The radiation shielding door located in the main hall of the fusion device is a crucial component of the shielding arrangementss, requiring excellent neutron and photon shielding performance, while maintaining structural strength, seismic stability and airtightness to safeguard both the surrounding environment and personnel safety. Based on a composite shielding structure consisting of a 50 mm carbon steel inner layer, a 500 mm boron-doped polyethylene middle layer, and a 250 mm carbon steel outer layer, a Monte Carlo particle transport program (cosRMC) was used to simulate the shielding performance of the door. The results show that the shielding performance of the radiation shielding door exceeds that of a 1 000 mm thick concrete wall, satisfying the required shielding performance criteria. Considering the functional characteristics of the large radiation shielding door, which has a significant weight and undergoes frequent opening and closing, hinge, drive and locking mechanisms were designed. Finite element analysis performed using ANSYS software indicates that, under normal operating conditions, the main structural components exhibit maximum equivalent stresses of 25.9 MPa, maximum shear stresses of 13.8 MPa, and maximum deformations of 0.048 mm; under seismic design basis conditions, the corresponding values are 30.2 MPa, 16.1 MPa, and 0.052 mm, respectively.All values meet thd design requirements.

The performance optimization of radiation shielding materials remains a central focus in the field of radiation protection. Traditional approaches to shielding material design have relied heavily on extensive experimental data and empirical knowledge, which is not only time-consuming and costly, but also cannot guarantee identification of globally optimal solutions. This study proposes a strategy combining the Aquila Optimizer (AO) with a BP neural network to optimize the composition of X-ray shielding materials and achieve efficient shielding across different X-ray energy segments. Initially, Monte Carlo simulations are employed to establish an X-ray tube model. Functional elements featuring complementary K-absorption edge characteristics are screened via the XCOM program, and Monte Carlo calculations determine the shielding rate for various elemental proportions. Subsequently, a BP neural network is employed to model the non-linear mapping between input parameters (elemental composition) and output parameters (shielding performance). SHAP (SHapley Additive Explanations) interpretability is applied to quantify the contribution of each element to the shielding rate. The AO algorithm is subsequently employed to determine the optimal elemental proportion scheme. Finally, Monte Carlo simulations are utilized for performance testing and comparative analysis of the optimized composition. Results indicate that for the composition W∶Bi∶Gd∶Sm∶SEBS= 0.018 8∶0.261 0∶0.058 1∶0.162 1∶0.500 0, a shielding rate of 75.51% is achieved at 100 kV tube voltage, with a material density of 1.600 7 g/cm³. Additionally, an in-depth investigation of optimal functional element proportions for different energy segments was performed. This method demonstrates significant innovation and effectiveness, thus enriching the computational methods for research, development, and application optimization of composite shielding materials.

During nuclear accidents, accurate estimation of source term parameters is a critical component of nclear accident consequence assessment and emergency response. However, during the transient phase of an accident, timely acquisition of monitoring data is frequently hindered by conditions including high temperature, intense radiation, and instrumentation failure. Therefore, this paper proposes two deep learning-based integrated models for source term inversion to enhance both accuracy and robustness. The first model is an LSTM-Transformer hybrid architecture, which fully utilizes the advantages of the Long Short-Term Memory (LSTM) network in handling long-term dependencies, non-stationarity, and gradient stability in time series,thereby effectively capturing the dynamic evolution of radioactive nuclide dispersion. It also introduces the Transformer structure to leverage self-attention to modelglobal dependencies between different time steps in the sequence, significantly enhancing long-range contextual awareness . The second model is a dual-branch fusion LSTM model (Dual-Branch LSTM with Fusion, DBL-LSTM), which employs two parallel LSTM networks to independently extract features from gamma dose rate measurements and meteorological parameters, followed by feature-level integration via a fusion LSTM model for comprehensive modeling. This approach improves the capability for joint analysis of multi-source data and enhances adaptability to meteorological variability and noise. The Optuna automated hyperparameter optimization algorithm is integrated to further improve inversion accuracy and reduce uncertainties in manual parameter tuning. Based on the simulated data generated by the International Radioactive Assessment System (InterRAS), the release rates of three short-lived nuclides, Sr-91, Te-132, and I-131, were estimated, with a single LSTM model serving as the baseline comparator and the Mean Absolute Percentage Error (MAPE) as the evaluation indicator. Results indicate that the single LSTM achieved MAPE values of 15.10%, 8.20%, and 27.33% for ⁹¹Sr, ¹³²Te, and ¹³¹I respectively; the LSTM-Transformer model reduced these to 13.90%, 5.63%, and 25.63%; and the DBL-LSTM achieved further improvements, yielding MAPEs of 13.34%, 4.93%, and 24.21%.The integrated models demonstrate consistent improvements in both inversion accuracy and robustness relative to the single LSTM model, highlighting their potential for application in nuclear accident emergency scenarios.

To address the challenge of accurately determining the directional dose equivalent rate $\dot{H}^{\prime}$(0.07) in β-γ mixed radiation fields, a novel measurement methodology is proposed. This approach integrates measurements from ambient dose equivalent rate monitors, directional dose equivalent rate monitors, β spectrometers, and γ spectrometers. A β-γ mixed standard radiation field was established to verify this method.Following application of the methodology proposed in this study, the relative error between the measured $\dot{H}^{\prime}$(0.07, 45°) value and the reference value traceable to a certified standard instrument was reduced from 13.0% to 2.69%, verifying the method’s reliability.In field applications, $\dot{H}^{\prime}$(0.07) was measured and data were processed at 14 locations exposed to β-γ mixed radiation fields during the No.106 overhaul of Hainan nuclear power plant. The measured H·'(0.07)$\dot{H}$^*(10) ratios ranged from 2.06 to 29.1, with a mean value of 9.64—indicating that $\dot{H}^{\prime}$(0.07) was, on average, approximately one order of magnitude greater than $\dot{H}$^*(10). Accurate quantification of $\dot{H}^{\prime}$(0.07) in β-γ mixed fields is therefore essential for reliable assessment of occupational exposure risk and provides critical input for radiation protection decision-making in nuclear facilities.

This work evaluates the applicability of the Leading Nuclide Correlation (LNC) method to a specific low-background large-object radioactive contamination monitor, using the MCNP5 code. A MCNP model was developed to determine the key parameters required for the LNC method. Considering actual nuclear power plant conditions, the F8 tally was employed to calculate detector count rates across varied scenarios—including different materials, densities, filling fractions, and nuclide distributions. The activity values obtained from the LNC method and the equipment’s built-in algorithm were then compared against the reference values. The results indicate that the relative deviation associated with the LNC method remained consistently within ±15% across all experimental conditions, whereas the built-in algorithm yielded relative deviations as high as -77.5% under extreme scenarios, demonstrating the superior performance of the LNC method. This study confirms the technical applicability of the LNC method for clearance measurements of radioactive solid waste at nuclear power plants using this type of equipment.

A hierarchical assembly rapid prototyping method was proposed on the construction of radiation-equivalent physical phantoms. Polylactic acid (PLA), a customized bone-equivalent material, and polyvinyl chloride (PVC) foam were employed to simulate soft tissue, bone tissue, and lung tissue, respectively. Soft tissue, bone tissue, and lung tissue components are fabricated with high geometric fidelity through the integrated application of fused deposition modeling (FDM), investment casting, and computer numerical control (CNC) machining.Performance verification was conducted with respect to CT value consistency, geometric accuracy, and prototyping efficiency. The results indicate that the proposedmethod achieves excellent structural fidelity, tissue equivalence, and precise control over manufacturing cycle, thereby establishing a reliable benchmark platform for radiation dose simulation, radiotherapy validation, and radiation protection assessment.

To enable quantitative characterization of particle transport performance in stack effluent sampling nozzles, a shrouded sampling nozzle was designed in compliance with ISO 2889,"Sampling of airborne radioactive materials from stacks and ventilation ducts of nuclear facilities",and a transport ratio testing method was developed, and experimentally validated. Results show that at the design sampling flow rate of 100 L/min and free-stream velocities ranging from 8.0 to 12.0 m/s, the transport ratio remained within 1.00 to 1.10, meeting the ISO-recommended range of 0.80 to 1.30. Good agreement was observed between experimental data and theoretical predictions, with relative errors below 10%. These results indicate that the proposed nozzle exhibits stable particle transport behavior under the target operating conditions, fully complying with the acceptance criteria.

Radioactive monitoring of atmospheric aerosols in Chengdu was performed using a high-volume continuous sampling method. Six radionuclides (⁷Be, ²²⁸Ra, ¹³⁷Cs, ⁴⁰K,²¹⁰Pb and ¹³¹I) were quantified, and their activity concentrations fell within the background levels. The activity concentrations of ⁴⁰K, ²¹⁰Pb, and ²²⁸Ra exhibit similar trends to those of PM_2.5 and PM₁₀ concentrations, but are opposite to the trend of O₃ concentration. In contrast, ⁷Be demonstrates a negligible correlation with all three atmospheric parameters.This study enhances the foundational dataset on radionuclide concentrations in atmospheric aerosols in Chengdu,providing a scientific basis for the prevention and control of radioactive air pollution in the region.

To assess radioactive levels in non-uranium metal mines in Shanxi Province and evaluate their potential occupational health implications, 9 metal mines were selected in this study. Cumulative radon concentrations were measured in both mine shafts andground-level workplaces using the CR-39 solid track detector system. Results indicated that radon concentrations in the underground air of 5 metal mines exceeded 1 000 Bq/m³.Variance analysis revealed statistically significant differences in radon concentrations between 3 of the mines and the remaining 6 mines (P<0.05). High radon concentrations in the underground air of three mines were observed at the ore crushing sites and locations exhibiting groundwater seepage and water inrush.The findings of this survey and measurement indicate elevated radon concentrations in certain non-uranium metal mines, warranting the implementation of targeted mitigation measures to reduce miners' risk of radon-induced lung cancer.