Diffuser Profile - Convex or Concave?

[hardingfv32]

In the main volume of the diffuser, after the throat, is velocity decreasing and pressure increasing because of the increase in volume?

IF correct, is this SPECIFIC region producing a loss of downforce?

Brian

[Tozza Mazza]

In the main volume of the diffuser, after the throat, is velocity decreasing and pressure increasing because of the increase in volume?

IF correct, is this SPECIFIC region producing a loss of downforce?

Brian

That would depend on the pressure above the diffuser.
If you have a higher pressure above the diffuser, than below it, you will still be creating some downforce, although the smaller this difference, the less downforce is being generated (basic description I know, other factors may also being affecting this).

It could be that that area of the diffuser is the area being targeted by the EBD.

2012 will be an interesting year for the diffuser. With no DDD or EBD, it's going to be producing less downforce, and so its shape will have a larger impact.

Tom.

[ringo]

@ringo
I didn't really understand the different top surface subject.
Maybe you mean the different angle of the diffusers between those 2 images? If yes, this has nothing to do with the analysis, the voncave one is just rotated on the post processor. The flow direction is parallel to the throat on both examples.

No, not the angle. The shape at the top surface has to be the same to have a fair comparison.

For example you can expect to get the same results here:

Correct?

Now what if for the very same throat shape you chage the top surface to this:


Obviously you would get differing results between the two, even though the have the same thoat shape.

This would give invalid results. You would have the same exact diffuser shape, but different top surface, and you would conclude from the results one is better than the other becuase it has less drag and more downforce.

This is why i was saying you should do the experiment over, but make everything but the diffuser throat the same.
Like this:

This is the only way you can isolate the performance due to the shape only.
This is a purer experiment. Forget about tyres and wings for now. The comparison is with diffuser shape, nothing else.

[DRCorsa]

I got it now, ringo.
In my analysis, there is no air going over the top surface, only under it. So, i think that the conparison is fair.

Look at the image:

[ringo]

I see,

but what's on the outside, still air?

Your results are probably correct in terms of which one has less drag etc. but you just don't know if it is in the correct proportion.
Also you may not be sure if the still air on the outside is affecting your results; is being pulled by the diffuser flow etc.

Anyway, for argument's sake could you change your model and run the test again?
This time with a floor 4 times longer than the diffuser and a common top surface. Do an external flow simulation.

[Crucial_Xtreme]

Has anyone tested(CFD) the RBR starter hole with the two air ducts in the top of the floor? No other teams have jumped on this to my knowledge and I'm wondering how much this helps the RB7. They obviously don't do it for no reason and it's helping otherwise why do it. I was just wondering if any of you guys had tried to incorporate this feature into your design?

Also does anyone know if said air ducts at top of floor for starter hole will be allowed in 2012 Regs?

[hardingfv32]

Note the issues with the diffuser model above. You actually think they are going to get the starter hole done correctly? Do you have some dimensions to provide them with?

Brian

[Tozza Mazza]

I think that DR Corsa's findings are probably still right, although the exact figures are actually off, due to what Ringo is describing.

Any aero guys out there who can tell us what influence the length/height ratio of the diffuser has on the design of the diffuser?

Is it to do with stopping flow separation, and reducing stall that the design is concave, because a convex diffuser would stall in the box given in the technical regulations.

[hardingfv32]

I researched the use of a concave shapes for a day and found that there is going to be serious flow separation under all the normal aero examples that I found. Unless you can provide some method of keeping the flow attached, I am comfortable saying that the use of a concave diffuser shape is NOT about an aggressive shape that does NOT have flow separation.

The one thing I found is that the size of the separation bubble created by the concave shape will vary with changes in the Reynolds number.

So how does excessive flow separation fit into your models? What other benefit is worth more than attached flow to make this option viable?

Brian

[shelly]

@brian: have you found examples of concave shapes with strong 3d effects?

[hardingfv32]

Actually the best I could find was flow over a stepped surface. This is the closest I could get to the shape we are seeing at the throat of the concave diffuser. I do appreciate that we don't know how much of a radius is being used at the throat.

I found a number of studies of flat roofs that show tremendous separation at with say a 15 deg slope. How are we going to remain attach with this concave shape if that is the case with a 15 deg sloped flat roof?

I think there is a compromise being made between separation and something else with the concave shape that is a net gain of something that we do not appreciate at this point.

I need new descriptors for more searches. Concave is getting me nothing.

Brian

[BreezyRacer]

Actually the best I could find was flow over a stepped surface. This is the closest I could get to the shape we are seeing at the throat of the concave diffuser. I do appreciate that we don't know how much of a radius is being used at the throat.

I found a number of studies of flat roofs that show tremendous separation at with say a 15 deg slope. How are we going to remain attach with this concave shape if that is the case with a 15 deg sloped flat roof?

I think there is a compromise being made between separation and something else with the concave shape that is a net gain of something that we do not appreciate at this point.

I need new descriptors for more searches. Concave is getting me nothing.

Brian

15 degrees is indeed a pretty radical angle. Try 10 to 12 degrees and put a gurney on the end of it if you can. The diffusers you see in F1 are using the beam and rear wing to reduce flow separation, though like I said before, all of this stuff is interrelated carefully to make it actually work.

[DRCorsa]

I see,

but what's on the outside, still air?

Your results are probably correct in terms of which one has less drag etc. but you just don't know if it is in the correct proportion.
Also you may not be sure if the still air on the outside is affecting your results; is being pulled by the diffuser flow etc.

Anyway, for argument's sake could you change your model and run the test again?
This time with a floor 4 times longer than the diffuser and a common top surface. Do an external flow simulation.

I did it and the picture is more or less the same.
For the concave, downforce and drag levels are lower, there is the suction peak at the kink line and there is notably less downforce and stall after the kink line.

CONCAVE

Downforce: 1473N
Drag: 317N







CONVEX

Downforce: 2077N
Drag: 396N

[hardingfv32]

As DRCorsa summarized and demonstrated:

"For the concave, downforce and drag levels are lower, there is the suction peak at the kink line and there is notably less downforce and stall after the kink line."

So why would you use what on the surface looks like an inferior roof (floor) shape?

AND don't bring in the exhaust placement as this shape has been used for many years while the exhaust placement has been constantly changing.

I can't see where it would cause a restriction to the shape above diffuser.

Does this have something to do with the low pressure at the exit of the diffuser reacting better to the roof surface being flat as was suggested earlier? Nothing to do with pumping the floor.

Should this have been demonstrated with DRCorsa's simulation?

Brian

[ringo]

I see,

but what's on the outside, still air?

Your results are probably correct in terms of which one has less drag etc. but you just don't know if it is in the correct proportion.
Also you may not be sure if the still air on the outside is affecting your results; is being pulled by the diffuser flow etc.

Anyway, for argument's sake could you change your model and run the test again?
This time with a floor 4 times longer than the diffuser and a common top surface. Do an external flow simulation.

I did it and the picture is more or less the same.
For the concave, downforce and drag levels are lower, there is the suction peak at the kink line and there is notably less downforce and stall after the kink line.

CONCAVE

Downforce: 1473N
Drag: 317N





CONVEX

Downforce: 2077N
Drag: 396N

These comparisons say it all really. One is stalled with the separation bubble. The other has smooth flow.
I agree with these results, and it was puzzling me when i did a similar experiment.

So unless we are looking at the F1 diffusers the wrong way or there is something that we are not realizing.

What is left to make this more accurate is to apply a surface roughness to the ground. Find out how rough asphalt is, then apply a moving surface to ground in your model. It wont change much however, but it would be interesting nonetheless.