It’s been a while since some guys developed a moveable printing bed for a delta printer (sorry, text is in German only…but pics are still interesting: https://www.heise.de/make/meldung/3D-Drucker-mit-beweglicher-Druckplatte-3594506.html ). Any ideas if there’s a printer with this feature available, yet?
What keywords do I have to use?
The problem is not the printer - There is no slicer able to use this!
shouldn’t be too hard to write a slicer for it.
@Henner_Zeller please go ahead - the mechanics is simple (i already have a 3DOF plattform usable as printer bed). I mean lets take a look at how many parameter is cura using… may i ask how many 6DOF robots you have programmed?
I have programmed 8 axis robots and built motion-control systems for it. Hardest in this case is probably collision avoidance with the printhead when we have to tilt the print a lot.
exactly, it would be simple to have a sphere and just follow the normal vectors - but when having multiple objects with travel moves … Also printer G-code (as far i have seen) uses only XYZ and not IJK ABC UVW (some are supported) but no slicer is using them. Even most CNC machines using paths along the XY plane and not in Z.
Then you need a structure to build upon - which is easy if you can build your path on the previous layer but in this case that layer need to be a 3D surface. So would be amazing if you can write a working slicer for that!
Collision avoidance is one issue, another is figuring out what exactly the slicer is supposed to be doing with the rotary axes at any given time. Regular 2.5D slicing uses a deterministic algorithm with only one possible path solution given the input settings. 5D slicing has infinite path solutions.
I suspect that here the solutions are also limited as we essentially have a build-platform that tilts, but probably not much more than 60 degrees (as opposed to full rotation), which is totally fine for the task at hand, and hopefully limits the choices the slicer has to consider.
@Henner_Zeller first, what problem(s) are we trying to solve by tilting the bed? What do you actually want the slicer to DO with tilt?
Once you allow rotation DOFs, “layer” slices can vary in thickness, angle, and contour, along with nozzle orientation relative to the object. This is more complex than 5D milling where material removal is material removal. So you have a lot of DECISIONS requiring some kind of geometry recognition and optimization function.
Plus the possible machine-specific issues of collisions, working space definition, kinematic singularities, and “kinks” in the work space where the system can’t get from one position/orientation to another without stopping extrusion.
Basically a 5D slicer is going to end up being fairly application-specific AND fairly hardware-specific. There’s no reason for hobbyists to open-source develop that. Particularly when there’s already various solutions to creating arbitrary print geometry, such as soluble support or powder-bed printing.
Well, you need the tilt to be able to print overhangs without needing support material. Also you can make nicer layer lines in particular in these steep regions.
There are a few constraints that need to be configurable, but it will not be much more hardware dependent than current slicers.
There is a good reason to open-source develop that, as this is the easiest way to distribute things and discuss the concepts and math. Also, if it was not open source, it will be a dead end as nobody would build machines that can do that.
Also, writing CAM-software is fun and sharing it is more fun.
@Henner_Zeller why don’t you try it and let us know how it goes 
There are dozens of five-axis printers made by university teams and other very smart people, and none of them do anything more interesting than run one-off hand-scripted demo gcode.
Yes, I might write one actually, as I have some project upcoming where I need one.
i remember there was this cool non planar slicing https://plus.google.com/+NicholasSeward/posts/FFzY1EcFgkn … never heard again from it. (and this doesn’t need any special hardware) probably it is impossible to get an equaly nice end result
@Ulrich_Baer that was a demo… hard to generalize
@Ryan_Carlyle they also had this extrem demo - and as they change layerhight within a layer this is difficult to manage. I also have just seen that Cura has adaptive layer as experimental feature - works well so far.
https://plus.google.com/photos/photo/104209901845806192943/6113999525595345698?sqid=100932458296758031193&ssid=6b5c3af1-47d1-4914-ac53-7e74a6378dd2