|  
    (1) 
A faster procedure to glue and cut blocks of scintillators is to
      assemble them first into 50+cm assemblies, held by clips, and then
put glue along the centers of  the sections, and then cut them apart.
 
        tray_new.stl
 
         tray_new_cap.stl
 
      |   
   | 
    |  
    (2) 
Cut files (dxf):
 
cutfile_a
 
cutfile_b
 
cutfile_c
 
cutfile_d
      |   
   | 
    |  
    (3) 
Put a stripe of gel superglue next to each clamp.  
         Cure for  few minutes
      |   
   | 
    |  
    (3) 
Cut with on the Boss LS 3655 laser cutter, at 95% power (of 150W)
at 10 mm/s
 
Outgassing outside
 
   
      |   
   | 
    |  
    (4) 
Start of the 48x50 cm panel.
 
We need a stiff, lightweight base:
 honeycomb core panel
 
  
    How to cut carbon-fiber honeycomb panels
      |   
   | 
    |  
    (5) 
Tray for layer A. The prototype is 48x50 cm, and will be printed
in sections, since the printer has a working area of 20x20 cm.
 
  STL for layer a bottom left
  STL for layer a bottom left
  STL for layer a bottom left
      |   
   | 
    |  
    (6) 
The cross ridges hold the bottom bars in place. The fibers
from upstream bars pass underneath. This is better than trying to hold the
fibers down in the channels with tape, as I did in the previous prototype.
      |   
   | 
    |  
    (7) 
Bridges in red. They are 1/4 of the width, the tray sections 
     (grey, green, yellow) are ~1/3 of the width (cuts at 16, 32 cm from 
     the origin.
 
 
 
 layer a bridge 6.0 cm
 
 layer a bridge 12.5 cm
 
 layer a bridge 5.5 cm
      |    
   | 
    |  
    (8) 
Scintillating bars are glued in bundles of 16, then laser-cut
into blocks. 4 layers (A,B,C,D) × 6 blocks long × 12 blocks wide
× 16 bars = 4600 channels for the 48×50cm prototype.
      |   
   | 
    |  
    (9) 
Beginning construction of layer A; The trays are mounted on a 1/8" honeycomb core panel.
 From the left 3, 6, 9, 12.5, 14, 5.5cm blocks.
The even blocks are raised on bridges. The bridges for 12.5 and 5.5 are 
supported by mylar, which lays over the grooves where the fibers go.
      |   
   | 
    |  
    (10) 
Removing the '6' block shows the bridge. It is shorter than 6cm to allow
the fibers from '3' to bend down into their groove.
 
Removing the bridge shows 2 fibers going into their groove:
 
 
      |   
   | 
    |  
      (11) 
	Middle and right sections done, with small tweaks: ridges to keep blocks
	from sliding
	sideways increased from 0.3→0.6 mm. Added mising ridge at 3cm.
	 
       WLS fiber connector
     
	   | 
    | 
      (12) 
	New connectors: there is no reason to maintain the triangular pattern
	in the connectors as I did before. The piece that joined the two in the
	previous prototypes was too weak and tended to split. Here I line up
	all fibers, spaced 2mm apart. The total width
	of the connector is 39mm (to fit at 40mm pitch), and the walls of the
	joiner are sturdier. The joiners stack on top of each other, with a
	height corresponding to the total height of a layer (11.9mm). You can
	see half-circular cutouts, so the stacks can be screwed down.
       
	linker.stl
	 
	connector_v3.stl
	  |  
	      
   | 
    | 
      (13) 
	I printed a few to see how the fibers might want to be routed. Here
	you see some  A input conectors, some with
	wls fibers coming in, joiners B, some stacked.
	and 
	C output conncetors, no clear fibers attached.
	The wls fibers pass through the open space on the top of the joiners,
	which makes the assemblies want to be tilted.
	 
	Top view.
	 
	Note that we want 6 of these assemblies, here 3 are shown.
      |  
   | 
    | 
      (14) 
	For the next round, I will therefore tilt the assemblies, and also
	add open space below for the outgoing clear fibers.
       
	Next, flip the tilt angle, so that the wls fibers come in along the bottom. This way you can assemble the scintillators, insert the wls fibers and connect them to the joiners. Later, you can connect the outgoing fiber connectors.
	  |  
	  
    
       
       
    | 
    | 
      (15) 
	Tilted by 8°, for a ~5cm pitch between the six connector groups.
       
	connector_angled.stl
      |  
   | 
    | 
      (16) 
	If the connectors are tapered as shown, there is just enough space to route the 1mm fibers along the bottom. The gap there needs to be 3mm (once), 2mm (twice) and 1mm (twice). The last one needs no gap. 
      |  								    | 
    | 
      (17) 
	The layers are paired (AB, CD) and the pairs are spaced 5cm apart.
	 
	Even though the connectors and fibers now fit inside of 12mm, I cannot
	get my fingers in between layers A and B to connect output fibers
	(same for C,D).
       
	Note: the y-scale is stretched by ×5
      |  
   | 
    | 
      (18) 
	But if I flip A and C, maybe there is finger space to connect outgoing fibers to layers B and C, but likely not.
	 
	So during installation, one would mount AB first, and connect outgoing fibers on the two sides, next mount CD in a temporary position such that there is enough space to connect the outgoing fibers to surfaces C and D, and finally move CD into the final position close to AB. 
	  | 
	  
  
	   
     | 
   | 
    | 
      (19) 
      Back one step: How to connect wls fibers to the tilted joiners? In
      the top drawing, you see the 1mm wls fiber needs to go from the far end
      of the scintillstor bar into the connector which inserts to the center
      of the joiner, which needs to be mounted to the board. The fibers are
      stiff, and come in groups of 16, so I can't'
      back off the connector in order to insert it into the joiner.
       
	[A] 
      If I split the support board, and separate the parts  so that I can
	connect the  fibers, they get pulled out several mm. Then I have to
	slide them back together. But there are 6×16
  fibers, and I can probably not slide all of them in at the same time.
       
  [B] 	
  So I leave the joiners loose until the fibers are inserted, and mount them
  onto the board later.
  
	  |  
	  
       | 
	| 
	    (20)
	 |  
	  
       | 
	| 
	    (21)
	 |  
	  
       | 
	
	    (22)
	    Start test assembly:
	      - cut 16 fibers to length
	    (see 13 here)
	      - epoxy fibers into connector
	      - print new polishing blocks
	      - polish the ends
	    (see 14, 15 here)
	    
polisher
	  |  
       | 
	| 
	    (24)
	    First block:
	    remove the upstream connectors (this works only because the nylon screws bend).
	    
	 |  
   | 
    | 
    (25)
You have to pass the fibers (in order) underneath the bridges. This is tedious.
In the process I discovered that 2 of the 16 fibers were 1.1mm, and the other
14 were 8.8-9.2mm diameter. The thicker fibers will not pass, and I had to
cut them off.
	    
	 |  
       | 
	| 
	    (26)
	    The connector is inserted into the joiner first, and then
	    screwed down.
	     
	      Note the 2 cut fibers.
	  |  
   | 
    
      (27)
       
      This is how the assembly of one 16-fiber row of blocks goes:
       
       A: 16 fibers for block 1 get routed under
      the bridges in the base plate and inserted into block 1. Screw down
      connector 1.
       
      
       B: Fibers for block 3 get routed and inserted
      in block 3, connector 3 screwed down, over the fibers from 1.
       
      
       C: Same for block 5. over the fibers from
      1 and 3.
      
 
  
       
       
       
       D:Place block 6. When fiting, put connector
      joiners 4 and 2 in place.
       
       E: Place block 4. Fibers pass over the top of
      block 6.
       
       F: Place block 2. Fibers go over 6 and 4. 
       
       
       
       
     |   
       | 
	
	  (28) 
	  Once the connectors snap in place, they are hard to loosen. I added a pair of holes, and made a spreader tool to facilitate easy loosening of the connectors.
	   
	  
	    connector tapered input with holes
	   
	  
	    connector tapered output with holes
 
	   
	 |   
   | 
    
	  (29) 
      I made a 80/20 frame, with clips to hold the 1/8" carbon fiber boards. I also printed H-channels to join the boards.
	   
	  
	    H-channel panel joiner
	   
	  
	    80/20 clip
	   
	  
	    6-32 tap cap. The connector joiners are screwed down on
	  a 4×4cm grid of holes tapped in the carbon.
	   
	  
	    6-32 tap guide
	   
	   
     |   
	  
 | 
  
	  (30) 
    2 rows of layer A finished. Fibers that pass over blocks 4 and 6 are
    held with mylar 1mm-high brackets.
     
   |   
   | 
    
	  (31) 
      The first third of the top cover of layer A. Various nubs hold down the blocks, slipping
      in between the fibers. The flange will be drilled to hold it down
      with 6-32 screws.
       
       layeratop_left_bot.stl
       layeratop_left_mid.stl
       layeratop_left_top.stl
       layeratop_mid_bot.stl
       layeratop_mid_mid.stl
       layeratop_mid_top.stl
       layeratop_right_bot.stl
       layeratop_right_mid.stl
       layeratop_right_top.stl
     |   
       | 
	 
    	 
 | 
  
	  (32) 
    Prepare for tray B. I made a parametrized version of the tray, so
	  I can quickly make A, B, C and D.
        A   B   C   D
  -------------------
  dd1  30  30  30  60 
  dd2  90  90  90 120 
  dd3 180 180 150 210 
  dd4 305 305 240 340 
  dd5 445 445 375 500 
  dd6 500 500 500 500 
For now, we  need the rgb parts:
  b_right_bot.stl
  b_right_mid.stl
  b_right_top.stl
   |  
   | 
    
	  (33) 
      4 groups of 6 in place. Only  the bottom two have fibers.
     |   
	
 	
   | 
    
	  (34) 
      Top covers in place
      
  
      |   
	
	
   | 
    
	  (35) 
      Partial top: 3 covers held in place with a carbon fiber panel. The panel is not the full width, it was a left over piece. Needs to be replaced
      
      
	 
       
	Now layer A can be flipped over and layer B will be constructed on the
	other side.
	  |   
	
	   | 
	    
	      
  
	      continue on the next page
	 |