Clean up UT code for tracking

The UT code contains several duplications and legacy parts. Idea is to reduce it to only what is needed and clean up old code.

• Remove UTHit and everything connected with it such as UTHitHandler. Move everything to PrUTHits. #305 (closed)

•  Clarify the use of the Geometry Caches, which exist here and here. Move it to LHCb?

• Clarify numbers here and later in the file. Are all still needed?

•  Find the best approach to go from a row-column pair to a global sector number (lookup table or computation as here ?) Connected with: Rec#147

• Streamlinings and simplifications (started in Rec!3421 (merged) and in !4124 (merged))

@bkhanji @dovombru @mveghel

changed the description

@bkhanji I would need your help to understand a few things:

• What are these numbers? https://gitlab.cern.ch/lhcb/Rec/-/blob/master/Pr/PrKernel/include/PrKernel/UTGeomCache.h#L24 I combined them here into MaxSectorsAllBoards, but I am not sure this is the right name
• Similar question: What is this number here? Is it all sectors in the UT, or ...?

@xuyuan, @tskwarni please correct me if I am wrong, but from what I understand:

What are these numbers? .... NSectorPerBoard

Boards refer to Tell40 boards, one single TELL40 board read 6 "sensors" (a.i six groups of 4-ASICs).

Similar question: What is this number here? Is it all sectors in the UT, or ...?

That is right, it is the total number of sectors or a "group of 4-ASICs"(==512 channels) in UT. one group of 4-ASICs represents one single sensor.

@bkhanji Thanks. A follow up: If I have 6 4-ASICs, and 240 boards as written here, I get 1440 4-ASICs.

However, if I look at the total number of 4-ASICs (as per the second calculation), I get: 2304. Where is this difference coming from (or what am I doing wrong)?

There is also a complication from the central sensors(which I forgot to mention above): we have 5 sensors in the most central region instead of 4 and they are read by 8-ASIC not 4-ASIC. I am trying to reason the relation between the MaxSectorsAllBoards and NSectors, maybe others are faster to figure that out

@decianm The real number of total 4-ASIC is 1048.

For TELL1 RawBank (BankVersion v4), it has 240 boards, while for TELL40 RawBank (BankVersion v5) it is 216 boards. The "240" left here only for the backward compatibility.

The 2304 is the one if you multiply everything, 2side4layer9stave2face8module*2subsector. This "2304" is like a maximum possible size of the sector container, and it would be empty, for example, for some staves, like staveID=1, no modules with ID=3 on faceID=1 and no module with ID=4 pn faceID=0. And in many outer region, there are no sectors with subsectorID=1.

@xuyuan Ok, thanks, that makes sense!

Just for a reference, the actual UT readout segmentation, together with the new UTChannelID, is documented by Matt Rudolph in this EDMS document: 2604784 https://edms.cern.ch/document/2604784/1

Unfortunately I do not seem to be allowed to access it.

For me I just have to "sign in" with cern login (upper right corner) to access edms documents, but perhaps edms access does not come automatically for every cern user. I have extracted the document and put it at http://hep.phy.syr.edu/~tomasz/ut-software-numbering.pdf so anybody can easily access it.

changed the description