Random Wind Tunnel and Smoke Pictures Thread

[freebeard]

I won't be reading it (there's firewood to be stacked) so maybe you can tell us: How do they accommodate the difference in Reynolds number.

[kach22i]

2014 Chevrolet Camaro Z/28 is Designed for Downforce
2014 Chevrolet Camaro Z/28 is Designed for Downforce

Racing-inspired design elements enhance high-speed stability for faster laps

[aerohead]

Originally Posted by freebeard

This is why I'm intrigued by OpenVDB

OpenVDB

It's open source, being folded into Blender, and you might say it holds a fractal array of voxels (technically a fast shallow and sparsely populated binary tree). It's currently used by, e.g., Dreamworks SKG to make fire, smoke and water in Hollywood movies.

The neat thing is each voxel can have have non-spatial attributes like color, opacity, velocity or pressure. I'm sure what we want could be built on Blenders animation engine.

Anything useful to us would have to rely on the full Navier-Stokes Equation.It uses a spherical vector coordinate system and each cell's performance is dependent upon the pressure and velocity of all cells surrounding it.
It requires a supercomputer to run it and takes almost 2-days to solve a single iteration.
No desktop computer can do this yet.And if you call the companies who sell it,you better be able to prove that your a deep-pocket corporation or well-granted university before they'll even talk to you about their products.

[aerohead]

Originally Posted by kach22i

I'm assuming the laminar airflow over the top of the P51 wing would have quite the velocity, in a dive at least.

Notice the tiny turntable built into the wing.

Not sure if this disk would be adjustable in flight, or even how testing data would be accumulate.

Guess I have to read the paper.

http://history.nasa.gov/SP-4302/ch2.3.htm





Reminds me of testing shapes/bodies on the roof of cars.

They'd need some sort of 3-axis mechanical load cell inside the wind to measure pitch, yaw and roll,or detect flutter.
Since very sharp leading edges and the 'all-moving-tail' was the SOLUTION for supersonic flight,NACA may have just been concentrating on transonic shockwave/variable center-of-pressure related control surface issues.
Many test pilots died in near-supersonic dives when they lost all pitch control as shockwaves moved the wing's center -of-pressure and blocked airflow over control surfaces.They quite literally could not pull out of the dive.

[freebeard]

It requires a supercomputer to run it and takes almost 2-days to solve a single iteration. No desktop computer can do this yet.

What I'm saying is, that was then, this is now.

"It's open source, being folded into Blender, and you might say it... blah, blah blah."

The company that made it (Dreamworks SKG) doesn't [only] sell it, they Open Sourced it. Current versions of Blender — a free and Open Source program — are beginning to implement it.

...the full Navier-Stokes Equation.It uses a spherical vector coordinate system and each cell's performance is dependent upon the pressure and velocity of all cells surrounding it.

OpenVDB itself is a bucket that hold the data, a hypothetical CFD program would 'peek and poke' it. But because of the way the data is structured nearest-neighbor calculations are simplified and implementing Navier-Stokes can be done in any language, probably Python. There are nifty features that reduce the computational overhead. Calculating actual XYZ coordinates is the last step!

It's also over my head. Someone else is going to have to extend Blender's animation engine.

[aerohead]

Originally Posted by freebeard

What I'm saying is, that was then, this is now.

"It's open source, being folded into Blender, and you might say it... blah, blah blah."

The company that made it (Dreamworks SKG) doesn't [only] sell it, they Open Sourced it. Current versions of Blender — a free and Open Source program — are beginning to implement it.



OpenVDB itself is a bucket that hold the data, a hypothetical CFD program would 'peek and poke' it. But because of the way the data is structured nearest-neighbor calculations are simplified and implementing Navier-Stokes can be done in any language, probably Python. There are nifty features that reduce the computational overhead. Calculating actual XYZ coordinates is the last step!

It's also over my head. Someone else is going to have to extend Blender's animation engine.

I'll have to catch up to this at some point I guess.
For now,all I know, Is that CFD capable of solving bluff-body,3-D flow,in ground proximity,and providing results equal in accuracy to full-scale wind tunnel testing,can only be afforded by the likes of a multi-national corporation today.And the conditions I explained with respect to the CFD generated output are current within a couple of months.

[freebeard]

https://www.google.com/search?q=Blender+CFD

It will take a while for uptake in Blender and then people finding use for it. Someday we'll look back at OpenFOAM and laugh.

[kach22i]

I did not want to start a new thread just to post an image of this aerodynamic oddity. Therefore I'm just posting it here in this thread.

Sure looks like something which would need to be tested in a wind tunnel.

I know we have discussed this concept before, but have been unable to retrieve that old thread.

Air Tube - for cooling (Pre-A/C).
2016 Orphan Car Show by George Kachadoorian | Photobucket

[freebeard]

Swamp cooler. It's a standard accessory at VW shows.

It uses ram air to spin a fiber cylinder to pick up water from the bottom, then the air flows through the damp fibers for evaporative cooling. The better ones have an electric fan booster so you don't become dependent on airflow around the A-pillar.

They don't work well on older VWs without flow-through ventilation or pop-out back windows. The only place for the air to exit is the driver's door window. Here's one with plenty of flow-through.

[kach22i]

Reduzierung des Luftwiderstandes - volle Wirkung erst mit regenerativem Bremsen




Link first found in the forum in this thread:
http://ecomodder.com/forum/showthrea...n-27186-3.html

Aerodynamik mit Wolf-Heinrich Hucho


http://aerowolf.com/