mark kibort

Just discussing here Scott. If im thinking incorrectly, i would like to know.
but as far as what i know, you cant create higher velocity venturi effect with the diffuser, no more than you can with the exit nossle of a venturi. Do you not agree? sure there are some youtubers and that website that says it can, but just because it is said on the internet , doesnt make it true. if it is, i for one would love to know how.

PPo

I have been reading about diffuser design and underbody design for over a year now... because I am a car geek.

The only real hardcore truths that are ubiquitous (IMHO) are almost all found in Carrol Smith books:
-The airflow to the rear diffuser needs to be clean and ample. As a result, do not bother with a large diffuser if you can't feed it lots of clean air.
-Separation and dirty air intrusion severely impacts the benefit of a diffuser.
-A diffuser alone, without proper consideration, understanding, and design of the rear spoiler/rear wing, is a useless and trivial exercise... but it can still look cool.
-Many diffusers are designed to meet the restrictions of the rules. Just because you see it on a race car, doesn't mean it's the best design.

This next lesson I learned from RaceCar magazine, and ample research on the web:

-Unless you can spend many many hours in a wind tunnel, the maximum design angle of a functional diffuser is 7 degrees from a flat road plane. Your maximum angle should be reduced if your car has any kind of pitch, dive, or roll. The amount to reduce the angle is based on voodoo, guess work, and reading tea leaves unless you call Ernie at Mantis and stick your car in a wind tunnel. Boundary layer adhesion in a diffuser takes a lot of time to regain, so if you hit a bump in a corner, and you have an aggressive working diffuser well tuned with your wing/spoiler, you better be damn good at car control.

I looked at many, many race car/street car underbodies... then I looked at the Ferrari Enzo. The Enzo makes it clear what to do. Direct clean flowing air into the diffuser, make the diffuser as long and shallow as possible while still increasing volume as much as possible (yes, I said it right, find the balance).

Most of the BS you see on cars is just to impress kids and girlfriends. If you have 4 sections in a diffuser, you either have a whole whack of nice flowing clean air under the car, or your wasting your money. Most flat bottom race cars, after radiator cooling, brake duct cooling, having parts stick out past the floor bottom, and trying to avoid dirty flow from the side of the car, can only handle two diffuser sections effectively. So non-wind tunnel tested bolt on "track tested" diffusers are a waste of carbon fibre.

But I could be wrong...

I am reading all your posts intently as I plan to build a 944 splitter, semi-tunnel, diffuser, spoiler combination for ChumpCar racing. If work goes well, I will work with Ernie (as I live in Toronto) to test a few options...

Just my $0.02. Do you want change back?

mark kibort

actually, that was worth more!

bottomline, when you read things like the site that Winders posted, and one he posted 5 years ago on another same topic thread, the gross generalizations and claims are a little misleading. a diffuser is just that... it diffuses the flow. the flow slows, and the pressure rises. thats all its supposed to do. In the "web" articles, they tallk of " throat of the diffuser" and " high speed , low pressure , high downforce in that area" what the heck is that?? well, ill telll you , its the venturi and they dont clarify that it is much further up stream... jut until the mouth of the venturi.... and you are left to think that the diffuser does some air acceleration. it doesnt. you cant change the flow upstream with the downstream component, unless you block or restrict the flow. you need a venturi like structure ahead of the diffuser or it wont have a thing to do with downforce.

Most under body set ups are flat.. they get a little pressure reduction by the air accelerating under the body, because it's usually joined by side higher pressure zone flow. the most effective ways to get downforce is with the diversion of the air through hood vents. (and wheel well vents). that with a good wing in the rear and maybe a diffuser for drag reduction and you are probably 90% there for a street car conversion.

so, i think it would be interesting for you to build some venturi tunnel and a nice diffuser for the chump car and see how it works... i bet it would work well!

ace37

This is a one hour, sloppy adaptation of a 2D study I built for a college class five or so years ago. I had used a photo and CATIA to get a rough 2D center cross section of my Nissan 370Z Nismo. I then made a second version that was identical in every way except the spoiler was removed for comparison. I threw both into a CFD study using 100mph car/air velocity and a rolling road that matched the car speed.

The original mesh was terrible as my computer was weak and it was good enough to meet the assignment, so I reran it today with a mesh that was slightly less atrocious than the original. The mesh is shown and gets down to 1mm at the car surface. Convergence conditions were sloppy as I essentially ran it a few minutes on a nice modern machine and let the dust settle on energy/conservation/etc without trying to optimize or clean anything up. It's 2D anyway, so it's got a lot of assumptions and won't be very accurate to the real 3D car no matter what I do. For that reason I also didn't bother to sum up the downforce or drag.

Still, despite it being essentially a hackjob, it's free, available to us today, and on topic. The geometry/mesh is shown first, then the pressure and velocity contours at the car in English (American anyway) units:

ace37

Looks like a lot of fun!

mark kibort

good stuff.. but look how sensitive the simulation is. you cut out the 2d outline of a picture.. due to the picture shape, (not actual shape due to angle of photograph), you get a very very slight reduction in space under the toward the rear... since the simulation has to go on that change of dimension, look at how the speeds increase and pressure drop right at that point, but in reality, there is no such reduction of body height vs the ground. this has to do with the angle of the picture, and the trace... the rear bulge of the wheel flares, make this anomoly, which skews data in that region.. also because the picture was taking a few degrees higher than center

here is what porsche produced for a rake change:

333pg333

I totally agree about the concept of the rear diffuser being a lot more complex than many would think. Basically, what a few people in here have said already. If you don't have CFD, Tunnel or flat bottom, it's almost certainly not worth putting some bolt on thing under the rear of your car. That's why we went away from trying to 'capture' the air with very low rockers to funnel out the back. Better to get rid of that dirty air sooner rather than later.

mark kibort

I was up last night thinking about your tests and experiences. if you think about you set up and results, you can see that what you saw might have been due to the splitter working too well up front which gave you a loose car. you opened it up, reducing the front bite and giving the car more balance, rather than the diffuser creating any downforce. again, a diffuser, as positioned near the rear axle doesnt seem to be able to create any downforce, as this is the area where the air is expanding in volume, decelerating (not accelerating) and the pressure is rising. if it wasnt going faster than oncoming air to begin with, then the only thing the diffuser will do is straighten the flow as it exits the underbody to meet the air coming from the top of the rear of the car.

If anyone has theory to the contrary, i would love to hear it. As Patrick says, it is a complicated set of factors, but what we think is not often proof of what is happening. Going over a bunch of experments, did back in school, along with the equations, i still see nothing that indicates the diffuser can effect upstream flow.

what is interesting is the diveplane efffect on the underbody flow leaking out the sides. that i would like to see in windtunnel flow or in theory or simulation.

Mark

Krokodil

Mark,

This is why I suggest you need to think differently and at a molecular level. The diffuser does create downforce (please accept this for the moment) and it does so by accelerating the airflow under the car across the apex of the diffuser (this is why the location of the apex matters). It also requires the air at the forward throat to accelerate (hopefully near the front axle) providing front downforce.

However, your question is essentially how does the diffuser cause the air to accelerate? I am not an expert, but I will do my best to parrot (and simplify)my buddy until I can get permission to post his excellent emails:

As the fixed mass of air in the constrained volume under the car enters the larger volume of the diffuser it requires more mass to match the mass flow of the air downstream of the diffuser. To add this mass the air in the diffuser attracts, or pulls on, the air upstream of the car as it moves under the splitter. This attraction happens at a molecular level. And yes, as I understand it, this requires the air under the car to be moving faster (relative to the car) than the car relative to the ground.

Here is a small extract from one of the emails that I am pasting as it explicitly references downforce (please trust that he has the credentials):

"You could add a packer to tighten up the radius of the forward diffuser apex. I wouldn't go any more aggressive on the second upsweep, but a tighter forward one would increase the rear bias to the downforce contribution from the diffuser.

The exhaust blowing idea is a great one, but I'd pull the tips of the tail pipes forward a bit, so that the exhaust jet blows along some length of the endplate. It'll stop air spilling around the endplate and into the diffuser, which will be primarily a drag benefit, but there could be a small downforce gain, too."

Here is a link to a basic article until I can find the papers I am looking for: http://www.racecar-engineering.com/t...-aerodynamics/

BTW, his estimate re: my diffuser was that it probably provides 50-70# of DF at 120 mph with 70-80% of this at the rear. No, this is not very much, but it is 25% or of what is produced by the wing and essentially drag free.

Her is a picture of the diffuser before it was fitted. That is not my car in the picture, but it does provide a scale reference:



Cheers,

PPo

Interesting...

I am under the impression that flat bottom race cars are a product of rules rather than an application of aero design. In fact, it is my understanding that flat bottom cars produce that evil thing called lift... so getting rid of the dirty air sooner is probably a good thing.

So, I am contemplating a tunnel/diffuser design for the 944 series cars that starts just after the front passenger compartment area (around the mid jack point). It would consist of two 6-10 inch tunnels on either side of the exhaust/torque tube channel, and graduates at about 5-6 degrees away from the road plane, and eventually exits out the back. I am not too concerned about "feeding" these tunnels as there will be enough air from a relatively flat bottom ahead of them, but I am concerned about dirty flow hurting the inlet air flow. So I was thinking about small 1/2 inch channels leading up to the tunnels. (But those may create turbulence and ruin the whole thing!). I also planned to run these tunnels close to the transmission, just clearing the drive axels. The trick part is re-routing the exhaust. Not sure I want a side exit exhaust yet, so I may split the exhaust and route it over the channels and exit the exhaust between the rear wheels and the new diffuser tunnel... but then the idea of an exhaust blown diffuser comes a possibility, and I say screw it... it would be much easier to lose 30 lbs of fat to gain the speed advantage the home diffuser kit would bring...

By the way Mark, here is a clear example of a diffuser impacting the flow ahead of the car:

91K-Man

Hi All, I just wish you guys could be in Tunnel. A lot of what is discussed , will be cleared up when you get that oportunity. We all have BOOKS about this or that . I sat in a class for a week,, at SAE about wind tunnel testing. Well , when get in tunnel and actually blow some wind , your thinking will change.

Rear diffusers, yes I have tested several in house designed ones. A Yes ,they work, but you need to turn them on.

No tunnel , good guessing, in the tunnel , you get an answer.

mark kibort

haha funny!

yes, in that case the diffuser will effect the rear, (which was the front) of the car

as far as flat bottoms go... they do indeed, create downforce if very close to the ground. however, in the same way a plane can fly upside down and level, so can the flat bottom create lift, if the angle is increased just slightly above level. this is why tunnel and venting technology is so important for safety so this is less of a problem.

again, the diffuser is really for the merging of the air from the bottom to the rear of the car where it will meet and feed the top car air flow and rear of the car turbulence area. remember, there is no need for "diffusers "on a wing because the wing ends in a sharp point... we cant have cars that look like wings, but they keep on trying and are getting close.

if you look at the science and theory of how a diffuser works and what it does, it all makes sense.. However, im going to take a look at what Duane is saying and keep an open mind and bounce it off others i worked with in the field.


i have a scale tunnel so i can see the same effects on model versions or the concepts of venturi effect or diffusers.
the theory is sound for venturis, and this is taking a lot of that theory and putting it to use... what worked on the diffuser in the tunnel. when you say, turned it on , you mean got a lot of air through it? what was the result? lower drag? did you see frontal changes in downforce before the diffuser?

mark kibort

first of all , lets get the definition of the diffuser, because the picture you posted is a combination of things. you have a curved structure and i assume you refernce the "apex "as the highest point as it sits on the ground next to the car. actually, you have tunnels feeding the structure before the "apex and converging. this is not a diffuser, this is a venturi both horizontally and verically (above the ground) so yes, air passing through this region would have a lower pressure.

Now, your friend or (maybe just your interpretation) seems to be bending the laws of conservation of momentum, just a little with his statement.the mass flow doesnt change, it cant, but the speed and pressure do. this is how a venturi works... air speeds up at the converged area and pressure drops mass flow stays constant, and at the diverged area (the diffuser) the speed slows down, and pressure goes up (mass flow still is constant). that is the area where the diffuser starts heading up up toward the rear as bolted on the car.

with me so far?

Now, as far as your expert friend is concerned.. Im trusting his credentials here but what he is saying is fundamentally incorrect. maybe send him this comment from me to see if he can clarify. this is basic aero 101 stuff, so much might be lost in his description. Just like in a venturi, the air flow entering the venturi is at a fixed speed, pressure and volume.. the air flow hits a decreasing radius and the speed increases, while the volume decreases as your friend says.. However, the mass flow is retained constant due to the pressure (density ) going down , and speed going up... so, no change in mass flow. the reverse happens at the diffuser or exit of the venturi. the volume goes up, and the speed slows down as the pressure rises... again, mass flow is retained constant based on the law of the conservation of momentum.
also, the flow average speed cannot be faster linearally than the ground, as the only reason it speeds up, is its traveling a longer distance around the radius of the tunnels or converged area vs the ground, the speed returns to road speed on the exit. so i guess , technically, yes ,it does travel faster in the converged areas.

I think what he is quoted as saying, is that if you reduce the front "upsweep" you make the front aero less effective, putting more air to be accelerated at the rear of the apex of the diffuser entrance. keep in mind, this is NOT the diffuser, the diffuser starts at the point where the air is decelerated and pressure rising. (terminology, even though the manufacturer calls the entire piece a diffuser)

so, my take on what you and he has said is that the "apex" area is a converged region narrowing down the volume of the air flow. (not the difusser part) this accelerates the air, speeds it up and creates downforce... this is all true.. i understand. BUT, the diffuser is after the "apex" thats where the piece guides air to the rear of the car and the speeds slow, pressure rises, etc

I think it was your interpretation that the mass flow might change with the volume change, not his.. this is not happening in the system as it doesnt in a traditional venturi either. But, now i understand what you are saying and how it works. thanks for the clarification and info. makes me think about how to create one for the frontal region OR maybe fix a underbody gap that is unexpectedly createing lift under the car at the front axles.... thanks!
mark

Krokodil

Mark,

Edit - Changed my mind - I am out now.

Cheers,

91K-Man

Ok, how big is your scale tunnel ?