Exhaust Blown Floor - Forward Exhaust Exit

[Giblet]

There is a reason we don't see any reports about Ringo, because he doesn't make things personal. He simply has a stance, wants to proof it, and a wants to discuss it. If you don't keep some kind of focus of scope, you'll rarely complete any good analysis, and often, even following the wrong tunnel until the end can yield the path to the right tunnel.

Heated discussion are often centered around his opinions and data, but he mostly stays level and clear, and rarely if ever personal in a negative way.

I don't know if he is right or wrong, and I honestly don't care, but I do know he knows how to contribute and discus a conversation.

People deserve recognition as much as some deserve warnings and moderation.

[shelly]

I do not see how ringo discusses in a constructive way when he keeps telling that people trying to demonstrate a different point of view "are clowns" or "do not have any idea".

[hardingfv32]

It is not personal attach to challenge the information presented. One also does not have to have a better theory or proof before doing so.

I believe there is a false sense of some kind of higher order of understanding of the problem when one uses a CFD model. We have no idea of the presenter's expertise or the many assumption that go into the model. Mostly, the CFD represents an excellent visual illustration of what the presenter is trying to demonstrate. There is no way to now if it is correct without some other higher order of testing.

So back to the subject: What makes the side pods work in the grand scheme of the aero design?

Brian

[strad]

Just in case someone thought I was not serious or arguing...I ask again of Ringo...
Doesn't the air coming in the front..under the splitter push the exhaust to the outsides? I think this is what forms the seal and gives them a semi ground effects situation.

[ringo]

Ermm, grid convergence?

In CFD, if you can not afford the time, then you are wasting your time IMO.

Grid size and density is by far the biggest and most important variable, and grid convergence is the only way of knowing whether a mesh is suitable or unsuitable apart from how accurate the result is ofc.

You chose to converge to a property. Convergence is not a validation of a theory.
Convergence only means the computer's iterations around a certain property have leveled off.
Convergence is not related to the mesh. It's related to the data. It can also guide the calculation.

What makes the side pods work in the grand scheme of the aero design?

I don't really know, and i think some of the top F1 guys don't really know it fully.
When you say work, you mean to aid something or mitigate something?

Doesn't the air coming in the front..under the splitter push the exhaust to the outsides? I think this is what forms the seal and gives them a semi ground effects situation.

Well i think,( don't take my word for it!) The splitter air does not push the exhausts, it justs act as a barrier in some areas. The pipe angle and oncoming determines the exhaust path.

[ringo]

It is not personal attach to challenge the information presented. One also does not have to have a better theory or proof before doing so.

I believe there is a false sense of some kind of higher order of understanding of the problem when one uses a CFD model. We have no idea of the presenter's expertise or the many assumption that go into the model. Mostly, the CFD represents an excellent visual illustration of what the presenter is trying to demonstrate. There is no way to now if it is correct without some other higher order of testing.

Yep it's not guaranteed to be right. But it's open to criticism of a similar method.

There is no false sense of higher order. The CFD is not used blindly. Remember i tried to force the thing to flow under the diffuser of the car; many times.
The closest it came was to go near the sides with about 20% of the exhuast squeezing in there. Some flows in at very high speeds.
After that is seem apparent that something else was happening.

I would be glad to be proven wrong, as it would be a learning experience. But so far there has been no solid hypothesis to say the exhaust skirt is invalid.

I think those in the media like scarbs and the other grid walkers need to ask other teams, beside Renault, what they think, and suggest to them the exhuast is acting like a skirt to see what they say.

[shelly]

Ermm, grid convergence?

In CFD, if you can not afford the time, then you are wasting your time IMO.

Grid size and density is by far the biggest and most important variable, and grid convergence is the only way of knowing whether a mesh is suitable or unsuitable apart from how accurate the result is ofc.

You chose to converge to a property. Convergence is not a validation of a theory.
Convergence only means the computer's iterations around a certain property have leveled off.
Convergence is not related to the mesh. It's related to the data. It can also guide the calculation.

ringo you are wrong. You have grid convergence when results do not change with further refinement of the mesh: it is a criterion to establish if your mesh captures sgnificant flow features.

[Ferraripilot]

Ringo, you have some excellent points in your assessment of this, but my question is sort of obvious. Why on Earth would MB be swapping to this configuration if it is not optimal? I completely understand your points and agree with them otherwise though.

[ringo]

Their biggest problem from last year is balance in the car. Maybe it's the same problem with the short wheel base W02.
The front exhaust best feature is the centralized downforce. Maybe that's what attracting them.

They could be desparate for anything to gain ground as well. Their radiator layout makes it easier to pass the pipes underneath too.

[ringo]

ringo you are wrong. You have grid convergence when results do not change with further refinement of the mesh: it is a criterion to establish if your mesh captures sgnificant flow features.

I don't see the difference with what you are saying, other than you say the mesh converges. That doesn't happen in my case. My mesh doesn't change during calculation.
It refines before calculation starts, and i can view the mesh before it runs.
Refinement during calculation is not really convergence.
You said it yourself, it's a criterion example pressure, force, velocity.

[hardingfv32]

MB has no idea that this system better. It is highly unlikely that this type of system tests well in the wind tunnel. You need to make a calculated decision about whether it is worth testing. MB has nothing loss at this point of their car's development. With limited testing and no one else committing to the system, why not go for the brass ring?

Brian

[PlatinumZealot]

Ermm, grid convergence?

In CFD, if you can not afford the time, then you are wasting your time IMO.

Grid size and density is by far the biggest and most important variable, and grid convergence is the only way of knowing whether a mesh is suitable or unsuitable apart from how accurate the result is ofc.

You chose to converge to a property. Convergence is not a validation of a theory.
Convergence only means the computer's iterations around a certain property have leveled off.
Convergence is not related to the mesh. It's related to the data. It can also guide the calculation.

ringo you are wrong. You have grid convergence when results do not change with further refinement of the mesh: it is a criterion to establish if your mesh captures sgnificant flow features.

Mesh convergence is not really affecting the convergence of the properties that you seek. You set the mesh convergence level to "X" subdivisions maximum. After the mesh converges the flow calculation can still go on, and it usually does, and so the convergence of the flow it self is not directly related. So yes the mesh converges, but that is inevitable based on your CFD settings. The convergence of a certain pattern of flow is what I think he means, since that is depending on whatever the flow is doing, turbulence, transient flow etc.

I usually check if the mesh is capturing the details before the real calculations begin. I cannot imagine that F1 teams would check the mesh every single time they run a calculaiton. You just know after some experience how to set your mesh.

I think the whole point of this is say whether his CFD is accurate? The question is how accurate do we need it to be to show that the Renault exaust is meant to form a skirt? That is the whole crux of the argument right?

[hardingfv32]

"Computational Fluid Dynamics (CFD) falls short in predicting underbody conditions, as the vortices generated by the elements in front of the underbody are not reproduced accurately with the current generation of CFD codes. The current state of computer technology does not
allow complete solving of the governing instantaneous Navier-Stokes equations under normal engineering circumstances, due to the extremely fine meshes required in order to incorporate the smallest eddy vortices generated by the geometry. This Direct Numerical Simulation would thus require huge computational resources and calculation times that would be unacceptable for engineering purposes."

Vortices in favorable pressure gradients
Assessment of Vortex behavior in Formula 1 Underbody conditions

C.K. van Steenbergen
November 2004

So have things improved in the last 7 years? How do we know your systems are up to the task?

Brian

[ringo]

How do we know our "inside" sources are being honest?

What happened to those guys?

[flyboy2160]

"Computational Fluid Dynamics (CFD) falls short in predicting underbody conditions, as the vortices generated by the elements in front of the underbody are not reproduced accurately with the current generation of CFD codes. The current state of computer technology does not
allow complete solving of the governing instantaneous Navier-Stokes equations under normal engineering circumstances, due to the extremely fine meshes required in order to incorporate the smallest eddy vortices generated by the geometry. This Direct Numerical Simulation would thus require huge computational resources and calculation times that would be unacceptable for engineering purposes."

Vortices in favorable pressure gradients
Assessment of Vortex behavior in Formula 1 Underbody conditions

C.K. van Steenbergen
November 2004


....So have things improved in the last 7 years?...

Brian

Well, dudes, the state of the art in CFD has evolved. EFD has introduced adaptive meshing, which allows the CFD mesh to "automatically" become finer locally depending on solution differences between adjacent cells. Thus you only get very fine meshing where you really need it.

I've been able to see vortex swirling / reversed / separated flow running on an HP workstation with the results verified/matched against zillion dollar wind tunnel data and flight data. The fine meshing went as small as ~.02" to .03"

I admit I've never run adaptive mesh CFD a car - only aero-optical distortion - but I don't see why this revolutionary technique wouldn't work on an F1 model.