Diffuser Design

[iamuwere]

I drive a small formula car in SCCA competition. Top speeds of mid 130's with good flat cornering. We are allowed a flat bottom (55" width through the sidepods) with no more than a 1" deviation in height across the floor pan. Area before the front axle and after the rear axle are free. The obvious area to work with is the rear diffuser design.

I not yet found any good rules for diffuser design. Currently there is about a 10deg rise over approx 24" length. One center strake (straight) with straigt side plates.

Any good rules on diffuser design or good references with practical application?

Thanks,
jim

 

[Dave K]

I used

http://www.mulsannescorner.com/

for reference, and Joseph Katz book on racecar aerodynamics.

7deg slope seems to work best for us.

-Dave

http://www.moslerauto.com

 

[perkins690]

Is 130 mi/hr? or km/hr?. In any case, the speeds should be 131ft/sec or 36m/sec respectively. I assume you are trying to create "downforce" or negative lift, the area you have available to work is not that big to really create a significant effect, besides, you should consider the distance between your floorpan and the ground; since the air flow is passing through an open area, has little chance to be laminar and speeds are variable,I assume you have 120mm ground clearance (I might be wrong), 17deg. like most secondaries on a venturi will create the most beneficial gradient.Again, I hate to make assumptions on unspecified data, but, most of those little cars don't run fast enough to benefit from any aerodynamic design advantage. I hope I was clear and of any help.
Good luck!
I race 125 shifterKarts myself.

 

[iamuwere]

yes, sorry for the lack of data-130mph. Ride height is usually between 40-50mm. 120mm would be like have of my Cg.

So far-7-17 deg. Is there a limit to length (beyond class rules limitations)?

 

[pvw]

I am involved with a few cars that run at these speeds and slower and downforce is a significant advantage in high speed sweepers. Sorry no info on diffuser design thats why I am reading but if you can test with tuffts and a video car and a manometer strake. Check paul vandenbergs book racecar engineering out for instrumentation ideas you can make an adjustable diffuser (suck it and see). I race a hillclimb special (500cc kart on nitro methanol) that runs at an average speed of 85kmh 150kph top and the wings are worth 2 seconds in a 40second lap as acceleration is better when I'im not airborne I am looking for more downforce as drag hasn't been a problem I'm no where near terminal speeds. From my reading you may find 2 types of diffuser that will work a long slow one that could benefit from blowing or a short sharp one but it will need an external flap close to the exit to activate it ? Any real designers out there? good luck

 

[dnky]

I ran a difusser on my '81 corolla, in the scca valvoline run offs. I must agree on matching degree slope to the speeds that you expect to run at. I ended up with 12 degrees over 48 inches on a severely weight reduced car. It all depends on where you will need the down force the most, or where it will be most effective for your application. I can only say that i also agree that literature, and testing are the best way to find what works for you. Trial and error sometimes is the only way to fine tune to your specific needs. The literature should get you in the ball park, then practice is all up to you. It did take me several tries, and one pretty embarassing failure to get it right

 

[sreid]

7 degrees is the generally accepted number but cars have a significant negative pressure at the rear which helps the air flow.

Another way to improve the negative pressure under a flat bottomed car (and hardly ever mentioned) is the curve of the side pods as viewed fron the top. The curve should be kind of air foil shaped so the air flow creates a negative pressure along the side pod. This greatly reduces air leakage into the under body.

 

[sreid]

Just the pictures are worth a look.

http://www.mulsannescorner.com/nissanp35story.html

http://alcatraz.net/ysuzuka/

 

[btrueblood]

7 degrees? Or use splitter vanes:

http://naca.larc.nasa.gov/reports/1958/naca-tn-4309/naca-tn-4309.pdf

 

[sreid]

Splitter vanes work in a closed diffuser but in a race car the top of the diffuser duct must "see" the ground so the low pressure in the duct can contribute to the down force at the rear of the car.

 

[btrueblood]

"..the top of the duct must see the ground.."?

The vanes are attached to the car also, thus pressure forces are transmitted back to the vehicle frame. The goal of a splitter vane is to prevent stalled/seperated airflow due to high divergence angles. I'd assume the same goal applies for a racecar underbody diffuser - get the airflow back near ambient without stalling/seperating the flow, thus minimizing drag.

I'm more interested in hearing about whether the moving ground boundary condition affects the divergence angle achievable without stalling; I'd think you could use steeper angles (14 degrees), but that's just from seat of the pants, not any CFD or test data.

There is an ASME paper out there from the '50's that gives a wealth of good design info. on 2d diffusers; the achievable lengths vs. divergence angle vs. inlet Reynold's number.

I like the idea of tufts, and a video camera either on the car or in a chase car. It would be pretty difficult to get reliable manometry data on a moving vehicle, if you go this route, use solid state pressure transducers, and a cheap solid-state recorder (the kind that plug in to a PC via USB cable).

 

[sreid]

So what's the net lift of air flow in a closed tube?

 

[sreid]

Actually, in the fourth picture from the bottom (the bottom of the wind tunnel model), there are clearly vertical aero devices in the diffusers.

http://www.mulsannescorner.com/nissanp35story-2.html

 

[btrueblood]

"So what's the net lift of air flow in a closed tube?"

If the flow turns, there are net forces transferred to the tube walls. This is why water mains have anchor blocks at elbows. Your arguments would seem to state that biplanes are a waste of time, since one or the other lifting surface is doing no work. Or that turbines can't work (at least the ones with continuous tip rings).

Also, you state that "..low pressure in the duct can contribute to the down force at the rear of the car". The goal of a diffuser is to provide a smooth transition from low pressure back to atmospheric, ideally without seperating or stalling the flow. Thus, a diffuser creates higher pressures, not lower.

 

[Dave K]

Racecar diffusers are "open bottom" due to the rules not allowing "tunnels", or underbody "wings". Vertical strakes (turning/straightening vanes) are permitted though.

-Dave
Everything should be designed as simple as possible, but not simpler.

 

[sreid]

Let me try the question of "see the ground" this way.

You have a closed box with 1000 pigeons inside. You beat on the box and get them all flying. Does the box get ligher or heavier?

Same problem but the bottom of the box is a screen (because of pigeon guano, of course). Is the box heavier or lighter with the pigeons flying?

Positive pressure in the diffuser? Certainly more positive than the pressure in the ground effects throat but still lower than atmospheric.

 

[btrueblood]

Gunman,

Thanks, that's a good answer.

 

[btrueblood]

Sreid,

Yes, positive-going pressure.

Screw the pigeons. A turning or expanding/contracting flow in a tube will generate external forces, this is 1st law of thermodynamics stuff. Tell me a pipe can't generate external forces, and I'll give you a one-word reply that proves you wrong.

 

[sreid]

Btrueblood,

I did read the paper you posted on diffuser design and while interesting but I'm not certain it could help generating down force in a race car.

The original ground effect cars (Lotus) had essentually wing section shaped side tunnels. It was quickly found that for the race car case, one got more down force by making the tunnel in (sort of) three parts 1) an inlet nozzel 2) a flat section and 3) a diffusior section. This is all well documented in the literature.

When rule changes outlawed ground effects, the aero guys didn't pack up and go home. If airflow under the car could be used in some way to generate down force, they used it. For flat bottomed cars an inlet nozle was still used (clearly, Mr. tech inspector, the car nose must be radiused to ride over bumps) followed by the flat bottom (usually silghtly diverging) followed by diffusors (the flat bottom rule usually only applied up to the engine bay).

In the paper you cite, the closed diffusor experiments used cannalizing plates to make several channels each of which have a smaller divergence angle than the entire channel. It is intuitively obvious that this will work, the research was to find out how much and where.

In race cars the diffusor expansion mainly happens in the vertical direction. While the essential purpose is to avoid flow separation and tunnel stall, the negative pressure in the diffusor still has a significant down force contribution.

Horizontal "channelizers" in the diffusors would certainly work to improve the flow separation but only the last channel that sees the ground would now contribute to down force. This is a poor trade off for "go fast" reason.

While I agree with your general statement that flow can generate forces, I think you would agree that while flow in a jet engine can generate whopping axial forces, the net radial forces are zero.

 

[pvw]

SReid,Btrueblood,
Thanks guy's for the links good reading and the pictures are worth a thousand words. I have only had access to an early type tunnel open wheeler that uses skirts (ex maclaren designer). The sports cars in the link are more advanced.
I would like to add the following comments. When I was talking about a monometer strake I actually did mean with pressure sensors but you will be suprised what can be done with a water manometer I have a friend that uses one to find high pressure area's on sports cars for new airboxes.
I think part of the picture about Strakes is being missed and that is that the vertical strakes still contribute to allowing a steeper floor angle, especially on flat bottom cars.(its still an rate of change of area thing)
The flow under non skirt cars is effected by a vortex coming down the side of the car and if it can be made to roll out from the bottom it can actualy creat a vacum under the car. I believe they can generate more downforce now than when they had skirts but it comes at a price that is very low ride hights with more pitch sensitivity.
Don't forget the Rear tires that also create a narrowing and turbulence leading into the diffuser.
I remember an artical from a williams designer who admitted that the rotation of the driveshafts going through their diffuser actual created more downforce than a clean tunnel because it helped with reattachment. (Active arodynamic device?)
You are right about a long throat gives more downforce but you must use vortex generation or exhaust activation to keep it attached on the steeper diffuser or a sucking device such as a low mounted wing.
The last comment is that a wing in the throat of a diffuser or a multi element wing generates down force even though there is interference.The wake of a car is at lower pressure than atmosphere!
Thanks again for the links any others would be appreciated