skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Highly Enriched uranium Metal Spheres Surrounded by Various Reflectors

Abstract

A series of experiments was performed at the Los Alamos critical assembly facility in the early 1950s to determine the critical mass of highly enriched uranium spheres surrounded by thin reflectors of various materials. The objective of these experiments was to obtain a precision graph of the critical mass of highly enriched uranium metal as a function of reflector thickness and to generate transport cross sections for the reflector material. Thirteen configurations are described and evaluated under ICSBEP identifier, HEU-MET-FAST-085; two with 1.98-inch-thick and 4.158-inch-thick copper reflectors, two with 2- and 4-inch-thick cast iron reflectors, one with a 1.945-inch-thick nickel reflector, two with 1.88- and 2.02-inch-thick nickel-copper-zinc alloy reflectors, one with a 1.81-inch-thick thorium reflector, two with 2-inch-thick and 4-inch-thick tungsten alloy reflectors, two with 2-inch-thick and 4.075-inch-thick zinc reflectors, and one with a 2-inch-thick tungsten alloy reflector surrounded by a 2-inch-thick cast iron reflector. All configurations were slightly subcritical with measured multiplications ranging from 20 to 162. Analyses were performed in which uncertainty associated with six different parameters was evaluated; namely, extrapolation to uranium critical mass, uranium density, 235U enrichment, reflector density, reflector thickness, and reflector impurities were considered. Uncertainty in cast-iron alloying elements was also considered when appropriate.more » In addition to the idealizations made by the experimenters, two simplifications were also made to the benchmark models that resulted in a small bias and additional uncertainty. First of all, since impurities in core and reflector materials are only estimated, they are not included in the benchmark models. Secondly, the room, support structure, and other possible surrounding equipment were not included in the model. Bias values that result from these two simplifications were determined and associated uncertainty in the bias values were included in the overall uncertainty in benchmark keff values. Bias values range from 0.0021 ?k low to 0.0016 ?k high. Overall uncertainties range from ? 0.0023 to ? 0.0064. Major contributors to the overall uncertainty include uncertainty in the extrapolation to the uranium critical mass and the uranium density. Results are summarized in the following figure. The 3 configurations described and evaluated in HEU-MET-FAST-085 are judged to be acceptable for use as criticality safety benchmark experiments and should be valuable integral benchmarks for nuclear data testing of the various reflector materials. Details of the benchmark models, uncertainty analyses, and final results are given in this paper.« less

Authors:
; ;
Publication Date:
Research Org.:
Idaho National Laboratory (INL)
Sponsoring Org.:
DOE - NE
OSTI Identifier:
912457
Report Number(s):
INL/CON-06-11656
TRN: US200801%%887
DOE Contract Number:
DE-AC07-99ID-13727
Resource Type:
Conference
Resource Relation:
Conference: Conference on Nuclear Criticality Safety,St. Petersberg, Russia,05/28/2007,06/01/2007
Country of Publication:
United States
Language:
English
Subject:
73 - NUCLEAR PHYSICS AND RADIATION PHYSICS; ACCURACY; ALLOYS; BENCHMARKS; CAST IRON; COPPER; CRITICAL MASS; CRITICALITY; CROSS SECTIONS; EXTRAPOLATION; HIGHLY ENRICHED URANIUM; IMPURITIES; NICKEL; SAFETY; THICKNESS; THORIUM; TUNGSTEN ALLOYS; URANIUM; ZINC; 235U; HEU-MET-FAST-085; ICSBEP

Citation Formats

Jessica Feener, J. Blair Briggs, and Leland Montierth. Highly Enriched uranium Metal Spheres Surrounded by Various Reflectors. United States: N. p., 2007. Web.
Jessica Feener, J. Blair Briggs, & Leland Montierth. Highly Enriched uranium Metal Spheres Surrounded by Various Reflectors. United States.
Jessica Feener, J. Blair Briggs, and Leland Montierth. Tue . "Highly Enriched uranium Metal Spheres Surrounded by Various Reflectors". United States. doi:. https://www.osti.gov/servlets/purl/912457.
@article{osti_912457,
title = {Highly Enriched uranium Metal Spheres Surrounded by Various Reflectors},
author = {Jessica Feener and J. Blair Briggs and Leland Montierth},
abstractNote = {A series of experiments was performed at the Los Alamos critical assembly facility in the early 1950s to determine the critical mass of highly enriched uranium spheres surrounded by thin reflectors of various materials. The objective of these experiments was to obtain a precision graph of the critical mass of highly enriched uranium metal as a function of reflector thickness and to generate transport cross sections for the reflector material. Thirteen configurations are described and evaluated under ICSBEP identifier, HEU-MET-FAST-085; two with 1.98-inch-thick and 4.158-inch-thick copper reflectors, two with 2- and 4-inch-thick cast iron reflectors, one with a 1.945-inch-thick nickel reflector, two with 1.88- and 2.02-inch-thick nickel-copper-zinc alloy reflectors, one with a 1.81-inch-thick thorium reflector, two with 2-inch-thick and 4-inch-thick tungsten alloy reflectors, two with 2-inch-thick and 4.075-inch-thick zinc reflectors, and one with a 2-inch-thick tungsten alloy reflector surrounded by a 2-inch-thick cast iron reflector. All configurations were slightly subcritical with measured multiplications ranging from 20 to 162. Analyses were performed in which uncertainty associated with six different parameters was evaluated; namely, extrapolation to uranium critical mass, uranium density, 235U enrichment, reflector density, reflector thickness, and reflector impurities were considered. Uncertainty in cast-iron alloying elements was also considered when appropriate. In addition to the idealizations made by the experimenters, two simplifications were also made to the benchmark models that resulted in a small bias and additional uncertainty. First of all, since impurities in core and reflector materials are only estimated, they are not included in the benchmark models. Secondly, the room, support structure, and other possible surrounding equipment were not included in the model. Bias values that result from these two simplifications were determined and associated uncertainty in the bias values were included in the overall uncertainty in benchmark keff values. Bias values range from 0.0021 ?k low to 0.0016 ?k high. Overall uncertainties range from ? 0.0023 to ? 0.0064. Major contributors to the overall uncertainty include uncertainty in the extrapolation to the uranium critical mass and the uranium density. Results are summarized in the following figure. The 3 configurations described and evaluated in HEU-MET-FAST-085 are judged to be acceptable for use as criticality safety benchmark experiments and should be valuable integral benchmarks for nuclear data testing of the various reflector materials. Details of the benchmark models, uncertainty analyses, and final results are given in this paper.},
doi = {},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Tue May 01 00:00:00 EDT 2007},
month = {Tue May 01 00:00:00 EDT 2007}
}

Conference:
Other availability
Please see Document Availability for additional information on obtaining the full-text document. Library patrons may search WorldCat to identify libraries that hold this conference proceeding.

Save / Share:
  • A series of experiments was performed at Los Alamos Scientific Laboratory in 1958 to determine critical masses of cylinders of Oralloy (Oy) reflected by a number of materials. The experiments were all performed on the Comet Universal Critical Assembly Machine, and consisted of discs of highly enriched uranium (93.3 wt.% 235U) reflected by half-inch and one-inch-thick cylindrical shells of various reflector materials. The experiments were performed by members of Group N-2, particularly K. W. Gallup, G. E. Hansen, H. C. Paxton, and R. H. White. This experiment was intended to ascertain critical masses for criticality safety purposes, as well asmore » to compare neutron transport cross sections to those obtained from danger coefficient measurements with the Topsy Oralloy-Tuballoy reflected and Godiva unreflected critical assemblies. The reflector materials examined in this series of experiments are as follows: magnesium, titanium, aluminum, graphite, mild steel, nickel, copper, cobalt, molybdenum, natural uranium, tungsten, beryllium, aluminum oxide, molybdenum carbide, and polythene (polyethylene). Also included are two special configurations of composite beryllium and iron reflectors. Analyses were performed in which uncertainty associated with six different parameters was evaluated; namely, extrapolation to the uranium critical mass, uranium density, 235U enrichment, reflector density, reflector thickness, and reflector impurities. In addition to the idealizations made by the experimenters (removal of the platen and diaphragm), two simplifications were also made to the benchmark models that resulted in a small bias and additional uncertainty. First of all, since impurities in core and reflector materials are only estimated, they are not included in the benchmark models. Secondly, the room, support structure, and other possible surrounding equipment were not included in the model. Bias values that result from these two simplifications were determined and associated uncertainty in the bias values were included in the overall uncertainty in benchmark keff values. Bias values were very small, ranging from 0.0004 ?k low to 0.0007 ?k low. Overall uncertainties range from ? 0.0018 to ? 0.0030. Major contributors to the overall uncertainty include uncertainty in the extrapolation to the uranium critical mass and the uranium density. Results are summarized in Figure 1. Figure 1. Experimental, Benchmark-Model, and MCNP/KENO Calculated Results The 32 configurations described and evaluated under ICSBEP Identifier HEU-MET-FAST-084 are judged to be acceptable for use as criticality safety benchmark experiments and should be valuable integral benchmarks for nuclear data testing of the various reflector materials. Details of the benchmark models, uncertainty analyses, and final results are given in this paper.« less
  • A variety of critical experiments were constructed of enriched uranium metal during the 1960s and 1970s at the Oak Ridge Critical Experiments Facility in support of criticality safety operations at the Y-12 Plant. The purposes of these experiments included the evaluation of storage, casting, and handling limits for the Y-12 Plant and providing data for verification of calculation methods and cross-sections for nuclear criticality safety applications. These included solid cylinders of various diameters, annuli of various inner and outer diameters, two and three interacting cylinders of various diameters, and graphite and polyethylene reflected cylinders and annuli. Of the hundreds ofmore » delayed critical experiments, experiments of uranium metal annuli with and without polyethylene reflectors and with the central void region either empty or filled with polyethylene were evaluated under ICSBEP Identifier HEU-MET-FAST-076. The outer diameter of the uranium annuli varied from 9 to 15 inches in two-inch increments. In addition, there were uranium metal cylinders with diameters varying from 7 to 15 inches with complete reflection and reflection on one flat surface to simulate floor reflection. Most of the experiments were performed between February 1964 and April 1964. Five partially reflected (reflected on the top only) experiments were assembled in November 1967, but are judged by the evaluators not to be of benchmark quality. Twenty-four of the twenty-five experiments have been determined to have fast spectra. The only exception has a mixed spectrum. Analyses were performed in which uncertainty associated with five different parameters associated with the uranium parts and three associated with the polyethylene parts was evaluated. Included were uranium mass, height, diameter, isotopic content, and impurity content and polyethylene mass, diameter, and impurity content. There were additional uncertainties associated with assembly alignment, support structure, and the value for ßeff. In addition to the idealizations made by the experimenters (removal of a diaphragm), a few simplifications were also made to the benchmark models that resulted in a small bias and additional uncertainty. Simplifications included omission of the support structure, possible surrounding equipment, and the walls, floor, and ceiling of the experimental cell. Bias values that result from these simplifications were determined and associated uncertainty in the bias values were included in the overall uncertainty in benchmark keff values. Bias values ranged from 0.0002 ?k to 0.0093 ?k below the experimental value. Overall uncertainties range from ? 0.0002 to ? 0.0011. Major contributors to the overall uncertainty include uncertainty in the support structure and the polyethylene parts. A comparison of experimental, benchmark-model, and MCNP-model keff values is shown in Figure 1. The experimental keff values are derived from the original reactivities reported by the principal experimentalist. The benchmark-model keff values are the experimental keff values adjusted to account for biases that were introduced by removing the support structure and surroundings. The MCNP-model keff values are simply the values found from MCNP calculations using the benchmark specifications and ENDF/B-VI cross-section data. Figure 1. Comparison of Experimental, Benchmark-Model and MCNP-Model keff value. Calculated results for most of the experiments are« less
  • An additional experiment has been performed using the recently cast 6-kg {sup 237}Np sphere. The experiment consisted of surrounding the neptunium sphere with highly enriched uranium and an iron reflector. The purpose of the critical experiment is to provide additional criticality data that can be used to validate criticality safety evaluations involving the deposition of neptunium. It is well known that {sup 237}Np is primarily produced by successive neutron capture events in {sup 235}U or through the (n, 2n) reaction in {sup 238}U. These nuclear reactions lead to the production of {sup 237}U, which decays by beta emission into {supmore » 237}Np. In addition, in the spent fuel, {sup 241}Am decays by alpha emission into {sup 237}Np. Because {sup 237}Np is a threshold fissioner, the best reflectors for critical systems containing neptunium are those materials that exhibit good neutron scattering properties such as low carbon steel (99 wt % Fe). In this experiment, the iron reflector reduced the amount of uranium used in the critical experiment and increased the importance of the neptunium sphere.« less
  • Prompt neutron decay at delayed criticality was measured by Oak Ridge National Laboratory for uranium-reflected highly enriched uranium (HEU) and Pu metal spheres (FLATTOP), for an unreflected Pu metal (4.5% {sup 240}Pu) sphere (JEZEBEL) at Los Alamos National Laboratory (LANL) and for an unreflected HEU metal sphere at Oak Ridge Critical Experiments Facility. The average prompt neutron decay constants from hundreds of Rossi-{alpha} and randomly pulsed neutron measurements with {sup 252}Cf at delayed criticality are as follows: 3.8458 {+-} 0.0016 x 10{sup 5} s{sup -1}, 2.2139 {+-} 0.0022 x 10{sup 5} s{sup -1}, 6.3126 {+-} 0.0100 x 10{sup 5} s{supmore » -1}, and 1.1061 {+-} 0.0009 x 10{sup 6} s{sup -1}, respectively. These values agree with previous measurements by LANL for FLATTOP, JEZEBEL, and GODIVA I as follows: 3.82 {+-} 0.02 x 10{sup 5} s{sup -1} for a uranium core; 2.14 {+-} 0.05 x 10{sup 5} s{sup -1} and 2.29 x 10{sup 5} s{sup -1} (uncertainty not reported) for a plutonium core; 6.4 {+-} 0.1 x 10{sup 5} s{sup -1}, and 1.1 {+-} 0.1 x 10{sup 6} s{sup -1}, respectively, but have smaller uncertainties because of the larger number of measurements. For the FLATTOP and JEZEBEL assemblies, the measurements agree with calculations. Traditionally, the calculated decay constants for the bare uranium metal sphere GODIVA I and the Oak Ridge Uranium Metal Sphere were higher than experimental by {approx}10%. Other energy-dependent quantities for the bare uranium sphere agree within 1%.« less
  • The Hansen-Roach 16-group cross-section library has been validated for use in pure uranium metal systems by modeling the Godiva critical assembly using the neutronics transport theory code ONEDANT to perform effective multiplication factor (k{sub eff}) calculations. The cross-section library used contains data for 118 isotopes (34 unique elements), including the revised cross sections for {sup 235}U and {sup 238}U. The Godiva critical assembly is a 17.4-cm sphere composed of 93.7 wt% {sup 235}U, 1.0 wt% {sup 234}U, and 5.3 wt% {sup 238}U with an effective homogeneous density of 18.7 g/cm{sup 3}.