ULTIMATE CHALLENGE

[negfrequency]

ok, a lot of people i've talked to have made a very good point, and i want to see if it is consistent with all CAD programs. So i have decided to start this thread to challenge anyone to this.

OBJECTIVE:
Model a soccer ball or baseball, using the CAD program of your choice.

REASON FOR DIFFICAULTY:
Baseballs, soccerballs and similar objects are extremely difficault in CAD programs because FIRST: a lack of surfaces. Working on one is not easy, even with workplanes. TWO: You will find tools similar to rectangular or circular pattern are difficault or ineffective to use, as the pattern on the ball is consistent in all directions, and ends up the same on opposite poles. THREE: A decal on a revolution doesnt cut it. Im talking surfaces here.

If anyone DOES manage to complete this task, post the picture in this forum for all to see your incredible feat. I will also praise you. GL!

[Ian Mackenzie]

Well, someone's done it (not me):

http://www.cadcourse.com/winston/SoccerBall.html

I stumbled on the above link while looking for a picture of a soccer ball as reference; I don't have TurboCAD, so I can't open the tutorial, but the result looks good...

[vadyr]

does 3ds max count...?

[Greg Needel]

OK so i really don't want to do this because i know that it will take some time but i came up with a "simple" way of doing it.



If you create a sphere in the cad of your choice, then make a datum plane tangent to the sphere. make a new sketch of the pentagon or hexagon shape on the centerline of the sphere and extrude the outer portion away, you will be left with a pentagon/hexagon rod with compound curves on the ends. extrude away so you are left with a pent/hex block with 1 curve edge and 1 flat. If you then chamfer the edges to the proper degree (not sure what it is), and fillet the corners of the part you will have a building block for the ball. You will need to make both a pent and a hex part. Create a new assembly, and assemble the pieces together......done deal.

now that i typed that out i may try it when i have a few seconds. The only choice now is the software (pro-e, inventor, or solid works).


Greg

[negfrequency]

Whoa nice. So creating all of the hexagons separately and then assembling them. that would totally work, just it would be a pain and take forever to do like you said. Also, soccer balls have hexagons AND pentagons, and sizing them correctly and angling them would be quite a pain in the buttox. Nevertheless, i didnt think of it, so points.

in fact i think ill go do it right now since thats the best idea ive heard, even though i should be writing a thesis.....

So that solves the soccer ball problem, but what about the baseball, or even a golfball which have far more complicated patterns?

[Andy A.]

Originally Posted by negfrequency

Whoa nice. So creating all of the hexagons separately and then assembling them. that would totally work, just it would be a pain and take forever to do like you said. Also, soccer balls have hexagons AND pentagons, and sizing them correctly and angling them would be quite a pain in the buttox. Nevertheless, i didnt think of it, so points.

in fact i think ill go do it right now since thats the best idea ive heard, even though i should be writing a thesis.....

So that solves the soccer ball problem, but what about the baseball, or even a golfball which have far more complicated patterns?

This method would actually be fairly quick. Once you have the two basic pieces, the remaining challenge is assembling it with the proper coincident mates, and it should more or less build it's self. All the pieces are identical, so I don't see any reason why you would have to make them separately. Just use the same pieces over and over and reorient them.

Greg's idea on how to build the two pieces is pretty clever.

As for a baseball, just how difficult do you want to make it? A baseball's cover is two pieces stitched together, but it forms a very good sphere not including the stitching. The two pieces of hide are, as far as I can tell, identical. You could model the stitches, but it would be more time consuming the difficult.

You've got me on the golf ball. Short of a whole lot of Boolean operations (subtracting the dimples from a sphere) and careful construction geometry, I can't think of an easy way to do it. With out a golf ball at hand, I can't be sure there isn't some obvious pattern to the dimples. I've always assumed there was no pattern to it, but I could be wrong.

Now a basket ball, with all the little bumps, that would be a chore. And a resource hog. Try running that through a fluid simulator with a full mesh!

-Andy A.

[Nuttyman54]

Originally Posted by Andy A.

You've got me on the golf ball. Short of a whole lot of Boolean operations (subtracting the dimples from a sphere) and careful construction geometry, I can't think of an easy way to do it. With out a golf ball at hand, I can't be sure there isn't some obvious pattern to the dimples. I've always assumed there was no pattern to it, but I could be wrong.
-Andy A.

There most definitely IS a pattern. The dimples appear to be placed in a hexagonal arrangement. I don't know about any other programs, but the array and 3darray functions in AutoCAD should be able to handle this.

[Andy A.]

Originally Posted by Nuttyman54

There most definitely IS a pattern. The dimples appear to be placed in a hexagonal arrangement. I don't know about any other programs, but the array and 3darray functions in AutoCAD should be able to handle this.

Your right. I have not held a golf ball up close in years, and for some reason I thought that the dimples where spaced further apart and randomized.

I do know that the dimples serve to better the aerodynamics of the ball. Exactly how, I'm still unclear.

-Andy A.

[Greg Needel]

Originally Posted by Andy A.

Your right. I have not held a golf ball up close in years, and for some reason I thought that the dimples where spaced further apart and randomized.

I do know that the dimples serve to better the aerodynamics of the ball. Exactly how, I'm still unclear.

-Andy A.

this site has some good info on why the dimples are there

http://fi.edu/wright/again/wings.avk...r/golf-01.html

[Greg Perkins]

So this took me about 20-30 minutes to assemble, and for the last 40 minutes I've been diddling around with surfaces and Inventor Studio. Thanks Petek! If you've got inventor, it's really easy to do, just 2 parts and 1 assembly was all it took.

list of parts:
20 hex shape, .625" flats all around
12 Pentagon shape, .625" flats all around

When you go to constrain your parts, make sure you constrain (create in your mind a relative top and bottom to each part) to the bottom edge of each peice. that way when it bends into a icosahedron it will all fit.

Good luck!

[negfrequency]

I've been trying to put one together with actually curved surfaces. ill post it when im done i think that it will take little longer.

meanwhile CONGRADULATIONS. I would have never thought of some of these methods, and you have all earned my praise. If i find anything else thats difficault or another way to do this I'll post it on this forum.

[John Gutmann]

Originally Posted by Greg Needel

OK so i really don't want to do this because i know that it will take some time but i came up with a "simple" way of doing it.



If you create a sphere in the cad of your choice, then make a datum plane tangent to the sphere. make a new sketch of the pentagon or hexagon shape on the centerline of the sphere and extrude the outer portion away, you will be left with a pentagon/hexagon rod with compound curves on the ends. extrude away so you are left with a pent/hex block with 1 curve edge and 1 flat. If you then chamfer the edges to the proper degree (not sure what it is), and fillet the corners of the part you will have a building block for the ball. You will need to make both a pent and a hex part. Create a new assembly, and assemble the pieces together......done deal.

now that i typed that out i may try it when i have a few seconds. The only choice now is the software (pro-e, inventor, or solid works).


Greg

That isnt the correct geometry, on a real soccer ball they cut out pentagons then they are curved. They are not cut away from a curved surface.

I would say to do it maybe make a sphere
make a work plane in the center, then offset 2 work planes from it in either direction along one axis. Make an angle work plane on the same axis and repeat. Then to actually cut it out make a sketch on each work plane, slice graphics then project geometry the edge that you get from slice graphics.

First off about the offsetting of the work planes. You need to go measure a soccer ball with a piece of string to find out the distance between the to parallel edges the penta/hexgons. then use math to find out what that dimension is if that line - line distance is put on an arc and then find out how far apart the edges are.

When your done with all that^^

Make a new sketch facing all the work planes and then project geometry the 5 / 6 lines onto the one sketch, then extrude cut the sphere away.

lastly chamfer the inside edges and fillet the outside ones for ease of assembly. When I have access to auto desk for a while in study hall I will model it up


OR....after a second thought....

Using somewhat of that method^^calculate the small circle you can fit that all into(penta gon / hexagon. then figure out the depth of that sector(or is it a section i cant remember the term). Makes a sketch>revolve it. Then on the flat part draw the hexagon from the calculated distances and extrude then do finishing steps.

[petek]

First step is to analyze the problem.

Building the basic buckyball model isn't so bad - just very repetitious! If you want to add curvature to the faces, I'd probably use a loft from the planar shape. [hint: you only need one linear and two angular dimensions and one type of constraint]

It took me about half an hour to create this much: