Fractals? We're talking about fractals? Yes, yes we are.
Recently, I took the family to the Kalamazoo Valley Museum to see their Fractals as Art exhibit...along with all of the other great stuff they have there. I have always thought fractals were pretty interesting, and seeing the results in various forms is always pretty compelling. My kids (5 yr old, and 2 yr old) didn't care much about the math behind them...they just like the visual results. I thought I would use their interest as an example of discussing math theory with them. (Note: at this point my wife said "why do you always have to 'nerd up' a good thing?")
After seeing all of the visual results of fractals, I decided to work out some fractal based designs in Inventor 2012. I am not the smartest guy around...so, I wanted to take at some of the more basic fractal concepts. My first go at it wasn't going to be a Julia Set fractal. I am very confident I could do a Menger Sponge or Sierpenski Triangle in Inventor (or AutoCAD for that matter). Really, I wanted to take a look at doing something very familiar in a different way.
So, I decided on the Koch Snowflake.
The basic idea is a curve without tangents, constructable from elementary geometry with finite volume and infinite parameter. Now, the question was: could I do this in 3D? (and could I do it in a few hours before my son goes to bed, so he can see it).
Now, what I ended up making was not truly a 3D version of Koch Snowflake. A Koch divides each segment in thirds and inserts the original shape (a triangle). I was using triangle based pyramids..and decided to invert the next 'layer' as half the size of the previous pyramid. I hope that makes sense.
After doing some quick layout sketching (thanks Mr. Gascho for my geometry education), I got to work.
In Inventor, I just went with a simple process:
1 - Model triangle base pyramid (via loft to point)
2 - Insert work axis
3 - Create offset work plane for next pyramid creation
4 - repeat as often as needed
Once the initial geometry was built, it was just a matter of performing the necessary Circular Patterns around the existing work Axis.
Now, this is just the start. I could have kept going and added the additional fractal'ation to these areas (again, infinite parameter)...but, the little man was happy with what he saw.
...and I didn't want to risk 'nerding it up' with him too much.
-lw
As fractals are 'rules' based they lend themselves to automation. When do we get the VBA generated version?
Posted by: Duncan Anderson | 01/11/2012 at 12:06 PM
I smell an iLogic takeoff here....
Posted by: Carl Smith | 01/11/2012 at 12:14 PM
SWEET! Yes iLogic! And iHeart this idea!! I will find time to do this with my students for sure!
Posted by: J.Schmidt | 01/11/2012 at 12:20 PM
my thoughts were around iLogic. again, i was under a time crunch last night...so, just modeled my concept. but, iLogic is the next step for sure.
Posted by: Autodesk Manufacturing Toolbox | 01/11/2012 at 12:37 PM
This is exactly the kind of thing I've been looking into lately!; This looks like the kind of approach that could be used to create variable density fill for objects destined to be 3D printed?
Posted by: AlexFielder | 01/11/2012 at 03:31 PM
Just a quick update...as i continued the fractal design method to the model, it gets closer and closer to spherical form.
...which makes sense given the model lends itself toward infinite surface area.
I thought it was interesting anyway.
Posted by: Autodesk Manufacturing Toolbox | 01/13/2012 at 09:36 AM
I made a pentaflake in SolidWorks (2008). It is my interpretation of a 3D fractal: http://www.charlesculp.com/solidworks/PentaFlake_500.jpg
Posted by: Charlesculp | 01/16/2012 at 09:50 AM
That's gorgeous!
My attempt at a CAD-based fractal is here. http://www.youtube.com/watch?v=0POQYrwHtG8&feature=g-upl&context=G27b8417AUAAAAAAAFAA
-Blake
Posted by: Blake Courter | 01/16/2012 at 12:53 PM