Applicability of macro 3D printing for nanomachine prototyping

From apm
Jump to: navigation, search

Q: Is there any chance that results …

  • … form prototyping at the macroscale (with plastics or metals) could be eventually transferable …
  • … to nanoscale target designs (out of gemstone like compounds) …

… without needing to completely restart from scratch?

And that despite of …

  • … 3D printed plastic (or metals) for macroscale prototyping and …
  • … macroscale physics …

… being very different from …

… respectively.

In other words:
Q: Is there any point at all in attempting some macroscale prototyping of nanoscale machinery that we cannot yet build (state 2021 …)?
Doing some future-backward development and attaining some more theoretical overhang.

A: This page comes to the conclusion that …

  • … yes, results of macroscale prototyping accessible today should be transferable to advanced cog-and-gear nanomachinery once it becomes experimentally accessible and …
  • … yes, some eventually useful future-backward development is should really be possible here.

Well, given that the known (and on this page summarized) constraints are known and understood by the designer.

This page is not only about

For more on that see this pages section: #Areas of design

Basic check of applicability – by employing conservative design

The critical question to ask here is:
How can we say that a nanoscale target design will work, given that we get a macroscale prototype to work?

The answer in one term is: Conservative design.
That is: We must not employ some characteristics of physics that is …

  • … available at the macroscale but is …
  • … not available at the nanoscale.

So, where would be the areas where we'd run risk on using
performance properties of actual macroscale prototypes that
would not be available at the nanoscale?

Spoiler: We won't find much.

For the sake of finding that out it is useful to construct:

  • (A) imaginary macroscale machinery with the (high) performance characteristics of gemstone based cog-and-gear nanomachinery.
  • (B) imaginary nanoscale machinery with the (low) performance characteristics of macroscale prototyping machinery

Construction of these imaginary machineries for conservative design can be done by employing

To be on the safe side actual prototype machinery must strictly under-perform the imaginary machinery in
either (A) or (B) – one implying the same for the other.

Especially performing the exercise of (A) makes it painfully obvious that
Macroscale style cog-and-gear machinery performs significantly better rather than worse when
operating under the conditions of nanoscale physics rather than macroscale physics.
Even if we can barely experimentally test it with current day experimental limitations.
Related:

Conservative choice of material for prototyping

Conservative design regarding relative deflection amplitudes

The goal: Keeping acceleration induced relative deflection amplitudes of macroscale prototype designs
from becoming smaller than what nanoscale target designs could ever achieve.

The choice of macroscale prototyping material must be such that
it does not perform better against suppressing relative deflections from accelerations,
than an eventual nanoscale target design eventually can.

While material stiffness falls massively with making machinery smaller
(making even diamond softer than the softest jelly in terms of stiffness)
the forces from dynamic accelerations and static tensions scale exactly the same way!
(thus that ultra soft "diamond jelly" gets barely deflected)

Quantitatively both stiffnesses and forces scale quadratically.
That is: Halving size of machinery quarters both

  • the stiffnesses and
  • the forces (from both dynamic accelerations and static tensions)
  • (side-note: forces from gravity and magnetism drop much more rapidly)

So with stiffnesses and forces both falling equally fast
in terms of resistance against relative deflection
our imaginary macroscale machinery with nanoscale physics properties
(that we must not exceed in performance with our macroscale prototypes)
would perform exactly identically to macroscale prototype machinery made out of the same material.
In short: Relative deflection amplitudes stay the same across all scales. They are scale invariant.

This effectively means that:
In order to not get any false positive macroscale prototypes
our macroscale prototypes must not come close to or exceed the stiffness of hard gemstones.

This is easy to adhere to.
It's fulfilled for plastics and even metals as prototyping material.
Regarding acceleration induced relative deflection amplitude concerns
everything macroscale prototyped with 3D printed plastic
will work with absolute certainty when transferred over to gemstone at the nanoscale.

Note: All of the above should apply equally to ultimate material strengths.

Conservative design regarding resonances (ringing) and its amplitudes

The choice of macroscale prototyping material must be such that
it does not perform better in suppressing mechanical resonances (ringing)
than eventual nanoscale target designs will ever be able to.

Atomically precise gemstone structures typically have a very high Q-factor. Well, unless deliberately designed against this.
3D printed plastic in macroscale prototypes typically has much much higher damping (lower Q factor) as the baseline.
So one might worry that the prototypes might outperform (in terms of damping) what target designs could ever offer.
Especially for rather fast moving elements this might be a concern.

But then again:

  • Deflections amplitudes in plastic prototypes are much much bigger due to the much much lower material stiffness of plastics compared to gemstone. And ...
  • Resonance amplitudes are also largely amended/alleviated by the desired deliberate slowdown at the lowest assembly level for reduction of frictional energy dissipation.

– Slow assembly motions at the smallest scales likely will allow for designs that avoid overswinging and ringing all-together.
– Faster transport motions and and faster mid-size-microscale (re)assembly motions may be more critical.

Bending

What about bending?
We have superelasticity in the nanoscale target design.
So for macroscale prototyping we must not chose materials that are reversibly bendable for more than >10% bendable.

This is pretty easy to adhere to.
Just design clips and such that they don't exceed this massive strain values.

Leaving conservative design thoughts:
It's rather difficult to exploit even a small part of this potential in macroscale prototypes.

  • Many 3D printed plastics are brittle and like to break on much less bending than 10% strain
  • Materials that do allow repeatedly for 10% strain are typically rubbers with abysmally low stiffness
  • Many 3D printed materials don't like to be put unter permanent strain => material creep and stress cracking

What comes closest in macroscale technology to crystolecule superlasticity may be nickel titanium alloys (aka nitinol – also called superelastic).
But that is not suitable for maximally cheap prototyping on hundreds and hundreds of parts.

Effects of lack of material creep

Structures under permanent tension-load or permanent bending-load may be a lot more problematic in macroscale prototype designs.
This is especially problematic prototyping systems that as a whole are under a permanent baseload.
Concrete example: The ReChain frame systems.

Presence and avoidance of overengineering from excessive conservative design

Overengineering in macroscale prototypes can come from:

  • excessive conservativeness
  • fundamental limitations of the physics of the macroscale – (sometimes circumventible by careful cheating)

Such overengineering may lead to designs that:

  • would work perfectly well, but
  • are not exactly what one would want to go for

So taking the conservative approach to far
(which is not much of a constraint anyway, as we've seen above)
is not a good strategy either.

Effect of the lack of superlubricating sleeve bearings at the macroscale

Due to:

Bearings in macroscale prototypes may not be implementable in the same
compact and elegant sliding sleeve bearing fashion as in the target systems.

  • Macroscale ball bearings are a bad match for nanoscale systems (Just as they are for 3D printing) but
  • cone roller gear bearings are a good match. While perhaps excessively overengineered they would certainly work at the nanoscale too.

Alternatively for prototyping one might want to opt for
normal ball bearings as stand-in make-believe vitamins.
That are supposed to be replaced with superlubricating sleeve beraings in the nanoscale target designs.

Overall faking superlubricating sleeve bearings with as few as possible standard bearings may be the best choice??

Effect of the lack of the VdW force at the macroscale

VdW forces are not available in macroscale prototype designs. That is a huge problem.
Adding form closure to all pick-and-place interaction leads to significant over-engineered designs.
Given presence of VdW forces is thought about properly,
Fully form closing system designs would certainly work without changes though.

Forces from gravity and magnetism
(which both are not available in the nanoscale target designs, and thus should not be relied on according to basic conservative design rules)
can be (ab)used to qualitatively fake VdW forces to some degree.
There are some complications though.

Complications include:

  • Gravity is anisotropic, and constant – very different to VdW forces which are isotropic and very short ranged
  • Magnetism has polarity, a still a somewhat different force falloff characteristic, and requires magnets as vitamins
  • Magnetism does not allow for static levitation (Ernshaws law) whereas this is kinda possible with VdW forces

Using electrostatic forces or surfaces tension forces to fake VdW forces may be an option but would require
smaller prototyping than what is possible with filament based 3D printing.

Overall faking VdW forces with magnets may be the best option??

Summary of conservative design considerations

Summarized conservativity check:
Note that what we want is for the prototype to be worse in its limitations than the target design. Across the whole board.
This is limiting but makes it much more likely for the target design to actually work with little to no changes.

Unlike the materials of the nanoscale target systems
the macroscale prototyping materials (e.g. plastics) have …

  • … much lower resistance again deflections – CHECK
  • … no superelasticity – CHECK
  • … material creep – CHECK
  • … wear damage – CHECK
  • … no superlubricity – CHECK – this is only friction per unit bearing area – total friction follows the opposite trend – see "overall system" below
  • … no VdW force – CHECK – but this is too much conservative design =>
    => fake VdW force with magnetism and maybe gravity – but don't "overfake" it.

Overall system:

Overall macroscale style machinery just performs way better across the board
when scaled down too the nanoscale and made atomically precise.
Contrary to what is often falsely assumed.

Areas of design

Atom aware bulk limit

An example for a design at the lower bulk limit (a basic gas tight bellow)

FDM 3D printing (FDM ... fused deposition modelling i.e. printing with molten plastic from a nozzle) works nicely in the atom aware bulk limit (that means beside other things exact locations of atoms remain unspecified but the main crystallographic angles and orientations are preferrably used).

General:

  • keeping all surfaces parallel to the main crystallographic planes avoids introduction of irregular steps when the bulk model is refined to an atomistic one
  • the ~45° overhang limit for FDM (glue-gun) 3D-printing -> similar situation in nanoscale => scaled down models stay makeable
  • macro assembly using no sticking force at all only shape locking => assembly of nanoanalogous models may waste energy but will work

Things to note on emulation of nanoscale VdW sticking for macroscale models.

  • Note that there is always a sucking force when two coplanar flat surfaces contact each other. It points toward the sliding direction in which the contacting area grows fastest. (wiki-TODO: add image)
  • Avoid contacting two large planar or complementary in direct approach - (high force energy wasting snap) - try to slide them onto each other instead. (1D to 0D locking can be done near reversibly (wiki-TODO: make a demo))
  • Gravity can emulate VdW sucking but only downwards (anisotropically)
  • Tiny magnets can emulate VdW force (albeit in a very limited way) - additional manual assembly effort
  • Shaking can emulate thermal motion as long as no free energy is extracted
  • macroscale friction (from spring force) to emulate VdW forces - bad idea - very different behaviour
  • macroscale friction (from tight clearances) to emulate VdW forces - very bad idea - hard to predict friction strength
  • macroscale ratcheting (from spring force) - very good idea (this is actually friction on the nanoscale)
    - designs should aim to make all energy recuperable i.e. no click sounds - unlock move lock

Atomistic modelling

Erik K. Drexlers superlubricating "big bearing" - atomistically modelled - This is a photo of a 3D printed model. See: http://www.thingiverse.com/thing:631715

Atomistic models (i.e. where the atom locations are specified and visible in the model) are more useful for visualization advertisement and educative demonstration of principle since the character of the short range force are not accurately recreatable at the macro-scale.

Forces

Magnets can provide a very very crude approximation that may in some cases suffice to demonstrate a principle like e.g. superlubrication, soft force-field gear-meshing, a combination of both or other things. There's also some external work to hands on model some early stage nanotechnology protein folding for FDM printed models. They call their models peppytides. Template:WikiTodo

Looks

Full color gypsum powder inkjet style 3D printers (a service offered by various printing services)
are ideal for reasonably small parts (smaller than what is possible with filament based 3D printers)
that are meant only for visualization. Not for recreation of plausible forces.

Side benefits

Development of specific prototype systems like

might be useful for concrete more or less near term macroscale applications too.

Frames for just about anything:
Quickly re-composable frames for all sorts of sane and insane things

Space applications: Frames and perhaps self replication
Quickly re-composable structures in space made – eventually from laser sintered metal powders (mined from metallic asteroids)

Extending on existing self replication in the context of 3D printing: Self replicating RepRap 3D printers by definition make some of their own plastic parts.
But they do not assemble the parts they make to more RepRap printers. That task is still up to human hands.
Pick and place robotic devicees capable of assembling both:

  • more RepRap 3D printers and
  • more copies of those same devices

Would be interesting at least. Practicality seems questionable.
It does very much not seem that it could fit into the envelope of a typical 3D printer
at least given crude filament based 3D prints as base-parts.

Keep being clear about the ulterior motivations:
A focus in prototyping for nanomachinery will have quite different priorities in the details.
So trying to sell the idea under the context of a different more near term macroscale product
while still keeping the ulterior motive of designing for nanomachinery prototyping is likely a pretty bad idea.
Unreasonable design choices will raise eyebrows.
(wiki-TODO: Thus clear up the ulterior motivations on the RepRap wiki page about the RepRec & ReChain projects)

On open develoment:
A forkable (coopyleft) approach should be the right way to go here.
So if someone wants to develop in a direction with a very different target "product" they can do so.
And improvements there can be ported back.

Related

The author of this Wiki (about) conducts a meta project that aims to build up a collection of 3D-printable 3D-models (mainly in atom aware bulk limit) that will hopefully turn out to be useful in the development and understanding of advanced nanofactories.
See main article: The DAPMAT demo project

External links

Retrieved from "http://apm.bplaced.net/w/index.php?title=Applicability_of_macro_3D_printing_for_nanomachine_prototyping&oldid=15882"