A very interesting theory about drag
 

https://www.youtube.com/watch?v=Z3Gn2HGwYuE

This video came about after I sent a request to several top car aerodynamicists (ones I'd came into contact with through working on my book) about the relationship between pressures measured on the side of the car, and the base (ie wake) pressure.

My measurements show that pressures at the rear edges of cars don't always match wake pressures (although the match seems pretty good on squareback shapes). So what actually is the connection between (say) side pressures and base pressures? If we understand that more clearly, we (as amateurs) can then develop approaches to reduce base pressure and so drag.

Dr Adrian Gaylard (Jaguar Land Rover) came back with an intriguing response that I'd never thought of before. He referenced Hoerner's classic book on aero drag, linked to a more modern paper and put that all in the context of his own understandings.

This video looks at that particular theory - I may later do others on what the other aerodynamicists told me.

My gut feeling is that this theory helps explain some aspects of base pressure - as all the aerodynamicists told me, there is a lot going on, and so one theory is unlikely to explain everything!

It does, however, make sense in the context of the measurements I have made in the last month on cars as diverse as a Skoda Roomster squareback, W212 Mercedes notchback, Gen I Honda Insight fastback and Tesla Model 3 fastback.

Thick boundary layer -> less base drag (via less jet pump affect)

Can the affect also be measured as a pressure increase on the back of the car? Did you have sensors on the back of the Tesla?

And related... Same idea apply to vortex generators? That is, they upset/disturb the jet pump action?

No I didn't think until afterwards that I should also have measured base pressure on the Tesla. But I can easily disturb side flow on another car and see if side pressures decrease and base pressures decrease.

Despite the claim often being made, I've never seen any evidence that vortex generators work in altering base pressure. They certainly can work in putting energy back into the boundary layer and causing better flow attachment around corners, but that's a completely different thing.

Thank you for posting the video.

Is this thicker boundary layer air responsible for mitigating or jumping over the different angle between rear window glass and trunk on a saloon/sedan?

This intersection is usually accused of interrupting attached flow, however in wind tunnels with smoke the flow looks smooth, angle differences seem to disappear - especially with increased distance from surface.

Where are the greatest pressure differences measured at the rear of the car?

1. Between top and bottom

2. Between top and side

3. Between side and bottom

Only 4-planes of interaction on a rectangular section car/truck, correct?

I suspect the "Jet Pump" phenomena is greatest under the car where it is squeezed and partially contained by ground plane.

I have not read the PDF in the link below, but it could be related to this topic.

Review on Aerodynamic Drag Reduction of Vehicles
https://www.researchgate.net/figure/...fig1_330661575
http://forums.pelicanparts.com/uploa...1609857516.jpg

http://forums.pelicanparts.com/uploa...1609857516.jpg

I also found this paper regarding side spoilers. Looks like they are able to CFD pressures off the surface, and these pressure zones have a shape to them.

The boundary layer(s) are also illustrated in a way I have not seen before.


Influence of Side Spoilers on the Aerodynamic Properties of a Sports Car
https://www.mdpi.com/1996-1073/12/24/4697/htm
http://forums.pelicanparts.com/uploa...1609858812.jpg
http://forums.pelicanparts.com/uploa...1609858812.jpg

The first paper just looks like a grab-bag of everything they could find research on - ie a non analytical literature search.

The second paper, though, is quite fascinating - active spoilers located on the front bumper side and A pillar. But neither was good for drag, and their argument that the spoilers could help control the car in cross-winds overlooks the fact that the spoilers would have to be able to move extremely fast. (In the paper they get over that problem by talking about a constant crosswind - but that's not very real world.)

Re the boundary layer and:

Is this thicker boundary layer air responsible for mitigating or jumping over the different angle between rear window glass and trunk on a saloon/sedan?

No, in fact the thicker the boundary layer, the less likely air is to follow changes in shape. That's why vortex generators, that put moving flow energy back into the boundary layer, can cause better flow attachment around corners.

Re:

Where are the greatest pressure differences measured at the rear of the car?

There is almost no tech lit available on this, and I haven't done that series of measurements. My guess is that the answer would be heavily dependent on the shape and underside smoothness of the car.

My disappointment was that the digram you showed and ask us to imagine 3D showed the flow over and under the car but it was all about side pressures.

There is a silent minority at Ecomodder who hold that the air is standing still and it's the vehicle that is moving. ;)

Which is to my point. If you have contact with Dr. A. Gaylard ask him this: All talk about boundary layer references free air. The underbody is a plenum with four open sides and two [relatively] moving walls. I know that when testing motor oil (analogizing from engine theory, I know) what they use is two concentric cylinders with an oil film between. The shear forces can tear oil molecules apart.


So the condition in the underbody should be shear forces instead of boundary layer. Or, I'm wrong. Whichever.

I have an idea for easy 3D diagrams, maybe I can prepare an example.

Because the pressures above and below the car depend on a lot of other factors besides boundary layer thickness, and so to introduce top / bottom would have caused confusion in terms of the overall point being made.

As I said in the video, I may cover the other theories of the influences on base pressure in further videos (and they would include, for example, trailing edge pressures on the upper surface that would in turn involve body shape, vortices, etc).

You are implying that the boundary layer under the car extends between the underbody and the road. (That is, it's the shear forces that are most germane.) This is not the case in any modern car that has a smooth underfloor. It may, however, be the case with old cars with rough underbodies that have a very low ride height.

I have bought multiple pitot tubes so that I can directly measure boundary layer thickness, but I doubt I'll be doing that under the car (too easy to lose them over a bump).

Okay, but I wasn't thinking about a rough underbody. Assume the underbody is as rough as the pavement.

There will be a gradient away from the underbody and the road, with some amount of turbulent mixing at the center. IIRC test show an optimal ride height [for a given design]. Normally assumed to be exposure of the tire faces?

I'm working on that graphic demo. Something presentable seems achievable, else it's not the good idea I'd thought.

Ablative extensions to the pitot tube?

Assume the boundary layer against the smooth car underside growing to, what, 30mm? I don't think that 'boundary layer' in this context makes sense for the ground (because the ground boundary layer might be a few metres thick, and the ambient air is ostensibly still ie not moving like the air past the car). That would seem to me to give about 70mm (typical car with 100mm ground clearance) of 'free air'.

But you do make a very interesting point. I've not seen, that I can think of, the velocity gradient plotted under the car on a vertical axis. Does anyone know of such data?

:)

I watched a car run over a cat once. It bounced up and down twice. (and then got up and walked away)

One could attach a flexible forward-facing tube to a trailing skid to get close to the ground plane.

I'm no help otherwise. :)

Excellent points.

Regarding moving air Vs moving object, I've tried to discipline myself several times on this topic but continue to fall short. I can get half way there if I try, but only for short periods of time.

1. When imagining pressures at the front of the car I can successfully imagine the air being still and the car slamming into the air molecules.

2. When imagining drag pulling back on the moving body/car parasitic, vortexes, and low pressure described in opening post of this thread.

All other times I revert into thinking the air is moving, and my obsession with wind tunnel images with smoke doesn't help me there. :o

Regarding air pressures under the car, just use them so they aid handling at speed. Accept that it will be there, and manage it.


Car Aerodynamics Basics and How-To Design Tips cont…
https://www.buildyourownracecar.com/...-and-design/4/
https://www.buildyourownracecar.com/...turiTunnel.png
The pressure differentials at stern of moving vehicle result in unwanted turbulence and vortex formation.

I think information is out there, we just gotta remember where we have posted it before. :cool:

drag
 

Hucho offered :
1) The pressure drag ( Dp ) = the integration of force components in the flow direction, resulting from the pressure distribution.
2) The pressure drag would include vortex drag and rearward suction.
3) The wake pressure would be impacted by local streamline pressures at the separation point ( line ) at top, sides, underbody, wheels, vortices, downwash, turbulence, and the geometry and surface angles at separation locations.
4) Notchback wakes would be the most challenging, as these cars have the most complex shapes.

data?
 

I looked, and for the most part, it appears that the road surface is basically treated as if it has zero boundary layer, only the bottom of the vehicle, and then only if it has a smooth underside.
Wind tunnels will suction off any boundary layer reaching a test vehicle, or lift the vehicle up even with the top of it.
' lifting of the vehicle by the amount of the displacement thickness of the ground boundary layer ( in the wind tunnel ) has no effect upon the flow around the vehicle and the effective forces.' Hucho, page-412.

Makes sense from the 'car moves, not the air' perspective.

The device I cited in #7 is called a en.wikipedia.org/wiki/Rheometer.

Rheo
 

Seems an appropriate name.
GM's current 'magnetorheological' shock absorbers fit that nomenclature perfectly.:thumbup:

That's actually not a good way of thinking about it, because quite a few front-facing surfaces develop thrust (eg leading edge of bonnet/hood, top of windscreen). Instead it's much better to think of what speed the air is going at. The slower it is going, the higher the pressure; the faster it is going, the lower the pressure.

Yes, but don't forget that attached flow on the rear parts of the car can cause drag as well.

Only these two schematics:
https://ecomodder.com/forum/member-v...106-171821.jpg
(Hucho, 4th ed.)

https://ecomodder.com/forum/member-v...106-171846.jpg
(Scibor-Rylski/Sykes, 2nd ed.)