Jump to content


Photo
- - - - -

Flat-floor+diffuser vs inverted aerofoil: which ground effects is best?


  • Please log in to reply
20 replies to this topic

#1 V8 Fireworks

V8 Fireworks
  • Member

  • 10,497 posts
  • Joined: June 06

Posted 25 January 2018 - 12:09

Conventional wisdom in racing comments seems to be that venturi tunnels / inverted wing is superior than a flat floor with a diffuser for producing downforce and optimising aerodynamic efficiency -- but, is it really?  :confused:

 

Any ideas?  Assuming of course, that sliding skirts cannot be used on safety grounds.

 

I find it curious that many unregulated categories choose to use a flat floor with a diffuser, or a combination of contraction ramp, flat floor and diffuser ramps like the Panoz Champcar, and few one-make open wheel (or other) categories use "inverted aerofoil" style "venturi tunnels".   It would be lovely to fire up the CFD  (or wind tunnel) and compare some different configurations, but it is beyond my scope sadly.

 

My understanding is that it best to use contraction ramps that push the air out to the sides, like the Panoz (or bargeboards are used instead in flat-floor-rule F1), and a diffuser design that brings the air intowards the centre-line initially, before diverging it as wide as possible for maximum expansion.  Correct?  The early inverted wing cars seemed to use very "two-dimensional" designs, which perhaps could be improved with modern three-dimensional thinking -- maybe the Panoz already incorporates these enhancements?

 

62j0yh3.jpg

Panoz DP01, inlet to floor

 

One would think it would be better to divide the tunnel (or diffuser) into many channels of smaller span to be able to increase the angle of expansion which it can tolerate before stalling. 

 

Yet, Adrain Newey seems to have used the inverted aerofoil design for his Aston Martin AM-RB001, and it seems to use no strakes to divide up the underbody into steeper aspect ratio sections.

 

"Racing comments" logic seems to be that inverted wing /venturi tunnels is better -- yet if true how come modern flat floor cars have so much more downforce than inverted wing era cars?  :confused:   Ultimately is it all the same thing?

 

4870199087_d433365640_b.jpg

 

Lotus T125 diffuser(s), note the low aspect ratio of the diffuser outlets

 

IMG_20120522_190018.jpg

 

This home racecar builder has used a similar design to the Lotus T125 (presumably for ease of construction, compared to inverted aerofoil).

 

 

aton-redbull-857.jpg

 

Aston Martin AM-RB001 inverted aerofoil

 

aton-redbull-858.jpg

 

Rear view of Aston Martin AM-RB001 inverted wing car

 

c931a4d22fff00be0750e809dbe8e4a3--lotus-

 

Classic inverted wing era ground effects car

 

 

outline_topimg02.jpg

 

 Traditional inside wheelbase inverted wing underbody of Swift Formula Nippon car

 

 

One question:  Does it really matter if the tunnel (expansion) is inside the wheelbase like the Lotus 79 or the Swift Nippon car?  Is (in the case of one-make formulas not bound by regulations) a flat floor inside the wheelbase, that has low pressure over it created by a large diffuser that goes way beyond the rear axle -- much like the Lotus T125 track day car -- just as good (or even better) for downforce, centre-of-pressure and aerodynamic efficiency compared to inverted wing (whilst, of course, allowing much more space to place ancillary components near the engine than the inverted wing)?

 

Perhaps so many one-make junior categories are simply following the styling of F1 cars, without considering alternative designs -- or perhaps, it just doesn't matter or make much difference.  :confused:

 

Edit:

 

ap10thingstosee-a4b071b77772617f.jpg

Dallara DW12, underbody with "venturi tunnels" -- no inlet ramp unlike earlier Panoz or Lola though

 

The tunnels of a Dallara DW12 (I hope Mr. Castroneves was OK in this crash).  I take it this style of tunnel, moves the point of most negative pressure further upstream towards the centre of the car, compared to flat-floor+diffuser, where the lowest pressure is at the beginning of the diffuser and hence essentially acting mainly on the rear axle (as below), rather than acting neutrally on the car.

 

diffuser_Prototipe.jpg

 

LMP1 underbody pressure coefficent, http://www.formula1-...r_Prototipe.jpg

 

Edit 2:

 

ground-effect-0.jpg

Wing car by Scarbs https://www.f1techni...opic.php?t=2996

 

Scarbs has this concept for a full-length "wing car".  Would this be the "ideal" configuration for ground effects (as Newey is using for the AM-RB001)?

 

I notice here Scarbs extends the expansion beyond the rear axle, whereas ground effects era F1 cars rarely choose to do so.    :confused:  Was it regulation that limited ground effect era F1 cars to have their inverted wing inside the wheelbase, or they choose that for some other reason?  :confused:


Edited by V8 Fireworks, 25 January 2018 - 19:36.


Advertisement

#2 Nemo1965

Nemo1965
  • Member

  • 5,511 posts
  • Joined: October 12

Posted 26 January 2018 - 11:59

Interesting question... I was thinking about that as well, after reading Adrian Newey's new book and his thoughts on aerodynamics of the underside. 

 

The full body (more than full length, right) wing car of Scarbs is very interesting but but would it not be even more full when inside of the two venturi's there would be a venturi from the from the front of the nose? Like a venturi channel between the two venturi's

 



#3 Charlieman

Charlieman
  • Member

  • 1,665 posts
  • Joined: October 09

Posted 26 January 2018 - 14:43

What is best for what?

 

What is best for a "cheap to run" F3 or F4 car? What is best for a two/three seat road sports car which may become the form or body shape of a low production sports racer? And what does "best" mean?

 

V8 Fireworks asks: '"Racing comments" logic seems to be that inverted wing /venturi tunnels is better -- yet if true how come modern flat floor cars have so much more downforce than inverted wing era cars?'

 

Modern flat floor cars work because F1 teams spend millions of pounds to do so. We don't know whether venturi/side pod cars would create more/less downforce because racing car designers stopped developing the years ago. Years ago, rule makers determined that limits on front/rear wing placement, track and air box height would slow F1 cars. Yeah. The theory remains the same, that speed and quality of racing can be controlled by F1 flat floors, limits on front wing end plates and barge boards, whatever else they think of as a restriction that won't work.

 

Cars in the junior categories have flat floors because -- well it's the way systems work. Nobody questions whether it is appropriate for cars to be made that way -- they are just seen as little F1 cars. Cars in the junior categories need aero things because drivers need to learn how to use them. But not a lot of them.



#4 MatsNorway

MatsNorway
  • Member

  • 2,730 posts
  • Joined: December 09

Posted 26 January 2018 - 17:24

Is there even a principal (correct my spelling please) difference between a tunnel and a diffuser?  Your moving air by the aide of the ground. The double decker diffuser is the only one where i can see a difference in function.

 

I think vanes is almost always beneficial. They serve as separators for "different" tunnels. And once the tunnel becomes complex in its shape you have to have vanes to guide the air and get the maximum out of it. Without vanes some diffusers probably stall/stop working as intended.

 

On a car with suspension they probably would also help minimize loss pr. mini "diffuser"  when the car is rolling in the corner.


Edited by MatsNorway, 26 January 2018 - 17:36.


#5 Charlieman

Charlieman
  • Member

  • 1,665 posts
  • Joined: October 09

Posted 26 January 2018 - 18:04

Is there even a principal (correct my spelling please) difference between a tunnel and a diffuser?  Your moving air by the aide of the ground. The double decker diffuser is the only one where i can see a difference in function.

 

I think vanes is almost always beneficial. They serve as separators for "different" tunnels. And once the tunnel becomes complex in its shape you have to have vanes to guide the air and get the maximum out of it. Without vanes some diffusers probably stall/stop working as intended.

 

On a car with suspension they probably would also help minimize loss pr. mini "diffuser"  when the car is rolling in the corner.

 A principle is something like a physical concept. And a Principal with a big P is like a headmaster. But that is not what you asked. Generous smile.

 

Mats asks about tunnels and diffusers, without mentioning venturi and wings.

 

Wings are easy to explain -- things in the air. A venturi or reverse wing at load altitude (perhaps they are different things?) are the foundations of underfloor aerodynamics. The Lotus story about the 78 and 79 cars was about telling stories that weren't completely true.



#6 MatsNorway

MatsNorway
  • Member

  • 2,730 posts
  • Joined: December 09

Posted 26 January 2018 - 18:20

What word is more apt to use then?

 

Surely calling a ventury a upside wing is a gross oversimplification. For instance they stall at different angles and tunnels++ makes far more downforce for the drag you get.


Edited by MatsNorway, 26 January 2018 - 18:20.


#7 Charlieman

Charlieman
  • Member

  • 1,665 posts
  • Joined: October 09

Posted 26 January 2018 - 18:33

It is best described as under floor aero dynamics.

 

The reverse aero wing existed as a concept before the Lotus 78. And when Lotus made it into a venturi, they were using science that had been discovered before them. 



#8 gruntguru

gruntguru
  • Member

  • 6,691 posts
  • Joined: January 09

Posted 27 January 2018 - 03:43

For outright L/D the airfoil underbody would be superior. That's why many formulae (including F1) mandate a flat region.


Edited by gruntguru, 27 January 2018 - 03:44.


#9 Charlieman

Charlieman
  • Member

  • 1,665 posts
  • Joined: October 09

Posted 27 January 2018 - 14:49

For outright L/D the airfoil underbody would be superior. That's why many formulae (including F1) mandate a flat region.

One might look at a diffuser as a "broken venturi". It is one that is de-optimised by cutting the front off.

 

The principles go back to Bernoulli and the understanding that the pressure of fluid (and fluid flow) reduces if the volume is expanded. On that basis, a reverse wing at low altitude, a venturi or a diffuser might be considered different ways, with different efficiencies, to do the same thing.

 

However, the way that diffusers have been exploited -- engine exhaust blowing, flaps and gizmos to reduce pressure at the diffuser exit -- means that it is not enough to have the greatest area or form. It's why I think the field is best described as under floor aero dynamics.



#10 saudoso

saudoso
  • Member

  • 6,741 posts
  • Joined: March 04

Posted 29 January 2018 - 01:52

The biggest volume you can expand the flow over, the best it is. Flat floor/difuser isn't a design choice, is a rule limitation.

#11 Charlieman

Charlieman
  • Member

  • 1,665 posts
  • Joined: October 09

Posted 29 January 2018 - 13:58

The biggest volume you can expand the flow over, the best it is. Flat floor/difuser isn't a design choice, is a rule limitation.

A designer is dealing with a cubic (expansion volume) to square (under floor area) ratio, and that ratio always makes rules of thumb tricky. Plus you have leakage, which is related to perimeter.

 

I agree, of course, that motor sport rules require designers to create unusual solutions. To me, the problem with all under floor aero solutions is the limited applicability beyond motor sport. I'm a sceptic so I can see examples when they might work for a supercar and circumstances when things could go horribly wrong. 

 

When considering how an F1 aero technology might be back-ported to junior categories (if thought the right thing to do) or back to F1, the side pod venturi solution is most practical to implement. Venturi entry and floor areas can be regulated. Side pods form part of the deformable crash structure. Venturi entry and exit designs can be regulated -- I think!

 

When starting this debate, V8 Fireworks wrote: "Assuming of course, that sliding skirts cannot be used on safety grounds."

 

If I remember my history correctly, sliding skirts were banned because the cornering speed of F1 cars exceeded the expectations of late 1970s and early 1980s rule makers. Skirts were controlled (ineffectually) to slow progress. As an object, skirts were not considered dangerous to drivers and spectators although a few left the car unexpectedly. But there is no reason to make skirts evil whilst allowing other technology.



#12 PayasYouRace

PayasYouRace
  • RC Forum Host

  • 18,517 posts
  • Joined: January 10

Posted 29 January 2018 - 14:05

There's a lot of questions in that OP.

 

As has been asked, better for what?

 

One thing I can briefly add is that you rightly point out that early GE cars had very two dimensional tunnels. My university project 10 years ago was to take a car with such sidepods, and I added fillets to the inside of the tunnels to make them 3 dimensional. It increased L/D quite significantly.

 

More later... (hopefully)



#13 Charlieman

Charlieman
  • Member

  • 1,665 posts
  • Joined: October 09

Posted 29 January 2018 - 14:27

My university project 10 years ago was to take a car with such sidepods, and I added fillets to the inside of the tunnels to make them 3 dimensional. It increased L/D quite significantly.

To make the exit of the venturi more like a trumpet than a tunnel?



#14 saudoso

saudoso
  • Member

  • 6,741 posts
  • Joined: March 04

Posted 29 January 2018 - 15:05

I've been taking part in a racing car design/ CFD competition.
What I've learned over a year is that ut all depends.
Something works great, then you change another - downstream or upstrem, it doesn't matter - and that nice thing you had works no longer.
So there isn't really a straightforward anwser.

#15 PayasYouRace

PayasYouRace
  • RC Forum Host

  • 18,517 posts
  • Joined: January 10

Posted 29 January 2018 - 15:11

To make the exit of the venturi more like a trumpet than a tunnel?

Yes I suppose. Like a square section trumpet.

One of the keys was controlling the inlet size as well as the throat and exit. Too much air coming in the inlet and the Venturi gets overloaded and stalls.

Based on a conversation I was lucky enough to have with Willem Toet, I drew inspiration from his Benettons of the 80s. I put in a sort of shaped divider at the now blocked off entrance to the side pod, and then placed a barge board to feed it with air from the nose rather than the dirty wheel wake. It worked well.

ETA: I was in a four man team, I was responsible for underfloor, one was rear wing, one was front wing and one was bargeboards. What made it extra interesting was that we chose not to replace the sidepods, so we had the self imposed restriction of a fixed exit to the venturis, though we could make a new diffuser for the central channel of the car.

Our project was an annual cycle and there was obviously no chance to develop my ideas further myself because by then I was graduated and the following year’s students took over,

#16 PayasYouRace

PayasYouRace
  • RC Forum Host

  • 18,517 posts
  • Joined: January 10

Posted 29 January 2018 - 15:29

To answer another of the OP’s questions, the reason modern flat floor cars produce so much more downforce than the old wing cars is because modern cars use the flat floor in a much more sophisticated way. There was some discussion of the BLAT Indycars recently, and that was a forerunner to today’s cars in many ways. Basically, you use the rake of the chassis to create a diffuser out of the whole thing, and then seal the sides with vortices generated upstream. With modern computational and experimental methods that weren’t available back in the late 70s/early 80s, the downforce sapping turbulence and separation can be controlled.

Cars like the Lotus 79 and Williams FW07 are very crude by today’s standards.

#17 PayasYouRace

PayasYouRace
  • RC Forum Host

  • 18,517 posts
  • Joined: January 10

Posted 29 January 2018 - 15:32

To address the point of the Aston AM-RB001: It represents very much a clean sheet design. It’s underbody has been carefully shaped to eliminate the strakes and separators that are seen on many cars. Those are there to control the flow in less than ideal shaped areas. They add a lot of drag, and would be avoided if possible.

#18 Charlieman

Charlieman
  • Member

  • 1,665 posts
  • Joined: October 09

Posted 29 January 2018 - 15:49

To answer another of the OP’s questions, the reason modern flat floor cars produce so much more downforce than the old wing cars is because modern cars use the flat floor in a much more sophisticated way. 

Sort of agree. Just a guess.

 

Basically, you use the rake of the chassis to create a diffuser out of the whole thing, and then seal the sides with vortices generated upstream. With modern computational and experimental methods that weren’t available back in the late 70s/early 80s, the downforce sapping turbulence and separation can be controlled.

 

One quality of vortices is that their behaviour is unpredictable. You make a guess at how they swirl. Is your model right or wrong, and how can you test it? Does the swirl go my way?



#19 PayasYouRace

PayasYouRace
  • RC Forum Host

  • 18,517 posts
  • Joined: January 10

Posted 29 January 2018 - 16:21

Sort of agree. Just a guess.
 

 
One quality of vortices is that their behaviour is unpredictable. You make a guess at how they swirl. Is your model right or wrong, and how can you test it? Does the swirl go my way?


A vortex isn’t unpredictable, it’s just harder to predict. Even then you can make educated guesses. The swirl will to the direction predicted by the high and low pressure surfaces of the vortex generator. A steady vortex is just another flow pattern and can be used with great effect.

Modern CFD has come a long way since even I was using it a decade ago and that helps a lot.

Advertisement

#20 Charlieman

Charlieman
  • Member

  • 1,665 posts
  • Joined: October 09

Posted 29 January 2018 - 16:54

A vortex isn’t unpredictable, it’s just harder to predict. Even then you can make educated guesses. 

Thanks for your alternative description! I'm not being snide.



#21 PayasYouRace

PayasYouRace
  • RC Forum Host

  • 18,517 posts
  • Joined: January 10

Posted 30 January 2018 - 08:12

Thanks for your alternative description! I'm not being snide.

 

I didn't think you were being snide. It's one of those things that are hard to grasp because there's a lot of similar but different concepts in aerodynamics.