[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: angular effect



Hi Matt,

I was mostly just curious to see what the Monte Carlo gave for this. A 
KLOE paper had stated that the sampling frequency (number of layers the 
shower crosses) will decrease as sin(theta) decreases and should 
contribute to the energy resolution proprotionally to 1/sin(theta). I 
wanted to see if I or others understood this. (The beam test data shows 
this somewhat but not clearly.) And then maybe put in to HDGeant if needed.

And you're right about how the plots were labelled. I should have stated 
this. The deposited energy sampling fraction seems to go like 
1/sqrt(sin(theta)) and the photon energy sampling fraction goes like 
1/sin(theta). So from what I understand from this is that from the 
sampling frequency alone we lose 1/sqrt(sin(theta)) and we lose from 
shower leakeage out the front another factor of 1/sqrt(sin(theta)). 
However, the resolution in both cases seem to go like 1/sin(theta) so 
both contributions seem to add together, most likely in quadrature, 
which would sort of make sense. Or am I out to lunch on this?

-Blake

Matthew Shepherd wrote:
>
> Hi Blake,
>
> These plots are interesting and certainly something we should explore 
> further.  It looks like the energy dependent (floor in this fit) is 
> still dominant.  I assume the difference between the plots is fraction 
> of energy in module vs. fraction of incident energy.  I'm afraid that 
> this effect is probably correlated with the energy dependence, so we 
> might not model it correctly by taking slices.  We may want to 
> consider a coarse 2-D fit to sigma_f / f.  If we come up with a crude 
> parametrization, this can be implemented easily in the MC by making 
> the assumption that photons originated from the target and thus the 
> true z position (recorded in the MC) can be used to determine the 
> angle.  (It is also possible that the true incident angle can be 
> passed up in the MC, but this is beyond my skill level!)
>
> -Matt
>
>
> On Jul 18, 2007, at 4:33 PM, Blake Leverington wrote:
>
>> Hi Matt,
>>
>> I ran a simulation of the module for a 150 MeV (one for 650 MeV is 
>> still running) with a flat distribution over sin(theta) to see the 
>> effect on the sampling fraction since the sampling frequency 
>> decreases with sin(theta). I've attached a couple of plots that show 
>> the effect.
>>
>> -Blake<2p_samfracEmod_sintheta.gif><2p_samfrac_sintheta.gif>
>