sPHENIX HCal discussion

[Edward Kistenev <kistenev AT bnl.gov>]

John,

I am sure that the disadvantages of a thin calorimeter for jet measurements are so obvious that massless calorimeter is indefensible. All my recipes for corrections are for single particles, for jets on underlying central event background they are unlikely to help much, Pt threshold will be pushed up and upper limit will go down (the infamous bin migration). And sPHENIX needs to be the first to recognize it. So my suggestion would be to avoid speaking of calorimeter structure built of Al. To me it would be sufficient to say that for the purposes of this document we considered replacing HInner with support structure for inner detectors and that “value” analysis pushes us to chose hardened Al as the best and cheapest material for such a structure. We may further say that this structure will occupy so-and-so many cm of radial space and cost ….$. That’s it, the remaining space should be used to present the arguments for scope recovery from your mail.  

Edward

On Feb 6, 2018, at 11:30 AM, John Lajoie <lajoie AT iastate.edu> wrote:

Folks,

    I think we are tying ourselves in knots a bit unnecessarily, mainly because we are trying to find a way to sell something to the committee that we don't really want. Whenever you find yourself acting against your interests in that way, it's time to step back.

    Stefan has very carefully pointed out that the current design w/o the inner HCAL is shorter than our original specifications, and rightly so, because we can't leave that hanging out there to be shot at.

    CD-1 is a review of a *conceptual* design. What I would do at this point in time is the following:

    We acknowledge both that the 5.5 interaction length depth was part of the original requirements, and that descoping has left us with a combined calorimeter that at the present time is shorter than that. We are actively evaluating the degree to which this will degrade the performance of the detector *and* at the same time we are aggressively working to restore lost scope (and depth) to the sPHENIX calorimetry system.

    This has the dual advantage both of not being bullshit and actually being true. In effect, this basically writes a committee recommendation for them, certainly for the Director's Review.  We will absolutely have to have some evolution of this response in terms of simulations (and maybe new collaborators) by the end of May, but I'm OK with that.  If you think about it, if the committee writes a recommendation that strongly encourages us to fine new collaborators to restore scope, that might actually help in finding new collaborators and getting them to commit!

That's my take,
John Lajoie

On 2/5/2018 10:03 PM, Edward Kistenev wrote:

And last - Liang Xue at GSU was the one person who spent time to develop initial steps for LCG based energy reconstruction in sPHENIX calorimeters. Unfortunately he left physics but his presentations are all in sPHENIX archives. Time permits - have a look.

On Feb 5, 2018, at 10:58 PM, Edward Kistenev <kistenev AT bnl.gov> wrote:

PS. Stephan, 

there is no literature which may answer the question about leakage to better then 10%. Reading through your references you will probably find that the coefficients in the formulas approximating averages are particle mass dependent and that the dependence is not just because the energy we measure  is kinetic, not the total. Showers have different spectra and particle composition at different depth, below 100GeV measured energy distribution for a constant momenta is absolutely nongaussian and response linearity is only in dreams.  What is interesting is that with all the problems we have with the shower fluctuations the amount of leakage is rather easy to compute even on the event-by-event basis (and correct). It has an rms equal to its value (almost poissonial), but the value itself has little dependence on particle mass and CG fluctuations (showers in the tails are all “equal”) are easy to control/compensate if calorimeter is longitudinally segmented. In essence this is how sPHENIX calorimeter was designed. It was supposed to consist of three longitudinal sections with HInner and HOuter (if creatively used) offering one extra section each (total 5). This is because the neighboring towers overlap. If you know the “line of flight” - energies in overlapping towers correspond to specific (changing event-by-event) shower localizations which are computable and usable for global CG computations. You may use tracking data to define the “line-of-flight” or use “in calorimeter tracking” iteratively. The non compensating nature of calorimeter can be handled in a similar way resulting in rather gaussian final energy distribution. All this is yet to be implemented but few words along these lines will not hurt the CDR. 

Edward

On Feb 5, 2018, at 10:19 PM, Edward Kistenev <kistenev AT bnl.gov> wrote:

Here is my beloved free pocket calculator addressing your problems down to better then 10%  (created at a time immemorial - BW). 

http://www.slac.stanford.edu/comp/physics/matprop.html

On Feb 5, 2018, at 9:37 PM, Stefan Bathe <bathe AT bnl.gov> wrote:

Dear Edward, John, and Jamie,

Which book is that, Edward?  It would be nice to be able to look up the references.

For 100 GeV (just to stay with my earlier example) I get 6.2 lambda from the first and 7.2 lambda from the second formula.  They are not within 10 % of each other nor within 10 % of the measurements for Fe I quoted earlier.

I agree we won’t see 100 GeV jets in AuAu.  I had to pick one energy to compare the numbers, and the kinematic limit seemed to me a convenient upper limit.  For 70 GeV all numbers will be about 0.2 lambda smaller.

Regards,

Stefan

-- 

---------------------------------------------------------------------------------
Stefan Bathe

Professor of Physics
Baruch College, CUNY

Baruch:                                     BNL:
17 Lexington Ave                      Bldg. 510
office 940                                  office 2-229
phone 646-660-6272                phone 631-344-8490
----------------------------------------------------------------------------------

On Feb 5, 2018, at 6:29 PM, Edouard Kistenev <kistenev AT bnl.gov> wrote:

10% approximation always looked fine to me

<PastedGraphic-1.tiff>

On Feb 5, 2018, at 3:23 PM, Stefan Bathe <bathe AT bnl.gov> wrote:

Dear All,

I find some inconsistencies with how many nuclear interaction lengths (lambda) are required to contain 95 % of the energy of a hadronic shower (L(95%)):

1) The CDR says 

L(95%) > 5.5 lambda 

in the introductory section of the HCal.  No energy is quoted.  So let’s assume 100 GeV pions as proxies for jets at kinematic limit for RHIC HI.

2) [WI00] (Fig 2.37, attached) gives 

L(95%) @ 100 GeV = 6.0 lambda

N.B.:  the reference is [AB81]!

3) [AB81] gives 

L(95%) @ 100 GeV: 87.5 cm Fe = 5.15 lambda (table 4)

contradicting Wigmans!

4) [HO78b] gives 

L(95%) @ 100 GeV: 82 cm Fe = 4.8 lambda (Fig. 10)

5) [KL91] gives

L(95%) @ 100 GeV: 82 cm Fe = 4.8 lambda (parameterization)

I’m inclined to dismiss Wigmans since the plot misrepresents the quoted reference.  Does anybody have better information?  Or maybe I’m misunderstanding something?

Regards,

Stefan

references:

—

[WI00] Wigmans, Calorimetry, Oxford, 2000” (p. 87, Fig. 2.37)

(see attachment)

--

[HO78b] M.Holder et al., Nucl.Instr.Meth.,151,69 (1978),
Performance of a Magnetized Total Absorption Calorimeter Between 15-GeV and 140-GeV
5 cm Fe sampling
L(95%) @ 100 GeV: 82 cm Fe = 4.8 lambda (from plot with data points and fit in paper; or parameterization in Kleinknecht textbook)

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=5&cad=rja&uact=8&ved=0ahUKEwju8M2mxY_ZAhWHyoMKHTAyAgQQFgg_MAQ&url="http%3A%2F%2Fcds.cern.ch%2Frecord%2F879171%2Ffiles%2Fep113_001.pdf&usg=AOvVaw18tk97fLX9RJZVIoNU1SVM

---

[AB81] Nucl.Instr.Meth.,180,429 (1981) 
The response and resolution of an iron-scintillator calorimeter for hadronic and electromagnetic showers between 10 GeV and 140 GeV
2 cm Fe sampling
L(95%) @ 100 GeV: 87.5 cm Fe = 5.15 lambda (table)
comments:
- interaction required in first 37.5 cm of iron; possible bias
- referenced in Wigmans, but I cannot reproduce Wigmans plot from data in paper

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=8&cad=rja&uact=8&ved=0ahUKEwjRoonEyI_ZAhVs4YMKHd-dCaUQFghTMAc&url="https%3A%2F%2Fcds.cern.ch%2Frecord%2F134124&usg=AOvVaw0ZxhJB8UKofvCs-ni1UPaX

—

[KL92] "Kleinknecht, Detektoren fuer Teilchenstrahlung, Teubner, 1992”, I find the following parameterization:
L(95%) = [9.4 ln E(GeV) + 39] cm Fe.  With lambda = 17.1 cm (Fe)

also references [BL82] H. Bluemer, Diplomarbeit Dortmund, 1982

<Wigmans2.37.JPG>

-- 

---------------------------------------------------------------------------------
Stefan Bathe

Professor of Physics
Baruch College, CUNY

Baruch:                                     BNL:
17 Lexington Ave                      Bldg. 510
office 940                                  office 2-229
phone 646-660-6272                phone 631-344-8490
----------------------------------------------------------------------------------

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

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

--

John Lajoie

Professor of Physics

Iowa State University

 

(515) 294-6952

lajoie AT iastate.edu

Contact me: john.lajoie

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