ANS is committed to advancing, fostering, and promoting the development and application of nuclear sciences and technologies to benefit society.
Explore the many uses for nuclear science and its impact on energy, the environment, healthcare, food, and more.
Division Spotlight
Mathematics & Computation
Division members promote the advancement of mathematical and computational methods for solving problems arising in all disciplines encompassed by the Society. They place particular emphasis on numerical techniques for efficient computer applications to aid in the dissemination, integration, and proper use of computer codes, including preparation of computational benchmark and development of standards for computing practices, and to encourage the development on new computer codes and broaden their use.
Meeting Spotlight
Utility Working Conference and Vendor Technology Expo (UWC 2024)
August 4–7, 2024
Marco Island, FL|JW Marriott Marco Island
Standards Program
The Standards Committee is responsible for the development and maintenance of voluntary consensus standards that address the design, analysis, and operation of components, systems, and facilities related to the application of nuclear science and technology. Find out What’s New, check out the Standards Store, or Get Involved today!
Latest Magazine Issues
Jul 2024
Jan 2024
Latest Journal Issues
Nuclear Science and Engineering
August 2024
Nuclear Technology
Fusion Science and Technology
Latest News
ARPA-E announces $40 million to develop transmutation technologies for UNF
The Department of Energy’s Advanced Research Projects Agency–Energy (ARPA-E) announced $40 million in funding to develop cutting-edge technologies to enable the transmutation of used nuclear fuel into less-radioactive substances. According to ARPA-E, the new initiative addresses one of the agency’s core goals as outlined by Congress: to provide transformative solutions to improve the management, cleanup, and disposal of radioactive waste and spent nuclear fuel.
Mahmoud Yaseen, Xu Wu
Nuclear Science and Engineering | Volume 197 | Number 5 | May 2023 | Pages 947-966
Technical Paper | doi.org/10.1080/00295639.2022.2123203
Articles are hosted by Taylor and Francis Online.
Recent performance breakthroughs in artificial intelligence (AI) and machine learning (ML), especially advances in deep learning, the availability of powerful and easy-to-use ML libraries (e.g., scikit-learn, TensorFlow, PyTorch), and increasing computational power, have led to unprecedented interest in AI/ML among nuclear engineers. For physics-based computational models, verification, validation, and uncertainty quantification (VVUQ) processes have been very widely investigated, and many methodologies have been developed. However, VVUQ of ML models has been relatively less studied, especially in nuclear engineering. This work focuses on uncertainty quantification (UQ) of ML models as a preliminary step of ML VVUQ, more specifically Deep Neural Networks (DNNs) because they are the most widely used supervised ML algorithm for both regression and classification tasks. This work aims at quantifying the prediction or approximation uncertainties of DNNs when they are used as surrogate models for expensive physical models. Three techniques for UQ of DNNs are compared, namely, Monte Carlo Dropout (MCD), Deep Ensembles (DE), and Bayesian Neural Networks (BNNs). Two nuclear engineering examples are used to benchmark these methods: (1) time-dependent fission gas release data using the Bison code and (2) void fraction simulation based on the Boiling Water Reactor Full-size Fine-Mesh Bundle Tests (BFBT) benchmark using the TRACE code. It is found that the three methods typically require different DNN architectures and hyperparameters to optimize their performance. The UQ results also depend on the amount of training data available and the nature of the data. Overall, all three methods can provide reasonable estimations of the approximation uncertainties. The uncertainties are generally smaller when the mean predictions are close to the test data while the BNN methods usually produce larger uncertainties than MCD and DE.