Ferrari F2012 CFD Visualization Video

[miqi23]

even if the wheel is spinning, which I doubt, it still doesn't account for the exhausts, which do have a non-negligible effect.

So you see a big difference in the flow of the illustrations above?

I can live without the exhaust. Do the research, this is as good as you are going to find on the current cars.

Brian

Hang on a second there.. Those two illustrations are totally irrelevant to when they are on a car. Secondly, the analysis is totally simplified and the flow in the rotating case looks wrong and the streamlines are not giving the complete picture - do you have centre line velocity slices?

[hardingfv32]

Those two illustrations are totally irrelevant to when they are on a car. Secondly, the analysis is totally simplified and the flow in the rotating case looks wrong and the streamlines are not giving the complete picture - do you have centre line velocity slices?

Exactly why are they totally irrelevant?

Yes, these are very isolated but very accurate. This is how you do research when you do not have a complete model of the car or the computing power to run such a sim.

This is a 94 page paper that clearly lays out the test methodology and CFD details. There is no one on this forum that will be able to challenge the results. Have a read before you question the results.

Brian

[Lycoming]

the problem is, that picture is not the 94 page article.

I have to say, it does not look like what I would expect having seen streamlines around a rotating wheel in a wind tunnel from Katz's Racecar Aerodynamics. in that example, there is a large area of disturbed flow behind and in front of the tyre and flow seperates at the top.

I'm not saying that CFD is wrong. I don't have enough context to say something like that. But its not like what I have seen from other sources, in which there is a very large and very noticeable difference between a rotating and stationary tyre.

[Crucial_Xtreme]

the problem is, that picture is not the 94 page article.

I have to say, it does not look like what I would expect having seen streamlines around a rotating wheel in a wind tunnel from Katz's Racecar Aerodynamics. in that example, there is a large area of disturbed flow behind and in front of the tyre and flow seperates at the top.

I'm not saying that CFD is wrong. I don't have enough context to say something like that. But its not like what I have seen from other sources, in which there is a very large and very noticeable difference between a rotating and stationary tyre.

Yeah I didn't post the topic because I thought it was an accurate representation of the F2012 but I thought it's somewhat close and would be interesting to here what people thought of what was happening, whether accurate or not. I didn't expect a big debate about it's legitimacy.

[hardingfv32]

would be interesting to here what people thought of what was happening, whether accurate or not.

1) Is it of value for evaluating flow around the radiator inlet, barge board? An amazing amount of flow is channeled out side of the rear wheels. I would have thought that the coke bottle side pods would have allowed more of this flow inside towards the rear wing.

2) Did you watch the stepped nose video? That seem accurate in shape and configuration. The flow goes all to hell very rapidly after hitting the ramp.

Brian

[Jersey Tom]

94 page report? So what? There are 800+ page textbooks which I'd challenge on completeness. Length, name.. these mean very little.

In any event this isn't my battle. But more importantly.. so what?! It's effectively an artist rendition CAD model which has no validation, paired with a CFD tool for which we see no validation.

All it is, is a cool looking video with curvy colorful lines.

[strad]

All it is, is a cool looking video with curvy colorful lines.

I totally agree..It is not very accurate but it's something cool to have on your phone.

[miqi23]

Those two illustrations are totally irrelevant to when they are on a car. Secondly, the analysis is totally simplified and the flow in the rotating case looks wrong and the streamlines are not giving the complete picture - do you have centre line velocity slices?

Exactly why are they totally irrelevant?

Yes, these are very isolated but very accurate. This is how you do research when you do not have a complete model of the car or the computing power to run such a sim.

This is a 94 page paper that clearly lays out the test methodology and CFD details. There is no one on this forum that will be able to challenge the results. Have a read before you question the results.

Brian

They are irrelevant in a sense that you are saying that those images show a flow similar to the one on the car and since there is not much difference between the two flows (in isolation) it does not matter if the wheel is rotating or not on the car.

Well, there is a huge difference when the wheel is stationary and rotating when on a car. The CFD visualisation on that website does not show things clearly and there is a proper clutter of streamlines all over the place.

[hardingfv32]

Well, there is a huge difference when the wheel is stationary and rotating when on a car.

Can you document this 'huge difference' difference or is it a gut feeling?

1) I the flow streams in the images I provided and those in the video both represent a similar technique of representing flow. There could be a reduction in detail because of the size of the video.

The images I provided came from a MSc Thesis from Cranfield University and supervised by Philip Rubini. Look him up and check out his and the departments credentials. You will be hard pressed to challenge this paper.

2) Assuming the images that I provided are correct, how do think flow around the car is going make a difference in the stationary vs rolling images?

Brian

[hardingfv32]

I totally agree..It is not very accurate but it's something cool to have on your phone.

Just to gage how good the CFD sim has to be before you can use it: Is the HRT team CFD sim good enough or do you require McLaren?

Do you have to have a complete understanding of all the variables used?

Brian.

[hollus]

It would be nice if you learnt to agree to disagree.
You have made your point loud and clear, and everybody heard it and understood it. They might or might not share it. Please stop hammering the point as that won't help it get in. It is not a nail and members are not wood.
Others have also made their points of view clear, and you probably understood them too. You might or might not agree with them.
Now, the idea in a forum like this is not to convince everybody else that you are right. You might be or not, I don't know. But with this many people here one has to show respect for other people's point of view. Shouting "again" many times and louder and louder does not make an idea more convincing, only more tiresome.
I think that a forum should be about exchanging ideas, not about winning arguments. But obviously many people disagree with me.

Now for a reasoned argument that has not been posted 3 times just in different words: The highest point in the wheel is rotating against the flow at twice the speed of the car (the bottom of the wheel is stationary). Where this surface (presumably with its own boundary layer) moving forward at 400Km/h meets air moving backwards at 200Km/h one would expect at least a little bit of high pressure, I guess.

Edit: I made a mistake there by using two different frames of reference. In the frame where the top of the wheel is moving a 400Km/h the external flow is stationary, so the relative speed is only 400Km/h, not 600.

[hardingfv32]

miqi23 Made a claim that the images are irrelevant. I wish to understand why. I am up for learning something about CFD today.

The highest point in the wheel is rotating against the flow at twice the speed of the car (the bottom of the wheel is stationary). Where this surface (presumably with its own boundary layer) moving forward at 400Km/h meets air moving backwards at 200Km/h one would expect at least a little bit of high pressure, I guess.

Well in fact that is exactly the case! The paper clearly shows that happening in pressure gradient images, but it does not translate/transfer into the ribbon stream images which is what we are discussing. The program that developed the pressure gradient images is the same program that develops the ribbon streams. So I think it is safe to say that your concern is not well founded.

PM me with your Email and I will provide the PFD.

Brian

[Jersey Tom]

I totally agree..It is not very accurate but it's something cool to have on your phone.

Just to gage how good the CFD sim has to be before you can use it: Is the HRT team CFD sim good enough or do you require McLaren?

Do you have to have a complete understanding of all the variables used?

Brian.

1. It's not the real car's geometry. We don't know if the model is generally off by 0.1" or 1.0"

2. We don't know if this CFD tool is great or junk. Need to have some sort of validation case study to go against and at least understand what it does.

[miqi23]

Well, there is a huge difference when the wheel is stationary and rotating when on a car.

Can you document this 'huge difference' difference or is it a gut feeling?

1) I the flow streams in the images I provided and those in the video both represent a similar technique of representing flow. There could be a reduction in detail because of the size of the video.

The images I provided came from a MSc Thesis from Cranfield University and supervised by Philip Rubini. Look him up and check out his and the departments credentials. You will be hard pressed to challenge this paper.

2) Assuming the images that I provided are correct, how do think flow around the car is going make a difference in the stationary vs rolling images?

Brian

This 'huge difference' is a fact and not a gut feeling.

Open your horizon a bit and look at it this way - (in very simple terms) when the wheel is on a car it is interacting with the front wing. A lot of effort is made to manage the tyre wake when its rotating. Getting that right is the most difficult part of F1 CFD because the tyres not only rotate, they deform as well, which creates a complex transient phenomenon. If we ignore the transient deformation for now and just concentrate on stiff tyre rotation (say carbon fibre wheels), the front wing vortices would still interact with the tyres.

Now, look at the Ferrari front wing for instance. All those add-ons you see are there for a purpose. Cutting a long story short and looking at the relevant bit for now, the separation point on a rotating tyre when on a car is way forwards (not the case when in isolation) and this results in a huge wake. That huge wake forms a unique flow signature down stream of the car and if you get that wrong in CFD, your down stream design would be wrong - simple!

I hope you get the point why your isolation cases are irrelevant.