Revisit 4 Axis Mendal and Polar Rep Rap

I started a new thread in this forum and New Repstraps and Post Mendal since the 4 Axis Mendal and Polar Rep Rap topics have not seen posts in some time.

Adding another rotational component would make overhanging parts possible although gravity could become the enemy.

Even though the following hardware design may not be feasible it does make it easier to explain the additional rotational component: Envision a polar driven "phonograph arm system" [vinyl albums] which pivots the arm and the arm also contains a linear actuator to move the print head fore and aft from the point of rotation. Now add an additional stepper motor which rotates the arm and print head perpendicular to the arm or envision rotating the print head on the end of the arm. This additional rotational component added to the Polar Rep Rap would accomplish the dream of the author of the 4 Axis Mendal topic.

I realize that this exact "phonograph arm system" concept would be difficult to actually implement because of the difficulty in hardware design: slop, accuracy, strength, etc. I have read the other topics under the Polar Rep Rap topic [dual arm and linkages, etc.] which are probably better hardware design but I am explaining using the "phonograph arm system" because it is easier to envision the additional rotational component.

Obviously this would also make the mathematics and control software more complicated. Converting a part design from XYZ Cartesian Coordinates to polar coordinates should be somewhat straightforward but that is easy to say from someone who is not actually having to do the work.

I'm not a robot builder, but i follow Rep Rap, as i think its THE future.


I think overall to really create complex parts, the main base on which your objects stands should be more complex.
It should be able to move in XY (the printer may move in Z).
the platform should be rotatable around Z (you can print circles but a rotating platform is more precise).
Besides rotating around Z it should be rotatable around Y axis.

The reason for this is if you would print a human starting from the feed in upward position with its arms hanging.
One wouldnt be able to 3D print the hanging arms.
You might think then well then lets lay this model on his back, but then the same problem would happen for the fingers.
The solution would be printing the shape upside down, standing on his head.
So the male shape has a solution but other shapes dont have a solution for this.

To be universal - do it all machine
I think the best way would be a floor part that can move XY rotate Y and X, and be able to grab the object to take it into a different position..hands would be ideal for this,..but robotic arm / hands might be a bit too complex...
Or the oposite, the object stands still but a highly agile arm prints in any direction at any xyz location.
Its not simple, and its i gues more likely cheap models wouldnt be able to do this.
Its also very complex since moving means often getting of course, how to keep track of exact positions. (pulse lasers for local gps like positioning?)

I definitely think this forum has too many mechanical engineers and not enough chemists. Surely it would be easier to simply have 2 print heads riding the same x carriage, one printing your part and the other some sort of soluble, or lower melting, or otherwise removeable support medium. Hell print wax as the support, then you drop it in boiling water for a few minutes and use It again!

This rotating your build platform on he y axis stuff sounds insane!

you can have multiple extruders. its up in thingiverse

and you can print a human upside down, starting from the head, just the finished part will be upside down on the platform.

and spinning the extruder around Y wouldn't work half the time; another part of the carriage or extruder would get in the way*.

current high precision ideas include magnetic linear encoders and optical mice*.

IMO, the extruder should move in all 3 axes relative to a non-moving part. the reasoning behind this being that moving the part would require a machine larger than 8 times the build envolope (assuming part moves along all 3 axes). lets take a darwin versus a cupcake. the darwin accomplishes a larger build area:actual machine size*.

* wild theories/from what I gather from memory of thingiverse comments