[Halld-offline] Normalization of simulated events

Fri Oct 17 07:51:26 EDT 2014

Hello Sean and all,

I was also concerned about the plan to try to simulate an amorphous
radiator by rocking the diamond to push the primary peak up near the
endpoint. This is not guaranteed to work because there are many peaks in
the diamond formfactor and it is surprisingly difficult to find an
orientation without any coherent peaks visible in the spectrum. Keep in
mind that there are 2 angles needed to orient a diamond, and you are only
specifying one variable when you specify the primary coherent edge, so
there is an algorithm I am using to guess a good set of angles to make a
clean (2,2,0) peak appear where you want it and pretty much nothing else.
Of course you can never get rid of the multiples (4,4,0) and (6,6,0) etc,
but I mean no other q vectors showing up with significant amplitude.

This algorithm works well below 95% of the endpoint, but if you try to rock
the crystal way up past 99% of the endpoint then invariably you will see
other q vectors making visible peaks in the spectrum down below 90%. I have
attached below the spectrum that you get using the configuration you
propose for commissioning simulations with Epeak=0.999*E0. This is not an
amorphous radiator spectrum. The peaks you see are not from the primary
2,2,0 reflection, probably some odd reflection like 3,3,5 that has a
smaller coherent gain, but there are a lot of them.

I recommend that you push the Epeak up past E0 to the unphysical region.
Then the algorithm turns off coherent generation altogether and you get
pure atomic bremsstrahlung. I attach that spectrum below for comparison.

-Richard Jones

On Thu, Oct 16, 2014 at 1:42 PM, Sean Dobbs <s-dobbs at northwestern.edu>
wrote:

> Richard,
>
> Sure, that works.  For the purposes of comparing with previous work, I am
> taking the third argument to the BEAM card to be 0.0012, and let's use the 10
> µm Al radiator.
>
> Can I expect that these rates will scale linearly as a function of the
> thickness of the radiator, or is it more complicated than that?
>
> Thank you.
>
> ---Sean
>
> On Thu, Oct 16, 2014 at 11:41 AM, Richard Jones <richard.t.jones at uconn.edu
> > wrote:
>
>>  Sean,
>>
>>  I suggest the following procedure:
>>
>>    1. You decide what the lower bound on the photon energy spectrum you
>>    want to simulate should be, and set this value as the third argument in the
>>    BEAM control card.
>>    2. Tell me what you have chosen, and what radiator to use, and I will
>>    run my normalization code for a standard current of 100nA.
>>    3. I will tell you how many sim events happen per second of live time.
>>
>> -Richard J.
>>
>> On Thu, Oct 16, 2014 at 11:59 AM, Sean Dobbs <s-dobbs at northwestern.edu>
>> wrote:
>>
>>> Richard,
>>>
>>>  Yes, the BEAM simulations are more relevant for calculating the rates
>>> in the detectors, so it would be better to talk about normalizing them.  I
>>> haven't fully grappled with that calculation.  It's been suggested to me
>>> before to normalize based on an expected rate of photons at the endpoint of
>>> the spectrum, but that doesn't seem very precise to me.  What do you
>>> suggest?
>>>
>>>  Thanks,
>>> Sean
>>>
>>> On Thu, Oct 16, 2014 at 6:36 AM, Richard Jones <
>>> richard.t.jones at uconn.edu> wrote:
>>>
>>>>  Sean,
>>>>
>>>>  In your message, it looks like you are computing a dimensionless
>>>> factor. What does this mean? I don't understand this use of the radiation
>>>> lengths of plastic target thickness. Why is this relevant, when you are
>>>> simulating hadronic interactions in the target? A normalization factor
>>>> should have dimensions like 1/Coulombs, or 1/s at a given current.
>>>>
>>>>  Maybe we should start of asking what you want to calculate, in
>>>> precise terms. Trying to normalize bggen simulations doesn't make a whole
>>>> lot of sense if you are interested in something like total rates in
>>>> counters, because it only includes the hadronic interactions in the target
>>>> from the high-energy part of the beam spectrum. If you want to normalize a
>>>> run with BEAM simulations, then the issue of normalization becomes more
>>>> relevant, but there is no role for plastic target thickness in that
>>>> normalization calculation.
>>>>
>>>>  -Richard J.
>>>>
>>>> On Wed, Oct 15, 2014 at 7:10 PM, Sean Dobbs <s-dobbs at northwestern.edu>
>>>> wrote:
>>>>
>>>>> Hi Richard,
>>>>>
>>>>>  To follow up on the discussion in the Offline meeting today, here's
>>>>> how I've been normalizing event rates for this fall's running.
>>>>>
>>>>>  I've been simulating hadronic photoproduction events using bggen in
>>>>> the range E(gamma) = 0.15 - 10 GeV.  To get an incoherent photon beam, I've
>>>>> set the coherent peak to be near the electron beam energy, e.g., E(beam) =
>>>>> 10 GeV, and E(coherent peak) = 9.999 GeV.
>>>>>
>>>>>  Then I've rescaled the overall event rate from bggen by the beam
>>>>> current, the thickness of the radiator, and the thickness of the target.
>>>>> (Note that the collimator aperture is set to 5mm).
>>>>>
>>>>>  For example, to go from the nominal settings of
>>>>> - beam current = 2250 nA, diamond radiator of 1.7E-4 RL, and 30 cm
>>>>> hydrogen target (3.3% RL)
>>>>> to commissioning settings of
>>>>> - beam current = 100 nA, aluminum radiator of 1.12E-4 RL (10 um
>>>>> thickness), and 2mm thick plastic target (0.25% RL)
>>>>> I applied the factor
>>>>>
>>>>>  (100 nA/2250 nA) x (1.12E-4 / 1.7E-4) x (0.25 / 3.3) = 2.22E-3
>>>>>
>>>>>  Is this correct?
>>>>>
>>>>>  Of course, to accurately get the rates from the EM background, one
>>>>> needs to simulate the beam photons.
>>>>>
>>>>>
>>>>> Cheers,
>>>>> Sean
>>>>>
>>>>>
>>>>>  --
>>>>> Sean Dobbs
>>>>> Department of Physics & Astronomy
>>>>> Northwestern University
>>>>> phone: 847-467-2826
>>>>>
>>>>
>>>>
>>>
>>>
>>>  --
>>> Sean Dobbs
>>> Department of Physics & Astronomy
>>> Northwestern University
>>> phone: 847-467-2826
>>>
>>
>>
>
>
> --
> Sean Dobbs
> Department of Physics & Astronomy
> Northwestern University
> phone: 847-467-2826
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20141017/36d6f12b/attachment-0002.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: E0_10_Epeak_1050.gif
Type: image/gif
Size: 4807 bytes
Desc: not available
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20141017/36d6f12b/attachment-0004.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: E0_10_Epeak_1050.eps
Type: application/postscript
Size: 41992 bytes
Desc: not available
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20141017/36d6f12b/attachment-0004.eps>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: E0_10_Epeak_999.gif
Type: image/gif
Size: 5274 bytes
Desc: not available
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20141017/36d6f12b/attachment-0005.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: E0_10_Epeak_999.eps
Type: application/postscript
Size: 43113 bytes
Desc: not available
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20141017/36d6f12b/attachment-0005.eps>