sPHENIX tracking discussion

[Ross Corliss <ross.corliss AT stonybrook.edu>]

Hi Tony,

Sorry for the delay.  punch list notes are at the bottom of this email, and slides posted to https://indico.bnl.gov/event/24992/attachments/57871/99388/distortions.qm2025.punchlist.pdf .

I've been more or less completely out of it yesterday and today, but I seem to be sending emails now, which is hopefulyl a good sign.

-Ross

In terms of observing distortions, what is the status and outlook?

- observations:

    - we can observe the laser flash on north and south CM.  It:

        - is consistent with our modeled static distortion, though currently this is done mostly by looking at the straight lines of the laminations.  

        - illuminates the majority of the CM, though the range includes saturated and dim regions.

        - sees background hits from the CM gold layer as well as 'foreground' hits from the aluminum

        - may show a relationship between arrival time and radius that matches other readout issues identified, but this has not been demonstrated yet.

In terms of distortion corrections, what is the status and outlook?

- static:

    - we will not have a measured correction from the direct lasers in time to be useful for QM.  (the lasers still need to be fully commissioned)

    - we have a model that matches decently well to data, but this needs more investigation

    - the modeled correction probably needs to have the TPC misalignment added in.  This should not be a dominant effect, and should be somethign that can be picked up by other corrections, but if we can properly account for it here we should.  We can't confirm this until we have better tools (see below)

    - we should look at how much the GEM voltages at the front face vary, both spatially and in time, and see if those are significant enough to affect the static correction.

    - since we will not have lasers, we need to develop more tools to assess the correction directly.  We have some of these (see earlier this summer), but some create circular-reasoning loops.  We have looked at several of the below, but need to revisit them now that tracking is substantially different than it was at the start of the summer:

        - low lumi , cosmics with field on

        - r, phi, or z residuals of all tracks in TPC-only

        - matching to primary vertex without silicon?

        - can we leverage mass peaks?

        - can we use the TPOT?

        - The CM is a good constraint, but that needs more development (see that section).

- (module edge):

    - this has been measured in field-on. 

    - we should check whether the effect has lumi or other time depencies

    - we should look at what the effect is in field-off

    - we should consider whether the effect should be rendered as a purely phi shift, or if it should be a shift in phi and R.  this has implications for the CM reco, but not much effect on track reco.

- average:

    - where are we on this?

- digital current:

    - no practical route to having this for QM.  needs lots of work to convert zs streamed data into a space charge map, and needs some validation of a map against actual data before we proceed.

- central membrane:

    - clustering algorithm exists for the CM clusters.

        - can perform clustering in r-phi-z

        - ID clusters and match them but it needs work to handle dynamic range and saturation

        - we need to finish matching and extraction of the distortion.  It is mostly done, but not completed(Ben, Chris)

        - we need to build a final 'field off truth' and 'field on zero spacecharge' truth

various issues (some overlap):

1. matching to inner silicon not settled, which may mean the static correction is not sufficient

2. origin of large z residuals, if really present, may be readout or may be a volumetric distortion.

    1. this can be investigated with CM

3. digital current is not available yet.

    1. we can build a test bed by using the long-readout, nonzs data and comparing to the zs, and study how the zs-streaming digital current will differ from the nzs values.  How precise will our result be, if we just use streaming data?

4. central membrane reco needs better clustering algorithm to find centroid in the presence of saturation and background

5. CM reco does not currently resolve distortions.  a combination of the distortions being small, and the clustering algorithm having large pulls depending on saturation.  

6. central membrane event-by-event reco needs to be streamlined and properly tested with event-by-event corrections

7. track-driven distortions need to be tested with real data and integration from the CM work

tl/dr: (not a complete list here yet)

must:

- verify the static correction is complete (to what level of accuracy/precision? we can accept some mismatch)  (Ross?)

- get the TPC local alignment  (Mariia?)

- make the CM reco robust to saturation  (Ben?)

- get global alignment to the point where the TPOT correction can be extracted  (Tony?)

- demonstrate the CM + average workflow works.  (Hugo?)

want:

- produce a metric of how well we can bracket residuals

if time permits:

- prepare for direct laser data

- build proper models

==========
Dr. Ross Corliss
Research Assistant Professor
Center for Frontiers in Nuclear Science
Stony Brook University
virtual office:  https://stonybrook.zoom.us/my/rossoffice?pwd=ZmZ2SlRIMVFvUUJwbUkyOVNVTmE5QT09

On Oct 10, 2024, at 2:53 PM, Anthony Frawley (via sphenix-tracking-l Mailing List) <sphenix-tracking-l AT lists.bnl.gov> wrote:

Hi All,

We have seen many plots in tracking meetings this week that will be of interest to the collaboration for the presentation on tracking status tomorrow. None of them are posted on the indico page!

Please send me copies of your slides as soon as possible, together with any updates you may have.

Thanks

Tony