Flow Illustrator - a simple windtunnel simulator

[Chris D.]

Originally Posted by Peakster

You know I was just curious :

hmmm makes me think about trying this..

[twolostminds]

This is some cool software!

YouTube - 2009 FEH Test 01

Uploaded a 2009 Ford Escape Hybrid to my youtube.

[azraelswrd]

Wow, this program is nice. I knew my car had bad drag but wow... was it really this bad? Yikes!

I'll have to look into that floating parallel spoiler idea.

[texanidiot25]

Originally Posted by i_am_socket

I'm getting some interesting movies using the high Re numbers.

I get a lot of this:


using 2,500,000. Is it supposed to look like that?

Lower the Dt number to around .01, it seems to spaz out with higher Dt numbers. Real psychodelic stuff.

[adrive7]

Originally Posted by i_am_socket

I'm getting some interesting movies using the high Re numbers.

I get a lot of this:


using 2,500,000. Is it supposed to look like that?

I would try adding white area all around the vehicle. Also, get rid of the tires so you can see the air flow under the car. If you increased the time step (dt) take it back down to around .01. It looks like you are seeing effects from the edges of the pictures, but I don't know why it would do that. Adding white area around it should get rid of that problem (if thats what it is)

The Reynolds number is fine.

[cfg83]

adrive7 -

Originally Posted by adrive7

I would try adding white area all around the vehicle. Also, get rid of the tires so you can see the air flow under the car. If you increased the time step (dt) take it back down to around .01. It looks like you are seeing effects from the edges of the pictures, but I don't know why it would do that. Adding white area around it should get rid of that problem (if thats what it is)

The Reynolds number is fine.

I agree. In my runs, I added a "road" below my "hover-car" and I put some empty space behind the car to analyze the wake that my mods created.

If you put the tires on, I think it is like having "Flintstone Roller" tires, because you are cutting off airflow below the vehicle. I *think* if you want to approximate tires, you could add shallow "bumps" where the tires are. Maybe the compromise is to approximate how much "area" the tires displace below the car and make the "bump" proportional to that. Does that make sense?

I used an Re of 2,500,000 in my second analysis, and I set it to 40 seconds. I also used a 2-bit Bitmap.

CarloSW2

[wagonman76]

This is pretty cool, but I can't play the videos either. My computer at work is the only one with high speed internet and I can't install anything, at least right now. Could someone give me a hand with this one? I copied a side view (pre-airdam) into Autocad and traced around it, then solid filled it, so it was black on white. Wheels off should simulate airflow under the majority of the car.

[donee]

Hi All,

This is all very interesting. But its been commented elsewhere what is usabilty of the results? One thing to look at is the flow at the rear. If there is flow upward, then extending the cam back is the thing to do. And if the flow is downward, then doing a diffuser, with some rear belly-panning is the the thing to do. What one wants to do is make the flow go straght back without upward or downward velocity components.

Can somebody do a Prius without wheels ?

[azraelswrd]

Wagonman

here

Values I used:

Re = 2,500,000
dt = 0.01
length = 30 sec
cutoff level = 0

[wagonman76]

Cool, thanks a bunch!