Diffuser / under body questions

[flynfrog]

Im desing an under body for our FSAE car ive read a few books on the subject but i cant seem to find any real answeres.


I under stand the basic shape should be that of an air foil but are there any guidelines as far as speed vs area vs camber and inlet and outlet angles

http://www.lazyforums.net/uploadfiles/undertray.JPG

this is what i have so far i do have room to make it a constant curve but not enough room to add hight to the center section

[miqi23]

Are you sure they would work at those slow speeds, however it can form a good project for your final year. Basically try to keep the flow attached and do several designs and choose the most effecient one, something producing the optimum downforce to drag ratio.

[NickT]

Sorry I can't find any specifics at the moment, but I would imagine a higher angle of attack than those in use currently, simply because of the lower speed of the car. However I would strogly recommend you do some testing as you need to make sure you still produce good down force at the extreems of pitch and yaw. This was critical to the success of the early F1 cars as if the centre of pressure moves around too much or you loose a significant amount of downforce at the limit then the driver will never have confidence in the car.

Check out the back issues from http://www.racetechmag.com/ they had a huge amount of coverage on Formula SAE

[flynfrog]

Sorry I can't find any specifics at the moment, but I would imagine a higher angle of attack than those in use currently, simply because of the lower speed of the car. However I would strogly recommend you do some testing as you need to make sure you still produce good down force at the extreems of pitch and yaw. This was critical to the success of the early F1 cars as if the centre of pressure moves around too much or you loose a significant amount of downforce at the limit then the driver will never have confidence in the car.

Check out the back issues from http://www.racetechmag.com/ they had a huge amount of coverage on Formula SAE

there will be plenty of testing i just need a good place to start the problem with a higer angle of attack is that you start to get lots of seperation and you are more likely to stall it out

that and i dont have clearence under the chassis

[kilcoo316]

I remember something about F3 cars having a diffuser angle of around 13 degrees... I think it was 13...

[flynfrog]

I remember something about F3 cars having a diffuser angle of around 13 degrees... I think it was 13...

my side tunnels are at 13 after checkin gwith a picture of a f3 car that seems about right

should i try a naca profile for the aspect ratio?

[Mikey_s]

FF,

just a couple of comments from a non-expert in FSAE, or aero, but I have a definite interest in this field. I had a quick look at the regs (but i wasn't going to read 124 pages!), so what I am about to say might be excluded somewhere. (I just did a quick search for undertray and diffuser)

From your picture you have a venturi at the front of the car and then the opposite shaped venturi as the diffuser. Without running CFD on the model I can imagine that you are going to get a high pressure area at the narrowest point of the "inlet" venturi which will work against you at the front end of the car. In a typical F1 vehicle the air is intriduced under the car by a splitter, a "sharp", straight line whch separates the flow in a nice clean fashion. Therafter the undertray slopes gently upwards thereby increasing the volume under the car which leads to a gradual negative pressure gradient. When you get to the back end of the car the diffuser then increases the volume dramatically, thereby leading to a large(r) pressure drop - providing it doesn't stall.

My 2 cents would be to suggest that you do away with the scooping venturi at the front (it's working against you), introduce a splitter, try and set the minimum ground clearance for the front end of the car and then slope the undertray towards the rear (are stepped undertrays legal?), retaining your diffuser at the back end. I did see in the rules that any "sharp" edge must have a radius of at least 3mm, so that you don't chop legs off if you hit someone , so it can't be too sharp

[luisandregg]

Maybe our friend Scarbs can help you with your splitter:
http://www.scarbsf1.com/Aero_frontal/index.html

Is that what you were talking about, Mickey?

[Mikey_s]

luisandregg,

yes, the splitter serves to a) limit the amount of air permitted under the car and then b) channel the air above the splitter out to the edge and (via the bargeboards) into the radiator pods.

Generation of d/f under the car is a function of limiting the mass of air allowed under the body and then increasing the volume (for the same mass of air) so that the pressure drops, thus generating a pressure differential between the top and bottom of the car = d/f.

Clearly air will try and sneak in from the sides to try and equalise the pressure difference, therefore in F1 they make the undertray as wide as possible and, in the old days, put skirts on the side to minimise ingress of air. This is also the reason to try and run the car as close to the ground as possible to minimise to possibility of air entering from undesired places. Additionally the cars nowadays have vortex generators which generate vortices to "seal" the edges... prolly a little advanced for FSAE cars.

You can generate much more downforce, much more quickly if you have a stepped floor, so the FIA banned these!

[Reca]

On the book “Dall’aerodinamica alla potenza in Formula 1” written by Enrico Benzing, there are examples of some curves optimised for diffuser, here the equations :

A1 : y(x) = .233 x + 1.084 10^-5 x^2 + 1.261 10^-5 x^3 – 2.735 10^-9 x^4
B1 : y(x) = .14 x + 1.282 10^-3 x^2 – 1.225 10^-5 x^3 + 1.401 10^-7 x^4
C1 : y(x) = 6.642 10^-2 x + 1.734 10^-3 x^2 – 2.15 10^-5 x^3 + 1.866 10^-7 x^4

A2 : y(x) = .131 x + 1.204 10^-3 x^2 - 1.725 10^-5 x^3 + 2.711 10^-7 x^4
B2 : y(x) = 5.203 10^-2 x + 2.51 10^-3 x^2 – 4.953 10^-5 x^3 + 4.725 10^-7 x^4
C2 : y(x) = 1.514 10^-2 x + 1.294 10^-3 x^2 – 2.544 10^-5 x^3 + 3.149 10^-7 x^4

All have a slope in the first part of about 5-7°. Then diffusers of the group 1 are more linear while group 2 have a more evident curvature. The letter identifies the height at the trailing edge in % of diffuser length, about 35% (A), 26-27% (B) and 20% (C).

Obviously for the inlet section you can be way more “brutal”, flow is accelerating there, no risk of separation.

Hope this helps as starting point.

From your picture you have a venturi at the front of the car and then the opposite shaped venturi as the diffuser. Without running CFD on the model I can imagine that you are going to get a high pressure area at the narrowest point of the "inlet" venturi which will work against you at the front end of the car. In a typical F1 vehicle the air is intriduced under the car by a splitter, a "sharp", straight line whch separates the flow in a nice clean fashion. Therafter the undertray slopes gently upwards thereby increasing the volume under the car which leads to a gradual negative pressure gradient. When you get to the back end of the car the diffuser then increases the volume dramatically, thereby leading to a large(r) pressure drop - providing it doesn't stall.

Mikey, you have it all backwards, aerodynamics doesn’t work based on P V = n R T hence = constant for a given temp. We are not talking about a closed environment so air can’t enter or get out hence the given number of gas molecules will occupy the space so you have a variation of density (hence) pressure depending by the variation of volume.

In aerodynamics mass conservation means that the mass of fluid entering in a given duct (real as a tunnel or virtual as a streamtube) has to be equal to the mass of fluid getting out => to any section of the duct you have to apply the relationship rho * area * v = constant, where rho is the density, Area is the cross sectional area of the element and v is the speed of the airflow.
At low speed (under Mach = 0.3) the flow is incompressible meaning that density is constant so you have => Area * v = constant.
Then where velocity is high pressure is low and viceversa (in a very ideal case, when a quantity of hypothesis are valid, then you can apply the famous Bernouilli equation p + 0.5 rho v^2 = constant).

Consequently in the narrowest point at the front, since the area is lower, velocity will be higher and consequently pressure lower. In the diffuser then area is increasing so speed is decreasing hence you have a positive pressure gradient leading to the increment of pressure from the low pressure under the car to the high pressure present at the trailing edge.

Read my post in the following thread, I explained with more details the basics of how the underfloor works : viewtopic.php?t=3277&start=15

[flynfrog]

thanks reca that was the reply i was waiting for

ill give those foils a try

[Mikey_s]

Reca,

I said at the outset of my post I am not expert in aerodynamics, but I am a scientist and do not see anything in your post that would (in principle) disagree with what I said in mine.

Of course I recognise that this is not a closed system - far from it in fact. I should know better than to try and argue aero with an aero expert, but the basic principle of sloped undertrays and diffusers is to create a low pressure zone under the vehicle. Dismantling the process into basic principles one could sat that the air molecules do not have any knowledge of bernoulli, or newton, they simply react to the action of the other air molecules around them (that much is chemistry and I am a chemist!).

It is also obvious that the mass balance must be equal unless air is being removed from the system, since fan cars are banned the mass entering the vehicle (from any direction) must equal the mass leaving the vehicle (in any direction).

I don’t dare to suggest that this applies to your Reca (lol), but a common misconception it to imagine the process as a static car and a moving airstream (which is one way to look at it). In reality, of course, it is the car that moves over a “static” air molecule which is then manipulated to generate a pressure differential on the top and bottom of the car. This is effected by virtue of the motion of the car as it clearly does not function when the car is not moving (unless it is a windy day!).

My basic point was that when the splitter splits the air, a molecule under the car will react according to local pressure differences and it will not speed up, or slow down unless the pressure (read number of other air molecules around it) is different – there is nothing “active” about the undertray of the car – the pressure differential arises as a consequence of the motion of the car and the change in relative volumes as the car moves past a given part of the track – in other words when the car is static a diffuser doesn’t work and there is no flow under the car.

I am not sure that you can separate the air velocity and pressure to say that the pressure is lower because the velocity is higher – again, the air molecule simply reacts to the forces around it. The fact that the diffuser “drives” the flow under the car is a result of it creating a pressure gradient which can only be a result of the change in volume. The pressure gradient can be described in elementary terms as the number of air molecules “behind” any given air molecule compared to the number of air molecules “in front of it”. If that is the case the molecule will move backwards with a rate determined by the number of collisions stopping it (you could describe it as diffusion for example! maybe that's where the name diffuser comes from!!).

If the flow in the diffuser should become turbulent then the pressure gradient becomes non-linear and the system stops working efficiently. The key to having an efficient diffuser is thus to ensure that it can accommodate the mass entering from the undertray whilst still maintaining an efficient pressure gradient.

I must do some work, but stripped back to the molecular level I don’t see a lot of difference between what you described and what I did. Pressure differentials result from a different mass (read number of molecules per unit volume) of air above the car relative to below the car – there is no other explanation at the fundamental level. The difference in density arises from the motion of the car and the change in volume created by the shape of the undertray/diffuser – the velocity and absolute pressures of the airstream at various positions are resultant – they are not what drives the system, the energy to create these differences comes from the motion of the vehicle (i.e. the engine).

[kilcoo316]

Mikey, what your discussing with stationary air/moving object compared to moving air/stationary object is Skaides(?) & Blasius boundary layers - fundamentally there is no difference to the flow.


As for molecular interactions, in subsonic flow, the pressure differences/perturbations are transmitted instantaneously to all other points of the flow - so the air will experience a local change in velocity [from zero to say Y] - but to the car, the differential will change from X m/s to X+Y m/s.


The total pressure of the air is lower as its speed raises [at Mach < 0.3]: Pt = Ps + Pd

Pt = Total pressure
Ps = Static pressure [given by atmospheric conditions]
Pd = - 1/2 rho * V^2 [Bernoulli]


The pressure differences result from differences in flow velocity, which comes from differences in area for the flow to pass through & conservation of mass.

[wowf1]

To reinforce Reca's post, the air entering the underbody at the front of the car will accelerate due to the area reduction (low ride height). Thus a low pressure region will be observed on the plan form area underneath the car.

Due to the increasing cross-sectional area of the diffuser, the flow will be slowed back down, and pressure will increase. This is a relatively simple principle of fluid dynamics. I believe the reason behind increasing the pressure before exiting the car is to reduce the drag caused by separation.

[Reca]

thanks reca that was the reply i was waiting for

ill give those foils a try

You’re welcome, let us know about the results.

I said at the outset of my post I am not expert in aerodynamics, but I am a scientist and do not see anything in your post that would (in principle) disagree with what I said in mine.

As a start what you say leads to a completely wrong pressure distribution on the car’s underfloor.
You said that at the narrowest point of the “venturi inlet” there’s high pressure, while pressure there is low.
You said that in the diffuser there’s a pressure drop while in the diffuser pressure increases.

Hardly an explanation can be right if it leads to results clearly in contrast with reality.
And it doesn’t matter if you are reasoning with car’s moving and air stationary or with car’s stationary and air moving, the resulting pressure distribution on the underfloor is the same. (unless you believe that thousands people working 24/7 on wind tunnels all over the world are just wasting time)