was lost when a side force of 9000 N acted on the tire.

Figure 17 shows the total pressure distribution close to the road surface with and without a side force acting on the tire. The results show that the position of the separation point on the outboard-side wall of the tire moves back significantly when a side force acts on the tire. This backwards shift of the separation point changes the circulation around the tire in the XY planes, and the tire wake which previously flowed to the outboard of the vehicle now flows under the vehicle.

The fact that this reduces the dynamic pressure underneath the vehicle, resulting in a decline in downforce, can be seen from the change in the static pressure underneath the vehicle when the tire goes from a state of no side force to one in which side force is acting.

[...]

http://i.imgur.com/8ZQkKCE.jpg

That's a wholesale change in flow characteristics, and it was all because of a deformation that measured a scant 20mm. What do you think will happen to a guesswork model with even more variance?

The results won't even be indicative of Formula One, much less a specific car, and no one will have learned anything in the process.

I fully acknowledge that I've been an asshole here, and I apologize for the troubles. But, I defy anyone to bear witness to a flagrant display of willful ignorance and not get pissed off.

The attempt to develop a progressive refined simulation of a real world aerodynamic device is not "willful ignorance". We have no Idea how far ChuckDanny will choose to take his simulation. Experiments don't just arrive fully developed. They are performed, their results often found wanting, and then are refined. This iterative process entire, and not just the end result mind you, advances knowledge and insight of the subject at hand.

Your comments, while well mixed with legitimate constructive criticism, also come with heavy doses of rude and dismissive intemperance. I wouldn't consider that acceptable, even if you had offered matching efforts at explore/demonstrating the object of the study. But you haven't even done that. ChuckDanny has invited you to a dinner party, and you are insulting the cook, staff, and host, and you haven't even brought a token gift of wine.

Get pissed off as you like, but be polite and don't piss on others. Even my three year-old can manage not to do that.

[monsi]

I am reminded of the old aphorism - all models are wrong ... some models are useful

Regardless of the goodness of fit of this work to the real W06, it is interesting.

[Cold Fussion]

The attempt to develop a progressive refined simulation of a real world aerodynamic device is not "willful ignorance". We have no Idea how far ChuckDanny will choose to take his simulation. Experiments don't just arrive fully developed. They are performed, their results often found wanting, and then are refined. This iterative process entire, and not just the end result mind you, advances knowledge and insight of the subject at hand.

The whole point he is making is that you cannot model sub system A and then say this is what the system is doing, when you have neglected sub systems b through z. The effects of system is dependent on everything apart of it, and removing any one part of it means you have modeled a different system.

[chuckdanny]

Of course, i may be wrong (again, one day i will make a point by trying hard) but what your are discussing is basically the difference between linear systems and non linear. Linear systems are just the sum of there parts(subsystems, considering a defined objective) while non linear are more than the sum of there part (or less), the superposition principle doesn't apply. Subsystems are interacting, action/reaction. But it doesn't mean that you can't design a non linear system. And to design you must have laws that gives directions, trends, differential equations.
I suppose there are 1st order phenomena and 2nd 3rd etc, wheel wing interaction is of 1st order(greater interaction) and i have a cell budget limitation.
What we have here is a black box but because we can see the components we have partial informations on them.
It could stay black though

[bhall II]

I was recently reminded of a pressure map that, while far from definitive, may represent the closest I'll ever get to validating my theory.

As a refresher...

Everything beyond the suction peak is, by definition, an adverse pressure gradient. Otherwise, there would be no flow separation.

Because they constitute downstream blockages that inevitably slow down upstream flow, i.e. increase static pressure, the front wheels contribute to the separation of the vortices by pushing the suction peak forward, which reduces the efficiency of the wing.

It is my contention that the application of steering lock more or less removes those blockages to at least some degree, allowing the suction peak to travel aft as air flow accelerates. This increases the efficiency of the wing.

I think that enhancing this transient property is the whole point of the updated wing. I also think it's next to impossible to model this transient property without data we will almost certainly never see.

Within the box, you can see that pressure increases in the same direction as flow, forming the adverse pressure gradient that eventually causes flow separation.

You can see the high-pressure areas in front of the wheels that reduce the efficiency of the wing and contribute to separation.

While there's no steering lock applied to the wheels, this pressure map is taken from a simulation that modeled the effect of crosswind on the car. The resulting flow pattern removes the downstream blockages caused by the wheels in much the same way as the application of steering lock, and you can see reduced pressure in those areas indicative of faster/greater air flow.

These things more or less inform my theory that the wider end plates on the new wing increase efficiency by reducing the steering angle change required to remove detrimental blockages.

[aleks_ader]

I was recently reminded of a pressure map that, while far from definitive, may represent the closest I'll ever get to validating my theory.

As a refresher...

Everything beyond the suction peak is, by definition, an adverse pressure gradient. Otherwise, there would be no flow separation.

Because they constitute downstream blockages that inevitably slow down upstream flow, i.e. increase static pressure, the front wheels contribute to the separation of the vortices by pushing the suction peak forward, which reduces the efficiency of the wing.

It is my contention that the application of steering lock more or less removes those blockages to at least some degree, allowing the suction peak to travel aft as air flow accelerates. This increases the efficiency of the wing.

I think that enhancing this transient property is the whole point of the updated wing. I also think it's next to impossible to model this transient property without data we will almost certainly never see.

http://i.imgur.com/zGd4q6B.jpg

Within the box, you can see that pressure increases in the same direction as flow, forming the adverse pressure gradient that eventually causes flow separation.

You can see the high-pressure areas in front of the wheels that reduce the efficiency of the wing and contribute to separation.

While there's no steering lock applied to the wheels, this pressure map is taken from a simulation that modeled the effect of crosswind on the car. The resulting flow pattern removes the downstream blockages caused by the wheels in much the same way as the application of steering lock, and you can see reduced pressure in those areas indicative of faster/greater air flow.

These things more or less inform my theory that the wider end plates on the new wing increase efficiency by reducing the steering angle change required to remove detrimental blockages.

Those pressure map you talking about. Could you describe source or any other vital simulation information (car, conditions, air, boundarys) form this image?

[bhall II]

It's from a video that details Sauber's wind tunnel. The pressure map reflects forces acting upon the rolling road as a result of the car's aerodynamics. Engineers have to compensate for those forces to prevent lifting the belt.

[tuj]

I admire the spirit of the experiment. Unfortunately without the exact specs of the wing to the 1mm tolerance or better, you will not get very good results most likely.



Consider that you need to know pretty much every parameter of the aerofoil to get the proper results and behavior. And these are simple, static, well-known NACA designs. It is likely that Mercedes has developed their own set of aerofoils through testing (CFD and wind tunnel and track testing). If these aerofoils are deflecting under load (which they probably are), if the tires are deforming under load (which they most certainly are), then the CFD simulation becomes *VERY* difficult to properly account for all the variables.

Again, I admire the spirit of the project and have always been a big fan or AeroBytes in RCE. But what I have learned reading AeroBytes is that very small changes in pitch, ride height, Gurney, dive planes, etc, can sometimes make HUGE differences in the -L/D numbers. The devil really is in the details with aero.