< week.9@martin.magal@fabacademy.2013

week 9: computer-controlled machining (Mar 20)

make something BIG

tools

hardware:

  • laser cutter
  • shopbot CNC

    • software:

  • ReconstructME
  • DAZ Studio 4
  • Rhino
  • Autodesk 123D Make
  • Sketchup
  • Autocad
  • partworks 3d
  • shopbot3

    • materials:

  • cardboard, ~2mm
  • plywood - 6mm
  • plywood - birch 16mm
  • expanded polystyrene - 150mm

    • model design

    this time the idea was to make a chair/armchair based on models of human figures posed with different body positions.

    first attempt

    at first i tried scanning myself and another student with ReconstructME. but the resilting model was full of holes. we tried scanning while we were static and anorher student turned around with the sensor. Very bad results, the software would only reconstruct part of the bodies.
    we tried turning me around in an office chair and keeping the sensor relatively in a fixed position (only going up and down), sligtly better results.
    we concluded that probably our PC wasn't up to the task, since calculations are probably done partly on thr graphic card's hardware and ours wasn't good enough.

    second attempt

    i then decided to model the figures in DAZ Studio 4, which starts with a template body named genesis:

    i added another body and posed them to get a posture that would allow a relatively comfort sitting, something likje this:

    i imported the figures into 123D Make, after some fixing in Rhino, and under construction technique i chose stacked slices:

    i decide to make first a scaled model, 1:10, in cardboard, to check the design in a physical form:


    although i liked the design,it had some flaws that convinced me to choose another direction: it was too wasteful in terms of material (it required between 4-6 plywood boards of 1200x2400mm) and maybe too strong or "kinky" in terms of concept.

    third attempt

    this time i went for a more economic and neutral concept, a new table design for our workshop:



    as one can see, i designed 2 options, the first one with the legs more centrically positioned, but probably less stable.
    the second one with the legs more spaced out.

    both designs are similar in the manner in which the 3 legs interlock with each other, which is NOT in a 90 degrees angle, PLUS 3 legs interlock at the same place, significantly weakening this spot.
    i hope the reinforcing HEX above will take care of that.

    in the next batch of pictures you an see the results of a scale model i did with cardboard cut in the laser cutter.
    i don't like the design at all )-:








    ANOTHER model design yet

    this time, a completely different subject: mill a mold of one of the hearing bells of my funal project.
    the mold material will be EPS, that's expanded polystyrene. the thickness of the material was 150mm.
    the mold was to be cast eventually in a composite material.

    the design options

    the first one was too steep, would get in the way of the mill's collet.

    this one will do!

    milling the mold

    lets start with a video that shows just how close the collet gets to the material.
    you must take that into account when designing the model!



    this is the control box. you can see the speed of the spindle, 12,000 rpm.


    here you turn it ON, both the red knob and the key.


    i didn't take into account that the pvc cover of the vacuum dust collector would get so much in the way.
    eventually i had to dismantle it, and most of the milling was done without the benefit of automatic dust collection.
    lets say that i had a LOT of cleaning to do afterwards.


    i hit the STOP button to stop the machine before taking the pvc cover apart, of course.



    the ROUGH phase was done in relatively big steps, to make the job quicker.


    after the ROUGH path is done, you load and run the FINISHING path file.





    you always have to take apart and clean the collet after a job is finished.