Model-based reconstruction for enhanced x-ray CT of dense tri-structural isotropic particles

Area of Science:

• Nuclear Engineering
• Materials Science
• Imaging Science

Background:

• Tri-structural isotropic (TRISO) fuel particles are crucial for advanced nuclear reactors.
• X-ray computed tomography (CT) is vital for TRISO particle characterization.
• Conventional CT methods face challenges like photon starvation and long scan times due to TRISO particle composition.

Purpose of the Study:

• To develop and apply model-based image reconstruction (MBIR) algorithms for improved CT imaging of TRISO fuel particles.
• To overcome limitations of traditional CT techniques, such as streak artifacts and noise.

Main Methods:

• Development and application of novel model-based image reconstruction (MBIR) algorithms.
• Comparison of MBIR performance against conventional Feldkamp-Davis-Kress (FDK) analytic reconstruction methods.
• Evaluation of image quality and acquisition time reduction using MBIR.

Main Results:

• MBIR algorithms significantly suppress artifacts and noise in CT reconstructions of TRISO particles.
• The proposed MBIR approach yields high-quality 3D volumes with minimal pre-processing.
• MBIR enables high-quality reconstructions using a fraction of conventional measurements, drastically reducing scan times.

Conclusions:

• MBIR offers a superior alternative to conventional CT reconstruction methods for TRISO fuel particle characterization.
• The developed MBIR algorithms facilitate faster and more accurate analysis of nuclear fuel components.
• This advancement supports the development of next-generation nuclear fuels through improved imaging capabilities.