In place assembly

From apm
Jump to: navigation, search

One can mount crystolecules into products by

  • either mounting them right in the final place where they are supposed to go
  • or pre-assemble some bigger parts and only afterwards mount those bigger pre-assembled parts in.

If cleanly separated ...

  • exclusively the first process would be done in the crystolecule assembly level and
  • exclusively the second process would be done in the microcomponent assembly level right above it
    (with significantly bigger manipulators - maybe ~32x)

... but practical reality might look different and blur the borders upwards such that a bit of the second process might also be done in the lower crystolecule assembly level.

Assambly level overlap - diminishing and vanishing at the bottom

Actually going down lower the assembly levels the overlap might first diminish and does then fully disappear. Going up towards the higher assembly levels more overlap might happen. Why?

In the bottommost assembly level:

  • (1) One wants to focus on standard elements (the standard crystolecules) because this level necessarily is highly specialized and barely programmable. Bigger assemblies quickly become extremely varied and nonstandard.
  • (2) The necessity for quicker movements (due to the larger size gap between base parts - that is atoms - to product parts - that is crystolecules) and stiffer manipulators (molecular mills) does not allow the geometric freedom to assemble the freshly mechanosynthesized with the same mechanisms (specialized mills do not have enough DOFs and range of motion and the gripping capabilities to do more complicated part assemblies)

As a side-note the second point also applies to todays (2018) 3D printing technology where
3D printer mechanics would be very bad for pick and placer part assembly (and vice versa).

Blurring upwards and downwards

  • Blurring upwards: The upper assembly level doing what actually the lower should do.
  • Blurring downwards: The lower assembly level doing what actually the upper should do.
    This links to the old outdated molecular assembler concept. Where the idea was that they would produce everything in place

with just one (or at best two) assembly levels involved. The main issue here (in this context) was inefficiency (the bigger the product gets the less surface you have left to work on even super complex dendridic retraction retraction channels don't solve the problem (See: "Fractal growth speedup limit") in case one really wants a fully solid block of product. Additionally there is the lack of specialization to standard parts that slows things down.


In todays 3D printing technology there's also the tendency to blur assembly levels downwards. This is simply because as of yet (2018) a second assembly level (a pick and place manipulator assembling pre-printed parts robotically and fully automated) above the bottommost one (the 3D printer) has not yet been introduced.

A concrete manifestation of the downward blurring of assembly levels in the 3D printing space is the hype for "in place printing" aka "printing complex assemblies in one piece". While this may work for some designs its important to recognize the limitations, and that in some cases (e.g. when there are very many parts - like in 3D printers themselves) it's actually just an inefficient hack to compensate for the lack of a higher up assembly level that is not called human hands. Plus there are issues with big necessary clearances and pretty much unautomatable support removal wich make getting well tensioned highly stiff assemblies very difficult (to say the least).

In short: In place printing sucks for any halfway serious machine design.

In place mechanosynthesis of bigger interlocking assemblies of crystolecules may face similar hardships as in place 3D printing.

Concrete discussion of the blurring of assembly levels

Blurring assembly levels downwards

  • In place mechanosynthesis:
    The mechanosynthesis assembly level is invading the dominion of the crystolecule assembly level above.
    This might not be such a good idea.
  • In place crystolecule assembly:
    The crystolecule assembly level is invading the dominion of the microcomponent assembly level above.
    This might be useful depending on the application case.
  • In place microcomponent assembly:
    The microcomponent assembly level is invading the dominion of the product(fragment) assembly level above.
    This is very likely a good idea. After all this is the eraliest point where convergent assembly can be stopped for a practical nanofactory. So if the convergent assembly hierarchy is indeed stopped here, there is simply no other option than (massively parallel) in place assembly.

Note that out of place assembly is the norm.
That is: In case the assembly levels are properly separated and there is no blurring, one has out of place assembly.

Meta-side-note: "Downward-blurring" is used here because what should be done at a specific assembly level is shoved down to the assembly level below. This may not be the best choice for naming. (wiki-TODO: check pros, cons, alternatives, ...)

Blurring assembly levels upwards

  • The crystolecule assembly level doing mechanosynthesis:
    This is maybe useful for very rare special parts. Or for fine after touches on standard parts.
  • The microcomponent assembly level doing crystolecule handling: ... Maybe this is not so useful?
  • The product(fragment) assembly level doing microcomponent handing: ??

Related