Difference between revisions of "Structural elements for nanofactories"

From apm
Jump to: navigation, search
m (tensegrity as subpoint)
(→‎Related: added link to Static rebar profile force circuit)
 
(One intermediate revision by the same user not shown)
Line 48: Line 48:
 
= Sparse space trusses and tensegrity structures =
 
= Sparse space trusses and tensegrity structures =
  
−
In contrast to a space frame a space truss only takes axial forces in the nodes. No torsion moments bending moments and shearing forces.  
+
In contrast to a space frame a space truss only takes axial forces in the nodes.
 +
No torsion moments bending moments and shearing forces occur at the truss nodes.  
  
 
Structures that have no degrees of freedom even when all vertices do not take any momenta (stiff space trusses) are of special interest.
 
Structures that have no degrees of freedom even when all vertices do not take any momenta (stiff space trusses) are of special interest.
Line 90: Line 91:
  
 
* [[connection method]]
 
* [[connection method]]
 +
* [[Static rebar profile force circuit]]
 
* [[Diamondoid molecular element]]s here also called crystolecules
 
* [[Diamondoid molecular element]]s here also called crystolecules
  

Latest revision as of 12:46, 31 January 2017

This article is a stub. It needs to be expanded.


Shape-lock-chain-core-reinforcement

Problem

A main goal for structural elements is to make them from reusable standard pieces. The smaller and the less complex the standard pieces are the more reusable they become.

The problem with just holding a lot of small pieces together merely by Van der Waals Force is that one may loose structural stiffness (TODO: to check quantitatively) and one introduces lots of potential failure points where at elevated temperatures the especially severe thermal vibration can break things up. Clipping connections (not "noisily" connected to save energy) may be a bit sturdier but also increase complexity and size of the parts and thus potentially make them less reusable.

Solution

The solution is to use shape locking as the means for connection. The resulting structures assembled by pure shape locking usually have low stiffness though. It turns out that it is possible to stiffen big assemblies made out of lots of small simple standard parts that are shape locked together by applying the principle of concrete reinforcement.
By ...

  • threading a (indirectly) shape locked chain through a stack of small profile segments
  • preventing the chain from sliding in the stack of profile segments by widening the chains starting point and
  • pulling the chain where it comes out the stack of profile segments thereby pushing back on the stack of profile segments

... one can stiffen a long rod-like structural element with arbitrary profile shape. A profile segment can be as simple as a tube but also have more intricate shape like e.g. a guide-rail. Stacks of wider profile segments can be tensioned by multiple chains - three will often make sense since three points define a plane - four make sense for cartesian symmetry. Special profile segments can be slid on e.g. thinner spacer segments or segments that add connection points for hinges.

To tension a profile stack one could use a wide variety of spanner designs. One possibility would be a simple screw driven by a worm-gear for high mechanical gain. The worm gear needs to be strongly spring locked. In this special design the chain should be non-circular such that it can locally take on the torsional load introduced by the screw.

Reusability

Without shape-lock-chain-core-reinforcement one needs many many truss-elements types of a very fine grained length spectrum. If an old macroscopic product is to be recycled in a new one where most of the truss lengths don't match the parts can't be reused.

When shape-lock-chain-core-reinforcement is used instead there is no longer a need to have a variety of unwieldy truss elements. Instead one only has a lot of small passivated crystolecules. And those are much more likely to be reusable in a new very different makro product. (See: Recycling)

Also long thin high aspect ratio truss elements may be a bit harder to handle robotically than small compact ones in a narrow size range.


Shape locking and spanning drive chains is an other issue (machine element).

Dense 2D-slabs and 3D-housing-blocks

Profile segments can be shaped such that when they are spanned in one axis they will pull together blocks in one or two other axes. This way one can fasten a huge block of very many quasi loose parts (held together only by VdW force) by only threading through tensioning chains in a single direction. Note that the pacts can not only be enclosed in the directions normal to the spanning direction (which obviously is easy) by they can also be stiffly spanned together in these directions. This is especially useful for advanced atomically precise products on the high performance end of the spectrum.

Sparse space trusses and tensegrity structures

In contrast to a space frame a space truss only takes axial forces in the nodes. No torsion moments bending moments and shearing forces occur at the truss nodes.

Structures that have no degrees of freedom even when all vertices do not take any momenta (stiff space trusses) are of special interest.

In this group falls the set of deltahedra (all faces are equilateral triangles). They can be intuitively explored by the usage of the popular geomag construction toy.

Vertices surrounded by six coplanar equilateral triangles have a different weaker character than other vertices (TODO: check if those are soft modes)

The tetrahedron is the most simple deltahedron. Note that:

  • Tetrahedra alone cant be stacked to build up linear non wavy trussworks.
  • Tetraherda cannot be used to fill space. This is what happens if you try: (link to wikimedia commons image)

In both cases octahedra need to be added.

  • stiff icosahedra can incompletely but infinitely tessellate space (leaving ok gaps) - they form rhombic dodecahedral super-structures. Octahedra can be interleaved.
  • (TODO: What are further highly symmetric stiff deltahedron based infinitely tessellatable truss structures?)

Truss nodes

Nodes of a space truss must not put any moments or forces on the connected members. Thus when the members rotate their virtual extension lines must always run through the nodes center point. Simple naive off center hinges do not provide this functionality. A proper truss nodes thus needs to be designed a little more elaborately such to allow two layers of different radius cylindrical sliding around the common node rotation point oriented normal to each other (combined this is spherical sliding).

When tensioning a segmented truss member the connections to the node points should be included in the force cicuit such that the node menḿber connections do not introduce uncontrolled lash.

A fully interlinked octett truss has twelve truss members pointing to every node (along the 110 directions) this can become quite crowded. Normally a less over-restrained fractal truss structure is preferrable.

Tensegrity

Non-redundant tensegrity base units are single point failure - if one element breaks the whole structure collapses. To avoid this redundant elements can be added. Bigger tensegrity structures can consist out of many independent base units. For stiff structures so called "soft modes" need to be avoided.

There are hollow spiral n-gonal prismatic structures that may be suitable as frames.

Related

external links

Retrieved from "http://apm.bplaced.net/w/index.php?title=Structural_elements_for_nanofactories&oldid=5426"