I'd like some help finding an online FEA application - Used for wind tunnel simulatio

[ItsMyFirst]

Hi everyone.

A few days ago a post prompted me to register and get involved! Neat eh?

Well it was someone who had some time on his hands and a workstation with FEA modeling software. He has a RAV4 and did a few
calculations. From what I can see the image in the display looks simple enough to warrent using, maybe I'm not knowledgeable, a
photograph for the outline of a vehicle.

Then inputting the size of the vehicle would lend dimensions to the outline of your car.

Here' s a suggestion.
If we take a photo showing exactly a good outline (the ride height in the photo), the front (showing the total area ) of vehicle:

then an FEA software might be able to model the CD (wind drag) of the car.

This would be exciting. As I know many people are excited to try
adding air dams or boattails to their cars and trucks.

Wouldn't using this software (if it were available as an online application)
be simple? Or with enough testing maybe it would at least yeild data showing
improvements or losses.

Could it be made possible for us to purchase a license or a service run by someone here to input our vehicle photos and get the CD ( air drag)
value for the vehicle and then using the FEA determine how much an air dam and other mods would / might yeild.

For you who have the time actually run the vehicle with / without the mod and see if the FEA would save time and help speed development of accessories for the vehicle.

I can visualize a database of vehicles with the gains or loss from adding various mods or designs.

Each designer can quickly try their design with the FEA and then for
those who will real world test the results would show the correlation.
With that, even if the FEA was wrong, we could develop a formula to
show how wrong and where it's wrong. That way the right answer
can be had.

I can see in my mind's eye people quickly coming up with designs to
try an beat another designer's idea.

It would be a really SUPER group project and well that's why I registered.

I'd be willing to share my experience in science.
I'm a scientist with a background for 30 years in electronics and chemistry and optics. My work was with materials (of all kinds).
Since I retired at about 46 years old I've spent time researching things of interest to me.

I'd like to buy or build or mod some electric vehicles and start posting real
world data (see the RechargeIT Experiement)

See google for RechargeIT Experiement

I posted a short note in Aerodynamics mentioning that I was asking
if an online FEA application would work out for us. I didn't mean to
repost the same thread. Thanks

Andre

[Bicycle Bob]

I think you'd have to input the whole shape, not just two views. Car builders have million-dollar software packages, and million-dollar wind tunnels to refine the FEA estimates.

So far, I've been trying to learn CAD off and on for 20 years.

[ItsMyFirst]

Originally Posted by Bicycle Bob

I think you'd have to input the whole shape, not just two views. Car builders have million-dollar software packages, and million-dollar wind tunnels to refine the FEA estimates.

So far, I've been trying to learn CAD off and on for 20 years.

Thanks and Hi.

I mentioned that I saw a post by an RAV4 owner .. which I'm sorry I
can't find again now. He had inputted only, I believe, the frontal
area, and a side view (I think) and he got a value for CD of about ( I think)
0.39 which was nearly identical to the published spec. That's when I
got excited that he could get good data from little input.

Anyway if I find the link, or if you happen to know it, it will be posted here.

Thanks and enjoy the fun

[RobertSmalls]

The data you get out will not be better than the data you put in. Here is a simple method I used to make a model of my Insight in Pro/E. I then added a tail extension to the model and got an idea of the size and shape of that tail extension.

I tried running the model through a few commercial CFD packages. Programs that output just a streamline illustration tend to be very simple to use, and I was able to get results (of unknown validity) from them. I also tried ANSYS 10, which can spit out a CdA for direct comparison between the established value for the car, the "before" model, and the "after" model. Unfortunately, I was unable to generate a valid mesh in ANSYS CFX.

Some programs can produce pressure and velocity distributions, which would be helpful for identifying problem areas on the model, IFF the model does not deviate from the car in aerodynamically significant ways. The underside of the car is a great example of something that's hard to model. The wheels ought to turn, the road ought to move, and the underbody should be fully modeled with brake cables, suspension, and exhaust hanging down. I would hesitate to make changes to the car based only on the result of a CFD simulation that uses an incomplete model of the car.



BotSapper took a 3DStudioMax model of an Insight, available on the Internet, and did this with it: http://ecomodder.com/forum/showthrea...tml#post168928

If you can convert a .3ds to a mesh for CFD, I'm sure I'm not the only one who would love to see Botsapper's model (with an aero tweak or two) run through the simulation with and without a boat tail.


If you'd like a copy of any of my .dxf files, I'll make them available to you.