acidmonkey
Guest
|
|
« on: December 12, 2009, 03:00:18 AM » |
|
I stumbled across some crossposted imaged of the 3D rendered Mandelbulb's and I knew I had to come here. I've been trying to figure out how to generate 3D fractals for a while now, but have never known where to start. I have access to a number of rapid prototyping FDM machines and I am extremely interested in attempting to use them to print off some physical fractal models. Structurally, these shapes are incredibly impressive and I believe they could have numerous applications in the real world.
Obviously the process behind this would be rather complex, but I'm wondering if this has been done yet or if there is any interest in it.
|
|
|
Logged
|
|
|
|
Snakehand
Guest
|
|
« Reply #1 on: December 12, 2009, 09:56:51 PM » |
|
I stumbled across some crossposted imaged of the 3D rendered Mandelbulb's and I knew I had to come here. I've been trying to figure out how to generate 3D fractals for a while now, but have never known where to start. I have access to a number of rapid prototyping FDM machines and I am extremely interested in attempting to use them to print off some physical fractal models. Structurally, these shapes are incredibly impressive and I believe they could have numerous applications in the real world.
Obviously the process behind this would be rather complex, but I'm wondering if this has been done yet or if there is any interest in it.
A friend of a friend has access to similar equipment, and I was thinking about asking him about what data fromats are used. If the data formats are reasonably simple, I would be happy to provid a dataset for 3D printing. Which 3D data formats does this equipment use as input?
|
|
|
Logged
|
|
|
|
|
acidmonkey
Guest
|
|
« Reply #3 on: December 13, 2009, 11:09:26 PM » |
|
It uses a 3D mesh format called a stereolithograph (.stl format).
I've never used Voxel data but it may work out. I can convert pretty much any sort of 3D modeling format into an .stl but getting it to that first step is what has been eluding me.
I'm also using a scripting language called Grasshopper. It works like an object oriented programming language but can be used to create generative 3D models. I'm definitely not a mathematics major though so I would need some help interpolating the formulas.
|
|
|
Logged
|
|
|
|
miner49er
|
|
« Reply #4 on: December 22, 2009, 02:14:57 PM » |
|
I would pay for a model of a mandelbulb. I would simply love to touch and feel one in my hands!
However, I don't think Rapid Prototype Machines produce very high resolution objects...
|
|
|
Logged
|
|
|
|
Snakehand
Guest
|
|
« Reply #5 on: December 22, 2009, 02:30:57 PM » |
|
It uses a 3D mesh format called a stereolithograph (.stl format).
I looked up the wiki article on .stl http://en.wikipedia.org/wiki/STL_(file_format) - and it seems that the data has to be in polygon form. The complexity of the mandelbulb makes it very hard to produce a good polygon model. Although one can always explore the subject further. How big a model can you make ( I assume ~20 cm ) - and how small details can the machine sculpt. (I am guessing 1/10 mm) - there is also the risk that the polygon data will be to complex for the machine to handle. Do you know of any limits to the polygon count ?
|
|
« Last Edit: December 22, 2009, 03:15:37 PM by Snakehand »
|
Logged
|
|
|
|
kram1032
|
|
« Reply #6 on: December 22, 2009, 02:47:47 PM » |
|
I'd think, the polycount isn't that problematic but rather the thin areas which could break already during processing.
Maybe you could choose a distance estimation, based on the minimum thickness allowed for printing...
|
|
|
Logged
|
|
|
|
David Makin
|
|
« Reply #7 on: December 22, 2009, 03:53:16 PM » |
|
Terry Gintz has rendered and exported a Mandelbulb to Bryce so it's definitely possible to create the required source data for a 3D print. However I think at the moment this can only be done for the Mandelbulb with his personal updated version of QuaSZ.
|
|
|
Logged
|
|
|
|
|
Hodge
Forums Newbie
Posts: 1
|
|
« Reply #9 on: December 24, 2009, 05:32:28 PM » |
|
The resolution of rapid prototyping machines is usually adequately high for a Mandelbub print, although one would have to be careful to eliminate fine filaments, and remnant resin might indeed be a problem. A straight forward method would be to generate a mesh model in a common format, eg 3ds, obj, VRML etc. Then, massage the mesh, reducing poly count and fine fillaments. Then convert to stl and try to print! Does anyone have a low poly mesh ( say 5000 to 10000 poly) to play with ?
|
|
|
Logged
|
|
|
|
guytron
Guest
|
|
« Reply #10 on: December 29, 2009, 08:37:30 PM » |
|
Laser sintering of metals or ceramics gives fine resolutions (sub millimeter) and high durabilities. You will need to render at some really high resolutions to get the fine details.
|
|
|
Logged
|
|
|
|
pabs
Guest
|
|
« Reply #11 on: January 06, 2010, 07:32:23 AM » |
|
oh man yes please print some 3d mandelbrots! that's so great! i'm not a code writer or mathematician but i love fractals and just found out recently about 3d printing! do it!
|
|
|
Logged
|
|
|
|
GQuinn
Forums Freshman
Posts: 19
|
|
« Reply #12 on: January 12, 2010, 01:30:51 PM » |
|
I stumbled across some crossposted imaged of the 3D rendered Mandelbulb's and I knew I had to come here. I've been trying to figure out how to generate 3D fractals for a while now, but have never known where to start. I have access to a number of rapid prototyping FDM machines and I am extremely interested in attempting to use them to print off some physical fractal models. Structurally, these shapes are incredibly impressive and I believe they could have numerous applications in the real world.
Obviously the process behind this would be rather complex, but I'm wondering if this has been done yet or if there is any interest in it.
Hi acidmonkey. I've just started a new thread with links to some photos of a 3d printed mandelbulb. http://www.fractalforums.com/the-3d-mandelbulb/3d-printed-mandelbulb-photos/As some people have pointed out, there are issues with accuracy, detail and fragile materials, but we have got a nice model to sit on top of the computer :-) Gary.
|
|
|
Logged
|
|
|
|
|