Mercedes GP MGP W01

[747heavy]

some good explainations and CFD animations are found here:

http://www.ibrii.com/n/x9j4Q

and in this thread in F1t forum here:

viewtopic.php?f=6&t=8028&start=120

[JohnsonsEvilTwin]

Much appreciated 747.

My understanding of F1 aero is(basic amongst other things)that its pretty much exact opposite of aeronautical aerodynamics.
Naturally in planes, more air travels over the top of the wing creating a pressure difference which in turn gives you "lift".
The steeper the angle the more the pressure difference, the more the lift but the higher the drag.

Turned upside down, its tantamount to "downforce"
As I said my understanding is fairly basic.

Regards the F-duct, Mercedes do indeed refer to it as such. Its air eminates from somewhere around the cockpit and trials all the way around the sidepods, through the rear wing supports to the duct on the rear wing.

Is it any wonder they are struggling to make it work?

[JohnsonsEvilTwin]

@ ESPImperium

This backs up what you are saying. Its 9 months old, but still gives you an idea.
Mercedes ranked most powerful, with Toyota last. The difference? 18hp
In addition, Renault have been allowed a further modification so the difference cannot be more than 10hp IMO
http://www.jamesallenonf1.com/2009/11/a ... e-in-2009/

[ESPImperium]

@ ESPImperium

This backs up what you are saying. Its 9 months old, but still gives you an idea.
Mercedes ranked most powerful, with Toyota last. The difference? 18hp
In addition, Renault have been allowed a further modification so the difference cannot be more than 10hp IMO
http://www.jamesallenonf1.com/2009/11/a ... e-in-2009/

I may not have a particular grasp on certain elements of the mechanichal and aero regs in relation to engineering, but my management of figures and general sports management is where my core strength lies.

The article that you have there is 9 months old now, and engines have increased in power, Mercedes has increased by arround 5-7 from that figure, Ferarri arround the same, Renault have put on arround 9-12hp, whitch leaves them still arround 10 down on the Mercedes. The thing is, Cristian Horner is bleating about 15hp, in real terms, its arround half he wants, just enough to given them arround 0.2 seconds a lap more pace. The thing is, the Mercedes guys arnt worried, as the Renault F1 Team hasnt been bleating on about this as well.

Mercedes arnt worried, they produce a quality product that is proven to win championships, the RS27 hasnt got a title to its name yet, as Alonso won the title in 2006 with the RS26 (A to D Series). The RS27 is completly diffrent to the RS26 mostly.

[ggajic]

Yes, but - more HP means more wasted heat - more cooling - less aero efficiency. 10-15Hp doesn`t seam too much. But if you consider thermal efficiency of F1 engines I think that they are at best around 0.18 or less. Good explanation why Red Bull is faster then Renault with same engine - or McLaren with more powerfull engine is quite simple - it is more aerodynamically efficient car..

[JohnsonsEvilTwin]

Newey could provide more or less the same dimensions that Red Bull has with a Mercedes engine.
The fact they had a lump fitted with the Mercedes exact dimensions late last year is testament to that.
Red Bull enquired, Mercedes were willing, but Mclaren rebuffed.

The aero concepts and the ability to stretch these concepts are the defining factors in F1 today. This is ably demonstrated by a comparison between the MP4-25 and the W01.
Or the Renault and the Red Bull for that matter.

Personally, I feel there needs to be a bigger trade off. Aero cannot dominate as it has over recent years. More focus needs to shift to real world engineering like engines and mechnical innovation rather than aero tweaks that have little or no relevance in the real world.

Further to the relevance of this thread, is it possible to calculate drag using CAD? It would be most interesting to compare the W01 to the Mclaren in terms of drag co-efficient, the Merc appears to have the properties of a barn door this year!

[747heavy]

Not CAD per se
You would need CFD, and a very very detailed model (CAD data are one part of this).
But like many other things it´s just a simulation, an appoximation of reality if you like.
It´s not 100% reality, but it get´s better and better - comes closer to reality every month/year.
It´s a function of calculating power (computer speed/memory) and model quality. Both are improving rapidly.

You can´t just take a picture or make some rough measurements and predict DF and drag, if this is what you wanted to ask.

[ggajic]

I think that most of F1 teams (including Mercedes GP) are using Ansys (since Ansys bought Fluent they dont have serious competition). CFD is actually using models designed in CAD (usually Pro Engineer or Catia/Solid Works) to calculate drag and visually show pressure/velocity fields. Computer power is required to solve Navier-Stokes equations since fluid continuum (in this case air) is separated in infinite number of elements. 10 or 15 years ago since computing power was not so strong as we have today this was still in progress and results were far from accurate. But with the way current GPUs are developing (remember that graphic card processors now have over one billion transistors and are much faster then CPUs) I expect that CFD calculations and models will become very very accurate. Of course, people like Newey will still use pen and paper and for basic ideas this will remain important for some time. But remember that computer in 1997. managed to beat chess master.. So it is just matter of time before CFD designed car becomes winning car...

[PlatinumZealot]

@ ESPImperium

This backs up what you are saying. Its 9 months old, but still gives you an idea.
Mercedes ranked most powerful, with Toyota last. The difference? 18hp
In addition, Renault have been allowed a further modification so the difference cannot be more than 10hp IMO
http://www.jamesallenonf1.com/2009/11/a ... e-in-2009/

It didn't say the Mercedes was the most powerful. I interpret it as the BMW being the most powerful then the Benz then Ferrari etc, because he listed BMW first. And this is can be expected, since the BMW was known to be the most powerful in the 10 cylinder days.

[PlatinumZealot]

Not CAD per se
You would need CFD, and a very very detailed model (CAD data are one part of this).
But like many other things it´s just a simulation, an appoximation of reality if you like.
It´s not 100% reality, but it get´s better and better - comes closer to reality every month/year.
It´s a function of calculating power (computer speed/memory) and model quality. Both are improving rapidly.

You can´t just take a picture or make some rough measurements and predict DF and drag, if this is what you wanted to ask.

I think if you have enough pictures and measurements you can do it of course.

An extreme example, Let us say that Mclaren have one of those portable 3D scanners in the pits, all they have to do is point it at that RedBull and (pretending that they could scan under the car and the nose etc), they can import the 3D scan into the CAD program and rebuild the surface and analyse it. The car is just a shape so I think highly detailed visual information is enough to crack the aero. If that doesn't work out then it means that RedBull has something else up their sleeves.

I am not really an aeroplane or history buff, but didn't the russians put much effort into spying on and mimicking the American aircraft and vice-versa. I am just saying I think reverse engineering is a worth while legitimate way to design.

[marcush.]

I think that most of F1 teams (including Mercedes GP) are using Ansys (since Ansys bought Fluent they dont have serious competition). CFD is actually using models designed in CAD (usually Pro Engineer or Catia/Solid Works) to calculate drag and visually show pressure/velocity fields. Computer power is required to solve Navier-Stokes equations since fluid continuum (in this case air) is separated in infinite number of elements. 10 or 15 years ago since computing power was not so strong as we have today this was still in progress and results were far from accurate. But with the way current GPUs are developing (remember that graphic card processors now have over one billion transistors and are much faster then CPUs) I expect that CFD calculations and models will become very very accurate. Of course, people like Newey will still use pen and paper and for basic ideas this will remain important for some time. But remember that computer in 1997. managed to beat chess master.. So it is just matter of time before CFD designed car becomes winning car...

cfd is an analysis tool not a design tool,if i´m not totally mistaken.the morphing of the car ,the incremental steps to optimise flow may be a somewhat iterative process you might be able to able to create and work down automatically ,but the first step will need an idea from an ingenious head ..who comes up with the idea of a wing that does deflect in a weird ways when at speed ...or a hole in the bodywork to create an airvolume inside the car to be filled to gain downforce..this is something a computerprogramme willnever be able to invent.
CAD as a engineering tool tends not to promote creativity far from it as it will certainly promote ways of design that are easily performed by the user.so in effect the CAD system will filter out or tweak the potential ideas for their feasibility to be put on screen.
I have often seen the CAD boffins not really able to visualize anymore. they draw up in plane and the pc does the third dimension...DMU takes away a lot from creative thinking in terms of packaging as everyone is occupied with juggling parts around to fit them in a tight contsraint.It does not lead itself to sophisticated ingenious solutions really.

[747heavy]

Not CAD per se
You would need CFD, and a very very detailed model (CAD data are one part of this).
But like many other things it´s just a simulation, an appoximation of reality if you like.
It´s not 100% reality, but it get´s better and better - comes closer to reality every month/year.
It´s a function of calculating power (computer speed/memory) and model quality. Both are improving rapidly.

You can´t just take a picture or make some rough measurements and predict DF and drag, if this is what you wanted to ask.

I think if you have enough pictures and measurements you can do it of course.

An extreme example, Let us say that Mclaren have one of those portable 3D scanners in the pits, all they have to do is point it at that RedBull and (pretending that they could scan under the car and the nose etc), they can import the 3D scan into the CAD program and rebuild the surface and analyse it. The car is just a shape so I think highly detailed visual information is enough to crack the aero. If that doesn't work out then it means that RedBull has something else up their sleeves.

I am not really an aeroplane or history buff, but didn't the russians put much effort into spying on and mimicking the American aircraft and vice-versa. I am just saying I think reverse engineering is a worth while legitimate way to design.

maybe you should send this sugestion to McLaren!!??
Sounds straight forward to me !!!

No seriously, you are right, and it is done in all walks of life/engineering and especially in F1.
But you need to understand a concept to copy it effective, not just an outside shape or appearence.

I may choose my words unwise I should have used "one photo" instead of "a photo".

Sure it helps when you get your hands on a CD/DVD with CAD files. I think McLaren and Toyota may know a little bit about that as well.

[mx_tifoso]

A few pages worth of posts were moved to the 'Mercedes GP' thread as they did not discuss the car itself, but the team as a whole. Please keep team or political topics out of the car threads, they shouldn't be in here as these threads are for the technical aspects of the cars, and nothing but the cars.

[marcush.]

from my experience i´d say you could possibly get a sound idea about outer shapes but as we have no possibility to scan the inner flows and especially not the detail design inside it will be just not possible to know or calculate these inner flows (radiator ducting,dd diffusser ducting) as detailed knowledge simply is not available and not visible(mabe you can find out about the radiator core flow characteristics because you happen to use the same supplier?)but we get into highly speculative areas there..and I think this is only possible to judge when you are actually in and close to the car at times(Grid).for us it is just to much error ...we have no idea how much time and effort teams spend to optimise radiator flows ...which can be influenced quite dramatically with detail design.

[n_anirudh]

Now that the weight distribution has been fixed, would Pirelli be forced to design around this??

Isnt there a chance for all cars ultimately having understeer or oversteer??