Search NASA⌕ Search

Engineering topics

Pelak, Robert A.

Publications and source records attributed to Pelak, Robert A..

PDQ Users Manual. Manual Version 2, for PDQ Code Version 1.20

PDQ is a tool for the management of the input and execution of batch jobs for simulation codes that use a text based input system. It accomplishes this goal by operating at two levels. First, it takes input file templates (commonly known at LANL as input deck templates) and creates multiple instantiations by performing substitutions of data from table files into symbols (variables) found in the template. Second, it provides commands to submit the created files to the SLURM batch system for execution. These two activities taken together produce a whole that is greater than the sum of its parts and provides an elegant way of executing studies across multiple similar simulations while minimizing the risk of typographical errors in the input files. PDQ was originally developed as a job management system called XVS by Jeff McAninch while he was at LANL. Besides the capabilities described here, XVS had many other features specific for interactions with particular simulation codes. After Jeff’s departure, maintenance of XVS was taken over by Rendell Carver; he added some new features as well as kept it functioning as the batch system at LANL was changed from LSF to MOAB to SLURM. In 2017, Rob Pelak decided to develop a different version that removed the additional features (many of which were rendered obsolete with the retirement of the simulation code or batch system that they supported) and produced a cleaner “bare bones” version of XVS. A few other behaviors of XVS that Rob found irksome were altered. Rob gave the resulting code a new name: PDQ. In 2022 Danielle McDermott developed a version that runs under Python 3.X. As suggested by Rob, she used the python2to3 utility to identify most changes. Given that PDQ continues to operate with Python version 2.7 we have advanced the version number to 1.20.

97 MATHEMATICS AND COMPUTING↗

Numerical Simulations of Non-Proliferation Experiments

We have implemented a set of numerical simulations to numerically investigate the outcomes of non-proliferation experiments. The high explosive (HE) in the simulations is in the form of a cylinder. The two-dimensional simulations are carried out in the r-z cylindrical coordinate, and the three-dimensional simulation is in the Cartesian coordinate. From the simulation, the cavity formed after detonation stops its expansion after about 0.15 sec. At the time the cavity stops expanding, the cavity shape seems quite spherical although the initial HE is in the shape of cylinder, but the radius in the vertical direction, which is about 19.7m, is a little less than the one in the horizonal direction, which is about 21.5m. Although there is no layer of air between the HE and tuff in real experiments, in addition to the simulations without air, we add a layer of air in one simulation to assess any possible influence of the air on the shape of the cavity.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗

Wedding Cake Experiments Design and Data Packages (V.2)

This report documents the design and data for four related experiments known as the Wedding Cake shots. It is a revised version of LA-UR-17-28352. The basic designs of these four shots are the same. A high-explosive plane-wave lens launches a flyer towards an acceptor material with a smaller diameter. This acceptor is either another piece of high-explosive or a piece of polyethylene. A detonation or shock wave propagates through this material and launches a second flyer. The velocities of both flyers are recorded using Photon Doppler Velocimetry (PDV). The overall shape of the shots is reminiscent of a tiered wedding cake and, consequently, these shots are know as the Wedding Cake shots. Details of the shot assembly are shown in the Wedding Cake Assembly Drawing and component drawings that accompany this report.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗

PDQ Users Manual Version 1 for PDQ Code Version 1.12

PDQ is a tool for the management of the input and execution of batch jobs for simulation codes that use a text based input system. It accomplishes this goal by operating at two levels. First, it takes input file templates (commonly known at LANL as input deck templates) and creates multiple instantiations by performing substitutions of data from table files into symbols (variables) found in the template. Second, it provides commands to submit the created files to the SLURM batch system for execution. These two activities taken together produce a whole that is greater than the sum of its parts and provides an elegant way of executing studies across multiple similar simulations while minimizing the risk of typographical errors in the input files.

97 MATHEMATICS AND COMPUTING↗

A Statistical Methodology for Quantifying Improvement in Predictive Capability: Key Features and Benefits

In References [1, 2, 3], we proposed a statistical methodology for quanti fying improvement in predictive capability. In this note, we summarize the key features and benefits of our approach. We begin with a brief summary of our methodology. Broadly speaking, our approach models the predictive process statistically at a high level. Ac cordingly, we assume the existence of a body of experiments, the results of which have been (or will be) predicted. We can then treat the discrepancy between a prediction and the corresponding experimental result as a datum and the collection of such data informs us about the statistics of the prediction process.

Engineering(42)↗

Using a Simple Binomial Model to Assess Improvement in Predictive Capability: Sequential Bayesian Inference, Hypothesis Testing, and Power Analysis

We present a Bayesian statistical methodology for identifying improvement in predictive simulations, including an analysis of the number of (presumably expensive) simulations that will need to be made in order to establish with a given level of confidence that an improvement has been observed. Our analysis assumes the ability to predict (or postdict) the same experiments with legacy and new simulation codes and uses a simple binomial model for the probability, {theta}, that, in an experiment chosen at random, the new code will provide a better prediction than the old. This model makes it possible to do statistical analysis with an absolute minimum of assumptions about the statistics of the quantities involved, at the price of discarding some potentially important information in the data. In particular, the analysis depends only on whether or not the new code predicts better than the old in any given experiment, and not on the magnitude of the improvement. We show how the posterior distribution for {theta} may be used, in a kind of Bayesian hypothesis testing, both to decide if an improvement has been observed and to quantify our confidence in that decision. We quantify the predictive probability that should be assigned, prior to taking any data, to the possibility of achieving a given level of confidence, as a function of sample size. We show how this predictive probability depends on the true value of {theta} and, in particular, how there will always be a region around {theta} = 1/2 where it is highly improbable that we will be able to identify an improvement in predictive capability, although the width of this region will shrink to zero as the sample size goes to infinity. We show how the posterior standard deviation may be used, as a kind of 'plan B metric' in the case that the analysis shows that {theta} is close to 1/2 and argue that such a plan B should generally be part of hypothesis testing. All the analysis presented in the paper is done with a general beta-function prior for {theta}, enabling sequential analysis in which a small number of new simulations may be done and the resulting posterior for {theta} used as a prior to inform the next stage of power analysis.

42 ENGINEERING↗