F1technical.net

That is certainly a reasonable explanation Shelly.

I have a fair amount of CFD experience, mainly with Fluent and even for simple geometries it is difficult to get what I would call an accurate picture, be it for numerics or even flow vis.

I realize we are not analysis Cd values here and are primarily looking at a qualitative approach but a more than sceptical view must be taken imo.

PNSD wrote:That is certainly a reasonable explanation Shelly.

I have a fair amount of CFD experience, mainly with Fluent and even for simple geometries it is difficult to get what I would call an accurate picture, be it for numerics or even flow vis.

I realize we are not analysis Cd values here and are primarily looking at a qualitative approach but a more than sceptical view must be taken imo.

Hi! Have you ever tried panel method?

I fear that in F1 case with bluff bodies and separation panel method is t erratic. I think Ferrari has used it back in the 80s

ringo wrote:

MIKEY_! wrote:Interesting. Think a similar pump like system could be used to suck air though the radiators? With the red steam being exhaust and the blue being radiator air. Radiators placed in the area between the 'inlet' and the 'suction chamber'.

Yep.

It's really old technology. Just not well known in the general public.
Most water treatment chlorinators have ejectors.

Don't think it's that uknown in general public. Most people have already seen spray guns and airbrushes. Jet pumps are quite common devices in industrial area too.

The problem is they need at least one converging nozzle in front of injector to accelerate the flow and create low pressure, and i don't see any converging nozzle in front of R31's exhaust pipe.

shelly wrote:I appreciate volarchico's post, but I see that ringo, you have started over again with the same attitude, which has lead to a long and sometimes inconclusive thread.
I'll try not to misbehave myself, so I will start trying to point out what I think without putting it like it's falling from the skies.

My first point is this: a home made cfd with limited computing resources does not give reliable results when we are trying to discern exhaust trajctories in a f1 car. Who does agree/disagree/ need explaination about this?

It's not even the problem of resources.
CFD is just a tool, and as with any tool most important thing is to have skills to use it.

I've recently build quite powerful computing cluster for mechanical faculty at out technical university, and witnessed hours lasting discussion with 2 Professors and 4 PhD's involved, which numerical models are best for solving given problem.

ringo wrote:It's really old technology. Just not well known in the general public.
Most water treatment chlorinators have ejectors.

We've all played with them at school when we used an artist's airbrush. You blow down a tube and the air flow sucks paint up from a bottle to create an aerosol spray.

However these examples are enclosed spaces. I presume that ringo is using it as an analogy for the FEE jet dragging the ambient air along, some of which will accelerate the underfloor air flow?

Would a better analogy be the downdraught you feel standing near a waterfall?

marekk wrote:

shelly wrote:I appreciate volarchico's post, but I see that ringo, you have started over again with the same attitude, which has lead to a long and sometimes inconclusive thread.
I'll try not to misbehave myself, so I will start trying to point out what I think without putting it like it's falling from the skies.

My first point is this: a home made cfd with limited computing resources does not give reliable results when we are trying to discern exhaust trajctories in a f1 car. Who does agree/disagree/ need explaination about this?

It's not even the problem of resources.
CFD is just a tool, and as with any tool most important thing is to have skills to use it.

I've recently build quite powerful computing cluster for mechanical faculty at out technical university, and witnessed hours lasting discussion with 2 Professors and 4 PhD's involved, which numerical models are best for solving given problem.

The problem with most professors and PhD's is that they would rather stand around for hours talking about how to do something than just going and doing it or trying it. I agree that academia does have its place when they are trying to find the 100% exact solution to a problem. But in the real world, an 80% "close enough" approach is often...well, close enough!

@shelly, I agree with your statement. Home-made CFD is probably not accurate enough to predict the exact flow trajectory. Maybe good enough for bulk motion and flow around the car, but nothing exact. To have a mesh refined enough to be accurate is computationally prohibitive for a personal computer.

One more observation not directed at anyone specific, but I have noticed on this forum it is VERY easy and prevalent to tell someone they are wrong, but it seems sparse those who actually spend the same energy to prove their own opinion is right.

The problem with most professors and PhD's is that they would rather stand around for hours talking about how to do something than just going and doing it or trying it

Those who can,,Do!...Those who can't TEACH!

volarchico

Ringo propose the theory of an ejector and I ask for details on how this is applied to the R31. It is my opinion that it is not correct and he has provided nothing to change that opinion.

I am not a F1 aerodynamicist. I will be lucky to have an opinion, much less prove it. But, I can evaluate someone else's theory fairly well, after a few questions, and tell if it holds water... in my opinion. How can you judge how much time one spends on these subjects. For me, with a limited skill set, it takes a lot of research just to question other peoples ideas. I am learning, which is about all you can expect from me.

Brian

I witnessed in the past few discussions between chemists, biologyst and mathematics guys, and couldn't agree more with your view, volarchico .
Those mechanical guys are far more down to the earth.
The point is - even if have free access to the K computer cluster in Kobe, we still lack models, knowledge and experience to use this tool.
We are free to speculate as long as this makes fun, but don't think anyone of us should pretend to give authoritative answers to such complex questions.

@Brian: when I said "not directed at anyone specific" I really meant it. The comment was meant to be more thought-provoking than anything. It is easy to tell someone they are wrong, but it is difficult to provide the true answer. I guess my personal feeling is that if you can't provide the correct answer, be very humble in telling someone they are wrong. There is nothing wrong is explaining someone else's faulty hypotheses to further discussion and make sure preposterous ideas don't gain traction but at least they are trying.

@marekk: Agreed. Authoritative answers from arm-chair F1 aerodynamicists on this forum are not possible. But it sure is fun to try and pretend.

volarchico

I understood you statement and was not disturbed by it. I just thought a more detailed comment about one member of this forum was in order.

In this case Ringo has voiced a theory which he cannot translate to actual physical traits on the R31.
This is not unusual.

Brian

Sonic59 wrote: Hi! Have you ever tried panel method?

With limited experience. Primarily looking at pressure distributions over sections. So 2d geometry.

As Shelly says, I'm not quite sure such a method would appropriate for this application?

shelly wrote:I appreciate volarchico's post, but I see that ringo, you have started over again with the same attitude, which has lead to a long and sometimes inconclusive thread.
I'll try not to misbehave myself, so I will start trying to point out what I think without putting it like it's falling from the skies.

My first point is this: a home made cfd with limited computing resources does not give reliable results when we are trying to discern exhaust trajctories in a f1 car. Who does agree/disagree/ need explaination about this?

For some reason i am put under the microscope with my opinion and it is rubish until proven realistic. Stop carrying feelings. I wasn't being rude to anyone.
I don't really understand what you mean by attitude either. I re read my post and i don't see anything offensive in it.

And saying homemade CFD is not accurate.
As said earlier the only limitation is the calculation time.
I didn't create a program for it to be homemade. I can't create these wonderfull software.
Secondly i'm just as qualified to run an accurate CFD as any other F1 engineer out there. I just don't have the work experience.

It would be nice if you could prove that "homemade" CFD is inaccurate.
So far it's holding up nicely.

richard_leeds wrote:

ringo wrote:It's really old technology. Just not well known in the general public.
Most water treatment chlorinators have ejectors.

We've all played with them at school when we used an artist's airbrush. You blow down a tube and the air flow sucks paint up from a bottle to create an aerosol spray.

However these examples are enclosed spaces. I presume that ringo is using it as an analogy for the FEE jet dragging the ambient air along, some of which will accelerate the underfloor air flow?

Would a better analogy be the downdraught you feel standing near a waterfall?

Maybe, but the ejector is accurate as the speed of the exhaust gas is much much greater than the under floor speed, as compared to a water fall.
A 2010 blown diffuser also works in a similar fashion. So does the ferrari exhuast shown.
The spray gun is a good example too.
The space is not really important, not as much as the interacting surfaces and the relative speeds, at least that is what i think.