Re: [fluka-discuss]: Dose rate at a point from Multiple sources

From: Raksha Rajput <rakshak_at_britatom.gov.in>
Date: Sat, 18 Apr 2020 12:20:00 +0530 (IST)

Dear Experts,
Thank you for the kind explanation.
I tried attempting the run with 20 cycles, errors are reduced. Am going in steps of 10 cycles, to see the errors getting reduced.

Here is my geometry:
Multiple cylindrical Co-60 sources arranged in a cylindrical cage inside the lead shield thickness of approx..26 cm all around.
The dose rates need to be calculated on the side surface of the shield as well as on the top surface.
Kindly suggest.

Thanks and regrads,
Raksha.

----- Original Message -----
From: answers_at_pcfluka.mi.infn.it
To: "Raksha Rajput" <rakshak_at_britatom.gov.in>
Cc: "paola sala" <paola.sala_at_mi.infn.it>, fluka-discuss_at_fluka.org
Sent: Thursday, April 16, 2020 7:24:37 PM
Subject: Re: [fluka-discuss]: Dose rate at a point from Multiple sources

Dear Raksha,
as you said, in the ##.lis file you find the error in percentage that
refers to the corresponding value in the matrix of values.

As an example, if you have 10 a.u. +/- 10% this is equivalent to (10+/-1) a.u.

Remember that most scoring cards in FLUKA automatically normalize the
results to the number of primary particles, i.e. your results are
independent on the number of primaries that you simulated.

There is no recipe for the optimal number of primaries; that depends on
the maximum uncertainty that you can accept in the results for your
purposes. However, keep in mind that the statistical uncertainties is
proportional to 1/(N)^(1/2) where N is the number of primaries.

It is usually recommended to run at least 5-8 cycles, to calculate the
mean values with reasonably small statistical uncertainties.

Regarding the biasing, from the results in your attachment, it seems that
the biasing has no effect. Indeed, the results from the first and the last
bin vary only by a factor of about 10. Therefore, the biasing you use (up
to 60000) increases significantly the running time of your simulations,
without appreciable gain in your results. In this case it would/may be
more convenient to run more cycles, instead (e.g. 100).

However, if you provide us with more detailed information about your
geometry we may support and suggest you better.

I hope this helps.



On Wed, 15 Apr 2020, Raksha Rajput wrote:

> Dear Experts,
>
> After running the input as suggested, I am getting the results. The doubts now are:
>
> How to interpret the statistical errors in the result ( 2*87 = 114 bins)? The errors range from 4 % to 21.5 % . How should the errors be compared for different number of primaries ? How to optimize the number of primaries and also no. of cycles ?
>
> I also tried Importance Biasing (three regions are biased, each is divided into 14 regions using Planes i.e., surface splitting) and a factor of 2.2 is used for increasing the importance.
> How to compare errors in 'with' and 'without' Bias results. I have attached the .lis files of the results. I have attempted it for the first time. Kindly suggest corrections.Please see the use of Biasing as follows:
>
> #if BIAS
> BIASING 0.0 1. BLKBODY _at_LASTREG 1.PRINT
> BIASING 2. 2.2 Sh1 Sh1 1.PRINT
> BIASING 2. 4.84 Sh2 Sh2 1.PRINT
> BIASING 2. 10.65 Sh3 Sh3 1.PRINT
> BIASING 2. 23.43 Sh4 Sh4 1.PRINT
> BIASING 2. 51.54 Sh5 Sh5 1.PRINT
> BIASING 2. 113.38 Sh6 Sh6 1.PRINT
> BIASING 2. 249.44 Sh7 Sh7 1.PRINT
> BIASING 2. 548.76 Sh8 Sh8 1.PRINT
> BIASING 2. 1207.27 Sh9 Sh9 1.PRINT
> BIASING 2. 2655.99 Sh10 Sh10 1.PRINT
> BIASING 2. 5843.18 Sh11 Sh11 1.PRINT
> BIASING 2. 12855. Sh12 Sh12 1.PRINT
> BIASING 2. 28281.01 Sh13 Sh13 1.PRINT
> BIASING 2. 62218.21 Sh14 Sh14 1.PRINT
> BIASING 2. 2.2 A7a A7a 1.PRINT
> BIASING 2. 4.84 A7b A7b 1.PRINT
> BIASING 2. 10.65 A7c A7c 1.PRINT
> BIASING 2. 23.43 A7d A7d 1.PRINT
> BIASING 2. 51.54 A7e A7e 1.PRINT
> BIASING 2. 113.38 A7f A7f 1.PRINT
> BIASING 2. 249.44 A7g A7g 1.PRINT
> BIASING 2. 548.76 A7h A7h 1.PRINT
> BIASING 2. 1207.27 A7i A7i 1.PRINT
> BIASING 2. 2655.99 A7j A7j 1.PRINT
> BIASING 2. 5843.18 A7k A7k 1.PRINT
> BIASING 2. 12855. A7l A7l 1.PRINT
> BIASING 2. 28281.01 A7m A7m 1.PRINT
> BIASING 2. 62218.21 A7n A7n 1.PRINT
> BIASING 2. 2.2 A14a A14a 1.PRINT
> BIASING 2. 4.84 A14b A14b 1.PRINT
> BIASING 2. 10.65 A14c A14c 1.PRINT
> BIASING 2. 23.43 A14d A14d 1.PRINT
> BIASING 2. 51.54 A14e A14e 1.PRINT
> BIASING 2. 113.38 A14f A14f 1.PRINT
> BIASING 2. 249.44 A14g A14g 1.PRINT
> BIASING 2. 548.76 A14h A14h 1.PRINT
> BIASING 2. 1207.27 A14i A14i 1.PRINT
> BIASING 2. 2655.99 A14j A14j 1.PRINT
> BIASING 2. 5843.18 A14k A14k 1.PRINT
> BIASING 2. 12855. A14l A14l 1.PRINT
> BIASING 2. 28281.01 A14m A14m 1.PRINT
> BIASING 2. 62218.21 A14n A14n 1.PRINT
> #endif
>
>
> Thanks in advance.
> Regards,
> Raksha.
>
>
>
> ----- Original Message -----
> From: "paola sala" <paola.sala_at_mi.infn.it>
> To: "Raksha Rajput" <rakshak_at_britatom.gov.in>
> Cc: answers_at_pcfluka.mi.infn.it, fluka-discuss_at_fluka.org
> Sent: Monday, April 13, 2020 12:41:31 PM
> Subject: Re: [fluka-discuss]: Dose rate at a point from Multiple sources
>
> Correct!
> Paola
>> Dear Experts,
>>
>> Just to confirm: whether the DOSE EQ will give the result in pSv/particle.
>> I am scoring in R-Z binning.
>> If I need to convert it for one Curie, i need to multiply it by (3.7E+10 *
>> 3600) , I will obtain it in pSv/h from one Curie source ?
>> Please advise.
>> Thanks and regards,
>> Raksha.
>>
>> ----- Original Message -----
>> From: answers_at_pcfluka.mi.infn.it
>> To: "Raksha Rajput" <rakshak_at_britatom.gov.in>
>> Cc: fluka-discuss_at_fluka.org
>> Sent: Friday, April 10, 2020 9:12:21 PM
>> Subject: Re: [fluka-discuss]: Dose rate at a point from Multiple sources
>>
>> Hello
>> In order to score DOSE-EQ you do not need to define specific regions, and
>> it is much better to leave air all around your shield.
>> The calculation of DOSE-EQ starts from particle fluences, not from
>> deposited energy. Particle fluences are convoluted with conversion factors
>> that depend on particle type and energy. The standard conversion
>> coefficients in fluka transform particle fluences in ambient dose
>> equivalent (H*(10)).
>> If you have cylindrical symmetry, You can define a cylindrical USERBIN
>> around your shield, with one or more inradial intervals and how many
>> longitudinal intervals as you need, then pick up the dose values in the
>> bin you need. If you do not have cylindrical symmetry, you can use
>> r-phi-z binning, or a cartesian binning.
>>
>> If you ask for DOSE-EQ, do not put water around, it will absorb the
>> radiation and change the dose value.
>>
>> And, do NOT use the AUXSCORE to select photons only: in reality, the
>> photons will produce electrons, that are the particles really depositing
>> energy. If you need dose you have to consider electrons. So, no AUXSCORE
>> card.
>>
>> If, instead, you really need dose in water, where dose is defined
>> as
>> deposited energy/ unit mass
>> then ask for DOSE in your region-by-region userbin.
>> AND do NOT use the AUXSCORE card, otherwise you will miss almost all
>> the deposited energy, because it is deposited by secondary electrons.
>> Photons do not deposit energy directly, it only happens in the simulation
>> because there are transport thresholds: photons below threshold
>> stop and their energy is counted as deposited there.
>>
>> Hope this helps
>> Paola
>>
>> On Thu, 9 Apr 2020, Raksha Rajput wrote:
>>
>>> Dear Experts,
>>>
>>> I have a cylindrical source cage, with Co-60 cylindrical source at 25
>>> positions, equidistant. I wish to calculate dose at different points on
>>> the
>>> surface of the lead shielding. How should I calculate the dose rate at a
>>> point ? Presently I am scoring USRBIN/DOSE-EQ with AUXSCORE/PHOTON in
>>> SPH
>>> bodies of 5 cm radius, touching the surface of my shielding flask.
>>>
>>> Referring to the link below, I got answers to my questions but only for
>>> point no.5, please throw some more light, or suggest the specific
>>> options in
>>> the USRBIN card. 
>>> https://pcfluka.mi.infn.it/FLUKA/web_archive/earchive/new-fluka-discuss/159
>>> 80.html
>>>
>>> Thanks and regards,
>>> Raksha.
>>>
>>>
>>>
>>>
>>
>>
>>
>> __________________________________________________________________________
>> You can manage unsubscription from this mailing list at
>> https://www.fluka.org/fluka.php?id=acc_info
>>
>
>
> Paola Sala
> INFN Milano
> tel. Milano +39-0250317374
> tel. CERN +41-227679148
>



__________________________________________________________________________
You can manage unsubscription from this mailing list at https://www.fluka.org/fluka.php?id=acc_info
Received on Sat Apr 18 2020 - 10:43:55 CEST

This archive was generated by hypermail 2.3.0 : Sat Apr 18 2020 - 10:44:04 CEST