LEGO Technic Chemistry Lab - The LEGO Molecular System

[Didumos69]

Yet another. Sorry, I can't stop .

[Didumos69]

This can actually be build, but so far I can't get a pattern out of it .

[agrof]

That part is my toy as well recently, too bad, that there is no matching 2 ended connector with 120° angle... I try to figure out some nice abstract molecule.

For the core You built (), I would only use this part: , and use only the external axles as connection. There must be an outcome, as all the angles are 120° in each directions.

EDIT: and here we go, LDD doesn't allow to connect them properly, but I see no reason, why it wouldn't work with real parts. Maybe Stud.io are more forgiving, and all connectors can be properly placed (even the black one in the core). Actually it can create 4 connections, but just tried, at least 3 connections works for sure all the time, leaving 1 free electron always. Choosing which one is free: one (or three? - haven't tested yet) more possibility for genesis, although it will be the pretty same structure in the end. Basically this atom has more reactivity potential to build up molecules.

Edited April 6, 2017 by agrof

[DrJB]

8 hours ago, J_C said:

This is turning into the most interesting topic on EB! Thank you for sharing. This made me to pull out old chemistry textbooks and start to read those again.

Agree fully, this is really addictive. Reminds me of when fractal and self-similar geometries started coming out ... Keep at it!

[Didumos69]

2 hours ago, agrof said:

EDIT: and here we go,

It still connects to 6 neighbors just like all the others so far and it troubles me that it doesn't fit perfectly, but it's intriguing nevertheless .

I finally managed to make one that has more than six connected neighbors; in this one a fully embedded molecule has 12 connected neighbors, each one connected with two connections:

[agrof]

Yummi, that is beautiful organic molecule!

I forgot this part exists: , it slides perfectly in place even at the tight space (the black one).

So it is much nicer now, with proper connections (still 1 always free as described), so far is my favorite!

EDIT - DISCLAIMER: 

This topic escalated much further than expected, so I decided to open a dedicated folder for my files. I updated all the links and pictures, but in quotes they will be invalid. Sorry for the inconvenience.

Edited April 6, 2017 by agrof

[ColletArrow]

This is an impressive thread. I can't think of anything right now, especially since I recently uninstalled LDD (it was distracting me from revision!), but when I get it back I may have a play with these myself. And the mentality of this certainly fits with the first line of your signature, @Didumos69!

[agrof]

Here, I tried to something more complex, and also protons  and neutrons  are represented.  (Still not correctly, as they are not equal in quantity with "electrons".)

I don't know how far can this structure be extended in the width, but for sure this module works...

... and can be stacked limitless in height / length, I just named it as protein.

I like how it turned out,  each module has double hexagonal connections horizontally, and also the layers are double connected vertically (one inner, one outer also hexagonal), so should be a strong structure.

EDIT: I could add two more row to the perimeter and keep it stackable (2 modules shown below), here is the LDD file for it.

 

 

Edited April 6, 2017 by agrof

[ColletArrow]

That's a really surprisingly versatile combination for what initially looked a bit odd. Good find!

[agrof]

Yes, and there are some more combinations, example.

The "secret" is, to use the same and/or multiplied angles. Here is used 120° and 180° - both are multiples of 60°.

On @Didumos69 neighbourhood there is 45° and 90° connections, previously we tested only 90° than 120° ones.

[Didumos69]

Interesting ideas @agrof, I will study them a little better after the weekend.

I hope to see more!

Edited April 7, 2017 by Didumos69

[aeh5040]

Awesome work, folks!  I particularly like this one.  Interesting that it consist entirely of squares, of two different sizes.

On 4/6/2017 at 7:59 AM, Didumos69 said:

 

[aeh5040]

I have a new challenge relating to this one, in the spirit of Didumos69's signature(!)  Notice that the basic unit has 6 unused pin holes, one in each connector, and these holes are accessible from both sides (unlike the others).  Is it possible to arrange several offset copies of the sparsified lattice (with appropriately chosen axle lengths) so that all these holes are filled (by axles from other copies going through them)?  Each axle would have to go through two holes.

On ‎4‎/‎5‎/‎2017 at 9:58 PM, aeh5040 said:

 

Edited April 12, 2017 by aeh5040

[DrJB]

Here is a periodic repetitive structure, built with only 2×4 bricks. Not as elegant as technic connectors, but uses a single part. This is on display at Chicago's MSI (Museum of Science & Industry)

 

[aeh5040]

On ‎4‎/‎7‎/‎2017 at 6:21 PM, aeh5040 said:

I have a new challenge relating to this one, in the spirit of Didumos69's signature(!)  Notice that the basic unit has 6 unused pin holes, one in each connector, and these holes are accessible from both sides (unlike the others).  Is it possible to arrange several offset copies of the sparsified lattice (with appropriately chosen axle lengths) so that all these holes are filled (by axles from other copies going through them)?  Each axle would have to go through two holes.

 

I have not succeeded in doing this, although I feel fairly sure it should be possible. Come on folks, help me!

Edited April 12, 2017 by aeh5040

[DrJB]

I tried to but could not visualize what is happening with the picture. The axles are 'offset' one stud from one cell to the next, correct? Maybe an LXF of the basic 'element' could help.

 

Maybe two exact lattices 'intertwined', just as we did the 'intersecting octahedra' in the sister-thread?

Edited December 17, 2017 by DrJB

[aeh5040]

5 hours ago, DrJB said:

I tried to but could not visualize what is happening with the picture. The axles are 'offset' one stud from one cell to the next, correct? Maybe an LXF of the basic 'element' could help.

 

Maybe two exact lattices 'intertwined', just as we ded the 'intersecting octahedra' in the sister-thread?

The offset from one unit to the next is 1/2 stud in one direction and 1 1/2 in another direction (and the third direction depends on the axle length of course). I'll see if I can make an lxf, although LDraw is my usual medium.

Intertwining does sound like the right idea, but I think there may need to be more than two lattices. One thing is for sure - each axle needs to end up going through exactly 2 further pinholes. This is because the number of units is 3 times the number of axles but each unit has 6 unused holes.

I tried offsetting copies in Mlcad, but it just got too confusing - I could not see war was happening nor whether things were colliding.

Edited April 13, 2017 by aeh5040

[DrJB]

I can view LDraw files too (just can't make them). Whichever is faster. All I need is a single 'node'.

 

Please ignore request. I somehow thought it was the same axle going through. I got it now. ScreenShot and LXF file below. Next, we need to replicate and connect ...

sqr-lattice.lxf

 

 

Edited April 13, 2017 by DrJB

[DrJB]

I think I got it ... how about this? I used 8M axles with end stop for ease of positioning in LDD. LXF file here

I essentially built two lattices then positioned them so the all pinholes are filled (It took more work in real life as LDD is finicky during alignment of large sets). There might be more elegant versions than this though.

 

 

Edited April 13, 2017 by DrJB

[aeh5040]

That looks very interesting, but it's not quite what I meant, although perhaps it is a helpful step.

I want each axle to start and end in an axle hole (as in the basic lattice), and pass through the other pin holes on the way.

It's possible what you've done could achieve this (in a much sparser way than I was imagining). I think this amounts to the question: if you continue every axle beyond its stop, will it meet another stop end-on before it runs into anything else (including another axle)?

[roland]

If you want to generate (huge?) repeating patterns using a single submodel you might want to look into scripting using my LDCad.

edit: I could write an example if you're interested.

Edited April 13, 2017 by roland

[DrJB]

8 hours ago, roland said:

If you want to generate (huge?) repeating patterns using a single submodel you might want to look into scripting using my LDCad.

edit: I could write an example if you're interested.

yes, please.

[J_C]

10 hours ago, roland said:

If you want to generate (huge?) repeating patterns using a single submodel you might want to look into scripting using my LDCad.

edit: I could write an example if you're interested.

Yes, please.

I love using your LDCad. I did not try scripts there. I would be very interested in some example. Thanks in advance.

[roland]

I'll try to setup something useful this weekend.

[roland]

did a quick test using this:

 

as a node, and this script:

function onRun()

  local ses=ldc.session()
  if not ses:isLinked() then
    return
  end
  
  local xDiff=120
  local yDiff=-69.282
  local zDiff=120
  
  local xCnt=10
  local yCnt=5
  local zCnt=10
  
  local subModelName='node.ldr'
  
  local sel=ses:getSelection()
  sel:remove()
  
  local posOri=ldc.matrix()
  local sf=ldc.subfile()
  
  for x=0,xCnt-1 do
    for y=0,yCnt-1 do
      for z=0,zCnt-1 do
        posOri:setPos(x*xDiff, y*2*yDiff+genTools.IF(x%2==0, yDiff, 0), z*zDiff)
        sel:add(sf:addNewRef(subModelName, 16, posOri))
      end
    end
  end
end

This is a very simple one, node might not even be confirm the 'rules' in this thread ?  But it being an examples and all :) Things will get more complicated if the node needs rotation etc.

It generated 500 nodes (6500 parts)

zipped test model + script  (needs LDCad 1.6 Beta 2):

http://www.melkert.net/action/download/patScript.zip