Aerodynamic streamline template - online tool (view streamlined extension on any car)

[Shortie771]

Woot! I did something useful for once

[KamperBob]

Originally Posted by Shortie771

Wow, great work! I wish I had that kind of skill with 3D imaging. I tried making a kammback for my Del Sol on Google Sketchup and I couldnt get the thing to work out at all. I couldnt figure out how to connect the kammback to the existing car model and I couldnt figure out why no matter what I did I couldnt make my kammback not have some tiny gap in it. I could keep making the gap smaller and smaller but I couldnt get it to completely close.

I finally gave up on it, but I would have really loved to finish that so I could run the model through a simulated wind tunnel.

Thanks. I banged my head a lot trying to make SketchUp extrude a 3D profile. Compound curvature is a real PITA. I ended up working with wire frames and carefully setting the number of arc segments. That basically gave me a well enough defined mesh to manually connect triangular patches. For me a bonus was prep for construction. The cross ribs were already defined so I could make bulkheads. (After losing my shop fabrication was suspended.)

[sgtlethargic]

Originally Posted by Shortie771

...

I do have GIMP, but I suck at it lol. So I just use the transparet selection feature in Paint. That is how I did the top view image. I could do the same with the side view image as well in a matter of minutes, but is it really necessary to make the template the foreground? If so, why?

I wasn't aware of the transparent selection feature in Paint. A quick search yielded this:
How to Make a Background Transparent in Paint | eHow.com

[Shortie771]

Finally made a virtual kammback for the Del Sol, following the aerodynamic template. Sadly to my suprise though, it didn't seem to do anything for my aerodynamics The high point on my car is just too far forward (right at the top of the windshield it looks like). The air separates from the car before it even makes it far enough back for the kammback to be useful.

My car is supposed to have a CD of .35 stock. The software shows .48 (with or without kammback). Maybe I am doing something wrong.

All results are at 60 MPH with the exact same settings.

Template:


Flow Lines:


Top View:

[aerohead]

Originally Posted by Shortie771

Finally made a virtual kammback for the Del Sol, following the aerodynamic template. Sadly to my suprise though, it didn't seem to do anything for my aerodynamics The high point on my car is just too far forward (right at the top of the windshield it looks like). The air separates from the car before it even makes it far enough back for the kammback to be useful.

My car is supposed to have a CD of .35 stock. The software shows .48 (with or without kammback). Maybe I am doing something wrong.

All results are at 60 MPH with the exact same settings.

Template:


Flow Lines:


Top View:

It may require a university-grade CFD program to even approach the ability to develop meaningful Cds.The big-boys run super-computers to get 'close',then hours in the wind tunnel with clay modellers to dial in the final contours.
In the past,General Motor's Electronic Data Systems made programs available to university teams lacking wind tunnel facilities on campus,to help with their solar racer designs.
I contacted a state sales rep for one of the software companies which markets CFD software to schools.Knowing I was only an individual,they never returned my call.
I'm given to understand that the 'vehicle' must be defined within a 3-D AUTOCAD environment,with the exterior features represented by thousands of individual panels developed from the wire-frame model.
A system of sub-programs sets the mathematical foundation for the full Navier-Stokes Equation of spherical coordinate system which exceeds celestial mechanics in it's complexity for solution.
The wake structure must be known a priori in order to solve the equation.Not the reverse.

[Shortie771]

Well this is a $15,000 program, luckily it's free for about another week... Other than that you lost me in the fourth paragraph. I know these models that I am using and editing are far from perfect, but I thought it would be it at least a bit closer.

[orangustang]

Originally Posted by Shortie771

Well this is a $15,000 program, luckily it's free for about another week... Other than that you lost me in the fourth paragraph. I know these models that I am using and editing are far from perfect, but I thought it would be it at least a bit closer.

Autodesk makes some amazing software that I will never ever be able to afford on my own. Unfortunately, like aerohead said, it may not be good enough in this case. Drag is extremely difficult to calculate for any but the simplest shapes, and often even very good numerical approximations are way off. There might be some parameters you can play with in that program that will yield some more reasonable results.

That being said, you may have started your kammback too low. I'm no aerodynamics expert (I just do the coding), but my understanding is the curve of the template at the 0° mark is supposed to lie tangent to the roof of the vehicle. When you did it in Paint, it looked right, but you have the template set too far forward in the later image. This is consistent with the results from the program, which appears to be telling you the kammback is dropping off too steeply. The program may be a lost cause, but I'd be interested to see what it tells you if you align it to the 0° mark.

Update regarding the IE bug: I haven't gotten around to really looking at it much. I have identified one problem, though there may be more, but I haven't really begun looking for a solution yet. I haven't forgotten you and other IE users, I just have a lot on my plate at the moment.

[Shortie771]

Originally Posted by orangustang

Update regarding the IE bug: I haven't gotten around to really looking at it much. I have identified one problem, though there may be more, but I haven't really begun looking for a solution yet. I haven't forgotten you and other IE users, I just have a lot on my plate at the moment.

Not a problem, life gets busy at times. Take all the time you need, thanks for the effort and work you have already put forward

As for the kammback, I will take the advice and edit the kammback when I get some free time. Hopefully that will help. Even if the Cd isn't correct, at least I have a baseline of what the program says my car is and I can judge off of that to decrease it. Even if its off, hopefully I can get a decent idea of the kammback design I will go with eventually. Was looking to jedi sol to make his kammback first, but that didn't happen so I guess I will have to do it

[aerohead]

Originally Posted by Shortie771

Not a problem, life gets busy at times. Take all the time you need, thanks for the effort and work you have already put forward

As for the kammback, I will take the advice and edit the kammback when I get some free time. Hopefully that will help. Even if the Cd isn't correct, at least I have a baseline of what the program says my car is and I can judge off of that to decrease it. Even if its off, hopefully I can get a decent idea of the kammback design I will go with eventually. Was looking to jedi sol to make his kammback first, but that didn't happen so I guess I will have to do it

I have some video of a Porsche 914 crossing El Mirage Dry Lake at speed.
The entire area behind the backlight is nothing but turbulence.The Del Sol has some sail panel that the Porsche doesn't have,but I suspect that adding some 'roof' back there would show at the pump.
The 1967 Lotus Europa has the same kind of 'back' but it's long enough for good reattachment and scores Cd 0.29 with its 'open-air' rear.
Take a peek at what Smart did with their Roadster Coupe' all-glass Kamm back.

[Xist]

Yes, it does look like you set the template too far forward. It seems like you have the roof peak at the 1.5° mark. The car seems to fit, but please retry with a Kamm back that fits with the roof peaking at 0°.