View source for Tensioning mechanism design

In advanced atomically precise products there is a need for [[structural elements for nanofactories|structural elements]].
As ist turns out a good method for constructing large structures from small pasts is to use the reinforcement method that is well known from concrete construction. compressing together a structure with a core under tension.

To reversibly tension an initially loose multi-part structure one needs a tensioning mechanism.
As a superficial investigation reveals there is quite a big design-space for tensioning mechanisms.
To find a near optimal solution for advanced atomically precise nanosystems (e.g. Nanofactories) this pages tries to distill the characteristics of tensioning mechanisms and judge them based on those characteristics.

'''Ńote: These spanners in conjunction with the other necessary elements for a space truss are also perfectly usable on the macroscale.'''
e.g. for the construction of today's (2016-07) 3D-printers or pick place robots.

= Basics =

To get a potential [[trapdoor misconception]] out of the way first.
In the macrocosm friction based self locking is often used to keep a spanner in its spanning position.
For example macroscopic non atomically precise screws, wedges, worm gears self lock by friction.
In the nanocosm though there is [[superlubrication]] in conjunction with loose knocḱing thermal motion (and limited screw lead smallness) which together makes self locking rather unattainable.

Thus other fixation methods need to be employed to keep spanners in their spanning positions.
These methods ultimately are sufficiently strong springs. (Shape locking is possible but this just defers the problem.)

= Characterisation of 1D rebar spanning elements =

The classification criteria listed here are not necessarily orthogonal. That is choosing an option regarding one criterion may severely influence the avalability of options in another.

== Location of the spanner in the ''force circuit'' ==

With ''force circuit'' I'll refer to the closed tension compression loop that is formed by spanner shape lock chain and profile segment shell.

=== double sided  - hull-pusher / chain-puller ===

If the spanned structure is a truss element connecting two point a double-sided spanner that can sit somewhere in the middle is needed.
a double sided spanner can either pull together the chain in the core while keeping the hull length constant or push apart the hull while keeping the core-chain length constant (or do both).

* A ''hull pusher'' keeps the structural length defined by the stretched core chain length. => Length adjustment parts need to be core chain parts. They are harder to access but well hidden and have very limited space for implementation.
* A ''core puller'' keeps the structural length defined by the compressed hull segment stack length. => length adjustment parts need to be hull segment parts. They are better accessible but also exposed and have a lot of space for implementation.

* A double sided simultaneous push pull tensioner is hard to construct and has questionable benefit. The structure length inder tension may become hard to predict thus it does not be usable for tensioning of truss elements that are supposed to keep their length when tensioned. 

=== one sided endpoint - endpoint spanner ===

To just hold bigger sized structures together firmly one sided spanners can be used.
They may be esier to implement

== Linearity ==

* Wedge, screw based and gear spanners are usually linear
* Lever based spanners are usually highly nonlinear

With nonlinear spanners one usually does not need pieces (core-chain or hull) that are adjustable in length under zero load.
With linear spanners it depends on their maximal displacement range whether such pieces are needed or not.
With linear spanners it is easier to set a defined load.

== Mechanical gain and Spanning displacement range ==

In ''minimal size crystolecule design'' screw leads can't be made very low thus simple screw based spanners have rather low mechanical gain. 
This can be improved by driving the screws with wormscrews.
The range maximum displacement range can be big enough to expose the first ''chain disassembly point''

Differential screws can have very high gains but have very low displacement range due to their large differential part displacement.

Singly staged wedge spanners have a rather low displacement range especially if the gain is high (assuming ~constant spanner size)
It's a bit better with doubly staged wedge spanners.

Lever based spanners have theoretically infinite gain at their dead-center position.
{todo|is this an essential singularity or not?}

== Size / compactness ==

Why it  matters:
Usually the spanner should be smaller than the assembly that is supposed to be spanned. {todo|elaborate on reasons}

* Most compact seem to be screw spanners.
* They are followed by worm driven screw spanners and by wedge spanners.
* Lever spanners are pretty bulky.
* Differential spanners are quite big and unconditionally need additional size adjustment elements
* Practical gear spanners seem to be quite big.

== Actuation method - direction and compensation ==

The only forces that a space truss is supposed to take well are forces coaxial to the truss element. 
Shear forces bending moments or torsion twists should be avoided.
Actually it is possible and not a bad idea to avoid axial forces too.
All of these forces and moments can be applied if they are locally compensated by the actuation-element / robotic end-effector that actuates the spanner.

* axial force => axial pinching / axial stretching
* torsion twists => axial contertwist ("two wrench method")
* bending moments => normal direction countertwist ("two wrench method") 
* bending force => normal direction pinching / stretching

The compensated actuations can be applied form one side of the spanner or two opposite sides of the spanner for even more stability at the cost of higher complexity.

Actually many macroscompic spanners do not provide dedicated means for local force compensation.
{todo|add examples}

== Fixation method ==

Since as mentioned before there is practically no frictional self locking in diamondoid AP machine elements a spring is needed to fix a spanner in place.

If a spring is too small and thus have too low of an energy barrier it runs the risk of jumping open by thermal motion.
Since a spanner mechanism of any kind is quite big relative to e.g. the small crystolecules that make up minimally sized chains segments a single spring can be made big enough such that it has a high enough energy barrier to make failure sufficiently unlikely.

There are two ways to include a spring that differ in the way the energy stored in the spring can be recuperated.

The bad one is an inaccessible clip-lock.
This is similar to the clip locks often seen in today's macroscopic plastic products just that given enough force it can be pulled apart again non-destructively (structurally reversible).
The holding force is limited (this is not so much a problem).
If one intends to recuperate the spring energy (force times displacement length / moment times displacement angle) it gets difficult to impossible.
Since the force/moment is not locally compensated it disperses widespread over the structure and the actuation manipulator assembly where it gets lots of opportunity for energy dispersion.

The better one is an accessible clip lock (e.g. something like a cloth pin) having the spring force/moment channeled through a defined path makes the energy reliably recuperable.

== displacement quantisation ==

Using a spring to lock the tensioner in its tensioned state means that tension can't be choosen in a continuum.
Depending on the chosen tensioner style the tension is adjustable in a smaller or greater number of states.
To gain finer locking quantization one can:
* either try to increase the mechanical gain of the spanner
* or try use differential plates (linear / circular)
* or use unloaded length adjustment elements (chain/hull) in conjunction to the spanner

= Pros and cons of different spanner types =

== Wedge spanners ==

* 0 medium compact
* + easy robotic actuation
* 0 linear
* - low mechanical gain
* - low displacement range
* (simple design / hard to assemble ?)

=== Doubly staged wedge spanners ===

* 0 medium compact
* + easy robotic actuation
* 0 linear
* + high mechanical gain
* -- very low displacement range
* (simple design / hard to assemble ?)

== Screw spanners ==

* ++ very compact
* 0 medium difficult for robotic actuation ?
* 0 linear
* - rough spanning adjustment quantization
* - low mechanical gain
* + high linear displacement

=== Worm-geared driven screw spanners '''(PROMISING)''' ===

* + compact
* 0 medium difficult for robotic actuation ?
* 0 linear
* + fine spanning adjustment quantization
* + high mechanical gain
* + high linear displacement
* (complex design)

== Differential screw spanners ==

* - bulky
* 0 medium difficult for robotic actuation ??
* 0 linear
* + very fine spanning adjustment quantization
* + very very high mechanical gain
* - very low linear displacement
* (complex design)

== Lever spanners ==

* - bulky
* - not easy to operate by robotic end effector ?
* 0 nonlinear
* - no spanning adjustment quantization
* + high mechanical gain
* + high linear displacement
* (simple design)
(Lever spanners are very good for macro-mechanical hand operation though!)

=== Eccentric tappet tensioner '''(PROMISING)''' ===

(commonly known as "quick-acting clamp" or "quick release skewer" from bicycles)

* + compact
* + easy to operate by both hand and robotic end effector
* 0 nonlinear
* + spanning adjustment quantization
* + medium to high mechanical gain
* - low linear displacement
* (simple design)

== Gear spanners ==

* -- very very bulky
* 0 medium difficult for robotic actuation ??
* 0 linear
* + wide range of possible no spanning adjustment quantizations
* + wide range of possible mechanical gain
* + possibly high linear displacement
* (maybe complex design)

= Related Parts =

* length adjustment parts (chain & hull segment) <<<
* node points
* intended breakage positions
* unintended disassembly stopping points