RP-ML readers are familiar with the additive and subtractive methods of
RP: respectively, the mass of the part increases or decreases during the
process. Logically there must be a third method where the mass stays the
same. The process would then be one of internal reorganization. Call
this internal RP.
Surely this is a famously complex category, as much of biology could be
viewed as internal RP.
Yet the simplest internal RP would be just a binary process: an
initially entirely solid plastic lump partitions itself into regions of
solid and void. Inherently this is a growth process. Internal voids
first form at minute size, and then grow to their final size and shape.
The mass of the part is constant, its volume increases as the form takes
shape.
The quintessential example is a kernel of popcorn. As it comes,
unpopped, we can think of it as a roughly spherical mass of void-free
thermoplastic.
Imagine popping a kernel in slow motion. Its shell ruptures and voids
appear within the mass. The voids grow as they collect steam that
diffuses to them. This inflation gradually transforms the initially
simple outer shape into something quite complex, imitating somewhat the
biological process of growth and differentiation.
Eventually the growth slows to a halt as steam formation cools the
thermoplastic to the point of stiffening, and the puffed kernal holds
its fabulous ultimate form.
It might seem that there is little we could control in such a
self-directed process. Yet there is one thing we can control entirely:
the initial location of all of the voids.
In classical theory a void (bubble) needs an infinite overpressure of
vapor to form in a defect-free liquid: there must always be some
nucleating defect. Larger defects nucleate bubbles at lower
overpressures and thus starve off smaller nuclei near by.
With a suitably pure material and a suitable density of nucleation we
can be guaranteed that bubbles will only form at man-made defects.
How might we program the defects? One method would be to put defects on
the surface of a thin tape that is then wound tightly in a roll and
heated in an oven to become a solid billet.
A more direct method would be to converge laser light to a focal point
below the surface of a material that is transparent, yet slightly grayed
with dispersed carbon particles. Isolated carbon particles in the focal
zone will get hot enough to locally decompose the resin into trapped
bubbles of vapor.
As with living things, it will not be possible to just pick any
arbitrary shape for what we make--the shapes will come partly from us
and partly from the laws of nature.
It may require a certain leap of imagination to be interested in this
method. Whatever we can make this way will certainly be different from
the mechanical kingdom we have now.
Jim Mallos
For more information about the rp-ml, see http://ltk.hut.fi/rp-ml/
This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 23:02:43 EEST