Abdelmeguid Abouellail

The study focuses on modeling a device designed for conducting experiments under low reactor neutron fluence rate conditions. The relevance of the work is driven by the challenges of accurately measuring low reactor neutron fluence rates, which are critical for the safety and efficiency of nuclear reactors and for conducting research in various scientific fields. To create a directed neutron flux, a horizontal collimation channel of the Prizm-AN irradiation device with a radioisotope capsule neutron source of the internal bounded neutron (IBN) type is being investigated. The paper presents the results of particle transport modeling using the PHITS and SOURCES-4C codes, which allowed for the evaluation of the generated neutron fluence rate parameters and the effective dose rate of radiation. The obtained results confirm the functionality of the proposed device and its compliance with radiation safety requirements, making it suitable for conducting experiments similar to those performed on reactor installations.

Surface inspection under large strains is a critical problem for both civil and mechanical engineering. Optical techniques can be used for its solution. Digital image correlation (DIC) is one of these promising optical techniques; it is used to measure displacement and strain fields without contact with a surface under study. Input parameters can directly affect the strain field calculation accuracy. The correlation method is a key factor in DIC calculation of a strain field, especially under large strains. In this work, three spatial and incremental methods are tested; their accuracy and applicability are estimated; their advantages and drawbacks are discussed. The results can be useful for the developments of software for non-destructive testing.

Modern personal computer machines with high-capacity processors allow the application of advanced techniques in the field of testing and inspection for civil and mechanical engineering. Digital Image Correlation is currently one of the most cost-effective and user-friendly optical and noncontact techniques for strain field measurements. Pydic is a free and open source Python code for local 2D Digital Image Correlation. It was developed by Damien André, SPCTS/ENSIL-ENSCI, Limoges, France. Pydic is based on OpenCV, a well-known open-source computer vision and machine learning library. In this work, the Pydic code was validated for strain field measurements using a set of speckled images provided by the SEM DIC challenge. The same set of speckled images was used to compute the strain field with GOM correlate software. An accuracy assessment was made by comparing the results of Pydic code and GOM correlate software; also, the results published in the DIC challenge report were considered in the comparison. The results show an acceptable agreement by visual comparison and the computed strain field range. The weaknesses of Pydic code were highlighted, and future work for code improvement was discussed.