Geodesic Dome Coordinates in RISA 3D

[garrettgman]

I'm trying to model a geodesic dome in RISA 3D but can't generate the coordinates of the nodes for the dome. Does anyone know of a coordinate generator? Or a simplified way to model a geodesic dome in RISA 3D?

 

[BAretired]

I don't know anything about RISA 3D, but coordinates are probably best handled using spherical coordinates, i.e. a vertical axis, a radius and two angles. If you want to translate these to cartesian coordinates, a simple transform is available.

BA

 

[JoshPlumSE]

It's actually much easier than it seems like it would be.

Model the arch using one of the "high level generation" utilities. Then copy-rotate that arch through 10 or 15 degrees or so. Draw in your plate elements to form one slice of the pie. Once that is done, it is as simple as using the copy rotate feature to form the rest of the dome.

 

[garrettgman]

@ JohsPlum
I've tried that copy-rotate feature with the "high level generation" arch. I can't get it to work right. The triangles on the dome don't line up with the coordinates when I copy-rotate arches.
I'm hoping there is a website or someone who has come up with equations that will generate the 3D coordinates automatically.

 

[Lutfi]

I am not familiar with Risa. I use STAAD. I would generate an arch; say the north arch, with as many nodes as you feel comfortable with. In STAAD, there is a command that allows you to offset the arch as many segments around the center point. Provide the number of offsets and the angle. May be I am oversimplifying the approach.

Also, a phone call to Risa technical support may give you an immediate answer.


Regards,
Lutfi

 

[OCI]

There are websites that can give you the coordinates you need. Geodesic domes are actually more complicated than you might think. There are many types and configurations.

 

[OCI]

http://www.desertdomes.com/

Here is one with a calculator

 

[OCI]

this might be a better site

http://geodome.sourceforge.net/

 

[CTSeng]

When the geometry gets complicated I'll draw a stick model in 3D in AutoCAD then export to DXF then import the DXF into RISA. You'll want to experiment with the way it works since you can bring in member shapes to match layers too.

 

[dhengr]

Garrettgman:
I think you were talking to Mr. RISA (JoshPlum, am I correct?) @ 12:45, on 23MAY11. You might want to hit him up, again, for a little more help.

BA:
“I don't know anything about RISA 3D, but....” Me too, but... There you go again, you provided a clean, simple, basic solution for the question asked. But, you didn’t offer any software that will do our elementary and preliminary thinking for us. I suspect that you and I would have one hell of a time even trying to deal with the program complexities, I know I do, given the way you and I approach design problems during the prelim. steps. We are all forced to get mired in the minutia of these programs, to the point that we forget what the original problem was, but we know it to 12 decimal places. And, the damn program doesn’t tell us that detail it analyzed can’t practically be fabricated or installed, it just gives us a high von Mises stress that we don’t understand either. It makes me wonder how anything ever got built before the advent of computers, and the complexities of them and the software they spawned became an end in themselves, not a simple means to an end. These programs can do amazing things, almost anything, problem is, mere mortals can’t figure out how to make them do it. When you overcome one idiosyncracy, that brings three more into play.

Makes you wonder how Walther Bauersfeld and Buckminster Fuller ever did their domes, before computers, or when computers were as large as the domes themselves and could do less calculatin than your cell phone can today.

 

[JStephen]

On the software I've used- it would amount to arraying the nodes and circumferential members around, then putting in the radial members manually (or doing one whole pie-shaped section and arraying it around). It may be easier to do it all flat, and then elevate nodes to the proper level.

And, I forget the details, but on the software I've used, if you drew a member between two random points in space, it tried to orient the member axis through the origin, rather than putting it vertical- in other words, check that your members are rotated right when you get it done.

 

[BAretired]

dhengr,

I think you have hit upon the truth. Whenever I have tried to use a Finite Element Analysis program, I find that I get bogged down in the idiosyncrasies of the program and lose track of the engineering aspects of the problem.

As the old Pennsylvania Dutch saying goes..."Ve get too soon oldt, und too late schmardt".

BA

 

[JoshPlumSE]

If you cannot get RISA's arc commands (circular or parabolic) to adequately model a cross section of your dome then bring in the basic geometry another way.

1) Set up an Excel spreadsheet with equations which define the 2D geometry of the arch. Then copy and paste the coordinates from Excel's spreadsheets into RISA's spreadsheet.... one of the advantages of being a spreadsheet based program.

2) Draw the arch in AutoCAD and use the DXF function to bring over the basic coordinates of the arc.

Once you've got a 2D slice of the arc modeled in RISA, then it's just a matter of using the center point of the arch with the copy-rotate feature to copy it through one slice of the pie. Then spend some time modeling the plate elements or members that will fully model that slice. Then use the copy rotate feature to copy that slice around to form the whole dome.

Also, if you guys have the current version of the program, just e-mail the support group (support@risatech.com) some basic drawings and we could probably provide more detailed guidance.

 

[garrettgman]

Thank you all for your assistance. I was able to get the model of the dome in AutoCAD and imported the DXF file into RISA. This worked like a charm. Thanks again.