🔬 physics

Assessment of the neutron fluence during operation of CEMHTI Cyclotron using Monte-Carlo codes - Simβ-AD study

The Simβ-AD study established a methodology for assessing radioactive waste activation at the CEMHTI Cyclotron by validating neutron fluence measurements from activation foils against four Monte-Carlo codes (FLUKA, MCNP6, PHITS, and RayXpert®), demonstrating good agreement between experimental data and simulations using simplified models.

Original authors: Jean-Michel HORODYNSKI, Abir HASSANI, Frédéric CHAPELLE, Nicolas ARBOR, Djokhar BETELGUERIEV, Lucia Victoria GARCIA GARCIA, Stéphane HIGUERET, Léo NOWAK, Cédric DOSSAT, Sébastien BOUILLON, David CHAUL

Published 2026-07-07

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Jean-Michel HORODYNSKI, Abir HASSANI, Frédéric CHAPELLE, Nicolas ARBOR, Djokhar BETELGUERIEV, Lucia Victoria GARCIA GARCIA, Stéphane HIGUERET, Léo NOWAK, Cédric DOSSAT, Sébastien BOUILLON, David CHAULIN, William HATE, Patrice RIFARD, Thierry SAUVAGE

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: Predicting the "Ghost" Left Behind

Imagine you have a very powerful machine (a cyclotron) that shoots tiny particles like protons and deuterons at a target. When these particles hit, they don't just stop; they create a chaotic shower of invisible "ghosts" called neutrons. These ghosts bounce around the room, hitting walls, magnets, and equipment, turning them slightly radioactive.

The main goal of this study was to figure out: Can our computer simulations accurately predict how "radioactive" these ghosts make the room?

The researchers wanted to build a reliable recipe (methodology) to measure this radioactivity without having to break everything apart later. This is crucial for managing radioactive waste safely.

The Cast of Characters

The Cyclotron: A giant particle accelerator at the CEMHTI facility in France. It's like a high-speed racetrack for tiny particles. It recently shut down, making it the perfect place to test ideas before the track is gone forever.
The "Ghosts" (Neutrons): Invisible particles created when the beam hits its target. They are the main culprits that make other materials radioactive.
The "Sticky Traps" (Activation Foils): The researchers placed small, thin sheets of metal (Gold, Scandium, Tantalum, and Terbium) in different spots around the room. Think of these as sticky flypaper. When the neutron ghosts fly by, they stick to the metal, changing it slightly. Later, the researchers measured how much "stickiness" (radioactivity) was on each sheet.
The "Crystal Balls" (Monte-Carlo Codes): These are four different super-computer programs (FLUKA, MCNP6, PHITS, and RayXpert). They are like weather forecasters. You feed them the map of the room and the speed of the particles, and they try to predict exactly how many ghosts will hit the sticky traps.

The Experiment: A Reality Check

The team ran the cyclotron with two types of beams (protons and deuterons) and placed their "sticky traps" in various locations: near the beam exit, near the magnets, and against the walls.

After the machine was turned off, they took the metal sheets to a lab and measured exactly how radioactive they had become. Then, they compared these real-world numbers with the predictions made by the four computer programs.

The Results: How Did the Crystal Balls Do?

The researchers found that the computer programs were actually quite good at guessing the outcome, but not perfect.

The Score: The computer predictions were usually within 50% to 150% of the real measurements. In the world of complex physics, this is considered a "good agreement." It's like a weather forecaster saying "it will rain between 1 and 3 inches," and it actually rains 2 inches.
The "Gold" Glitch: One of the computer programs (FLUKA) had a specific trouble with the Gold sheets. It seemed to guess the wrong amount of radioactivity because it wasn't sure how to handle a specific "state" of the gold atoms (like confusing a sleeping cat with an awake one). When they fixed this logic, the predictions got much better.
The "Corner" Problem: Near the beam exit (where the particles hit first), the predictions were a bit more scattered. This is like trying to predict how water splashes when you throw a stone in a pond; the ripples near the splash are chaotic and hard to model perfectly. The computers sometimes underestimated the low-energy "ghosts" in these tight corners.

Why This Matters (According to the Paper)

The study proves that you can use these computer codes to estimate how much radioactive waste a cyclotron facility will produce.

The "Beta" Problem: Some radioactive materials only emit a type of energy called "beta" radiation, which is hard to detect without destroying the object. The paper suggests that if we can accurately predict the neutron activity using these codes, we can estimate the "beta-only" waste without having to cut open the machinery.
The Takeaway: By combining the computer models with a few real-world measurements (the sticky traps), we can create a reliable method to manage radioactive waste from particle accelerators.

Summary in a Nutshell

The researchers tested four different computer programs to see if they could predict how much a room would become radioactive after a particle accelerator ran. They used real metal sheets as "sensors" to catch the invisible particles. The computers did a decent job (within a factor of 1.5), proving that we can use these digital simulations to plan for radioactive waste management, provided we know exactly how the machine behaves and fix a few small logic errors in the software.

Technical Summary: Assessment of Neutron Fluence during Operation of CEMHTI Cyclotron using Monte-Carlo Codes

Problem Statement
Effective radioactive waste management for cyclotron facilities requires accurate radiological characterization, particularly for $\beta$ -only radionuclides (e.g., $^{36}$ Cl, $^{49}$ V, $^{55}$ Fe, $^{63}$ Ni) which necessitate destructive measurement methods. The Sim $\beta$ -AD study was launched to establish a methodology combining Monte-Carlo (MC) codes and active neutron detectors to assess material activation within cyclotron facilities. The specific challenge addressed in this paper is the benchmarking of MC codes against experimental data to validate neutron fluence calculations and define the ratio between $\beta$ -only and $\gamma$ -emitting radionuclides. This work focuses on the CEMHTI Cyclotron facility, which operated from 1976 until its shutdown in 2023, providing a unique opportunity to validate models against real-world operational data before the facility's decommissioning.

Methodology
The study employed a comparative approach between experimental measurements and numerical simulations using four distinct Monte-Carlo codes: FLUKA, MCNP6, PHITS, and RayXpert.

Geometrical and Physical Modeling:
- A high-fidelity model of the CEMHTI bunker, experimental areas, cyclotron components (magnets, yoke, inflector), and targets was constructed based on blueprints and in-situ measurements.
- Two primary beam scenarios were modeled:
  1. Proton Beam: 16 MeV, 25 µA stopped in a Tantalum block, with a secondary loss on the deflector.
  2. Deuteron Beam: 17.5 MeV, 30 µA stopped in a Beryllium target.
- Specific physics settings were optimized for each code to enhance neutron production statistics, including the use of specific cross-section libraries (JEFF 3.3, TENDL-2019/2021/2024, ENDF/B-VIII.0) and variance reduction techniques (e.g., forced collisions, importance sampling).
Experimental Setup:
- Passive activation foils made of Gold (Au), Scandium (Sc), Tantalum (Ta), and Terbium (Tb) were placed at various locations (extraction, pipe, yoke, deflector, wall, door, magnet) during two irradiation campaigns in 2021.
- Foils were irradiated for 5 hours (protons) and 2 hours (deuterons).
- Post-irradiation, $\gamma$ -spectroscopy was performed using a 40% efficiency ISOCS detector to quantify the activity of $\gamma$ -emitting radionuclides ( $^{198}$ Au, $^{46}$ Sc, $^{182}$ Ta, $^{160}$ Tb).
Activation Calculations:
- Neutron fluences derived from the MC codes were used as input for offline activation calculations using FISPACT-II (with TENDL-2019 libraries) and a pre-version of the RayActive code.
- FLUKA also performed online activation calculations using its RESNUCLEI card.

Key Results
The study compared measured activities against calculated values, yielding the following findings:

Overall Agreement: Good agreement was achieved between measurements and calculations across all four codes, with ratios of measured-to-calculated activity falling within the 0.5–1.5 range.
Proton Beam:
- FLUKA online calculations showed good agreement except for Gold (Au) foils, where an arbitrary 50/50 ground-state/isomer ratio in the RESNUCLEI card led to discrepancies. This was resolved using FISPACT-II.
- Codes showed higher discrepancies for foils located near beam loss points (deflector, pipe), particularly for Au, Ta, and Tb, which are sensitive to low-energy neutrons. This suggests variations in the calculation of low-energy neutron fluence at small angles relative to the beam.
- RayXpert and MCNP6 exhibited similar trends, while PHITS followed FLUKA's trend in most cases.
Deuteron Beam:
- Calculations using FISPACT-II coupled with fluences from FLUKA, MCNP6, and PHITS showed good agreement.
- FLUKA online calculations significantly underestimated Au foil activity due to the isomer equilibrium issue.
- Other codes tended to overestimate activation, likely due to an overestimation of low-energy neutrons from the Be(d,n) reaction and subsequent thermalization in concrete walls.
Code Performance: RayXpert and MCNP6 showed higher discrepancies compared to other codes near beam-loss points, attributed to the underestimation of low-energy neutron fluence at small angles.

Significance and Claims
The paper asserts that neutron fluence estimation using Monte-Carlo codes, when paired with relatively accurate facility models, provides a reliable method for assessing the activation of passive detectors in cyclotron environments. The study demonstrates that:

Methodology Validation: The combination of MC codes and offline activation tools (FISPACT-II, RayActive) is a viable methodology for characterizing radioactive waste activation, achieving results within an acceptable performance range (0.5–1.5 ratio).
Uncertainty Sources: The primary source of uncertainty for the proton beam scenario is identified as the lack of precise data regarding beam loss intensity and shape within the accelerator.
Future Application: With the CEMHTI cyclotron now shut down, the validated methodology will be applied to compare numerical simulations with measurements taken on decommissioned cyclotron parts and concrete walls. This will facilitate further studies on specific differences between MC codes, cross-section libraries, and neutron thermalization behaviors, ultimately supporting the Sim $\beta$ -AD goal of defining $\beta$ -only to $\gamma$ -emitting radionuclide ratios for waste management.

The authors conclude that while improvements in beam parameter knowledge are necessary, the current results confirm the utility of these simulation tools for radiological characterization in accelerator facilities.