Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Breaking a surface of a Sphere in to Triangles? 1

Status
Not open for further replies.
Replies continue below

Recommended for you

That was actually done in another piece of software that gives you the ability to create a geosphere. Then you just create a pyramid on top of each triangle.

We have actually created the sphere in a few different pieces of software. I am just on a mission to try and do it in SW for my own little home project.

I beleive Jason did all those models in the Pic in Maya.

Stephen Getsy
Product Development Engineer
Silgan Plastics
 
Let me get this straight. you really want a geodesic sphere? So you do not want a sphere cut into triangles. You actually want a spherical looking thing made up of flat pieces - over 11k pieces. (Are you some kind of masochist? - just kidding.)

If you do a google search you will find a site that gives formulae to calculate the sizes for goedomes. (Been there a couple of years back but can't remember the ULR - it's mostly "fascinating" trivia by some guy with a geodome fetish. However I noticed that he is not a millionare yet.) Anyway it gives the lengths for the side members of the patches based on the sphere diameter and number of faces or some such. I am sure you could convert it to what you want. I used it to essentially do the same thing, but I was not as adventurous as you are (many times less patches). Also bear in mind he is making domes, not spheres, so he ends up with odd partial patches at the hemispherical edge that you will not need.

Also, if you look carefully at the patterns of pieces, you will find you can make a limited number ot subassemblies (possibly only two) and use some creative circular patterns in horizontal bands of the assembly.

John Richards Sr. Mech. Engr.
Rockwell Collins Flight Dynamics

There's no place like 127.0.0.1
 
A geodesic sphere consists of 20 hexagons and 12 pentagons. Each pentagon is surrounded by 5 hexagons. Both will have the same length per side. Mate these together and the sphere will form itself. You can then correlate the sphere radius to the lenght of each side.

If you then take each hexagon and divide it into 6 equal triangles, and divide each pentagon into 5 equal triangles, you now have 720 peices total. The two long legs of each new triangle will be equal, but longer than the original side so that the vertex will fall on the sphere surface.

If each triangle is now divided into 4 equal triangles, this yields 2,880 pieces. Again, the common vertex should fall on the sphere surface.
Take it yet another step and this will yield 11,520 pieces.

Does this make any sense? Once you understand it, you'll find its much easier to do than it is to explain.

Good luck


 
NHPilot
Close ... but no cigars. I just tried your method using unity as the side length of Pentagon & Hexagon, but could not get the 5th Hex to mate.

Have you succesfully tried this?

[cheers] from (the City of) Barrie, Ontario.

[ponder] If you choke a smurf, [smurf] what color does it turn?
 
It is ... how many people do you know that want to draw a Geodesic sphere? If there was a big demand for it SW would add it in. Remember SW is basically geared toward mechanical engineering ... not civil or architectural.

[cheers] from (the City of) Barrie, Ontario.

[ponder] If you choke a smurf, [smurf] what color does it turn?
 
Yes, I saw an overcontraint originally in SolidWorks 2001Plus. As you start mating tiles, only the first one will be fixed. All others should be underconstrained so as to 'find their own' place. The parts do have to be placed relatively close to were they will end up or SolidWorks starts to flip things around unexpectedly. Spheres like this came up a while ago as a question to which I sent in only one of many solutions, some time back in SolidDigital digest magazine.
Incidently, I designed a few of these, the first of which was done in Autocad in 3D wireframe. They have been used as integrating spheres from (20 inch diameter to 3 Meter diameter) to measure light output of large lamps.
 
NHPilot
Do you have a Pentagon with 5 Hexagons modeled (or sketched) & mated? If so would you please send it to the email in my profile. I would like to see what I am doing wrong! Thanks.

[cheers] from (the City of) Barrie, Ontario.

[ponder] If you choke a smurf, [smurf] what color does it turn?
 
I have a model of a Geo Sphere I down loaded from some were probable solidworks library I can’t remember exactly
It’s not a big model I could email it to anyone that’s interested
 
NHPilot
Thanks for the models ... I see you had the same problem that I did with the mates ... it should work with all sides being Coincident ... but it doesn't. Could be a SW bug/limitation caused by approximations. If you zoom in extremely close, you can see that edges/sketch lines/poiints do not actually coincide.

manxJim
Thanks also for the GeoSphere ... Great model, someone has put a lot of thought & effort into it. I look forward to disecting it.

Penguin221
You should get & disect these models from NHpilot & especially manxJim.
They should answer your original question.

[cheers] from (the City of) Barrie, Ontario.

[ponder] If you choke a smurf, [smurf] what color does it turn?
 
Penguin221
When you get it worked out to your liking,
please let us know. I would like to see your final config.
Good luck. In my spare time, I am also trying to work it out...just to see if I can.[pc3]
 
NHpilot and manxJim

Would you both mind emailing me your files?

Please make sure they are zipped, because our sever may delete them if they are not zipped.

My email address is

s.getsy@silganplastics.com

Thanks.

Stephen Getsy
Product Development Engineer
Silgan Plastics
 
penguin221,

I just checked out that website and got more and more interested the more I read along. Is it too late to be a part of the design work? How to get involved?

Thanks,
Brian
 
CorBlimeylimey,

'Thanks for the models ... I see you had the same problem that I did with the mates ... it should work with all sides being Coincident ... but it doesn't. Could be a SW bug/limitation caused by approximations. If you zoom in extremely close, you can see that edges/sketch lines/poiints do not actually coincide.'

I'm not sure what problems you mean. Do you have mate errors when you open my model? Are the parts not fully constrained?
Also, as far in as I can zoom, everything looks fine. Might you be seeing a video glitch?
 
Status
Not open for further replies.
Back
Top