sPHENIX EMCal discussion

[Craig Woody <woody AT bnl.gov>]

Anne, John, and others,
   I think we all know what issues we now face as a result of the devastation of the EMCAL design resulting from the descoping exercise (and yes, I think there is no other word for it other than devastating). From what I hear in the hallways, it seems we are going to be forced into the Scenario B design of the Descoping Document, which involves reducing the eta acceptance beyond what is assumed for Scenario A (which was +/- 0.6). I assume this will mean something ~ +/- 0.5, but I'm not sure if this is the fiducial acceptance or the actual acceptance. However, if it is the actual acceptance (which is what I presume was used for the cost estimate), this would then lead to a fiducial acceptance of something ~ +/- 0.4 (does this sound close to a familiar number of +/- 0.35 ? ). I think this will seriously impact our jet physics program, not only because of the reduced acceptance, but also because of the non-uniformity of calorimeter coverage over the full acceptance of +/- 1.1, as well as our Upsilon program. If we are forced to accept this change, then I think there is no way that we can justify building a 2D projective EMCAL. The physics arguments for improving the e/h rejection at large rapidity for the 2D projective design were marginal at best even at large rapidity, and I don't think we will ever be able to come up with a physics justification for requiring 2D projective at only central rapidity. The implication of this is that we simply do not need to build a new 2D projective prototype EMCAL to test at the next beam test at Fermilab if there won't be any EMCAL at large rapidity. It also means that we should abandon our R&D on the building 2D projective modules and concentrate on trying to improve the 1D projective modules we tested in our last prototype (and there were things that needed improving). That's not to say we should not complete the R&D on building the 2D projective to a certain stage where we can show that we could build the 2D modules if necessary, and I think we can in fact show that, but there would be no need to carry on the R&D beyond that. It also means that we should begin now redesigning the full EMCAL and all its sectors to accommodate the 1D projective modules. This should presumably be simpler than our current design that is based on 2D modules, but it will require a new engineering design.
  The second implication of Scenario B is the ganging of the EMCAL towers. This was left as two possible options, one being keeping the current segmentation and summing the four towers into one readout channel, and the other was to increase the physical segmentation into larger towers. The first scenario is clearly easier and does not require any further significant R&D on the light collection. However, even with that design, we still need to do some R&D to come up with a cost effective light guide, since the individually machined light guides used in our first prototype could not be used in the final calorimeter (or even the pre-production prototype) due to cost. I also share Edward's concerns about summing the channels together, but I think we can accept most of these consequences in the end. The other option of making larger towers would require, in my opinion, substantial R&D. Since the radial space for the light guides would presumably remain the same, we would need to build, test and measure the light collection from the larger towers. Mike Phipps did some nice simulations to begin to study this, but we would be obliged to actually measure it if we decided to go this way. Personally, I seriously doubt that we could design, build and test a complete set of new larger EMCAL towers for a new prototype on the time scale necessary to have it ready for a beam test in January (which would mean that the new prototype would need to be ready to ship by the first of January, just six months from now). Moreover, it would still have the problem of having to sum more SiPMs into the same readout channel, since we know for sure we would need to increase the number of SiPMs per tower on a larger tower design. However, the grim implication of accepting and implementing the first scenario of keeping our current design and ganging the 2x2 towers is that once we actually implement that, it is unlikely that we will ever recover the segmentation back again.  We would then wind up with ~ 5x5 cm towers, but using a very expensive material (tungsten) to achieve a 2.3 cm Moliere radius, and placing it at a radius 90 cm.  Does this make sense to anyone ?
   I agree with Anne and John that we need a firm, clear, written decision from the Lab Management that we are being forced to accept the Scenario B solution that is given in the Descoping Document. I was not directly involved in putting this document together, and I don't even know how the numbers that are in there were obtained, but from my perspective, it was supposed to show the consequences of implementing the various cost cutting options, and not meant to say that we advocated accepting any, much less all of them. I think the result of accepting the maximal amount of cuts in Scenario B will indeed seriously reduced the capabilities of the EMCAL, and so much so as to severely and adversely affect our physics program. I also agree with Edward that, given the resulting design, if you look at it, it simply doesn't make sense. Why are we building an HCAL out to +/- 1.1 in rapidity, with an EMCAL only out to +/- 0.5 using a very expensive technology that was chosen to give us a small Moliere radius so that we could do a good job on electron and photon shower separation, which now has a segmentation 4 times that ? Do we need really then the TPC to cover +/- 1.1 or could we live with a smaller acceptance for that as well ? Having just attended the CALOR conference and presented our current baseline design, which already raised a few eyebrows, I wouldn't want to go back there now and present the new design we are apparently coming up with.

Craig 
 
 

On 6/30/2016 4:24 PM, John Haggerty wrote:

Anne,

You're right, and of course I go to a lot of meetings so I have heard 
things, but I think we should get a written concurrence about the scope 
of work between the laboratory, the project, and the collaboration, and 
I think we need that on the timescale of the next few weeks so we can 
make good decisions on prototypes, R&D, and continued design.

On 6/30/16 3:25 PM, Sickles, Anne M wrote:

Hi John,

I agree your questions are good ones, but I was wondering what the
timescale for hearing back on the EMCal options considered in the cost
reduction exercise document.  There were a lot of big changes to the
EMCal considered, but I don’t know which of them we will need to do.
 Has there been any feedback?  I don’t think the EMCal group itself can
provide an answer to your first question:

- Do we even test an eta=1 calorimeter?  No need to test it if we're not
going to build it.

without that input.

Best,
Anne

On Jun 30, 2016, at 11:43 AM, John Haggerty <haggerty AT bnl.gov
<mailto:haggerty AT bnl.gov>> wrote:

You probably saw a very similar post to the HCAL list... the questions
for the EMCAL are the same, only harder:

Very soon, we need to agree to changes in the "reference design" of the
calorimeters engendered by the descoping exercise.

In particular, I think the next prototypes we test should be as close to
the final design as possible.   In the case of the EMCAL, there are some
real questions that we need to answer very soon:

- Do we even test an eta=1 calorimeter?  No need to test it if we're not
going to build it.

- Do we test "2x2" towers?  Do we have 4 little light guides with a lot
of corners, or bigger light guides?  I could imagine a single stuby
light guide could result in more uniform light collection, but it's not
at all obvious.

- Do we design preamps that can take 16 SiPM's?  This is a thing that
Steve would have to test before we did it, and it would have to be moved
to the front burner if we want it for January.

- Projectivity.  Enough said.

With the substantial changes to the EMCAL we're talking about, we may
also wish to consider delaying our appointment with the FTBF a month
or two.

Since not everyone attends every meeting, I think it's important that we
discuss this with a wide audience on the lists.  So fire away.

--
John Haggerty
email: haggerty AT bnl.gov <mailto:haggerty AT bnl.gov>
cell: 631 741 3358
_______________________________________________
Sphenix-emcal-l mailing list
Sphenix-emcal-l AT lists.bnl.gov <mailto:Sphenix-emcal-l AT lists.bnl.gov>
https://lists.bnl.gov/mailman/listinfo/sphenix-emcal-l

--------------------------------------------------------------
Anne Sickles
Assistant Professor, Department of Physics
University of Illinois at Urbana-Champaign
--------------------------------------------------------------