sPHENIX is a new detector at RHIC.

[Edouard Kistenev <kistenev AT bnl.gov>]

John,

in case if decision is to keep the passive parts of IHC “as is” and save on scintillators only the compromise solution which will allow to avoid systematics in jet data would be to use the IHC-magnet gap to install the layer of 2cm thick scintillating tiles built to produce signals from dphi x deta = 0.2x0.2 (1/16 of SiPM count, 1/4 of digitization channel count). They may even cover cylindrical surface (to reduce gap requirements) - IHEP (Protvino) has technology to injection mold tiles of 2x10x10 cm^3 which could further be coupled together to create more or less arbitrary shaped larger tiles. 

Edward

On Sep 14, 2017, at 4:07 PM, John Lajoie <lajoie AT iastate.edu> wrote:

Dear sPHENIX:

    I would just like to follow up Dave and Gunther's email with a bit more detail about the inner HCAL options that we are considering, and the simulations and engineering studies that are underway.

    In the default descoping option that Gunther and Dave outlined in their email, the bulk of the savings would come from the elimination of the inner HCAL. In fact, I see no way that the level of required cost savings could be achieved without essentially eliminating a detector system.  For the inner HCAL, this means that the focus needs to be on *how* the inner HCAL is eliminated, can it be done in a way that could potentially be recovered if additional money becomes available, and how do we minimize the impact on the physics capabilities of sPHENIX. 

    Let me run through a few of the options that have been proposed and let you know what we (the HCAL and Jet ToG, and the BNL engineers) are doing to evaluate them:

Eliminate the inner HCAL entirely:
---------------------------------------------
This option has no fallback - we would replace the inner HCAL with a frame to hold up the EMCAL and be done with it.  In this option the focus is on the dead material in the frame and the effect it would have on sPHENIX calorimetry.

The BNL engineers do not have a design for a frame, and I have been told that designing a frame and doing the FEA analysis would be a job that would take a couple of months.  Therefore we will not have a detailed design at the point at which we need to make a decision.  Instead, we are looking at options to try to bracket how much dead material the frame would represent:

The current stainless steel IHCAL has been engineered to hold the EMCAL, so from an engineering standpoint it would work fine.  It represents an upper limit on the amount of dead material and could likely be reduced.  However, we know from single particle simulations that one interaction length of dead material between the EMCAL and outer HCAL degrades the single particle resolution to the point that it no longer meets the sPHENIX specification.  This is NOT AN OPTION.

To bound the amount of material from the lower side, we could consider a cylinder of Al that would be strong enough to hold the EMCAL. The BNL engineers are currently calculating the thickness this cylinder would have to be.  We couldn't build a frame exactly like this, but it could be put into G4 and simulated.  In this sense it represents the minimum amount of material that would be required, and keeps that material azimuthally symmetric.

One we have the dimensions for the minimal cylinder, the plan is to introduce this into the G4 simulations and study the jet response (see below for more details).

We could consider making the inner HCAL/frame out of a less dense material, reducing the number of interaction lengths as compared to steel.  This brings up the next option.

Descope the Inner HCAL, but do it in a way it could be recovered:
--------------------------------------------------------------------------------------
In this option you would build the inner HCAL sector from a less dense material (Al). The hope is that if you can't find the money to instrument it you can live with the dead material, and if you can get the money you can recover some performance. An Al inner HCAL would be less than half an interaction length.  Single particle simulations from Abhisek and the Colorado group, shown at the last fortnightly meeting, seem to indicate that from the energy resolution standpoint this isn't so bad, so it's not obviously a non-starter.  However, this does not comes close to telling the full story of the effect for jets, and needs much more additional study.

There are a number of issues with an Al inner HCAL that need to be addressed:
-> FEA analysis needs to be done to ascertain if an Al inner HCAL could support the EMCAL. This is underway by Anatoli Gordeev at BNL.
-> The effect on hadron rejection needs to understood.  Sasha Lebedev at ISU is working on this.
-> The effect on jet observables needs to be understood.  The jet ToG is leading the way on this.  In this case we need to study *both* an instrumented inner HCAL, and an uninstrumented inner HCAL, to verify we could survive is the descoping is permanent.

The last item is very tricky. An Al IHCAL will leak more energy out the back - does this introduce a fragmentation bias? The shower may spread more in the inner HCAL and be wider overall.  How does this affect the underlying event subtraction? Note that all of these same issues must be addressed in the case of a frame as well - another, parallel set of simulations will be required for the "cylinder" described above.)

There's a key dividing line between the two options, and that is the amount of money you can claim as savings.  If you go with the first option (permanent descoping) then you save ~$4.5M, minus what you think you need for a frame.  How much would a frame cost?  There is no way to know without a detailed design, but estimates range from $100-$200k or $500-$800k, depending on who you ask. If you go with the second option you need to build the mechanical sectors to spec., which will cost about $1M, and the most you can save is $3.5M.   Obviously, as the frame gets more expensive it would make more sense to go for the second option and have the option to recover the scintillator and electronics.

At the last fortnightly meeting, Edward suggested a variation that would create a radially thinner IHCAL from a half an interaction length of steel, keeping the absorber material the same as in the current design.  (He also included a different scintillator and readout scheme, see his email in response to Gunther and Dave.)  This is basically a new detector design, but for descoping considerations it would still be a half an interaction length of dead material, and would include an air gap between the IHCAL and cryostat.  I have asked the machinists who built the IHCAL prototype to estimate the cost reduction for a simpler design for a half-sized inner HCAL and they estimate 25-30%. Without a design this is a bit of a guess, but it is on the same level as the cost reduction obtained by making the current inner HCAL design from Al, so just a frame of this type would still cost on the order of $1M.  In terms of the effect on the physics we should be able to get an idea from the Al IHCAL simulations (half an interaction length of material) and the cylinder (which will by definition have an air gap.)   There are NO PLANS for a dedicated set of simulations to otherwise address Edward's proposal.

I hope this helps the collaboration understand how the group of people working on this are framing the problem, and the simulations we are trying to complete on a short timescale. Clearly there will be some unanswered questions or imperfect information, and we can't explore all possible design options in the limited time available. We are trying to focus on a limited set of  concrete questions that we can give satisfactory answers to, and not a wide range of questions with only half-axx answers.

Regards,
John Lajoie

On 9/14/2017 9:12 AM, David Morrison wrote:

Dear Collaborators,

As you're likely aware, the project and the collaboration have been
asked by BNL ALD Berndt Mueller to provide by early October a scope for
the sPHENIX baseline detector which fits within $32M.  The current cost
estimate is about $38M.  The ALD anticipates meeting with DOE ONP in
late October to present the current scope and cost and discuss with the
them paths forward to realize the detector.

Everyone should anticipate that scenarios that address the cost cap set
by the ALD will decrease the physics capability of the detector to some
degree.  At the same time, we remain strongly committed to building the
complete detector and carrying out the full sPHENIX physics program.
Approaches for dealing with the tension between scope and funding as we
move forward include working with our collaborators to see about
opportunities for non-DOE contributions and by working to preserve
contingency funds to "buy back" as much capability as possible.

Between the executive council and the biweekly general meetings, we've
had a number of very constructive discussions about how best to respond
to this request – what scenarios should we study, what risks do we face,
what plots should we aim for?

We think the best strategy for addressing the ALD's $32M cost cap is to
focus on straightforward alterations of the current $38M detector.  The
menu of such changes identified so far include:

- no iHCal (just a non-detector support structure to hold up the EMCal)

- non-instrumented stainless steel iHCal

- non-instrumented aluminum iHCal

- limited pseudorapidity acceptance of EMCal (fewer W/SciFi blocks and
corresponding electronics)

- limited pseudorapidity acceptance of the oHCal (same steel structure,
but with fewer scintillator tiles and corresponding electronics)

Our current "default" scenario is no iHCal, coupled with limiting the
acceptance of the EMCal and oHCal until the $32M cap is met.  We think
this provides a clear path to restoring the full capabilities and also a
clear target for possible non-DOE contributions.

While we study the physics impacts of possible changes, the project is
working with the L2 managers to "scrub" costs and look for any savings
that can be readily justified.  If savings are identified, they would
lower the current $38M estimate and reduce the need for any descoping.
The current expectation is that this scrubbing exercise might identify a
few x $100k.

In thinking about all this, it's useful to keep in mind a key conclusion
of the August Director's review of the baseline detector, "The committee
believes that sPHENIX should be positioned for a successful CD-1 review
in the spring of 2018 if continued progress is made in development of
the technical, management and cost/schedule aspects of the project."
This is a really encouraging statement made by a panel of very
experienced people.  The date by which we get CD-1 approval has a big
effect on sPHENIX.  Subsequent steps in the project, including the year
in which sPHENIX would be ready to take data, follow rather directly
from that date.  Radical reimaginings of the detector or overly
optimistic cost scrubbing will likely delay the scheduling of the
official CD-1 review and generate a corresponding delay first physics.

Regards,
Dave and Gunther

--

John Lajoie

Professor of Physics

Iowa State University

 

(515) 294-6952

lajoie AT iastate.edu

Contact me: john.lajoie

_______________________________________________
sPHENIX-l mailing list
sPHENIX-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/sphenix-l