Difuser design and function

[Bill Musarra]

I have been studying lately about rear diffuser design and function.I run across several common threads. One that the design of the "fins" should direct the air away from(??) the center at an angle of approximately 7º.

Others say to keep the fins straight as seen on the F430 Ferrari.

I have several questions for now to get the ball rolling.

1. Should the fins go the full length of the diffuser
2. What is the ideal(or close to) upward angle from the bottom, and should it be a straight line or gentle curve. What is the stall angle???
3. Should the rear wheel wells be closed up, or close to that, on the inside to benefit the effect, or should a slanted fin(slanted toward the wheel away from the vertical) be placed at the wheel well to keep the flow away from the wheel well as above?
4. Should the underside of the body be extended out toward the wheel to eliminate air flow to the wheel well?
There are probably a lot more questions to ask, but this should start the discussion.

Bill

 

[Terry Oxandale]

I've heard the 7º reference many times as well, but I was thinking is was a limit in that beyond that, eddies and turbulance (stall?) can develop. It seems I've seen many that were designed at higher angles though, so I'm not sure. I've read that the nearer it is the the ground, the more effective they are and the greater the useable angle. I've seen the ones that splay outward as well as upward, but intuition tells me if you increase your area by going outward, then the vertical increase must become less, least you again, induce turbulance.

 

[Russ Noble]

Diffuser design really needs to be approached in a holistic manner looking at the overall airflow under the car. If you are looking for downforce with a flat bottom car, the diffuser is merely the rearward section of a basic venturi setup.

The front of which is the intake from the nose of the car which leads into a throat (the point of lowest ground clearance) behind which is the slightly raked flat bottom of the car, followed by the diffuser.

Do we have a resident aerodynamacist on this forum? Some expert input on application to a GT40 would be most worthwhile. There was quite a bit of research done on the aero for the 'new' Ford GT which might be worth looking at.

Does anyone have a link to any hard facts on the GT?

With regard to your questions Bill, below is only my opinion from recent reading.

1. I don't think it would do any harm to run them full length but they probably have less effect as they get smaller.

2. Testing (McBeath) of laboratory models (CFD) indicates an optimum angle of 12° -14°. In real life that will vary between applications, the frontal design of the car, ground clearance etc. If you have an associated rear wing, that will help 'drive' the diffuser and greater angles are possible. See the angles and wings on the back of F1 cars!

3. Anything that will stop turbulence from the wheel affecting the diffuser has got to be good. Flow from the outsides of the diffuser exit is often very erratic.

4. A 'shelf' close to the front of the rear wheel can act in a similar way to a front splitter, imparting some downforce in this high pressure area. This is done on some single seaters but I am unsure of its application to full bodied sports cars. There has to be a theory that it wouldn't do any harm, but would it do any good?


Iain Pretty has had some experience with enhancing the aero on his RF117. I'm not sure whether he was running diffusers though. Often what you see on race cars is not the absolute optimum setup, it is the optimum setup that will satisfy the usually restrictive rules with which the competitors/designers are forced to comply.

 

[Mike Trusty]

For conversation purposes here are a couple of pictures of the Ford GT.

 

[Jimmy P]

Bill, I recall a forum member who has experimented with a diffuser on a GT40 as he was involved in doing high speed runs in his car. Sadly, I can't recall his name. Might be worth a search though. I do recall him posting some nice photos too.

 

[Howard Jones]

This is a very good website for looking at pictures of various prototype's. I spend quite a bit of time here.

I include a link to get you all onto the site, and also has a bit of information on difuser's. There is much more here, you will need to just surf away.

Mulsanne's Corner: What is a Diffuser?

 

[Bill Musarra]

I still have a few questions.
This 7º that is talked about. Where is it. Is it in the curve to create the Bernoulli principle? In the angle upward? In my line of work, we deal with the principle all the time and I understand it well. We use it every day in the hospital in the delivery of medications and aerosols. But this 7º thing has me.
Next point is, what "form" should the diffuser take. On the Ford GT for example the upper plate is straight in its trajectory up to the bodywork. On others it is curved upward and meets the body work sort of like the Ferrari pic above. This seems to be more of an effort to make the rear look more appealing rather than functional. If the more true form is the straight line up to the body work, what is the maximum angle that should be considered?


Bill

 

[dennism]

I believe that the trick to the diffusers is to keep a constantly accelerating airflow attached within the diffuse to produce downforce
If the cross section area increase too rapidly, the flow will separate into turbulent flow with subsequent reduction in downforce
The reason that a flat top with increasingly wider spaced fins or regular fins with an increasing angle in the top would have the same effect since cross section area would be the same in either case
It would be interesting to see the effect on a Ford GT at speed with and without the diffuser
I would attach a picture of the Enzo diffusers that I have in my files but will have to learn how, its diffuser starts a short distance before the rear wheels
Also note that as the vehicle yaws, the diffusers cease to function properly, if at all
Dennis

 

[David]

I'm no expert but I've done a lot of reading up on the subject.

The vains/fins act to prevent the higher pressure air from 'outside' the diffuser from 'spilling' in. The reason for the curves rather than straight fins could be for many reasons, one could be the interference between the spoiler and diffuser. Another could be a lack of height to get enough angle on the diffuser.

You want to avoid flow separation, hence the smooth curves of the Ferrari and the limits on angles.

You can't get away from air 'spilling' into the wheel wells but you can do your best to keep it awayand 'vent' them afterwards. This is why the original GT40's had the vents above the lights. Also note the diffusers in front of the front wheels of the GT, the same applies to the rears.

See this enzo underbody (It is I believe a model but I think accurate).

Notice the diffusers ahead of both front and rear wheels and the way the rear wheels have 'fins' to separate the flow where possible. Also notice the big holes behind the front wheels to 'vent' the wheel arches. (There will be some for the rears too but don't know where they are)

 

[Cliff Beer]

We're missing something basic here..... The main reason the diffuser fins curve outward towards the rear is to create flow (rather than turbulence) in the wake of the rear tires.

As you can imagine, a really wide rear tire on a super exotic leaves a huge amount of turbulence in its wake. The diffuser is pinched inwards at the axle point due to the width of the rear wheels (not ideal). Thus, downstream (rearwards) of the rear wheels the diffuser is trying to lessen this turbulence by inserting attached flow into void. Why? Because all that turbulence created by the wide rear wheels tends to detach the flow across and up the diffuser at the rear, rendering it much less effective in producing downforce and steady flow.

I don't think there's any hard and rule about the angle. The amount the fins can bow (ie. the "cord line") outwards and still produce efficient/attached flow depends on a number of factors, the primary one of which is the expected rate of flow. You'll notice that the aerofoil on a plane designed to cruise at 175 mps is much "thicker" and comes with a cord line of a greater angle, then one designed to cruise at 1,075... Well, same principle here.

Regarding whether the rear wheel wells should be boxed or not, conventional wisdom is yes, absolutely. There are a few exceptions where the designers have tried to use that area as a component of the diffuser (not boxed) with the wishbones and CV axle designed to run in the flow, but it's very difficult to pull off properly.

Good luck!

 

[Russ Noble]

<SPAN style="FONT-FAMILY: Arial"><FONT size=3>

Also note that as the vehicle yaws, the diffusers cease to function properly, if at all<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com

Dennis

<o></o>
<o></o>
But down force, which is the point of all this, increases. Possibly depends on application though. <o></o>
<o></o>
It has been shown during wind tunnel tests on a Radical SR10 LMP sports prototype that downforce increased up to 4° yaw angle (approx 3% increase) then reduced and at 10° was approx the same as at 0°. However at 10° yaw, front down force increased 13% and rear reduced approx 5%. Part of this could possibly be attributed to flow separation in the diffuser but no doubt there are also other possible explanations. <o></o>
<o></o>
With regard to diffuser angle, that very much depends on throat area, which is essentially ground clearance. Small throat areas will only support small diffuser angles, large throat areas will allow greater angles. For a road car with a flat floor and 100mm ground clearance, from what I have read, a diffuser angle of about 14° could be achievable without significant flow separation. <o></o>
<o></o>
That brings up another point that I am unsure of. Is it better to have a smaller diffuser that runs underneath the suspension components giving a relatively clean flow, or is it more beneficicial to have a larger diffuser that the suspension links protrude into, as I have seen on some race cars? Would the gains of a larger diffuser more than offset the losses caused by the presence of the suspension components affecting the airflow?<o></o>

 

[dennism]

It's easy to trip laminar flow into turbulent, a rivet head is enough to do it
Converting turbulent into laminar is next to impossible considering the limited space that we have in our vehicles
The suspension components that are visible to underbody air flow just before the diffuser produce only turbulent flow entering the diffuser, limiting diffuser effectiveness
Ford and Ferrari both correctly keep the lower suspension links above the vehicle's flat bottom to isolate the diffuser from suspension turbulance
Dennis

 

[Tom]

If a chassis has the ability to accommodate a large volume tunnel, such as a 956, then the lower rear suspension could be left in the air stream.

However, very few of the cars we are building have the capacity to run a diffuser tunnel far forward since that would likely require a carbon fiber chassis, a very wide body, and use of engine/transaxle as a structural member, to build a significant volume. With our limitations, it would seem clear that the way to optimize the limited tunnel volume is to have the rear suspension tucked out of the airstream.

Also, with regards to the Radical, I suspect that it is required to have cambered side pod floor as per LeMans spec. When a car gets sideways, the camber acts as short diffuser, helping to prevent the car from lifting as it slides.

 

[Cliff Beer]

It's easy to trip laminar flow into turbulent, a rivet head is enough to do it
Converting turbulent into laminar is next to impossible considering the limited space that we have in our vehicles
The suspension components that are visible to underbody air flow just before the diffuser produce only turbulent flow entering the diffuser, limiting diffuser effectiveness
Ford and Ferrari both correctly keep the lower suspension links above the vehicle's flat bottom to isolate the diffuser from suspension turbulance
Dennis

Hi Dennis, the point was not to suggest that a diffuser can direct air outwards at the back and re-attach the flow in the wake of the tire. You're exactly right....obviously, that's not going to happen. I was merely pointing out that directing air which is flowing in a controlled and less turbulent manner into that area of high turbulence is beneficial overall for a number of reasons, a couple of which are mentioned above.

 

[dennism]

It would be interesting to run a small tuft test of the rear quarter of a GT that could show flow patterns adjacent the wheel and through and aft of the diffuser
The GT has enough similarity to a 40 that it could probably answer in better detail what we're looking at theoretically
GT40 aerodynamics have been somewhat of a passion with me for about 40 years, the original design as developed in competition was way ahead of almost everybody at that time
Remember that the Porsche 917 was a terror to drive until the Brits that had worked on the 40s put their experiences to get its aerodynamic issues resolved, Porsche didn't comprehend downforce when the 917 was developed and thought that the only secret to success was light weight and low frontal area
BTW, the GTs running in those mile drags often do over 250+mph and haven't had the aerodynamic stability problems that some of the Lambos have at that speed, their similarity to the 40s shows how good the original GT40 design was

 

[Howard Jones]

Most if not every GT40 chassis design has a horizontal chassis tube at the very bottom rear of the structure. This ties the tail section together and provides lateral stiffness to the rear sub frame.

The challenge would be to incorporate a diffuser that extends forward of this chassis member without a redesign of the chassis in this area. Moving the lower lateral chassis tube higher or incorporating a arcing upward piece to provide clearance will usually disturb rear gearbox mount locations as well as body mounts and other components in this area .

If this piece of chassis could be altered then we would have something to work with.

At some point I may attempt to address this area on my SLC, but with so much to finish in the basic construction along with the full development and setup cycle left to do after that it will be awhile.

 

[Bill Musarra]

Howard,
There are some 40s that meet that criteria now. The DRBs have square frame members that encase the trans and have up sloping members to the rear at the body line. These would be ideal in my mind for the diffuser. Take a look.

Here is a template of how the straight and the curved diffuser would look for first consideration. They would need modifying.

The first pic above shows that the bottom of the car is below most of the rear suspension and pieces could be fabed to take out much of the wheel well influence. On the rear clip a fiberglass panel drops down just inside the wheel and could be tied in. I have the openings in the wheel well tail section for venting already. I am not real sure I want to enclose the bottom of the engine bay yet, even with my 4 extra fans in there.

Bill

 

[David]

See my previous pic of the Enzo underbody and the way there is a shroud round the end of the transmission. Something very similar could work on the underside of the GT40

 

[Bill Musarra]

I think we have gotten some good responses on the question(s). It got me to thinking and I went back and reviewed the principles and looked at images that demonstrated the ideas we talked about. Some of the explanations were to me a little difficult to understand and a little hard to grasp. So I have decided to take a shot at "seeing what is going on", and simplify it. The whole concept of ground effects is just the Bernoulli effect in action on several fronts. With a diffuser, its purpose is to help keep the rear end planted on the pavement at speed. It allows us to do things with a very light car at speed that we could not do if we didn't use these ideas. I had to go back and reread the principle and I think a few of us are confusing some of it. So I thought a little down to earth talk would help grasp how this basically works. Here goes. Feel free to straighten me out if this is not so. I sourced some of the graphics from the internet and tried to give credit.

Think of three shapes.
The classic cylinder with the reduced diameter in the middle

<center>Bernoulli Concepts

A baseball.

Watts and Ferrer

The wing.

Eastlake and NASA Aerodynamics

Each will demonstrate what is going on with a diffuser. Remember I am trying to keep this simple OK?
Most people see the picture of the tube above and think that a drop in pressure is all that is going on. But it isn't. Simply put the air(fluid) that is flowing through the tube, when it reaches the restriction, speeds up in order for the same amount of air to pass through it as the larger part behind it. When it does, the pressure it uses to exert on the walls drops for two reasons. First is that, as it speeds up, the pressurehead on the front of the moving air is increased. Consequently the lateral wall pressure falls. Second, the air moving through the tube, at the surface of the tube is going slower than the air in the middle of the tube because of friction with the walls creating turbulence, which in turn creates pressure. This turbulence or pressure difference trys to equalize by drawing the walls into the faster moving air in the middle, it literally trys to collapse the walls. If you vented all three sections and installed a flow meter on each, the vent in the middle would be lower than the two ends. You can demonstrate this in your shop with a compressor and a pencil. With a blow gun on the end of the line blow the air over the top of the pencil. The boundry of the air going over the top of the pencil will create a turbulance that will literally pick the pencil up without any other force. Those of you with a simple mind like mine can see this in action in another way. With a Whooppee cushion. For those that don't know what a Whooppee cushion is PM me.
The comparison here is the curved lines of the fins on a diffuser.

Now lets look at the baseball thrown as a curve ball. For our purposes, we will be throwing a sinker ball. One that starts out level and curves down toward the ground as it approaches the batter. Imagine looking at the diffuser from the side, as if it is the top of the ball as it goes toward the batter. The top of the diffuser creates a high pressure zone next to its surface from the turbulence of the air being dragged along it. This makes the air pressure next to the ground lower in pressure and it tries to pull the diffuser toward the ground just like the baseball.

The last part is the wing. Imagine the wing upside down. This is basicly the rear of the car where the diffuser angles up to meet the body of the car. The angle causes the air to increase its speed which drops the pressure and it pulls the car to the ground(upside down lift). For those of you with the SLC with a wing. The wing amplifys the down force by being behind the car and increasing the speed of the air leaving the bottom of the car more and adding more down force.
I hope this gives an easier picture to think about. Like I said, if some of this is too far off for you let me know. I am not an engineer, and I have never played the part of one. So feel free to jump in if I am too far off for you. Some of this is a strech, but it is the basic idea I'm aiming at.

Bill

​

</center>

 

[Terry Oxandale]

What is the portion of down-force created specifically at the diffuser? The term "diffuser" indicates to me it is a device that facilitates the creation of down-force from the panels and venturies under the car, but not necessarily at the diffuser itself. By allowing the air to exit as cleanly (diffusing in an organized way into the chaotic turbulence behind the car) as is possible out the back of the venturi, the venturi is allowed to function most efficiently. If it doesn't work as planned, the higher pressure, slower air signals forward and thus reduces the effectiveness of the venturi.