I'd never heard of 5 axis 3D printing until I stumbled across this article.

@ekaggrat_singh_kalsi Good idea, that could be promising.

Anyone have experience w rhinodlic3r? Is this a path worth going down?

What about grasshopper? Brian uses it to do all sorts crazy things w gcode. It’s nodal, fir visual
programming algorithms. 5axismaker (Kickstarter people making a 5 axis multimachine) use it and there are done downloads here

http://forums.reprap.org/read.php?1,409829

Carl Bass showed me a “print” done on Autodesk 's welding robot. It was mind-blowingly good. He said the welder was s fancy off the shelf model but they didnt need the features. He thought a cheap welder would work. I saw the printer at Autodesk AU a couple weeks later. They have some skunkworks type software they are toying with. He said the challenge is cooling the print.

Anyway, I share this because 5 axis printing w plastic will lead to metal printing. Carl said their machine has a Cnc component to surface afterward. Sound familiar. Plastic will be easier and lead the way.

Brook

@Brook_Drumm i looked again at rhinosli3r . unfortunately what it does is turn nurbs into a mesh and then pass it to slic3r to do rest of the work. There was a project in native grasshopper 4 years ago called silkworm which was also promising but is dead now… Yes grasshopper is a good way forward . I have used it in the past to generate gcode for my sliced light experiments , but for 5 axis solving the math is a whole different topic which is beyond my understanding.

Looks very cool.
Designing the mechanics of a low loads 5axis machine is not the major challenge. 5axis milling cam is a very hands-on process with many decisions and machine knowledge is necessary. Often done as a set of simpler 3d operations rotated to different planes. No general solution exists.

Now, the major challenge for 5axis printing is to find a completely new slicing paradigm that can take almost any model and prepare paths with no more interaction from the user than a classic slicer .

I have previously proposed one such skeletonization scheme. Where reverse ordered printable shells are eroded away in in such fashion (selected via ambient occlusion ) that the print tool always can reach the intended area without collisions when replaying forward as printing. Though unproven to be a general solution, such strategy could possibly generate paths for a majority subset of models and probably all simple topologies.

Caveats include very complex geometric shell calculations with tons of undefined edge cases.

Before I pursued it further as I asked my self: what can 5axis print accomplish beyond 3d+support, 3d+wavy layers or 3d+turning?
If little more then it’s hell lot of work, though incredibly cool tech.

@Ryan_Carlyle As far as I can see, 5-axis machines need special software anyway if we want to make them easy to use - that requires writing a lot of stuff anyway, no matter if you start with something in existence or a completely fresh design.

@Brook_Drumm No wonder cooling is a problem when you use a welder… the temperatures needed for melding metal a a few orders of magnitude above that of your everyday plastics… :wink:

@Markus_Dieterle MachineKit can do 5-axis machine control today without any difficulty, you just need 5-axis gcode to feed it. It can also do things like take three-axis gcode and slave the AB tilt/rotate axes to the current direction of XYZ travel for kerf angle compensation. It’s kind of ridiculously powerful for stuff like this.

Tinyg guys have talked w me about flipping the 3d printed gcode by making all straight lines curves… flat lines become flat “curves”. They have planned a layer to translate any gcode into this so the surface can be all smooth curves. Non-trivial, I’m sure :wink: but they propose that if that is done, printing w 5 axis or machining w Cnc becomes very different. You can approach the surface from any angle or direction. It sounds like science fiction. But they are wicked smart. They are the only ones I know trying to merge 3d printing and gcode form the motion control angle. If they do not, that firmware will stand alone as a ubiquitous solution to any machine control.

Forgive my overly simplistic explanation. Hearing them tell it bends the mind.

Brook

@Brook_Drumm there’s a ton of great motion control stuff you can do if the path commands are smooth curves, like calculate centripetal acceleration. That enables a whole slew of improvements to the firmware acceleration algorithms, which simply are not possible with today’s GRBL-based firmwares that everybody uses. (Except for MachineKit anyway.) Imagine the elimination of all ringing. Not out of the question.

Join Illuminati now and embrace the light you http://seek.Money ,Power,Fame and wealth shall become yours in just three days. Are you a president,Actor,Actress,Footballer,Business persons, receive the light that never seizes…If interested to join contact our headquarters through WhatsApp +1(226)2103887 :small_red_triangle::eye::small_red_triangle:

REMEMBER DISTANCE IS NOT A BARRIER
:small_red_triangle::eye::small_red_triangle:

WhatsApp hail to +1(226)2103887 and say yes to your dreams.
Hail