Reproducing ICRP 144 Air Kerma Values for Monoenergetic Photons in FLUKA
Please provide the used software versions.
FLUKA: 4-3.4
Flair: 3.3-0.2
Description
Dear FLUKA Experts,
I am attempting to reproduce the free-in-air air kerma values (nGy·m³/Bq·h) for monoenergetic photons, as reported in ICRP Publication 144, using FLUKA. (files attached)
Geometry Setup:
A 30 cm diameter ICRU sphere is placed inside a coupling cylinder (CC) of 60 cm diameter and 200 cm height. This setup is enclosed within a hemisphere whose radius is based on the mean free path (MFP) of photons in air (e.g., ~480 m for 0.5 MeV photons, corresponding to 5 MFP). All regions are filled with air.
Due to the large source-to-target distance, ICRP 144 recommends a two-step simulation approach to improve computational efficiency.
According to ICRP 144: “To reduce the variance of the Monte Carlo simulations, the uniform source was reproduced by increasing the number of photons or electrons emitted per unit area and decreasing the Monte Carlo weight of photons or electrons as their emission point approached the coupling cylinder.”
I have followed this methodology as outlined below:
Step 1: Photon Sampling and Phase Space Generation
Photons are sampled uniformly within the hemisphere (radius = 5 MFP), and their phase space (position, energy, direction, weight, and particle type) is recorded at the surface of the coupling cylinder using the USERDUMP and mgdraw.f routine. To eliminate contributions from particles exiting the CC, the CC region is treated as a blackhole during this step.
For this study, I have used 0.5 MeV photons (radius = 480 m). The hemisphere is divided into 10 concentric layers for variance reduction.
Variance Reduction Logic (details in attached Excel sheet):
Natural probability × Initial weight = Biased probability × Biased weight
Natural probability = (Layer volume / Total hemisphere volume)
Initial weight = 1.0
Biased probability = User-defined
Biased weight = Calculated accordingly
Thus, more photons are sampled near the CC with higher biased probabilities but are assigned proportionally lower weights.
Step 2: Scoring Air Kerma Rate
The phase space data from Step 1 is used to irradiate the 30 cm sphere, where air kerma is scored. Initially, both photons (primary + bremsstrahlung) and secondary electrons were recorded at the CC surface. However, in Step 2, only photons were transported; secondary electrons were not considered.
Kerma Calculation Methods (see Excel RESULTS.xlsx):
- Energy Deposition (USRBIN)
- Fluence Method (USRTRACK)
◦ Converted to kerma using:
(a) Kerma-to-fluence conversion coefficients from ICRP 74 (see FLUENCE-1 sheet)
(b) μ/ρ approach using NIST database (see FLUENCE-2 sheet)
Results for 0.5 MeV photons:
• USRBIN card: 6.66E+04 nGy·m³/Bq·h
• USRTRACK card: ~7.90E+04 nGy·m³/Bq·h
• Reference value from ICRP 144: 0.109 nGy·m³/Bq·h
I have the following queries:
- Is the calculation in the attached Excel sheet correct? Could you please check and verify?
- Does the m³ unit in the result correspond to the volume of the hemisphere/environmental field? Should this volume be multiplied into the result?
- Is the variance reduction logic used in the source routine valid? If not, is there a recommended logic or biasing scheme suited for this type of source sampling?
- Do the Step 2 results (kerma scoring) require further normalization to match ICRP 144 values?
◦ If yes, how should the normalization be done—especially when multiple cycles and spawns are used for primaries in both Step 1 and Step 2?
◦ The .out file states: “Please remember that all results are normalized per unit weight.”
Could you please clarify how this affects the final result? - Is there a better approach to solve this problem in FLUKA?
Thank you for your time and support.
Best regards,
Rahul Roy
FILES.zip (4.4 MB)
Input files
Please upload all relevant files. (FLUKA input file, Flair project file, user routines, and data files)