sPHENIX HCal discussion

[Edouard Kistenev <kistenev AT bnl.gov>]

Hi, John, 

yes, we discussed it many times since we first began looking into adjusting 
T2K fiber pattern (with fiber exits on both ends of the long tile) to the 
needs of a single long tile with both ends exiting on one end of the tile. I 
do not remember any disagreement with rather simple design rules:
- bending diameter no less then 5cm;
- any luminous point in the tile (with few exceptions) is no further then 
2.5cm from fiber (~attenuation length for 420nm light in diffuse coated 
scintillator across extrusion direction);
- two ends of the fiber in the first version (with single SiPM per tower) 
exiting nearly parallel to the edge (to simplify bending and bundling); 
orthogonal to the edge and superimposed (with or without displacement needed 
to position fibers along the line at 45 degrees or at 90 degrees across the 
edge) in the current version;
- setting the gaps edge-fiber and fiber-to-fiber in 1:2 ratio everywhere 
outside bend area.

I never felt myself 100% comfortable with sharp corners in the forward tiles. 
They definitely violate my design rule setting 2.5cm limit to the distance 
from any luminous point in the tile to the fiber.    If they are source of 
the problem one may either suggest a different fiber pattern or switch from 
pointing to nonpointing geometry or use a different combination of multiple 
tiles with different geometries in a gap (effectively longitudinally 
segmenting tiles in the gap). The best solution to all uncertainties would be 
the G4 simulation for a number of tile-fiber configurations (must include an 
optical model), prototyping what is liked and testing different tiles with 
cosmic. 

At a time we had no tile lab in Colorado so we relied on published sources 
and experience, we have that lab now so I wander if we may ask Colorado to 
dig deeper into this subject and make a new set of rules for Richie before 
new tiles are produced (“stretching” probably decreased pressures to have 
forward configuration tested by the year’s end). 

Note that tiles in the attached pictures are test samples and not the tiles 
for the test beam. 


                Edward

PS. I have some components and proposal for the cosmic muons hodoscopic tile 
tester which may do tile scanning up to say 50 tiles at a time with muon 
impact position resolution better then 1cm. It will need 32 SiPM/PA channels 
and HBD readout (DRS4 even better - we may use fiber timing to improve 
position resolution). If there is an interest in building one - I suggest to 
discuss it and …..


> On Mar 10, 2016, at 9:36 AM, John Haggerty <haggerty AT bnl.gov> wrote:
> 
> Edward,
> 
> On 3/10/16 6:59 AM, Edward Kistenev wrote:
>> Richie,
>> do we have a decision to go back to 45 degrees or stay with superimposed
>> fibers. Attached are pictures of samples with fibers superimposed.
>> Expanding grooves for 45 degrees configuration will be simple but
>> irreversible.
>> We also need the whole insert design and at least an estimate for what
>> it will cost sPHENIX to order 200 pieces locally.
>> Edward
>> 
>> Edward Kistenev, PhD
>> PHENIX <http://www.phenix.bnl.gov> Physicist
> 
> Have we discussed somewhere the relative advantages and disadvantages of 
> light collection and fiber routing strategies?  If not, we should do that 
> soon and before we have Richie making drawings or Uniplast making tiles.
> 
> -- 
> John Haggerty
> email: haggerty AT bnl.gov
> cell: 631 741 3358