AmBe source normalization issue

Dear Francisco,
Thank you for your reply, am really sorry for replying late.

I tried doing what you suggested for photon (4.43 MeV in BEAM) and using USRBIN scored DOSE-EQ. But the output after normalising to the activity of my sample, i get a value of 0.99 microSv/h at 1m( while expected was 25 microSv/h. In the .out file, I can see the 99 % of the particles are escaping the system (must be because of high energy photon, minimum interaction with air…) - please refer input - inairambep0712.flair.
I am still not clear on scoring the gamma dose rate from Am-Be source. Please suggest.

Secondly, the neutron dose expected was 21 microSv/h ( answer to previous post) while I got 105 microSv/h (i/p = 2511ambe2b.flair - in previous post, so I wrote the obtained result is not close to expected result).

[ USRBDX o/p * neutron (photon) yield * flux to dose rate conversion factor (neutron or photon) ] = output in rem/h : kindly confirm is this correct ?

Thanking in advance, with regards,
inairambep0712.flair (3.7 KB)
inairambep0712_22.bnn.lis (632 Bytes)

Dear @raksha,

As a general rule for this and any other post, please provide all the info necessary to reproduce your results (e.g. the activity of your sample), and elaborate on the expected result (i.e. how or from where did you obtain it). This way we can address your issue more efficiently. Please provide these details.

Most of your photons escape, of course, because there is nothing they can interact with apart from air. But this is not a problem if you are interested in dose equivalent and/or particle fluences.

Maybe you are interested in using the CYLI-VOL SDUM option of the BEAMPOS card to define your photon’s initial positions more accurately. Also, it would be better to keep your source regions filled with their respective materials, and not with air.

With respect to the neutron dose, did you modify your simulation as proposed in the previous post? In particular, do you start your simulation directly from the neutron spectrum of your source? If you do not, I really encourage you to do so in order to obtain accurate results. And also, please point me to the source of your expected result.

Finally, the resulting units will of course depend on the units used in the different factors, which you do not specify.

Kind regards,

Dear Francisco,

Thank for your reply.

Please find attached the paper I am trying to reproduce the results of.
I mentioned the dose rates from Table 6 of this paper, the result of para 2.6 - Validation Experiments - second experiment (Schematic in Fig.4). I mentioned NDR of Table 6 with HDPE shield.

Yes, I did start my simulation directly with Neutron beam of 4.4 MeV in the case mentioned.

I would attempt the CYLIN-VOL source too.

With regards,
Raksha Rajput.
IGCAR_AmBe source container_radphyschem.2019.pdf (1.6 MB)

Dear @raksha,

Thank you for sharing the source of the data.

I took the liberty to heavily modify one of your input files to try to reproduce the paper results. Here some warnings:

  • The geometry used in the paper is not crystal clear to me. If we assume a cylindrical shape of the shielding and Fig. 4 to be a section of it, then I do not understand the statement in section 2.6 “…is housed in a HDPE container of height 47 cm and thickness 15 cm… ”. But anyway, I don’t think this is the origin of the issue.
  • Material compositions are not specified, which could bring some uncertainties to the table, but I don’t think this is the issue either.

Here you can find a tgz file 20211213_forum.tgz (25.4 KB) with your modified input file together with the spectra sampling routine and the ISO 8529-1:2001(E) Am-Be spectra as provided in this other topic. This will allow you and anyone interested to give it a try using a more realistic neutron spectrum, as I unsuccessfully suggested you in my reply to your previous post. To make it work, you need to compile it by going to …/CERN-FLUKA-spectra-sampling/src/ in your terminal and type “make” (if an error about the FLUKA directory shows up, change FLUKA_DIR inside the Makefile to point to the directory where you have FLUKA installed and try again). Then you must use the generated executable for those runs using neutrons as primary particles in the Run tab of Flair (select in “Exe:”). Please refer to Sampling from energy spectra for details.

In the input file:

  • Preprocessor directives allow you to chose between photons or neutrons as primary particles as well as a simulations considering the HDPE shielding or using air everywhere. Few examples are already setup in the Flair Run tab.
  • Importance biasing is included when the shielding is used to easily gain statistics outside it.
  • DOSE-EQ cylindrical scorings are used filtering the contribution of neutrons and photons. Note that in the simulation using neutrons as primary particles, we will also get a contribution to the dose by photons coming from neutron-induced reactions.
  • In the Flair Geometry tab, some layers has been created to plot the USRBIN results in top of the geometry. Some normalization is already included, which assumes the numbers stated in the paper you provided, in section 2.1: 4.4e7 neutrons/s and 3.3e7 gamma/s for a 740 GBq source. Then I just scaled this values for the 185 GBq and 37 GBq samples used in the paper and converted the pSv/pr provided by FLUKA to uSv/h. Once the normalized USRBIN is plotted, you can make use of the “info” tool in the Geometry tab of Flair to check the results in the locations of interest.
  • The source is a cylinder of 2 cm radius and 6 cm high.


  1. With neutrons as primary particles and air filling the whole geometry, we can reproduce the results of Tab. 5 in the paper.
  2. Combining the contribution to the dose rate of photons from the simulation using neutrons as primaries and the one using 4.43 gammas as primaries, we can reasonably reproduce the respective result in Table 6.
  3. Using neutrons as primaries and the HDPE shielding, I cannot reproduce the result in the paper either. I obtain about 120 uSv/h instead of the 21 uSv/h of the paper.

I encourage you and anyone interested to take a closer look to make sure I did not introduce some stupid error.

If none, I would recommend you to contact the corresponding author of the paper and see if he/she can shed some light. Special attention should be given to the scoring details: location and methodology. Unfortunately I am not an expert on MCNP so I cannot comment on the ring detector tally they mention to have used.

Hope this helps, do not hesitate to comment back.

Kind regards,

Dear Francisco,

Thank you very much for a detailed reply and thank you for your efforts to work on my input.

I shall be going though the modified input and ask doubts, if any.
(May I request the admin not to close this discussion for some time ? Am engaged in some other work, so may take time to revert. )

Thanks very much again, with regards,

Dear Francisco,

Continuing the discussion from AmBe source normalization issue:

I am sorry that I could not Run the modified input. I was able use ‘Make’ command correctly and attach it to Exe tab in Flair/Run. But while starting the RUN - there is an error (screenshot attached).
The input is saved in my usual working directory.

Please suggest.
Thank you, with regards,

Dear @raksha,

Sorry for the late reply, I have been on vacation for some time. Please change the queue used at the top of the Run tab (next to the number of cycles) from “tsp” to “default” and try again.

Kind regards,