[Halld-physics] draft text of the eta-Primakoff proposal update
Ashot Gasparian
gasparan at jlab.org
Tue Nov 30 18:52:49 EST 2010
Hi Matt,
I complitly agree with all you say below in your email.
They are all difficult and need to be worked out, BUT there is
one thing which may make our todays discussion much relaxed:
the 1% uncertainty in the Compton cross section is an overkill
statement and we need to change it for this proposal. It is left
from the PrimEx and our original proposal where we been looking
for 2% level measurement on eta decay rate. SINCE we have a new
relaxed error bar in this proposal, which is 3.2% in total then
the requirement for the Compton also SHOULD be on the similar
3.2% level.
If the 1% number is left in the proposal then we need to change it.
Hope this new corrected number makes much easier, though, I agree,
we need to look on ways to measure the detection efficiencies in
the experiment.
Thanks,
Ashot
.............................................................
Ashot Gasparian Phone:(336)285-2112 (NC A&T)
Professor of Physics
Physics Department (757)-269-7914 JLab
NC A&T State University Fax:(757)-269-6273 JLab
Greensboro, NC 27411 email: gasparan at jlab.org
.............................................................
On Tue, 30 Nov 2010, Matthew Shepherd wrote:
>
> Dear Ashot and Liping,
>
> First to Liping's note, I understand Compton is very well known QED process and it is fantastic to see that you did such a nice job using this in PrimEx data. Your plot is impressive -- it sounds like it was designed into the experiment from the start and worked very well.
>
> My concern is that you are proposing a measurement with GlueX that demands <1% efficiency in both comp-cal and FCAL and we do not have a single idea for how to achieve this in either detector, let alone a calibration concept that we have validated with MC. Moving comp-cal on a 2D table is a good start. At low intensity you may calibrate photon efficiency, but you need comp-cal for electrons. Interactions in detector material are completely different between photon and electron -- also interactions with detector material will be different as one moves off axis making beam based calibration tricky. The FCAL support structure and platform is about to be fabricated now. It is pretty much too late to consider 2D translation of FCAL.
>
> The precision of your measurement and hence the reason for making it depends crucially on this efficiency calibration. I just want to confirm that we really don't have a single idea for how to achieve <1% absolute efficiency. Furthermore it seems all the techniques that worked well in PrimEx in the past cannot be used in the GlueX setting. Is that correct? If so, how do we convince people that the experiment isn't a nonstarter?
>
> In the past most physics-driven efficiency calibrations I have done depend on being able to reconstruct some dominant exclusive channel by inferring the four-momentum of one of the particles from the missing four-momentum and then checking to see how many times the particle was actually "found." This is great in e+e- where the initial state is known very well, but I'm not sure anyone has thought about doing things like this in GlueX. Maybe other people with a lot of fixed target experience have ideas. I hope we find a way to do it, because 1% precision efficiency would be awesome -- right now, it seems pretty impossible though.
>
> Cheers,
>
> Matt
>
> On Nov 30, 2010, at 3:13 PM, Ashot Gasparian wrote:
>
>>
>> Hi Matt,
>>
>> I completely agree with all of your concerns about the efficiencies
>> and the acceptances to reach a 1% level in the proposed experiment.
>> One thing is sure, for the electron and photon detection efficiencies
>> we need to find ways to have direct measurements in the experiment.
>>
>> We did that in the PrimEx experiment with a direct measurement. We
>> move the HyCal in 2d around the low intensity beam with a pre-measured
>> tagging ratio. Also, we had a big solid lead glass detector behind of
>> HyCal. For the electron detection efficiencies, we used the pair
>> production process on HyCal.
>> The question is how we can do it in the proposed experiment?
>> We have in mind to have the CompCal on a 2d movable table. It is possible
>> to do that since it is a tiny detector. It does not require a big
>> investment. This will take care of the CompCal case. For the FCAL we need
>> to come up with some physics processes to do this kind of absolute
>> detection efficiency measurements. I do not have a ready answer yet. We
>> need to come up with some techniques.
>>
>> Ashot
>>
>>
>> .............................................................
>> Ashot Gasparian Phone:(336)285-2112 (NC A&T)
>> Professor of Physics
>> Physics Department (757)-269-7914 JLab
>> NC A&T State University Fax:(757)-269-6273 JLab
>> Greensboro, NC 27411 email: gasparan at jlab.org
>> .............................................................
>>
>>
>> On Tue, 30 Nov 2010, Matthew Shepherd wrote:
>>
>>>
>>> Hi Ashot,
>>>
>>> Regarding the rate, I checked my background calculations. Indeed the rate above 3 GeV for hits *in the FCAL* is very small -- about 5-6 orders of magnitude smaller than the total rate. Maybe there are more high-energy electrons due to pair conversions inside of the beam hole, but I wouldn't know without simulating.
>>>
>>> Regarding material and detection efficiency you have many additional things in the way including FDC packages. I don't see how to get 1% absolute efficiency on electron detection in the region of beam hole. How can this be validated with data? Also, I'm not sure how we can get 1% photon efficiency in the FCAL. One would need to know both the FCAL and comp-cal efficiencies to better than 1% to get a 1% measurement on the Compton cross section. (Your physics requirements likely drive the efficiency calibration in GlueX FCAL.)
>>>
>>> I agree you can control position and masking of the detector very precisely, but then you are going to depend on an MC to tell you the detection efficiency. Electron efficiency is a bit tricky because you have bremsstrahlung and material interactions. There is also probably some pileup due to the hundreds of megahertz of low-energy EM backgrounds in comp-cal. I think all of these things conspire to generate complex, non-gaussian energy resolution for electrons in the comp-cal. I'm not sure how to be certain of the efficiency of any kinematic cut at the level of 1%. With the photons there are conversions which can be predicted by MC, but then there is a systematic error.
>>>
>>> How are you calibrating these in PrimEx currently? Is there a physics process that can be used for the calibration or cross check. Note that using electron beams in Hall D is completely forbidden as best I understand.
>>>
>>> I don't mean to sound like a pessimist -- I'm just trying to understand how we plan to do this. In my experience elsewhere, getting absolute reconstruction efficiency for photons at the sub 1% level is very challenging and we had well-measured physics signals to calibrate the detector on.
>>>
>>> Cheers,
>>>
>>> Matt
>>>
>>>
>>> On Nov 30, 2010, at 12:30 PM, Ashot Gasparian wrote:
>>>
>>>>
>>>> Hi Matt,
>>>>
>>>> The total energy cut in HyCal was > 3 GeV, under this condition
>>>> the rate in HyCal was from 1 to 3 KHz and it made the trigger in the
>>>> past experiment. For sure, if we go down to 3 MeV this rate will be
>>>> many orders of magnitude higher.
>>>>
>>>> About the material between the target and FCAL/ComCal: that is right,
>>>> here in my calculations I assumed it is just a He gas (like in PrimEx).
>>>> I am putting some realistic number of material in my GEANT code now
>>>> to run it again. The expected result is that the co-planarity resolution
>>>> in Fig. 3 will go up for a few more degrees (not more). We can live with
>>>> that in the experiment.
>>>>
>>>> About the Compton Cross section and the accepatnce: Yes that is right,
>>>> since this process is for the verification (not for normalization) of
>>>> the primakoff cross section, we need it on the same accuracy level.
>>>> That is to say that we will need to know the acceptances on the same
>>>> level. Yes, there are difficulties on this measurements, but they are
>>>> all solvable. For the geometrical acceptances, the design of the CompCal
>>>> is to cover the central hole of the FCAL. That is we will have the
>>>> coordinates measured with a ~1 mm level. With that we would be able to
>>>> cut the more sensitive areas off from the geometrical acceptance. This
>>>> is how we do in our PrimEx experiment and we are in process of Compton
>>>> cross section publication.
>>>>
>>>> There was one more good question about the possible reflection of
>>>> electrons from the FCAL hole structure. Yes, we agree there is an issue
>>>> here to look, but there are several ways to come up with solutions also.
>>>> In our experiment, with this configuration, we will have several very
>>>> strong cuts for the Compton event selection: the energy conservation
>>>> is already a strong selections for those type of events (delta energy is
>>>> ~ 120 MeV). Further, the angle/energy correlation is another one.
>>>> From the other hand, This kind of situation we have in many places in our
>>>> experimental setup also. I agree, here it is more emphasized.
>>>>
>>>> About the electron detection efficiency: yes, agree it is not solved yet.
>>>> We do that in the PrimEx experiment by moving the HyCal in 3d ways around
>>>> the low intensity beam. There are some other ways we can do it with the
>>>> GlueX, we need to work on this part. You are right.
>>>>
>>>> Thanks,
>>>> Ashot
>>>>
>>>>
>>>> .............................................................
>>>> Ashot Gasparian Phone:(336)285-2112 (NC A&T)
>>>> Professor of Physics
>>>> Physics Department (757)-269-7914 JLab
>>>> NC A&T State University Fax:(757)-269-6273 JLab
>>>> Greensboro, NC 27411 email: gasparan at jlab.org
>>>> .............................................................
>>>>
>>>>
>>>> On Tue, 30 Nov 2010, Matthew Shepherd wrote:
>>>>
>>>>>
>>>>> Hi Ashot,
>>>>>
>>>>> I'm not too concerned about comp-cal radiation damage since we know
>>>> lead tungstate is tuff stuff.
>>>>>
>>>>> It seems like there is about 5-6 orders of magnitude difference
>>>> between my rate estimate and yours. Maybe the high energy showers
>>>> are really this rare, but it should be checked. I imagine GlueX
>>>> design has a lot more conversion material between the target and the
>>>> FCAL than was present in PrimEx.
>>>>>
>>>>> One other question: am I correct that that your cross section
>>>> measurement depends on measuring the Compton scattering rate to
>>>> the 1% level. This means that you need to know the both the acceptance
>>>> (through the FCAL beam hole) and the absolute detection efficiency of
>>>> these Compton events at the 1% level. Is it foreseen how to do this?
>>>> This seems like an extraordinary challenge to calibrate -- especially
>>>> the electron efficiency as the electron will be going through a region
>>>> of the GlueX detector (with the field off) that is not very well
>>>> understood.
>>>>>
>>>>> Cheers,
>>>>>
>>>>> Matt
>>>>>
>>>>>
>>>>> On Nov 30, 2010, at 11:29 AM, Ashot Gasparian wrote:
>>>>>
>>>>>>
>>>>>> Hi Matt,
>>>>>>
>>>>>> I probably forgot in my previous email that that number
>>>>>> (1-3 KHz) is for upper energy cut, (total energy deposition)
>>>>>>> than ~3 GeV. For sure, the rate will go up as we make this
>>>>>> threshold lower, but we are also planning to have a higher
>>>>>> energy region for the trigger.
>>>>>> Other than that I think your numbers are correct. As much as
>>>>>> the radiation is concerns, we do monitor each detector cell by
>>>>>> light monitoring system and also by the pion mass. We did not see
>>>>>> any sizable shift in gains during our experiments where we had
>>>>>> rather intensive photon beams.
>>>>>>
>>>>>> Thanks,
>>>>>> Ashot
>>>>>>
>>>>>>
>>>>>> .............................................................
>>>>>> Ashot Gasparian Phone:(336)285-2112 (NC A&T)
>>>>>> Professor of Physics
>>>>>> Physics Department (757)-269-7914 JLab
>>>>>> NC A&T State University Fax:(757)-269-6273 JLab
>>>>>> Greensboro, NC 27411 email: gasparan at jlab.org
>>>>>> .............................................................
>>>>>>
>>>>>>
>>>>>> On Tue, 30 Nov 2010, Matthew Shepherd wrote:
>>>>>>
>>>>>>>
>>>>>>> Hi Ashot,
>>>>>>>
>>>>>>> It seems that the total count rates you cite 1-3 kHz are orders of magnitude away from what we are used to thinking about.
>>>>>>>
>>>>>>> Take a look at the first plot on this page:
>>>>>>>
>>>>>>> http://www.jlab.org/Hall-D/software/wiki/index.php/FCAL_Backgrounds
>>>>>>>
>>>>>>> It is a bit dated, but should still be approximately correct. There you can see count rates *per block* in the region next to the beam hole of 3 MHz. It is significantly higher inside the beam hole and will probably be higher yet behind the calorimeter where you will get beam interactions with the walls of the hole in the FCAL.
>>>>>>>
>>>>>>> One very important note in this: we are counting energy deposition all the way down to 2 MeV. This is relevant for radiation damage, but not for tagging high energy particles like you would like. We need to check the rates at high energy.
>>>>>>>
>>>>>>> My naive estimate based on these numbers would be a total count rate in the CompCal of a couple hundred megahertz.
>>>>>>>
>>>>>>> -Matt
>>>>>>>
>>>>>>> On Nov 29, 2010, at 7:00 PM, Ashot Gasparian wrote:
>>>>>>>
>>>>>>>>
>>>>>>>> Hi Matt,
>>>>>>>>
>>>>>>>> Thanks for raising good questions. Let me try to come up with
>>>>>>>> my thoughts on your questions:
>>>>>>>>
>>>>>>>> 1) the suggested CompCal detector is actually a very small copy
>>>>>>>> of existing HyCal. It will have 16x16 PbWO4 crystal detectos
>>>>>>>> (with a size: 32x32 cm^2) and central 2x2 crystals removed
>>>>>>>> (that makes a 4x4 cm^2 hole in the center, like the HyCal is)
>>>>>>>> 2) the expected count rates: we had some MC results done in last
>>>>>>>> year for the CompCal detector also. My recollection is that we
>>>>>>>> did not finish it on the level to estimate the count rate. It is
>>>>>>>> in plan for the December, but we have a direct experimental data
>>>>>>>> from the two PromEx-I and PrimEx-II experiment. There we had a
>>>>>>>> similar detector with the similar beam hole and in intencive
>>>>>>>> photon beam: Ie=110 nA, on 10^4 r.l. radiator, no collimator on
>>>>>>>> beam, gives ~ 7x10^7 eq. photons/sec., 5 to 10 r.l. physics targets
>>>>>>>> in beam and the calorimeter located on ~7.5 m down from the targets.
>>>>>>>> Our rates in the proposal are also calculated for a similar beam
>>>>>>>> conditions. The total HyCal rate (for the sum) was from 1 KHZ to
>>>>>>>> 3 KHz depending on beam tune. With these rates we had it as the
>>>>>>>> primary trigger in the experiment with a good physics results.
>>>>>>>> The CompCal will be all the same only the outer dimension much
>>>>>>>> smaller. Based on our experimental reasults, I do not expect any rate
>>>>>>>> problem with this design. Though, we will try to finish the MC
>>>>>>>> simulations left from the last year.
>>>>>>>> 3) for the document for CompCal: I agree with you we need a draft of
>>>>>>>> technical design report for this detector. I will try to do that in
>>>>>>>> the comming weeks and post it in the GlueX wiki.
>>>>>>>>
>>>>>>>> Hope these are answering to your questions in some ways. Please let
>>>>>>>> me know if some of them needs more discussions.
>>>>>>>>
>>>>>>>> Thanks,
>>>>>>>> Ashot
>>>>>>>>
>>>>>>>>
>>>>>>>> .............................................................
>>>>>>>> Ashot Gasparian Phone:(336)285-2112 (NC A&T)
>>>>>>>> Professor of Physics
>>>>>>>> Physics Department (757)-269-7914 JLab
>>>>>>>> NC A&T State University Fax:(757)-269-6273 JLab
>>>>>>>> Greensboro, NC 27411 email: gasparan at jlab.org
>>>>>>>> .............................................................
>>>>>>>>
>>>>>>>>
>>>>>>>> On Mon, 29 Nov 2010, Matthew Shepherd wrote:
>>>>>>>>
>>>>>>>>>
>>>>>>>>> Hi Ashot,
>>>>>>>>>
>>>>>>>>> Do you have a document that outlines the details of CompCal
>>>>>>>> including simulations of the background? Has such a technique
>>>>>>>> been used for other experiments? What is the geometry of this
>>>>>>>> detector -- does it have a beam hole? I imagine count rates
>>>>>>>> for this detector must be incredibly high.
>>>>>>>>>
>>>>>>>>> -Matt
>>>>>>>>>
>>>>>>>>> On Nov 29, 2010, at 5:53 AM, Ashot Gasparian wrote:
>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> Dear Hall D Collaborators,
>>>>>>>>>>
>>>>>>>>>> We are required to submit a short update for the eta-Primakoff
>>>>>>>>>> proposal to this upcoming PAC37 for the beam time assignment and
>>>>>>>>>> scientific rating.
>>>>>>>>>> The proposal was approved a year ago by PAC35.
>>>>>>>>>>
>>>>>>>>>> The first draft of the suggested update is in the attachment of
>>>>>>>>>> this email. The submission deadline is on this Wednesday, so you
>>>>>>>>>> will have a few days to send your critical suggestions and questions
>>>>>>>>>> to us to make this document better for the submission.
>>>>>>>>>>
>>>>>>>>>> Thank you for your participation and valuable support,
>>>>>>>>>> Ashot, Liping
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> .............................................................
>>>>>>>>>> Ashot Gasparian Phone:(336)285-2112 (NC A&T)
>>>>>>>>>> Professor of Physics
>>>>>>>>>> Physics Department (757)-269-7914 JLab
>>>>>>>>>> NC A&T State University Fax:(757)-269-6273 JLab
>>>>>>>>>> Greensboro, NC 27411 email: gasparan at jlab.org
>>>>>>>>>> .............................................................
>>>>>>>>>> <eta_update_PAC37_v3.pdf>_______________________________________________
>>>>>>>>>> Halld-physics mailing list
>>>>>>>>>> Halld-physics at jlab.org
>>>>>>>>>> https://mailman.jlab.org/mailman/listinfo/halld-physics
>>>>>>>>>
>>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>
>>>>>
>>>
>>>
>
>
More information about the Halld-physics
mailing list