Diffuser Confusion

[vonk]

Something didn’t feel right about what was said in the media and on forums about the F1 diffuser technology. So I spelled out the following for myself, in an attempt to understand what’s going on. Check me out on this. But please no Techno-Babble. [-o<

A diffuser converts flow from high speed at low static pressure (pin) to low speed at high static pressure (pout).



Cut in half, we have the diffuser under a race car.



The flow at the outlet is at atmospheric pressure, therefore pin must be lower.



But, the FIA regulations dictate a flat floor, parallel to the ground with no protrusions, allowing no venturi effect to be created under the car. So, even under ideal conditions the static pressure under the car cannot be lower than atmospheric. Forcing exhaust gas under the floor of the car would actually lift it.

Some tricks can be played with the shape of side pods to create local low pressure regions at the edge of the floor,



or even an airfoil shaped interface between the reference and step planes.



However, none of these would reduce the inlet pressure to the diffuser to an extent that justifies the steep upper surfaces of current diffusers.

Actually, I believe, the true benefit of the diffuser is offered by its outside surface, that imparts upward momentum to the back flow from the bodywork. And there’s a lot of that back flow, which was previously just dumped into the wake.



What to do with the turbulent mess at ambient wake pressure inside the diffuser? Attempts have been made to clean it up by ducting available side flow into it. Adrian Newey must have been on vacation when Red Bull apparently tried exhaust jet-induction to increase flow through the diffuser. Of course, none of this can be “diffused”, as it all enters at atmospheric pressure. It only adds skin friction drag inside the diffuser, made worse by multiple guide fins.

I think I now understand why people are going to front exhausts. In the rear, exhaust exits as a pulsing hot jet, wasting all of its energy. If released in front, properly deflected, its energy could be made to mix exhaust with free stream in such a manner as to create a vortex that captures extra air and directs it onto the top surface of the diffuser, where it can be deflected upward to create extra down force. Not an easy task, with varying speed and throttle. (Hopefully, it won’t hit the face of the rear tire.)





Is Adrian Newey back from vacation?

[hardingfv32]

",,,even under ideal conditions the static pressure under the car cannot be lower than atmospheric"

I think this premiss is wrong.

http://www.racecar-engineering.com/tech ... odynamics/

Brian

[DaveW]

Your seductive analysis ignores the boundary layer, Vonk, which I think might be quite important in ground effect flow fields. Certainly, boundary layer "bleed" & a moving ground plane are considered to be essential elements of a good vehicle wind tunnel.

[vonk]

Your seductive analysis ignores the boundary layer, Vonk, which I think might be quite important in ground effect flow fields. Certainly, boundary layer "bleed" & a moving ground plane are considered to be essential elements of a good vehicle wind tunnel.

Funny you should mention that, Dave. Indeed, the boundary layer is important and it reinforces my argument. Here’s my take:

Imagine you’re riding in the car observing, by some magic means, what’s going on in the gap between the undertray and the pavement.


Figure 1. Vertical Section @ Centerline

You’d see air and pavement rushing toward you, entering the gap under the horizontal splitter. Because air is a viscous fluid, obeying the No-Slip Condition of fluid dynamics, a velocity gradient develops (2) and the relative motion between the undertray and pavement produces a shear drag between them. This, in turn, results in the development of a boundary layer on each (3) – first laminar, then turbulent. If the gap is narrow enough, the boundary layers will touch, causing vorticity across the gap (4). Also, the aforementioned shear drag, multiplied by the car’s speed, spells power transferred into the turbulent flow as heat.

In this turbulent mess, what happens to the static pressure (p0) in the gap? While Bernoulli’s equation

is actually only valid for non-viscous, laminar flow, it’s an acceptable approximation for the relationship between the static and dynamic pressure components of the flow at the entrance to the gap. The dynamic component,

represents the kinetic energy of the inflowing air, but in the gap, viscosity soon dissipates much of that kinetic energy into turbulence and heat. If the gap were two-dimensional, this would cause a constipation that raises the flow’s temperature and static pressure. Fortunately, the gap is not two-dimensional. It has a finite width. Where it ends laterally, it is exposed to free stream static pressure which communicates into the entire gap, causing some of the flow to leave the gap laterally.

So, now we have warmed, turbulent air entering the diffuser (5) at ambient static pressure and about half the free stream speed. What can the diffuser do but fill with turbulent flow at ambient pressure?

[Just_a_fan]

But, the FIA regulations dictate a flat floor

Wrong. The rules don't require the floor to be parallel to the ground. The floor must be coincidental with the step plane (or the reference plane as appropriate). The rules do not define the reference plane relative to the ground.

That's why the cars run with nose down rake.

[vonk]

But, the FIA regulations dictate a flat floor

Wrong. The rules don't require the floor to be parallel to the ground. The floor must be coincidental with the step plane (or the reference plane as appropriate). The rules do not define the reference plane relative to the ground.

That's why the cars run with nose down rake.

I stand corrected about the under-tray having to be parallel to the ground. But I believe the rake is to compensate for rear suspension deflection under heavy wing load. It would have little influence on under-car pressure anyway, because that space is laterally vented to the free stream.

[Just_a_fan]

But, the FIA regulations dictate a flat floor

Wrong. The rules don't require the floor to be parallel to the ground. The floor must be coincidental with the step plane (or the reference plane as appropriate). The rules do not define the reference plane relative to the ground.

That's why the cars run with nose down rake.

I stand corrected about the under-tray having to be parallel to the ground. But I believe the rake is to compensate for rear suspension deflection under heavy wing load. It would have little influence on under-car pressure anyway, because that space is laterally vented to the free stream.

But as the depth of the boundary layer increases below the car, so the ride height increases. One will help to cancel the effect of the other.

[vonk]

........ But I believe the rake is to compensate for rear suspension deflection under heavy wing load. It would have little influence on under-car pressure anyway, because that space is laterally vented to the free stream.

But as the depth of the boundary layer increases below the car, so the ride height increases. One will help to cancel the effect of the other.

Why would the ride height increase?

[horse]

I think the fallacy here is to think of the diffuser as an isolated part of the car. The whole car is an aerodynamic surface (not just the floor) that induces a circulation about itself that then increases the speed of flow beneath the car and therefore increases the dynamic pressure which reduces the static pressure.

The diffusers role in the generation of this circulation is to provide a trailing edge which will provide the upwash to set the system up.

[Just_a_fan]

........ But I believe the rake is to compensate for rear suspension deflection under heavy wing load. It would have little influence on under-car pressure anyway, because that space is laterally vented to the free stream.

But as the depth of the boundary layer increases below the car, so the ride height increases. One will help to cancel the effect of the other.

Why would the ride height increase?

Because that's what rake does - it changes the ride height from front to rear. The rear of the floor is further from the ground than the front of the floor.

There's a whole thread elsewhere about this (in the flexi wing topic I think)

[hardingfv32]

Why is it you guys never stay focused and come up with some clear conclusions? How do you ever learn anything.

The original premise stated in the first post is wrong. Doesn't anyone concur with that?

Here are some studies, while not directly on point, will help. I believe they both pertain to bluffed bodies, so flow over the actual top of the diffuser is not possible, thus voiding that theme as the main goal.

1) Experimental study of multiple-channel automotive underbody diffusers

http://pid.sagepub.com/content/224/7/865.full.pdf+html

2) Aerodynamic 2D Investigation of Race Car Diffuser

http://www.energy.aau.dk/GetAsset.actio ... Id=4620430

3) This study discusses ground clearance issue in the second part of the study.

Ground Effect Aerodynamics of Race Cars

http://eprints.soton.ac.uk/42969/1/GetPDFServlet.pdf

4) For you CFD admirers:

A CFD Investigation into Ground Eect Aerodynamics

http://www.lr.tudelft.nl/fileadmin/Facu ... aGenua.pdf


Brian

[horse]

The original premise stated in the first thread is wrong. Doesn't anyone concur with that?

This?

Actually, I believe, the true benefit of the diffuser is offered by its outside surface, that imparts upward momentum to the back flow from the bodywork.

Yes, this is totally wrong. I was trying to say that the diffuser operates as part of the car as a whole, not as an isolated object. It's not a pipe.

[hardingfv32]

1) This is the statement i am referring to: ",,,even under ideal conditions the static pressure under the car cannot be lower than atmospheric"

2) Fine... "the diffuser operates as part of the car as a whole...."

Then I propose a precise question to get back to the diffuser itself: What performance or properties does the diffuser bring to the total aero system? We need to understand the parts before we can integrate them into a larger system.

Brian

[vonk]

",,,even under ideal conditions the static pressure under the car cannot be lower than atmospheric"

I think this premiss is wrong.

http://www.racecar-engineering.com/tech ... odynamics/

Brian

Brian, my preceding posts explain how I arrived at that premise. The linked article is exactly what caused me to look into this. It just restates, without proof, many of the popular assumptions and believes that bother me. The existence of high speed, low pressure flow at the diffuser inlet is simply accepted, as is diffuser suction and on and on….

Why is it you guys never stay focused and come up with some clear conclusions? How do you ever learn anything.

The original premise stated in the first post is wrong. Doesn't anyone concur with that?

Here are some studies, while not directly on point, will help. I believe they both pertain to bluffed bodies, so flow over the actual top of the diffuser is not possible, thus voiding that theme as the main goal.

1) Experimental study of multiple-channel automotive underbody diffusers

http://pid.sagepub.com/content/224/7/865.full.pdf+html

2) Aerodynamic 2D Investigation of Race Car Diffuser

http://www.energy.aau.dk/GetAsset.actio ... Id=4620430

3) This study discusses ground clearance issue in the second part of the study.

Ground Effect Aerodynamics of Race Cars

http://eprints.soton.ac.uk/42969/1/GetPDFServlet.pdf

4) For you CFD admirers:

A CFD Investigation into Ground Eect Aerodynamics

http://www.lr.tudelft.nl/fileadmin/Facu ... aGenua.pdf


Brian

I scanned the linked reports. While interesting, they hardly simulate the real on track conditions of a full scale race car (that is not a closed system) where everything is open to the intrusion of free stream static pressure. BTW, I would have loved to be on the murder board reviewing the second study presentation.

[Gatecrasher]

The equations and talk of boundary layers and all that jazz are interesting but fundamentally if you take a fixed mass of air and increase the volume available, as the airflows, the pressure will drop thus sucking the car to the pavement. The bigger the increase in volume the lower the pressure. The higher the airflow then the less negative effect of atmospheric pressure "leaking" back into the rear of the diffuser, thus the added benefit for blown diffusers.