Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations MintJulep on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Trouble modeling geodesic dome

Status
Not open for further replies.

SKAR81

Mechanical
Feb 4, 2011
4
I am having difficulty mating parts to create a geodesic dome model. Due to the spherical nature of the dome, the lengths of the struts used to construct individual triangles cannot be fully described by a decimal number (irrational). I am therefore forced to use distance mates on some occasions and I have to arbitrarily choose a 'close enough' distance. Is there a way to mate 'as close as possible' or some other way of accomplishing this? Also is there an equivalent for angle mates?

Thanks Everyone
-Skar81
 
Replies continue below

Recommended for you

First I modeled with planes so I could find my swing and bevel angles for the true model. The irrational distance values obviously apply to the sides of the planes but I'm afraid of completing the true 3d model before I resolve the mating issue. (Only the parts assembled were needed to find all the pertinent angles)
 
 http://files.engineering.com/getfile.aspx?folder=093cbfa6-1e26-4990-aebe-2111d964bdd5&file=dome_pic.jpg
SKAR81,

How you model a geodesic dome depends massively on why you are modeling it. SolidWorks does an excellent job of modeling confusion and stupidity. If the cut is difficult for you in SolidWorks, it will be difficult for the carpenter or welder or whoever.

My first thought about geodesic domes in general is that they are space frames. Space frames are ideal for maximizing access for everything in and out of the frame. This goes for you, your hands, your tools, your family, your personal belongings, rain and snow, burglars, the police, damn foreigners, animals, etc. Space frames make great jungle gyms. As shelter, a complex space frame is a way to maximize the number of seams you have to seal, somehow. A good rule for any kind of structural enclosure is that you design your covers first, then figure out the structure underneath.

Do you have to hit the diameter bang on, or you can you design a standard triangle section with rational dimensions, and hit the "diameter" approximately?

If you can design a standard connecting bracket, you can connect square cut, standard lengths of lumber, extrusion or other structural sections.

A good trick with complex frames is to create a structure shaped "blob". Make it transparent. Set the material density as low as you can. I use 0.1kg/m^3. I wish SolidWorks took into account figments of people's imaginations.

Model the blob to the outside shape that you want. Add in functional details like doors, windows, access panels, etc. Locate your geodesic pieces to the blob.

Load your blob into Custom Properties. There is a click-box that will exclude it from your bill of materials.

Your triangle section is interesting. You are aware of course that the sides of your lumber cannot be perpendicular to the faces. You might be able to do this with a table saw or circular saw, but you will need some sort of jigging procedure to make up for tolerance stack-up.

Critter.gif
JHG
 
Thanks for the input drawoh. Some triangles have already been made and the concept has been proven in a scale model. What I really want to do is create a cad model. Even if I adjusted the diameter there would still be irrational numbers as each vertex points directly to the center of the sphere; hence a triple miter cut on each end of a strut. What I need to find is if it is possible to have solidworks define a "closest possible" mate or perhaps change the resolution of the entire model so that one thousandth of a unit is considered exact.

Thanks
 
I've used "close" within 0.010 mates when laying out (for instance) a fixed-centerline belt/pulley arrangement at some angle to an axis as compared to x-y decimal coordinates of the shaft centerlines for the pulley locations themselves.
It works fine and allows for metric/English units if need be.
(For instance, the belts/pulleys might be metric, but the pulley shaft centerlines, etc might be better to remain in English units)
I've also used this for Chains/Sprockets because the dimensions often don't come out to the nearest common decimal, but yet, the accumulated link(s) dimensions must be held.

If the triangle shape & dimensions present problems with mating; applying the "close to" modified mate to the geodesic form might be the way to go....as long as there aren't any further "stack-up" problems.

 
OK. I'm still confused. You want to model a geodesic dome. I got that much. Are you saying that you have an existing wooden triangle frame (shown in your pic above) that you want to stack up into an assembly?

Dan

Dan's Blog
 
Model as multi-body solid relying on geometry and very few dimensions needed - no calculations needed. SolidWorks will calculate the geometric solution out to as many decimal places as it uses.
Push the solid bodies out to assembly and Fix.
 
Would it be possible to model the framework as a weldment, and use the trim feature to add the triple miter cut to each strut?

Joe
SW Office 2008 SP5.0
P4 3.0Ghz 3GB
ATI FireGL X1
 
SKAR81,

The relationship between diameter and any linear measurement around a circle or sphere, is an irrational number. You are going to have a round-off error.

How accurate is your carpentery? ±1/16"? I can set my circular saw at home to within one degree. I suppose a good table saw can do somewhat better. All of these tolerances will accumulate as you add pieces to quite a substantial error. You are going to need some sort of surveying and jigging process.

Your round-off error is trivial.

Critter.gif
JHG
 
I think the problem the OP is facing is properly constraining a bunch of parts that aren't exact (just like the real world) and therefore tolerance stack-ups cause problems with SWx exact assembly constraints.

The solution is to model as multibody (and/or skeletal modeling techniques) and forget assembly constaints altogether (not needed).
 
rollupswx,

I think that modeling in SolidWorks is fairly easy. Define a centre axis. Insert the first element of each row. Create a circular array for the row.

Assembly of this thing is going to be a challenge.

Critter.gif
JHG
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor