Aeromods must transfer force to be effective

[Blacktree]

Originally Posted by Talos Woten

Apparently triangles are the simplest building block that keep their shape under a variety of loads from different angles.

To put it simply, triangles are the strongest shape. If you can take a shape and turn it into a collection of interconnected triangles, it will be significantly stronger.

[aerohead]

I spent some hours considering geodesic construction that I thought I'd share.
Only machines can define them
Specific surface feature identification and address location would be extremely complicated among humans
Only robots would be able to easily navigate the 'language' of the 'fractal' architecture
Only AUTOCAD could draw it
Only CAD-CAM could create tooling for it
It couldn't be 'stamped'
Glazings could not be installed, windshield, door glass, fixed quarter-windows, or backlight
I'm unsure how a body shop could repair it
I'm unsure how any cut-opening ( bonnet, door, boot ) could open and close without binding.
Even robotic painting would be problematic
Machine sanding would be impossible.
It couldn't be built from a mold.
Composite fabrics would not drape over it
Tooling would have to be in split sections, clamped together for layup, leaving 'flash' at all separation lines ( difficult to remove for priming and topcoats )
--------------------------------------------------------------------------------------
Aerodynamically, we might be able to get way with murder in the forebody, but once past the roof apex, the angled prismatic nature of the 'panel', at non-cyclic intervals, could introduce such a transverse flow vector contamination to the boundary layer, as to introduce such an adverse pressure gradient, that it would be impossible to maintain attached flow on top or sides.
This is already an established problem with 'smooth' compound-curvature panel airship bodies in a crosswind according to NACA ( NASA ).
Just some thoughts.

[freebeard]

Just some thoughts.

I appreciate them. Even the wrong ones.

but once past the roof apex, the angled prismatic nature of the 'panel', at non-cyclic intervals, could introduce such a transverse flow vector contamination to the boundary layer, as to introduce such an adverse pressure gradient, that it would be impossible to maintain attached flow on top or sides.

This is the basic question I would like answered. Vekke was no help. Dimples work, why not raked edges (cough! Cybertruck cough!)

Free and Open Source software is wonderful. Since 2.80 or so, Blender has had a Geodesic add-on. Here is the contextual menu.


https://i.stack.imgur.com/itv2a.jpg

Review the Object Types and Object Parameters. All the Platonic shapes for primitives; I'm not sure the difference between Squish and Eccentricity. You can add a Superellipse in two dimensions or all three.

Autodesk's parametric NURBs modelling is in contention with Blender's Geometry Node Editor.

[Talos Woten]

Originally Posted by freebeard

This is the basic question I would like answered. Vekke was no help. Dimples work, why not raked edges (cough! Cybertruck cough!)

This is an answerable question. It depends on the Reynolds number regime we are in and the size of the boundary layer. At 65 mph, the boundary layer near the front of the vehicle is around 1/4" and near the back is 1"+ (depending on location and how good the aero is). The general rule of thumb is that for subsonic, subcritical flow any imperfection less than half the boundary layer isn't significant, and that flow will reattach right over it. That's why panel gaps of 1/8" are fine for passenger cars, but high speed racers actually care about them.

As long as the angle of curvature of the geodesic is under the constraints above, flow will have no problems remaining attached. At slightly greater angles it might reattach later on the surface, the same way flow reattaches to the side of a car after being disrupted by the front wheels / mirrors. At larger angles, there is too much shadowing and the flow doesn't to reattach. At that point, being a geodesic / sharp shape has nothing to do with things; the airflow is being deflected too much.

Anyhoo, there's already been CFD done on the Cybertruck. It's decent. Crap compared to fully aero vehicles, but great comparted to regular trucks. Because the general "silhouette" is vaguely aerofoil-like, the angularness doesn't count against it too much.

[aerohead]

Originally Posted by freebeard

I appreciate them. Even the wrong ones.



This is the basic question I would like answered. Vekke was no help. Dimples work, why not raked edges (cough! Cybertruck cough!)

Free and Open Source software is wonderful. Since 2.80 or so, Blender has had a Geodesic add-on. Here is the contextual menu.


https://i.stack.imgur.com/itv2a.jpg

Review the Object Types and Object Parameters. All the Platonic shapes for primitives; I'm not sure the difference between Squish and Eccentricity. You can add a Superellipse in two dimensions or all three.

Autodesk's parametric NURBs modelling is in contention with Blender's Geometry Node Editor.

They don't work! Not on a full-scale automobile.
The only reason they're on a golf ball is because of Reynolds number. Even at 110-mph, coming off the clubhead of a driver or iron, the ball is too 'short' to achieve supercritical Rn.
Artificial surface roughness hastens the transition from a laminar boundary layer, to a turbulent boundary layer, the best thing for flow attachment and drag reduction.
The reason the golf ball has 'dimples' is that they don't stick 'out' and can't get knocked off. Sand glued to the ball would do just as well, but it wouldn't survive the rigors of of the game.
The stunt that Mythbusters pulled had to do with the crappy notchback design of the Ford. The only dimples that did any good to the Ford were the ones on the rear of the roof, preceding the top of the backlight, acting as vortex-generators, and helping with flow reattachment onto the trunklid / boot.
You cannot take the longitudinal flow on the aft-body of a vehicle and shove it sideways. You're creating the kind of pressure kink Hucho insists must be completely avoided. It's part of the fluid mechanics ground rules he hoped the readers would get from his 2nd Edition. It's 33.3% of what he hoped to get across.
I don't know anything about other editions.
------------------------------------------------------------------------------------
Please expand on what you mean by 'raked edges'. I'm not clear on what you're trying to convey.

[freebeard]

Thanks Talos Woten.

I'll contemplate the 1/4-1" range vs dihedral anglesI notice for a prolated shape, the more severe angles are at the front end. The rear would be inverted for a box cavity.

What do you think of this one. I made it for another thread in 2015 -- a bellmouth difusser.



One big vortex generator to order the wake.

edit:

You cannot take the longitudinal flow on the aft-body of a vehicle and shove it sideways.

I didn't say you can shove it.

[aerohead]

Originally Posted by freebeard

Thanks Talos Woten.

I'll contemplate the 1/4-1" range vs dihedral anglesI notice for a prolated shape, the more severe angles are at the front end. The rear would be inverted for a box cavity.

What do you think of this one. I made it for another thread in 2015 -- a bellmouth difusser.



One big vortex generator to order the wake.

edit:


I didn't say you can shove it.

It's the angled triangular facets which would impel the flow sideways off their optimum path.
The sharp angles around the corners are also 'spikes'.
They're also deflecting the flow 'away' from where the local streamline would 'go' if on a streamlined contour.
-------------------------------------------------------------------------------------
If you can imagine an air molecule above the body, as a roller lifter, the atmospheric pressure as the valve-spring tension, and the body as a camshaft lobe passing under the roller lifter; there is a limit to the 'contour' of the camshaft lobe, after which the spring tension is insufficient to hold the roller follower in contact against the cam. The roller 'floats'. It 'separates.'
--------------------------------------------------------------------------------------
The lowest pressure on the car is just ahead of the windshield header.
All the air would like to go there.
That's okay in the 'front' of the car.
It's the worst thing for the back of the car.
--------------------------------------------------------------------------------------
Beyond the roof apex, even on a perfectly streamlined body, the air is moving towards a higher pressure than where it's at, at the rear of the body.
The only thing that keeps the air moving rearwards is, shearing forces from the local streamline, just above the boundary layer 'strafing' kinetic energy into the boundary layer, 'pinning it down', like a machine gunner, against the body's surface.
If the body surface falls away too radically, the shearing force from above just richochets off the top of the boundary layer, without imparting any energy.
When this happens, the boundary shrugs its shoulders and says ---- it! It always wanted to go where the windshield was, and that's exactly what it attempts to do, rolling up into eddies as it tries to climb backwards up the back of the car, against the flow, then completely lifting off the surface as the local streamline attacks it, blooming into full turbulence.
The energy balance necessary to keep the flow attached is a very delicate affair.
'(C)hange in a pressure distribution is highly significant for the origin of drag.'
Hucho, 2nd-Ed, page-117.
' The pressure drag is explained by the deviations of the pressure distribution in actual flow.' Hucho, page-124.
' (I)t is very important to design a rear body surface which brings the divided streamlines smoothly together .' Hucho, page 61.
'(P)ressure drag is the largest component in the aerodynamic drag. Its minimization is the true objective of motor vehicle aerodynamics.' Hucho, p.119

[Talos Woten]

Originally Posted by freebeard

What do you think of this one. I made it for another thread in 2015 -- a bellmouth difusser.

One big vortex generator to order the wake.

An interesting idea. If you are heating the air in the center, this is effectively a turbine that produces thrust!

By the way, generally speaking, one wants to avoid sharp / cleaving edges at the front of a vehicle. The reason is because air rarely is coming on exactly straight along the main axis. Even modest crosswinds (10 mph) at highway speeds (65 mph) create angles that sharp edges shadow... and thus produce lots of counterproductive turbulence. If we look at real turbines:
https://storage.googleapis.com/mcp_2...-1920x1000.jpg
we can see that the leading rim is always a rounded shape. The reason why rounded fronts/noses are universal is because they still function well even when airflow is off angle to their primary axis.

The other reason why we want round noses instead of sharp edges is they also produce the lowest drag, when integrated by off angle. A way to produce low drag is to have the smallest area being "pushed back" against airflow. In a curved nose, that's basically a knife edge for an aerofoil and a tip for missile. If air comes off angle, the high pressure area is still small. But in a sharp / angular / flat surface, when air comes off angle, it often exposes some surface * some trig function that produces appreciable counterforce.

That's actually the biggest mistake I made when I first started my mods. I only thought in the two dimensions of the car silhouette, instead of the full 3D, let alone considering off axis flow. As Spock would say: "He is intelligent, but not experienced. His pattern indicates two-dimensional thinking."

As for the VGs, they produce turbulent flow and should be used with caution. The only use case where they are beneficial is when they reduce or prevent even worse turbulent flow. Their best use is to try to reattach separated flow earlier to a surface than it would otherwise. If we introduce a VG into a laminar situation, it actually hurts performance. So, if the backside of your toroidal turbine already has a smooth tail with minimal wake, it's uncertain injecting turbulent air into it would actually help.

[freebeard]

The other reason why we want round noses instead of sharp edges is they also produce the lowest drag, when integrated by off angle.

No need to sell me on round noses, I own a Superbeetle.

[Blacktree]

Originally Posted by Talos Woten

An interesting idea. If you are heating the air in the center, this is effectively a turbine that produces thrust!

A little off topic, but this is something I want to do with my project car. The idea is to make a rear diffuser, and route the engine exhaust into the front opening. For that car, fuel economy isn't really a concern. But less lift and/or drag would make it faster on the racetrack.