Frames, rails and cable trays
back to 'assembly and installation'
| (1) Pictures taken during a 2018 visit to the experiment More pictures from September 2022 2018: Isometric front view (right) and top view (below), from this EASM file. A selection was converted to an EASM file. The whole hall The diagonal wires (green) of the spider attach the front of the beampipe to points near the front and the back of the magnet (both top and bottom). The horizontal and vertical wires attach the back of the beam pipe to points in the center (z) plane, as well as to the back. In red is a frame where the horizontal wires attach.
| 
| ||||
| (2) There are 4 bars (green), which appear to be part of the magnet poles. These would make good attachment surfaces. However, the various spider frames and points are not mounted from these bars. Are we allowed to mount to these bars? |
| ||||
| (3) Top of the iron removed for clarity. In red are Y-shaped brackets which are anchored
in three points: points A are on the vertical walls, on the horizontal
center plane. Points B are near where the anchor points for the beam
pipe spider are. The angled sections are parallel to the magnet coil face.
| 
| ||||
| (4) Three chambers (black, measuring 1m high and 4m long), installed. The green arow points to one arm of the spider that holds the beam pipe near the front. The spider wires in the horizontal and vertical planes do not interfere. There are also diagonal wires going from the forward beam pipe attachment ring, and are anchored further towards the back. These are also cleared.
|
| ||||
| (5) 2022: Work resumes on the installation design. There is a possibility to have taller (>1m) panels at the wider end of the magnet. How tall? I added to the model the frame that holds the wires that support the beampipe, and the braces, since the braces limit how tall panels can be. | 
| ||||
(6)
| 
| ||||
| (9) The optimal tilt angle is when a normal to the plane intersects the beam axis. This is 74° wrt the horizontal plane. (Eric) 
|
| ||||
| (10) The 74° frame just clears the magnet brackets, and the brace of the spideweb frame limits the vertical space available for the detector panels
| 
| ||||
| (11) krakow_frames_feb23.ppt | 
| ||||
| (12) 7 panels (100,100,120,130,130,145,145) with fibers coming off the edges away from the center plane. The marks on the string are at 1m intervals On the back, the fiber bundles can travel straight to the downstream end | 
| ||||
| (13) The black panel is 90×350cm, defined approximately by the constraints on the rails at the narrow end of the magnet | 
| ||||
|
(14) Panels mounted onto the rails. Here, the fibers are routed to the top of the magnet. It takes a little over 6m from the panel exit to the top edge of the magnet. Add to that the distance to the individual panels (0.5 - 3.5m) plus the distance from the magnet edge to the rack, which means the longest fibers in the system are ~12m long (3m less for the top quadrant). | 
| ||||
|
(15) There is space below the magnet: Picture from the virtual walk-around → (Mar 23 note: this is an old view, there is more stuff there now)
| 
| ||||
| (16) Watch how I install the panels:
| 
| ||||
| (17) There is a set of trays (in this 10:1) model, they are 6×6cm), one for each panel. The wood bar is one of the panel rails. | 
| ||||
| (18) On the top of each panel, a rigid structure (blue) guides the fiber ribbons up and over the lip of the cable tray.
How big is this bundle of fiber ribbons? Each panel has
4(pairs of scintillator blocks) ×
12(blocks wide per layer) ×
4(layers per panel)
ribbons, each 1×8mm in cross section (
see here),
for a total cross section of 15.36cm2, or about
| 
| ||||
| (19) How big are these blue structures? At the end of the nominal panel end (100 - 145cm), about 12cm are needed for the connectors plus ribbon strain relief. Beyond that the ribbons curve up and over into the trays, here about 10cm up and 10cm over (perpendicular to the unit plane). mylar_80cm laser cut file Here is a 1:1 mockup of the end of a module (50×10cm), with 4×4×12 mylar ribbons coming out. How to route these 192 ribbons? | 
| ||||
| (20) Next layer goes over the top | 
| ||||
| (21) The ribbons curve and twist such that the surface that faces left in this picture end up facing down in the cable tray. | 
| ||||
| (22) The module (blue) is ~10cm wide, and the position of the cable tray is determined by the minimum bending radius of the fibers, here about 10cm. The next tray (see (17) above) has to clear the fibers, and here it is sketched in ~6cm above the lower one. | 
| ||||
| (23) Fiber ribbons at Saint-Gobin, which states "The typical bend radius is 30-40 times the diameter of the fiber", which would be 1.5-2cm in our case. An email from Steven Robare says "For [0.]5mm it is 10cm radius". Another link about minimum bending radius: "the bending radius of fiber should be more than 150 times the diameter of the fiber cladding". So for our 0.5mm fibers, this would be R>7.5cm, so to be safe, let's use 10cm radii. (In the pictures above, the radius is 5cm). | 
| ||||
| (24) With a bending radius of 10cm we really need to conserve space. Here the cable tray is rotated 90° (but the bending radius has not been changed yet).
| 
| ||||
| (25) Based on (23), I made a 17.5-cm radius quarter-tube. Ribbons run on the surface, and follow the orange cutout which also has a 17.5cm radius. Thus the ribbons reach the horizontal before going 90%deg; in the xy plane. | 
| ||||
| (26) Based on (23), I made a 17-cm radius quarter-tube. Ribbons run on the surface, and follow the orange cutout which also has a 17cm radius.
| 
| ||||
| (27) Given a minimum bending radius (here 17.5cm), this is the minimum space needed. If the orientation of the cable tray does not matter, the tray can in principle be moved along the green arc. | 
|
back to 'assembly and installation' Hubert van Hecke Last modified: Mar 2023