Project #: 22-053 | Year 1 of 2
Jessica Clayton,a Anemarie deYoung,b Anna Hayes-Sterbenz,b R. Jason Scharffa
aNew Mexico Operations; bLos Alamos National Laboratory (LANL)
Legacy nuclear data is used to validate current weapon design codes. Modelers working for Life Extension Programs (LEPs) consider results from many diagnostics, but face challenges and uncertainties due to how the analog data were recorded decades ago and information loss over many years. This project is an undertaking to unite nuclear diagnostic data from a large sampling of events. We assess relationships between different observable quantities using basic assumptions about fusion burn. We aim to build a consistent picture of the complicated burn process observed in multiple ways.
We collected nuclear diagnostic data from 23 legacy (pre-1992) tests for inclusion in this analysis. This sample represents a range of nuclear tests spanning a few decades, each which provides different input to current weapons designers about performance related to the unique design of each experiment. Because multiple diagnostics were used on each shot, the data from the diagnostics must therefore be consistent—as each diagnostic witnessed the same event.
The analysis began with PINEX (PINhole Experiment) images, each which provides a time-integrated image of the burn process, as measured in the line-of-sight. The first images used here already had basic processing performed on them (dewarping, transfer curve application, pinhole corrections); however, they had not been assessed for size in a uniform manner. While it would have been desirable to perform a batch processing on all images, it was necessary to individually align the images, remove stray calibration lights and artifacts, and carefully fit and remove the background from each. An Abel inversion was applied to each image, assuming cylindrical symmetry, thus allowing area and volume calculations for each shot.
The NUEX (NeUtron EXperiment) diagnostic was used to collect timing information related to neutron flux. Times were extracted from the oscilloscope traces that recorded NUEX signals.
Other parameters relevant to the nuclear tests were collected and compiled, a surprisingly tedious task. These included information about cable corrections, yield, weapon types, and amounts of input gases.
The data were assessed based on expected relationships from the generalized fusion burn equation. Multivariate plots allowed visualization of many parameters and their relationships. This work has been extended to FY 2023. We will add temperature to the assessments first. We will also utilize sensitivity analysis techniques to better understand how the size reconstructed from an image is affected by factors such as background subtraction, asymmetries, and misplacement of the symmetry axis. Optimization and machine learning schemes will be investigated with this small data set.
This project has already compiled a wealth of nuclear data into a single analysis. We have extracted timing (NUEX) and size (PINEX) data and used it to assess relationships with yield. We have found that background subtraction and asymmetries in the PINEX images warrant further study to understand how they affect the recorded data. While the data set is extremely limited, the diagnostic data from the 23 events studied here present a picture of how the PINEX and NUEX data can be tied together and help inform weapons modeling codes.
It has been 30 years since nuclear testing ceased in the U.S.; however, the nuclear data collected during those times remains and provides critical input to current weapons designers and modelers. This project was born out of a question posed to Nevada National Security Site (NNSS) scientists by LANL researchers. As the NNSS already provided PINEX image processing support, the knowledge base was available to extend the efforts and explore how multiple nuclear diagnostics tied together in one consistent physics picture. The results of the work are intended to inform designers and help them assess uncertainties in predictive capabilities regarding PINEX observables.
Publications, Technology Abstracts, Presentations/Posters
This work was presented in a classified session at the 5th Annual Reaction History workshop in September 2022 at the NNSS in North Las Vegas. This meeting was attended by partners working in nuclear event analysis from LANL and Lawrence Livermore National Laboratory. A classified report is also in progress with collaborators A. Hayes and A. deYoung to describe findings from this analysis.
This work was done by Mission Support and Test Services, LLC, under Contract No. DE-NA0003624 with the U.S. Department of Energy. DOE/NV/03624–1609.