NUREG/CR-0674
Benchmark Critical Experiments on Low-Enriched Uranium Oxide Systems With H/U = 0.77 Topical Report on Reference Critical Experiments
Rockwell International Energy Systems Group
Prepared for U. s. Nuclear Regulatory Commission
•
la NOTICE
This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, or any of their employees, makes any warranty, expressed or implied, or assumes any legal liability or responsibility for any third party’s use, or the results of such use, of any information, apparatus product or process disclosed in this report, or represents that its use by such third party would not infringe privately owned rights.
Available from National Technical Information Service Springfield, Virginia 22161 DISCLAIMER
This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency Thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof. DISCLAIMER
Portions of this document may be illegible in electronic image products. Images are produced from the best available original document.
NUREG/CR-0674 RFP-2895 Dist Code RC
BENCHMARK CRITICAL EXPERIMENTS ON
LOW-ENRICHED URANIUM OXIDE SYSTEMS
WITH H/U = 0.77
Topical Report on U. S. Nuclear Regulatory Commission Reference Critical Experiments
- NOTICE- This report was prepared as an account of work Grover Tuck sponsored by the United States Government. Neither the United States nor the United States Department of Inki Oh Energy, nor any of their employees, nor any of their contractors, subcontractors, or their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights. Rockwell International Energy Systems Group Rocky Flats Plant P. O. Box 464 Golden. Colorado 80401
Date Published; August 1979
Prepared for Division of Safeguards, Fuel Cycle, & Environmental Research Office of Nuclear Regulatory Research U. S. Nuclear Regulatory Commission Washington, D.C. 20555 under U.S. Department of Energy Contract DE-ACO4-76DP03533 NRC FIN NO.A1036
NUREG/CR-0674 RFP-2895 (iii)
ABSTRACT
Ten benchmark experiments were performed at the Critical
Mass Laboratory at Rockwell International’s Rocky Flats
Plant, Golden, Colorado, for the U. S. Nuclear Regulatory
Commission. They provide accurate criticality data for low-enriched damp uranium oxide (U3 0 g) systems. The core studied consisted of 152 mm cubical aluminum cans containing an average of 15,129 g of low-enriched (4.46% uranium oxide compacted to a density of 4.68 g/cm^ and with an H/U atomic ratio of 0.77. One hundred twenty five (125) of these cans were arranged in an ~ 770 mm cubical array. Since the oxide alone cannot be made critical in an array of this size, an enriched (~ 93% 235u) metal or solution "driver" was used to achieve criticality.
Measurements are reported for systems having the least practical reflection and for systems reflected by 254-mm- thick concrete or plastic. Under the three reflection con ditions, the mass of the uranium metal driver ranged from
29.87 kg to 33.54 kg for an oxide core of 1864.6 kg. For an oxide core of 1824.9 kg, the weight of the high concen tration (351.2 kg U/m^) solution driver varied from 14.07 kg to 16.14 kg, and the weight of the low concentration (86.4 kg
U/m^) solution driver from 12.4 kg to 14.0 kg.
NUREG/CR-0674 RFP-2895 (v)
SUMMARY
Hie benclni^rk critical experiments program Is sponsored by the Nuclear Regulatory Commlss^ion to prorlde accurate criticality data for low-enriched damp uranium oxide CUgOg) systems. Sfeny appilcations of these data to nuclear criti cality safety questions are useful In the nuclear Industry.
The benchmarfe critical parameters can provide a validated calculatlonal method to nuclear criticality safety engineers
for use as a guide for specifying a criticality limit.
The result of this experimental program Is the measure ment of ten critical configurations wherein both geometry and material parameters are specified.
Uranium oxide used In this program was enriched to 4.46 wt-% 2^^U, compacted to a density of 4.68 g/cm^, and packed
into 1.6-mm-thlck aluminum cans forming a 152 mm cube. Water was added to the oxide until an H/U atomic ratio of 0.77 was achieved. The experimental core studied was a 5 x 5 x 5 array of aluminum cans, each containing an average of
15,129 g of uranium oxide. Since the oxide alone cannot be made critical in this size array, a high-enriched
93% 23 5u) uranium driver, which replaced one to four cans near the core’s center, was used to achieve criti cality. The drivers were an enriched uranium metal sphere, high concentration (351.2 kg U/m^) solution, and low con centration (86.4 kg U/m^) solution. NUREG/CR-0674 RFP-2895 (vi)
Three reflector conditions were studied. The miniaally-
reflected case had the least practical reflector, made of
6 .35-mm-thick steel plate to contain the oxide cans. The other two reflector conditions had the critical configuration within thick-walled cubical shells composed of concrete or
plastic, both conmH!>n materials used in the nuclear industry-
For both reflectors, interior dimensions of the shell were
~ 770 nun, and the reflector thicknesses were ~ 254 mm for
the two materials.
All experiments were performed on a horizontal split
table. Each half of the table supported a portion of the
experimental assembly, and the critical approach was made
by decreasing the separation between the two portions of
the core.
Under the three reflection conditions, the critical core
separation ranged from 6.3 mm to 14.6 mm for the following
oxide and driver masses: the mass of the uranium metal
driver ranged from 29.87 kg to 33.54 kg for an oxide mass
of 1864.6 kg; the mass of the high concentration solution
driver ranged from 14.07 kg to 16.14 kg for an oxide mass
of 1824.9 kg; and the mass of the low concentration solu
tion driver ranged from 12.4 kg to 14 kg for an oxide mass
of 1824.9 kg. NUREG/CR-0674 RFP-2895 (vii)
TABLE OF CONTENTS
Page
ABSTRACT ------iii
SUMMARY ------V
LIST OF F I G U R E S ------ix
LIST OF TABLES ------xi
ACKNOWLEDGMENTS ------xiii
PREVIOUS REPORTS ------xv
INTRODUCTION ------1
EXPERIMENTAL PROCEDURE ------7
OXIDE C A N S ------13
Oxide Compaction ------13 Aluminum Cans ------13 Packing of Oxide and Water into Cans ------17 Material Description ------26 Determination of H/U Value ------32
METAL AND SOLUTION DRIVERS ------39
Metal Driver ------3 9 Solution Driver ------39
HORIZONTAL SPLIT TABLE ------53
REFLECTORS ------61
Concrete Reflector ------61 Plastic Reflector ------64 Steel and Environmental Reflectors ------70
CORE DIMENSIONS ------83
CRITICAL RESULTS ------93
Critical Parameters ------93 Concrete-Reflected Experiments ------95 Plastic-Reflected Experiments ------96 Minimally-Reflected Experiments ------98
DISCUSSION OF UNCERTAINTIES ------103
REFERENCES ------105
NUREG/CR-0674 RFP-2895 (ix)
LIST OF FIGURES
Figure Number Title
1 Experimental assembly of low-enriched oxide cans on the north and south halves of the horizontal split table in the concrete shells with the end reflectors removed.
2 Reciprocal multiplication with the table closed versus the mass of high concentration (351.2 kg U/m^) solution driver.
3 A compacted block of uranium oxide resting on a plastic sheet.
4 Sketch of the aluminxm cans which contain the uranium oxide,
5 A padked and sealed oxide can,
6 A photograph of the special oxide can packed and sealed.
7 Graph of oxide weight versus the number of days elapsed for each weighing since March 1, 1978.
8 Typical metal driver assembly and its supporting equipment.
9 An assembled uranium metal driver which will be positioned in the empty space in the oxide cans when the horizontal table is closed.
10 One pair of solution driver cans: one for the south half-table and the other for the north half-table.
11 The horizontal split table loaded for an experi ment but fully open.
12 The cross section of the south half of the horizontal split table.
13 The cross section of the north half of the horizontal split table.
14 Concrete reflector showing the dimensions. NUREG/CR-0674 RFP-2895 (x)
Figure Number Title
15 Plastic reflector showing the dimensions.
16 A photograph of the plastic reflector sitting on the floor.
17 Steel reflector showing the dimensions.
18(a) Experimental room looking south showing the east- west section.
18(b) Experimental room looking west showing the north- south section.
19 Description of core cuboid with a void for metal driver.
20 Description of core cuboid with a solution driver.
21 Configuration of core cuboid and reflector showing the elevation and plan views.
22 Minimally-reflected oxide cans with the uranium metal driver. NUREG/CR-0674 RFP-2895 (xi)
LIST OF TABLES
Table Number Title
1 Particle size of uranium oxide powder,
II Plastic materials used in packing oxide.
III Composition of non-fissile materials used in oxide cans.
IV Assay of uranium oxide in gram of uranium per gram of sample.
V Uranium isotopic enrichment used in oxide cans.
VI Average properties of oxide cans.
VII Properties of metal driver and its supporting materials.
Vlll Properties of uranium metal hemishells.
IX Solution driver can dimensions.
X Material composition of solution driver can.
XI Properties of uranyl nitrate solutions used as drivers.
XII Elemental composition of steel and stainless steel structural components of the horizontal split table.
Xlll Composition of concrete in weight-percent.
XIV Average weight-percent of plastic reflector material.
XV Average density of plastic reflector.
XVI Material composition of steel reflector.
XVII Composition of aluminum plate.
XVIII Core dimensions for the minimally-reflected, concrete-reflected, and plastic-reflected arrays of uranium oxide for the three different drivers, NUREG/CR-0674 RFP-2895 (xii)
XIX Principal critical paraaeters of ten 5x5x5 arrays of damp, compacted, low-enriched uranium oxide.
XX Weight and location of oxide cans as placed into the low-enriched oxide array with H/U — 0.77. NUREG/CR-0674 RFP-2895 (xiii)
ACKNOWLEDGMENTS
The authors gratefully acknowledge Robert E. Rothe,
W. R. (Bob) Sheets, and Douglas E. Payne for their assistance
in data collection. Many calculations important to the design
of the experiment performed by Deanne Pecora are also grate
fully recognized.
This work was performed for the U. S. Nuclear Regulatory
Commission, Office of Nuclear Regulatory Research, under
U. S. Department of Energy Contract DE-ACO4-76DP03533.
NUREG/CR-0674 RFP-2895 (XV )
PREVIOUS REPORTS
The following reports on the U. S^ Nuclear Regulatory Commission Reference Critical Experiments program have been issued.
RFP-NUREG-2481 - Quarterly progress report for period July-December 1975.
RFP-NUREG-2517 - Quarterly progress report for period January-March 1976.
RFP-NUREG-2555 - Quarterly progress report for period April-June 1976.
RFP-NUREG-2582 - Quarterly progress report for period July-September 1976.
RFP-NUREG-2622 - Quarterly progress report for period October-December 1976.
RFP-NUREG-2660 - Quarterly progress report for period January-March 1977,
RFP-NUREG-2690 - Quarterly progress report for period April-June 1977.
RFP-NUREG-2715 - Quarterly progress report for period July-September 1977.
RFP-NUREG-2746 - Quarterly progress report for period October-December 1977.
NUREG/CR-0096 - Quarterly progress report for period (RFP-2795) January-March 1978.
NUREG/CR-0297 - Quarterly progress report for period (RFP-2929) April-June 1978.
NUREG/CR-0499 - Quarterly progress report for period (RFP-2868) July-September 1978.
NUREG/CR-0642 - Quarterly progress report for period (RFP-2888) October-December 1978.
NUREG/CR-0041 - Topical Report, Benchmark Critical Experi- (RFP-2710) ments on High-Enriched Uranyl Nitrate Solution Systems. NUREG/CR-0674 RFP-2895 Page 1
BENCHMARK CRITICAL EXPERIMENTS ON LOW-ENRICHED URANIUM OXIDE SYSTEMS WITH H/U =0.77
INTRODUCTION
These benchmark experiments provide accurate critical data on low-enriched damp uranium oxide (UgOg) systems with three different drivers and under three different reflec
tions. These provide checkpoints against which computer code accuracy may be determined. The need for such benchmark data has been pointed out many times in the literature , especially for certain ranges of parameters.
For this first series of experiments, an H/U of 0.77 was used. The experimental series is continuing with increasing
H/U values and will be reported later. This first was the lowest practical H/U considering the available inventory,
the load limits on the existing critical mass laboratory equipment, and the fraction of the reactivity in the oxide.
The uranium was enriched to 4.46% 235u^ and the oxide compacted to a density of 4.68 ± 0.29 g/cm^. The compacted oxide was packed in thin-walled cubical aluminum cans 152 mm on a side. Water was added to the oxide until the H/U atomic ratio (including the hydrogen content contributed by the plastic bags surrounding the oxide pieces and other materials) was 0.77 ± 0.06. The core studied was a 5 x 5 x 5 array of oxide cans, forming an 770 mm cubical core. A 5 x 5 x 5 NUREG/CR-0674 RFP-2895 Page 2
array of oxide cans alone was not critical; hence, a high- enriched 93% 235^) uranium driver near the core’s center was used to attain criticality. One to four cans near the center were deleted to make room for the enriched uranium driver.
All experiments were performed on a horizontal split table. Each half of the table supported a portion of the assembly, and the critical approach was made by decreasing the separation between the two portions of the core (see
Figure 1).
Parameters varied in these experiments include the uranium masses of the three high-enriched uranium drivers and the reflector material. The driver was enriched (93.12%
235 u ) uranium metal, high concentration (351.2 kg U/m^) uranium solution, or low concentration (86.4 kg U/m^) uranium solution. The combined driver/oxide system has a neutron energy spectrum different from that of oxide alone. For (7 ) these assemblies, at least 50% of the fissions — occur in the oxide; hence, a calculation which reproduces the experi mental results must be able to calculate the contribution of the oxide. The three different drivers provide a check of cross section data by emphasizing different portions of the neutron energy spectrum.
The first two sections of this paper give the experimental procedures and describe the experimental oxide cans. The next NUREG/CR-0674 RFP-2895 Page 3
three sections detail composition and dimensions for all drivers, the horizontal split table, and the reflectors.
The final three sections present the core dimensions, the experimental results, and a discussion of uncertainties. NUREG/CR-0674 RFP-2895 Page 4
FIGURE 1
Experimental assembly of low-enriched oxide cans on the north and south halves of the horizontal split table in the concrete shells with the end reflectors removed. One solu tion driver can may be seen in the third layer of oxide cans on the north table occupying the space of two oxide cans. South Concrete Shell
iWSi
North Concrete Shell
i.
2: G SJ M O W \ G o *0 G w p I I TO to o tl 00 Ci to *0 Cn o r tP FIGURE 1
NUREG/CR-0674 RFP-2895 Page 7
EXPERIMENTAL PROCEDURE
All experimental work was performed on the horizontal
split table. This table consists of two half-tables, north
and south, which are closed during an experimental run. The
north table is first closed, and then the south table is
closed by increments toward the mating face of the north
table. Both tables are capable of fast opening in response
to a scram signal.
Once a given reflector was installed and aligned, the
table was loaded with the ~ 1900 kg of oxide, approximately
40% on the north table and approximately 60% on the south.
This was done following the manual assembly procedures, which involve plotting the inverse multiplication as a
function of the number of cans. With the completion of the
oxide loading, a final alignment of the core and reflector was required because the weight of the oxide on the
table may have slightly shifted the first alignment.
The next step was to assemble the driver to approximately half the calculated critical value. An experimental run was
then made by remotely closing the tables and plotting the appropriate inverse multiplication curves. At the completion of the run, the next driver mass addition was made and the run sequence repeated.
The data from several critical experimepts were used to select the correct driver mass for the driver/oxide NUREG/CR-0674 RFP-2895 Page 8
combination to be critical with the table nearly closed.
Figure 2 shows the inverse multiplication versus the high concentration (351,2 kg U/m^) solution driver mass. In the case of metal drivers, the choice was limited to available mass increments, resulting in a maximum table separation of
14.6 mm lor a critical data point. .Positive and negative period data were obtained for each critical point.
To measure tire table separation, small balls of a putty- like substance (Duxseal) were placed at points around the closing face of the reflector, the oxide cans, and, when used, the solution driver cans for the last subcritical run.
As the table closed on these small balls, they flattened to the closest separation between table halves achieved during that experiment. This putty-ball data was later related to the reading of differential transformers which monitor table closure during the final 33 mm.
After the critical run, the reflector end was removed and the exposed sides of the core measured. After obtaining one critical data point for each of three drivers, the oxide core was unloaded to change reflectors. During the unloading, measurements for the horizontal core dimensions were made for each layer of oxide cans.
For the plastic-reflected measurements, the north end reflector panel was raised 261 mm because of source-detector complications.^ The critical parameters, including the core NUREG/CR-0674 RFP-2895 Page 9
dimensions, are provided later in this report in the sections entitled Core Dimensions and Critical Results. NUREG/CR-0674 RFP-2895 Page 10
FIGURE 2
Reciprocal multiplication with the table closed versus the mass of high concentration (351o2 kg U/m^> solution driver. The 5x5x5 array of oxide is concrete-reflected. NUREG/CR-0674 RFP-2895 Page 11
CO
CO
CN
CM o> cn « CO !3 CO O 00 < i-i 2 Pm 2 O h- _J O (O CO
CM
O in n CN O CD CD O CD oasono anavihum NoiivoncJinniAi nvooadioaa
NUREG/CR-0674 RFP-2895 Page 13
OXIDE CANS
Oxide Compaction
The loose uranium oxide powder had a bulk density of
~ 1 g/cm^ and particle size as given in Table I. The powder was weighed into a small plastic bag, sealed with a very small paper-covered twist wire, and excess plastic cut off.
This bag of oxide was then placed in a 300,000 kg press and compacted. Each compacted block measured 74 ± 2 mm square with one corner rounded to match the can. A typical bag of oxide weighing 541 ± 1 g was pressed to an average thickness of 21 ± 1 mm, although ’’spring back" after pressing caused this to vary. A finished oxide block is shown in Figure 3.
In addition, some oxide blocks weighing 710 ± 1 g were pressed from bags of oxide prepared at the start of the program. The final compacted oxide density is 4.68 ± 0.29 g/cm^. The plastic bags used to contain the loose oxide, including a paper-covered soft iron wire twist-tab used to seal the bag, were embedded into the oxide.
Aluminum Cans
The aluminum cans in which the oxide blocks were packed measured 152.4 mm square by 150.8 mm high (outside dimen sions), and a flat aluminxim cover increased the outer height to 152.4 mm. The cans were deep drawn from type 1100 aliminum, 1.6 ± 0.1 mm thick. Small-radii curves on corners NUREG/CR-0674 RFP-2895 Page 14
TABLE I
Particle Size of Uranium Oxide Powder
Size Number* (um) Percentage
< 1 1.1 1-10 93.0 10 - 25 4.5 25 - 40 0.8 40 - 60 0.4 60 - 80 0.1 80 - 100 0.2 > 100 0.2
*Average of two samples. NUREG/CR-0674 RFP-2895 Page 15
FIGURE 3
A compacted block of uranium oxide resting on a plastic sheet. The plastic bag containing the loose oxide weighed
<^0.9 g and was not removed prior to compacting. Pieces of plastic were embedded in the finished surface. NUREG/CR-0674 RFP-2895 Page 16 NUREG/CR-0674 RFP-2895 Page 17
are necessary in forming the cans because of the process used. Figure 4 shows the dimensions of the aluminum can.
To permit later water injections, 6.35-mm-diameter holes were drilled in four vertical rows on two opposite faces for a total of 56 holes per can. See Figure 5.
Four special aluminum cans, designed to clear the metal driver support and the source removal equipment, were packed with oxide. These special cans were constructed *by modifying the 152.4 mm cube aluminum can to have a 37.5 mm wide by
99.5 mm deep slot (dashed lines in Figure 4). Thirty-two holes were drilled in four vertical rows on two opposite faces to permit later water injections in the 710 g oxide blocks used to pack these boxes. Each of the 32 holes was
6 .3-mm-diameter with two holes per oxide block. One similar hole was drilled in the top of the can to allow water to be added to the oxide fragments behind the slot. Figure 6 shows a photograph of the special aluminum oxide can.
Packing of Oxide and Water into Cans
Each compacted oxide block was put into a plastic bag, and excess bag material neatly folded under the block; then the block was placed in the can. This wrapping was intended to restrict movement of water between adjacent blocks. Each such bag weighed 1 g. The inner bags (previously described in the oxide compaction) were cut off just above the twist seal so they weighed ~ 0 . 9 g each. Hence, each block was NUREG/CR-0674 RFP-2895 Page 18
FIGURE 4
Sketch of the aluminum cans which contain the uranium oxide. The dashed lines represent a slot in the special can designed to clear the metal driver support and source control equipment. All dimensions are in millimeters. NUREG/CR-0674 RFP-2895 Page 19
Drwg. Not To Scale
i T 1.6
LID
152.4
I#- 54.3 ->| I#- 54.3 —►!
TT 1 II li 37.5A T II 1.6 II II 152.4 99.5 II II CAN II 150.8 L = = J , \ 8.3 7.9 Radius Radius Ti i Z 1.6 Plan View Elevation View
FIGURE 4 NUREG/CR-0674 RFP-2895 Page 20
FIGURE 5
A packed and sealed oxide can. All wall thicknesses are
1,6-nun-thick. The four rows of water injection holes are sealed with mylar tape, and the lid is sealed to the can with vinyl tape. NUREG/CR-0674 RFP-2895 Page 21
. -.U^- . .
j- - I
m
irf ■>■*'■ ■*-^- - • ^ 4 . * -
m
'^#1 NUREG/CR-0674 RFP-2895 Page 22
FIGURE 6
A photograph of the special oxide can, packed and sealed. NUREG/CR-0674 RFP-2895 Page 23
FIGURE 6 NUREG/CR-0674 RFP-2895 Page 24
surrounded by plastic bags totaling 1,9 g of plastic. Four
blocks completed a layer and seven layers completed a can.
After packing, the lid was taped on with vinyl tape.
The next step was to form holes in the oxide to provide
increased surface area for absorption of water. A length of
5-mm-diameter drill rod stock, pointed on the end, was forced
into the oxide through the holes in the can to a depth of
~ 70 mm deep. The absence of flutes on the drill rod mini mized the amount of material drawn to the surface during
the operation. In this way, two clean holes were made
nearly through each block in its mid-plane and trisecting
its width.
A syringe was used to inject a measured quantity of
water into each hole in the oxide. As a check on the water
added, the total water added to each can was weighed and
the can weighed before and after the addition of water.
The amount of water was calculated to yield an H/U atomic
ratio of 0.75 when its hydrogen was included with that of
the plastic bags, tape, and water initially in the oxide.
For cans containing twenty-eight 541 g blocks, this amount
of water was 274 g, and was uniformly distributed over all
holes. The holes were sealed with one layer of mylar tape.
Later, a second layer of tape was added for the plastic-
reflected experiments. Dimensions and weights of the plastic
bags and the two types of tape used for the oxide cans are
given in Table II. TABLE II
Plastic Materials Used in Packing Oxide
Plastic Bag Mylar Tape Vinyl Tape
contain seal watering seal cover Use individual holes on side to can oxide blocks of can
53.2 or 33.0 2 or 4 g Weight 3 g per can g per can’^ per can^
Width 169 mm 25 mm 25 mm
8 strips per 1 strip per 215 mm Length can, each can, 620 mm per bag^ 150 mm long long
0.017 mm Thickness (single 0.05 mm 0,14 mm thickness)
^33.0 g was used for the special oxide can„ tz: cj W ^2 g was used for the concrete-reflected and minimally- w reflected cases, and 4 g used for the plastic-reflected o w \ case, ►0 o P *0 so oc; I 1 ®The bag containing the oxide prior to pressing was only 0 to o 194 mm long because excess plastic had been trimmed 00 a to CO away. 01 Ol NUREG/CR-0674 RFP-2895 Page 26
Material Description
The measured composition for all nonfissile materials used with the oxide cans are given in Table III. The aluminum can and lid were made of type 1100 aluminum, and the "U"- piece in the special can was fabricated of type 5051 aluminum.
The aluminum densities are given in Table III.
Three batches of samples were assayed to determine the weight-fraction of uranium in the dry oxide. The average was
0.8449 ± 0.0008 g of uranium per g of sample. See Table IV.
The isotopic enrichment was measured on two occasions for composites of nine samples each. These results are shown in Table V, and the average measured values are assumed to apply to all experiments. The principal impurities in the uranium oxide in ppm (parts of impurity per million parts of uranium by weight) are: iron (312), silicon (128), boron
(<200), aluminum (37), copper (185), calcixim (18), cadmium
(< 5), chromium (128), magnesium (13), potassium (< 25), nickel (16), and phosphorus (< 70).
The uranium oxide was found to be gaining oxide weight with time. The average oxide weight gain for the period
March 1, 1978, to November 1, 1978, is 24.4 g per can with a standard deviation of ± 8.4. Figure 7 graphs the weight gain in grams versus the number of days elapsed for each weighing since March 1, 1978. The oxide experiments reported herein were performed between two and four months after
March 1, 1978. A test was conducted on four oxide cans to TABLE III
Composition of Nonfissile Materials Used in Oxide Cans
Composition (weight-percent)^ Element Plastic Mylar Vinyl Aluminum Can Aluminum Lid Aluminum "U"-Piece Bag Tape Tape (type 1100) (type 1100) (type 5051)
H 14.01 6.83 5.92 C 84.90 65.50 45. 91 N 0.16 0 1.2 27.02 10.82 Cl 25.73 Fe 0.5 0.4 0.7 Mg 0.01 0.01 0.01 Si 0.07 0.1 0.2 Ca 6.0 Mn 0.009 0.007 1.4 Ba 0.3 Cu 0.2 0.1 0.14 Ti 0.2 Cr 0.3 0.01 0.01 5Z! Zn 0.02 0.002 0.01 C3 W M Al^ 99.18 99.37 97. 54 O fO \ tp o Total 100.11 99.35 95.34® 100.0 100.0 100.0 p to tra I I 0 10 o Cg/cm3 ) 2.712 ± 0.002 2.715 ± 0.002 2.737 ± 0.002 00 o Density to «> <1 ^ Ol The measurement accuracy of aluminum is within ± 5%, and the others within ± 1%. ^Obtained by subtracting sum of other columns from 100%. ®The remainder is ash but not analyzed. CP5 T3 to O I W M O to 00 \ 00 to o to 1 o TABLE IV o <1 Assay of Uranium Oxide in Gram of Uranium per Gram of Dry Oxide
Number Date of Nature of Sample Assay Samples
11/15/76 9 dry, uncompacted, loose powder 0.8446 ± 0.0010
11/16/78 4 compacted, damp (H/U = 0.77) 0.8448 ± 0.0001
4/9/79 13 compacted, damp (H/U = 0.77) 0.8453 ± 0.0021
Average 0.8449 ± 0.0008 TABLE V
Isotopic Enrichment of Uranium in Oxide
Weight-Percent of Date Remarks Uranium Isotope Measured (analysis by) U-234 U-235 U-236 U-238
4/26/76 4.479 1st shipment, powder (NLO)’^
8/10/76 4.476 2nd shipment, powder (NLO)
11/10/78 0.030 4.465 0.086 95.419 H/U = 0.77 (RFP)®
3/12/79 0.030 4.425 0.090 95.455 H/U =0.77 (RFP)
4.461 Average 0.030 0.088 95.437 TOTAL = 100.016 wt-% ± 0.025
^National Lead Company of Ohio, Cincinnati, Ohio. ®Rocky Flats Plant, Golden, Colorado.
a: § w o ►d § ^ aq I I (D to O 00 0) to «T) ^ NUREG/CR-0674 RFP-2895 Page 30
FIGURE 7
Graph of oxide weight gain versus the number of days elapsed for each weighing since March 1, 1978. NUREG/CR-0674 RFP-2895 Page 31
CO CN
CN
CN (N
CN CO
CD
CO iU CD O Z O CO CO >- < CN o Pi cr g 111 M CO
CD 3
CO
CN
O CO o CO CN 00 CO CN CN CNCN CN
(B) NVO aacJ NIV3 1 H9 I3M 3QIX0 3DVB3AV NUREG/CR-0674 RFP-2895 Page 32
analyze the weight gain; and the weight gain was found to be attributed to an absorption of oxygen from the air by oxide powder, U0 ^ 2 .3 » which was not completely oxidized to U 3 O 8 .
This conclusion was further checked by a repeat of one of the earlier experiments to measure the change in reactivity over a period of time. See the section on Discussion of
Uncertainties.
Determination of H/U Value
To obtain an H/U atomic ratio for the damp oxide, the water and uranium in oxide plus the hydrogen from the plastic bag, vinyl tape, and mylar tape had to be determined.
[1] Hydrogen from Moisture Initially in Oxide
Five samples of dry, uncompacted oxide from each of the two shipments from the manufacturer were analyzed.
The average moisture content of these ten samples measured by thermogravimetric analysis (TGA) was 0.0027 ± 0.0013 g
H 2 O per g sample. From the initial water content, the hydrogen weight per can was 4.57 ± 2.20 g.
[ 2] Hydrogen from Water In.jected in Oxide
The injected water content.of the oxide was measured two ways. The first method (TGA) was used to measure total
(initial and injected) water content of one sample of damp oxide from each of 26 cans selected from the entire lot of cans. Fourteen cans were sampled by digging into one of the holes which had been used to inject water. The other twelve NUREG/CR-0674 RFP-2895 Page 33
samples came from the can side at right angles to the face receiving the water injection. These two sampling locations were selected to avoid biasing the data in case the injected water had not migrated uniformly throughout the oxide. The average water content was 0.0186 ± 0.0035 g H2 O per g of damp oxide.
The second method was to weigh the amount of water
injected into a can. The average weight of "dry" oxide
(initial moisture plus oxide) in packing each of 125 oxide cans was 15129.1 ± 33.7 g; and the average amount of water
injected into each was 273.2 ± 4.1 g. Therefore, the calcu
lated water content is 0.0177 ± 0.0003 g added H 2 O per g of damp oxide. The total water content for the damp oxide by
this second method is calculated to be 0.0204 ± 0.0013 g H 2 O per g sample after adding the initial water content.
An average of the two methods is taken as the best value of the total water content: 0.0195 ± 0.0019 g H2 O per g sample. From the total water content, the hydrogen weight per can was determined to be 33.61 ± 3.28 g. Hence, the hydrogen weight per can from the injected water was 29.04 ±
2.40 g.
[ 3 ] Hydrogen from Plastic Bags
From Table III, the weight-percent of hydrogen in the plastic bags is 14.01, and the total weight of the bags per can is 53,2 ± 2,66 as shown in Table II, It was assumed
that the weight uncertainty is 5%. Since major uncertainty NUREG/CR-0674 RFP-2895 Page 34
was resulted from items [l] and [2], the 5% has very little effect on the uncertainty of the H/U value. The hydrogen weight per can from the plastic bags is 7.45 ± 0.37 g.
[4] Hydrogen from Vinyl Tape
The total weight of the tape used in each can was
3 g, and the hydrogen weight-percent from Table III is 5.92.
Thus, the hydrogen weight per can from the vinyl tape is
0,178 ± 0.01 g, assuming a 5% weight uncertainty.
[ 5] Hydrogen from Mylar Tape
The hydrogen weight-percent in the mylar tape is
6.83 as presented in Table III, and the material used in each can is 2 g for the concrete-reflected and minimally-reflected cases, and 4 g for the plastic-reflected case. Therefore, the hydrogen weight per can from the mylar tape is 0.137 ±
0.006 g and 0.274 ± 0.012 g for the above cases, respectively.
Five-percent weight uncertainty was assumed.
Let Wj^ be each hydrogen weight, and Aw^ its uncer tainty, where i indicates the five different parts mentioned above. Since the hydrogen number density is
where A is Avogadro’s number and Mjj is the atomic mass of hydrogen; and the uranium number density is
TT = X z A Mu NUREG/CR-0674 RFP-2895 Page 35
where x is the weight of uranium oxide per can (15088.3 ±
33.9 g) and z is the uranium assay (0.8449 ± 0.0008), the
H/U atomic ratio per can is given by
H/U = w i I Mjj X z vi=l and its uncertainty is
5 6 (H/U) y (H/U) i=l L\i=l
Substituting the corresponding numerical values into the above equations, the H/U value is given as 0.77 ± 0.06 for all cases. The average properties (H/U value and oxide weight per can) of the oxide cans are listed in Table VI.
The H/U value for the special cans is similarly calculated and presented in Table VI.
A chemical analysis of the gas released during the TGA procedure was expected to show only water as the temperature was slowly increased to 500° C. Instead, COg gas was un expectedly discovered along with water at the temperature between 200° and 500° C. The weight of COg gas decreases the water weight obtained from the TGA test. The present results of the water content were based on the water per centages, as determined from three samples, to be 61.1%, W 52! c! TO ►O W O I t*l M O W 00 \ 0> to o Oi 50 I o TABLE VI O) -J »(>> Average Properties of Oxide Cans
Description Average Average Weight^ (g) of Can H/U Oxide^ Water Can^
oxide can 0.77 ± 0.06 15129.1 ± 33.7 273.2 ± 4.1 526.3 ± 3 . 2
special 0.73 ± 0.06 12253.2 ±76.7 2 1 0 . 6 ± 1.1 620.6 ± 1.5 oxide can
^ Quantity following the average is the standard deviation of the distribution, not an error estimate. ®Weight includes initial moisture. ®Weight measured after watering holes were drilled. NUREG/CR-0674 RFP-2895 Page 37
83.4% and 90.0% between 200° and 500° C. The percent of water released between 200° and 500° 0 was 46%. The uncer tainty due to the CO2 gas affects only this fraction of the weight. No such effect was noted below 200° C where 54% of the water was released.
The effect on of the uncertainty of the H/U value was investigated. The metal driver, concrete-reflected core was analyzed using the DTF-IV computer code with Hansen- (9) Roach sixteen-group cross sections — . For the ± 0.06 shown in Table VI for H/U, was ± 0.0028, which corresponds to about one-half of a sigma for a typical KENO^~^ calcu lation.
NUREG/CR-0674 RFP-2895 Page 39
METAL AND SOLUTION DRIVERS
Metal Driver
The metal driver was built of concentric, enriched uranium hemishells^— ^ as shown in Figure 8 . Each 3.28-mm-thick shell had a 7.14-mm-diameter mounting hole at the pole and four
3.17-mm-diameter holes near the equator for disassembly tools. Gaps for machining tolerance averaged 0.10 mm between the concentric shells. For contamination control purposes, the shells were greased with petroleum jelly during the assembly. The specifications for the enriched uranium shells, mounting bolt, driver mount, supporting rod for the driver mount, and petroleum jelly are given in Table VII.
Table VIII shows the dimensions and mass of the enriched uranium hemishells, which have a designated part number.
Figure 9 shows a photograph of a uranium metal driver mounted on the supporting rod.
Solution Driver
Two pairs of solution driver cans were fabricated for these experiments. These are made of type 304 stainleSs steel and have the shape of a rectangular box equivalent to two oxide cans side-by-side (see Figure 10). One pair was used for high concentration uranyl nitrate solution (351.2 kg
U/m^), and the other pair for low concentration solution
(86.4 kg U/m^>. Each pair consists of one can for the south NUREG/CR-0674 RFP-2895 Page 40
FIGURE 8
Typical metal driver assembly and its supporting equip ment. All dimensions are in millimeters. NUREG/CR-0674 RFP-2895 Page 41
T o leran ce G ap Drwg. Not To Scale
T h read ed Mounting Bolt 6 .3 5 0 D
— Uranium Components rQ - O u te r R adius n - Inner Radius
Driver M ount 3 1 .2 0 0 83.0 -- 5 0 .8
Supporting Rod 15.5 0 0
8 5 2 .0
FIGURE 8 NUREG/CR-0674 RFP-2895 Page 42
TABLE VII
Properties of Metal Driver and its Supporting Materials
Composition Type Element Remarks by Weight^
S34 U 1 .0 0 % individual enriched 2 3S U 93.19% component uranium 2 38 U 0.40% density = metal 2 38 U 5.41% 18.76 ± 0.06 g/cm3
carbon 85.0% material density = petroleum hydrogen 14.8% 0.816 g/cm^ Jelly aluminum 2 0 . 0 ppm average density‘s (trade name calcium 7 .0 ppm over gap = Petrolatum) copper 23.0 ppm 0.107 g/cm3
stainless chromium 17.4% steel nominal density = nickel 9.4% driver 7.90 g/cm3 iron 72.0% mount
stainless chromivim 18.6% steel nominal density = nickel 9.4% mounting 7.90 g/cm^ iron 72.0% bolt
carbon 0.152% aliminum 0 .0 0 1 % silicon 0 .0 1 2 % carbon chromium 0 .0 1 0 % steel manganese ~ 1% nominal density = supporting iron® 98.770% 7.82 g/cm3 rod nickel 0.015% copper 0.008% molybdenum 0.006% others® 0.027%
^ The normal accuracy is within ± 5%. ^Obtained by subtracting sum of other nine elements from 100%. ^Sum of other metallic elements as they exceed the limits of detection for emission spectroscopy. “^For example, for the 34.135 kg driver, the total weight of petroleum jelly was 5.5 g for the entire driver. NUREG/CR-0674 RFP-2895 Page 43
TABLE VIII
Properties of Uranium Metal Hemishells
Inside Outside Cumulative Part Mass* Radius Radius Mass* Number (mm) (mm) (kg) (kg)
1 0 . 0 2 0 . 0 1 0.296 0.296 2 0 . 0 2 0 . 0 1 0.296 0.592 3 20.13 23.37 0.176 0.768 4 20.13 23.38 0.176 0.944 5 23.47 26.70 0.233 1.177 6 23.47 26.70 0.234 1.411 7 26.80 30.03 0.302 1.713 8 26.79 30.03 0.301 2.014 9 30.13 33.35 0.376 2.390 10 30.12 33.36 0.375 2.765 11 33.44 36.70 0.465 3.230 1 2 33.44 36.70 0.465 3.695 13 36.80 40.02 0.555 4.250 14 36.80 40.02 0.554 4.804 15 40.17 43.38 0.653 5.457 16 40.16 43.38 0.651 6.108 17 43.46 46.70 0.766 6.874 18 43.46 46.70 0.766 7.640 19 46.79 50.04 0.889 8.529 20 46.78 50.04 0.890 9.419 21 50.17 53.37 1.004 10.423 2 2 50.13 53.36 1 . 0 1 1 11.434 23 53.46 56.69 1.147 12.581 24 53.46 56.69 1.149 13.730 25 56.79 60.03 1.288 15.018 26 56.79 60.01 1.286 16.304 27 60.11 63.35 1.445 17.749 28 60.12 63.34 1.441 19.190 29 63.45 66.71 1.612 20.802 30 63.34 66.70 1.612 22.414 31 66.78 70.02 1.779 24.193 32 66.79 70.03 1.777 25.970 33 70.06 73.30 1.949 27.919 34 70.10 73.34 1.951 29.870 35 73.42 76.66 2.135 32.005 36 73.43 76.66 2.130 34.135
*The nominal accuracy of mass is within ± 0.5%. NUREG/CR-0674 RFP-2895 Page 44
FIGURE 9
An assembled uranium metal driver which will be posi tioned in the «mpty space in the oxide cans when the horizontal table is closed. The cylinder attached to the top of the sphere is the source holder. NUREG/CR-0674 RFP-2895
■
f q ? • '4
FIGURE 9 NUREG/CR-0674 RFP-2895 Page 46
FIGURE 10
One pair of solution driver cans: one for the south half-table and the other for the north half-table. The fill hole for the north can is in shadow in the diagonal recessed notch. NUREG/CR-0674 RFP-2895 Page 47
f,
K NUREG/CR-0674 RFP-2895 Page 48
table and one can for the north table. Properties of these solution driver cans are listed in Tables IX and X. Samples of the uranyl nitrate solution were taken from the drivers over the course of the experiments. The laboratory analyses for the high and low concentration solutions are averaged and are given in Table XI. TABLE IX
Solution Driver Can Dimensions
High Concentration Low Concentration Driver Driver Property North South North South
Weight (g) 2700 2704 2684 2692
Wall Thickness (mm) 1.5 1.5 1.5 1.5
Inside Length (mm) 297.5 297.9 297.6 298.7
Inside Width (mm) 149.5 150.4 150.0 150.0
Inside Height (mm) 151.5 151.0 150.5 150.0
Inside Volume (mm^ ) 6734000 6761000 6714000 6716000
Usable Volume* (mm^ ) 6269000 6761000 6268000 6716000 12! C l 50 *Because of the hole on the edge, the south can cannot contain W solution to the top of the can. The measurement accuracy is O within ± 1 %. >0 o >0 50 I I 0) N O 00
TABLE X
Material Composition of Solution Driver Can
Weight-Percent by Element* Density Material (g/cm^) Fe Cr Mn Ni
Stainless Steel 70.50 18.70 1.50 9.30 7.894 ± 0.005 (type 304)
*The accuracy of weight-percent is within ± 5%. NUREG/CR-0674 RFP-2895 Page 51
TABLE XI
Properties of Uranyl Nitrate Solutions Used as Drivers
Uranium Solution Excess Total Concentration Density Nitric Acid Impurity (kg U/m3) (g/cm^) (molar) (ppm)
86.42 ± 0.20 1 . 1 2 0 1 ± 0 . 0 0 0 2 0.149 ± 0.003 1340
351.18 ± 0.57 1.4885 ± 0.0014 0.549 ± 0.015 1340
Uranium isotopic enrichment (wt-%): 1 . 0 2 2 ± 0.043 for ^34^ 93.172 ± 0.060 for 235u 0.434 ± 0.005 for 236u 5.372 ± 0.036 for 238u All uncertainties represent one standard deviation about the mean for multiple samples.
NUREG/CR-0674 RFP-2895 Page 53
HORIZONTAL SPLIT TABLE
The horizontal split table, which consists of two movable tables, is used for experiments. Once the fissile load has been assembled, the north half-table is moved to the closed position with the other half fully open. Measurements made during hand assembly assure the safety of this closure.
Closing the south half-table is the remotely-controlled % means of achieving criticality. Differential transformers are used to read out the final 33 mm of closing. The entire 1524 mm of table closure is monitored by a selsyn system. The table closing speed is adjustable to stay within reactivity addition rate limits. Both tables move away from center for reactivity removal under accident or scram conditions.
Figures 11, 12, and 13 show the horizontal split table.
The materials used in the split table are identified as a material name with designated numbers. The material composi tions are presented in Table Xll. NUREG/CR-0674 RFP-2895 Page 54
FIGURE 11
The horizontal split table loaded for an experiment but fully open. A spherical metal driver is located between the north and south tables. All dimensions are in millimeters. Drwg. Not To Scale
N o rth S o u th
O xide Cans R eflecto r
V oid ^ F o r Driver
Driver S u p p o rt 912
915 224 1857
1880 1884 ^ 10 ' 55 a S3 Vi M — ► 4 ^ 10 o Horizontal Table Base S3 \ ►0 O P S3 CTQ I I F lo o r (D to O 00 03 Table In Open Position 01 CO ^ 01 iP F I G U R E 11 NUREG/CR-0674 RFP-2895 Page 56
FIGURE 12
The cross section of the south half of the horizontal split table. All dimensions are in millimeters. NUREG/CR-0674 RFP-2895 Page 57
Drwg. Not To Scale
R eflecto r
South Table
1485 w jL
2 3 7 f# 1015 -►I W 5 ______--- T 10 2160
10 798
25
S7^ — 483 i F loor
F I G U R E 1 2 NUREG/CR-0674 RFP-2895 Page 58
FIGURE 13
The cross section of the north half of the horizontal split table. All dimensions are in millimeters. NUREG/CR-0674 RFP-2895 Page 59
Drwg. Not To Scale
R e flecto r
North Table
1485
1129
S5 l^- 203 178 287 163 ,
1162 S2
798
2160
S7
F lo o r
FIGURE 13 TO (D I H M Q 0> 00\ TABLE XII O 50 O Oi W Elemental Composition of Steel and Stainless Steel I Structural Components of the Horizontal Split Table o 05 »(Sk Use or Key to Type Weight-Percent by Element' Application Figures of All of Steel 11, 12, 13 Steel C Sn Si Cr Mn® Fe^ Ni Cu Mo Co s Others' 1.27-cm-thick mild 0.18 0.01 0.5 0.25 ~ 1 97.866 0.05 0.1 0.02 0.01 0.14 reflector Si steel base plate top of both stain 0.024 19.7 69.976 10.3 split table ®2 less halves steel 2 .5-cm-thick mild 0.235 0.01 0.02 0.14 ~ 1 98.535 0.015 0.015 0.012 0.01 0.008 plate, north S3 steel table only mild I-beam, south 0.238 0.01 0.14 0.09 ~ 1 98.258 0.12 0.1 0.013 0.02 0.019 table only S4 steel 5-cm-thick mild reflector 0.22 0.01 0.25 0.03 ~ 1 98.390 0.04 0.01 0.007 0.015 0.028 S5 steel support plate north table mild scram guide 0.63 0.1 0.1 0.01 1.1 97.885 0.1 0.01 0.02 0.045 ‘ Sg steel bars lower portions mild S7 0.25 0.1 0.1 0.1 0.7 98.536 0.1 0.01 0.02 0.04 0.044 of table steel
1 The nominal accuracy of weight-percent is within ± 5%. ®Mn exceeded the upper limit of detection for emission spectroscopy and is taken to be 1%, which assumption is consistent with experience and with three measured values determined by atomic absorption. ®Obtained by subtracting sum of other columns from 100%, ^Sum of other metallic elements as they exceed the limits of detection for emission spectroscopy. NUREG/CR-0674 RFP-2895 Page 61
REFLECTORS
Concrete Reflector
The concrete used in the experiments was chosen to represent a type of concrete (designated "03" in Reference
12) commonly used in the nuclear industry. The concrete reflector consists of four parts: two end panels and two shells. When the table was closed, the reflector was a thick-walled cubical shell 1290 mm along an exterior side with an ~ 780 mm cubical interior cavity. In addition to the concrete, the four parts combined contained -^3.0 kg of steel rebar. The two end panels each contained a total of
6.1 m, weighing 0.91 kg, of 4.9-mm-diameter steel rod welded in a square gridwork placed in the mid-plane of each panel during pouring. The north concrete shell contained a single steel loop weighing 0.69 kg (4.6 m long) in the middle of the shell, and the south shell contained two steel loops weighing 1.38 kg (total 9.27 m long). Figure 14 shows the dimensions of the reflector.
The concrete composition was determined by two methods.
The first composition was calculated by multiplying the weight-fraction of each ingredient in the concrete by the weight-percent of each element in that ingredient. The weight-percent was obtained from an elemental analysis of the sand, cement, and limestone, which were ingredients of the concrete. The second method was an elemental analysis NUREG/CR-0674 RFP-2895 Page 62
FIGURE 14
Concrete reflector showing the dimensions. The uncer
tainties are the standard deviations of measurements. All dimensions are in millimeters. NUREG/CR-0674 RFP-2895 . Page 63
254 ± 2
261 ± 2
255 ±2 o>
832 ± 1
255 ± 2
323 ± 2 QQ NUREG/CR-0674 RFP-2895 Page 64
of the well-cured concrete. Detailed analysis of the con crete was described in Reference 5. Table XIII gives the elemental analysis of the cured concrete as determined by
these two methods, with the average used in calculating atomic number densities. The analytical methods are listed in the table. For most methods, the nominal accuracy claimed
is about ± 5% of the amount present except for the Keldahl method at this nitrogen level: ± 100%. The density of the concrete is measured to be 2.234 ± 0.002 g/cm^. The total weight of the concrete reflector excluding bolts, nuts, and eyelets is 4070 kg.
Plastic Reflector
The methyl methacrylate (plastic) reflector resembled
the concrete reflector in size and shape. Each panel or shell of the thick-walled cubical shell was laminated of
4 to 16 different plastic sheets bolted together. Interior dimensions for each panel and shell can be found in Figure
15. Figure 16 shows a photograph of the plastic reflector.
As stated in the Experimental Procedure section, for the
three plastic-reflected measurements, the north reflector end panel was raised up 261 mm. For all other measurements,
this reflector end panel rested on the same plate as the
rest of the reflector pieces. (See Figure 16.)
The plastic reflector was made of non-fire retardant
common plexiglas [CH2 :C(CH3 )C0 2 CH2 ] and fire retardant TABLE XIII
Composition of Concrete in Weight-Percent
Analysis of Analysis of Element Average Method* Ingredients Cured Concrete
Hydrogen 0.59 0.57 0.58 CH Carbon 5.50 5.30 5.40 CH Nitrogen 0.00 0.05 0.025 K Oxygen 49.54 48.95 49.25 Difference Sodium 0.22 0.68 0.45 AA Magnesium 1.69 1.40 1.56 AA Aluminum , 2.05 2.37 2.21 AA, C Silicon 16.81 17.20 16.94 AA, C Sulfur 0.14 0.12 0.13 Eschka Potassium 0.13 1.42 0.77 AA Calcium 22.65 20.80 21.77 AA Titanixim 0.00 0.09 0.05 AA Iron 0.67 1.05 0.86 AA
Total 100.0 100.0 100.0
AA; atomic absorption O CH; carbon/hydrogen analyzer tJ ^ O K: Keldahl (t) t o o C: calorimetric 00 O) Ci 50 oi cn 0^ NUREG/CR-0674 RFP-2895 Page 66
FIGURE 15
Plastic reflector showing the dimensions. The uncer tainties are the standard deviations of measurements. All dimensions are in millimeters. NUREG/CR-0674 RFP-2895 PSfee 67
258 ± 4
'V
253 ± 1
253 + 1 D)
253 ± 2
FIGURE 16 A photograph of the plastic reflector sitting on the floor. The front reflector would be rotated 180° to be in the experimental position. NUREG/CR-0674 RFP-2895 Page 69 NUREG/CR-0674 RFP-2895 Page 70 plastic (designated "Plexiglas FI-3") sheets. Paper and glue were attached to each plastic sheet. The elemental composition of each of the four materials was determined by laboratory analysis. The weighted average weight-percent of elements for the plastic reflector was obtained by determining the weight-fraction of each material, multiplying each fraction by each weight-percent, and summing the products. The weighted average density for the plastic reflector was derived in a similar manner as for the elemental composition, except that the densities were multiplied by volume fractions. The weighted average elemental composition and density are given in Tables XIV and XV. The total weight of the plastic reflec tor excluding bolts, nuts, and eyelets is 2137 kg. Steel and Environmental Reflectors The steel reflector consists of a two-part box, one for the north table and the other for the south. Geometrically, the reflector is a 6.35-mm-thick cubical shell with an 780 mm interior cavity when the table is closed. The detailed dimensions for the reflector are given in Figure 17. The elemental composition obtained by chemical analysis is given in Table XVI. The most significant environmental reflectors in addition to the Steel frame and the horizontal split table were a large stainless steel and plexiglas hood-like enclosure and TABLE XIV Average Weight-Percent of Plastic Reflector Material Weight Weig it-Percent of Element* Material Fraction H C 0 Cl Hr P N fire retardant 0.7997 7.16 52.03 29.82 1. 81 7.10 1. 02 0.16 plastic non-fire retardant 0.1938 7.84 59. 59 32.23 plastic paper 0.0049 6.48 42.17 49.50 glue 0.0017 11.67 86.29 1.20 average value adjusted by 7. 28 53.51 30. 34 1. 45 5. 68 0. 82 0.13 weight fraction Cl S3 tn *The nominal accuracy is within ± 1%. o ?3 \ ► 0 ^ 0 p S3 CR3 I I (5 IS3 O 00 O) NUREG/CR-0674 RFP-2895 Page 73 TABLE XV Average Density of Plastic Reflector Volume Density Material Fraction (g/cm3) fire retardant 0.7835 1.284 ± 0.001 plastic non-fire retardant 0.2056 1.185 ± 0.001 plastic paper 0.0081 0.766 ± 0.001 glue 0.0028 0.728 ± 0.001 average value adjusted by 1.258 ± 0.001 volvime fraction NUREG/CR-0674 RFP-2895 Page 74 FIGURE 1 7 Steel reflector showing the dimensions. The uncertain ties are the standard deviations of measurements. All steels are 6.35-mm-thick. All dimensions are in millimeters. NUREG/CR-0674 RFP-2895 Page 75 o CH O M O) "9 S3 a d TO S3 (D 1 M o 00 \ OJ to o cn S3 1 o TABLE XVI o *4 Material Composition of Steel Reflector Steel Weight-Percent by Element Density Type C A1 Si Cr Mn Ni Cu Mo (g/cm3) mild 0.165 < 0.01 0.07 0.04 0.76 98.965 0.04 0.03 < 0.01 7.845 ± 0.005 steel ^ The nominal accuracy is within ± 5%. ®Obtained by subtracting sum of other columns from 100%. NUREG/CR-0674 RFP-2895 Page 77 the experimental room walls, ceiling, and floor. Figures 18(a) and 18(b) show the locations of the horizontal table and hood, and the dimensions of the experimental room. NUBEG/CR-0674 RFP-2895 Page 78 FIGURE 18(a) Experimental room looking south showing the east-west section. All dimensions are in millimeters. TOP610 CONCRETE 11280 Drwg. Not To Scale I— RAILING 9750 STAINLESS STEEL 3.2 694 889 CATWALK 1220 1220 0.635 710 3054 UJ -PLASTIC ^ 6.35 LU 4902 Horizontal Table Top 1981 1650 SOLUTION HOOD a 2159 798 203 (D M O 00 A <0 BOTTOM CONCRETE to 01 GROUND FIGURE 18(a) NUREG/CR-0674 RFP-2895 Page 80 FIGURE 18(b) Experimental room looking west showing the north-south section. All dimensions are in millimeters. TOP 610 CONCRETE 10670 Drwg. Not To Scale 3585 2443 RAILING ---- 44.6 O.D. 642 (Carbon Steel) 1738 2880 217 1220 1520 CATWALK - 127 (Carbon Steel) SOLUTION HOOD UJ UJ UJ 3.2 UJ (Table In Closed Position) STAINLESS STEEL Horizontal Table Top c3 W t*J 3764 1232 O O 798 203 *0 W CP5 I I (D M o 00 O) BOTTOM CONCRETE 00 M OI rfk GROUND FIGURE 18(b) NUREG/CR-0674 RFP-2895 Page 83 CORE DIMENSIONS As stated in the Experimental Procedure section, the core was measured to determine the critical dimensions. Figures 19 and 20 show the core configurations for the different drivers (see also the photograph in Figure 9). On the south part of the core, two oxide cans in the two top layers were slotted to allow for the removal of the source. In addition, the bottom two cans for the metal driver measurements also were slotted to accommodate the support rod for the driver. Each core was a 5 x 5 x 5 array of oxide cans with one or four cans removed for the driver. An aluminum (type 6061- T6) plate was used between each horizontal layer of oxide cans (a total of eight for the complete core). This plate was necessary to distribute the weight of the oxide cans, The aluminum plate was 1.6-mm-thick and the size of a 3 x 5 or 2 X 5 array of cans. The plates, weighing a total of 10.4 kg, are not shown on the figures but can be seen in Figure 9. The composition data for these plates are given in Table XVII. Table XVIII presents the critical parameters and core sizes of 5 x 5 x 5 arrays for the different reflectors. Figure -21 is the key to Table XVIII. NUREG/CR-0674 RFP-2895 Page 84 FIGURE 19 Description of core cuboid with a void for metal driver. These dimensions are used in Table XVIII. NUREG/CR-0674 RFP-2895 Page 85 t: o 05 H § M NUREG/CR-0674 RFP-2895 Page 86 FIGURE 20 Description of core cuboid with a solution driver. A second solution driver is located opposite the one shown. These dimensions are used in Table XVIII. NUREG/CR-0674 RFP-2895 Page 87 UJ o N CH o M LU CO ►0 w ^ P C3 crq W ( C I M 10 o 00 00 \ 00 «o o tn so I o TABLE XVII 05 o Composition of Aluminum Plate Weight-Percent by Element’^ Density Material Fe Mg Si Mn Cu Or Zn Al® (g/cm3) Aluminum 0.7 < 0.01 0.2 1.4 0.14 < 0.01 0.01 97.53 2.737 ± 0.002 (type 6061-T6) ^ The accuracy of weight-percent is within ± 5%. ® Obtained by subtracting sum of other columns from 100%. TABLE XVIII Core Dimensions for the Minimally-Reflected, Concrete-Reflected, and Plastic-Reflected Arrays of Uranium Oxide for the Three Different Drivers Steel Reflector Core (Minimally-Reflected) Concrete Reflector Plastic Reflector Dimension* High Low Uranium High Low Uranium High Low Uranium (mm ) Concentration Concentration Metal Concentration Concentration Metal Concentration Concentration Metal Solution Solution Sphere Solution Solution Sphere Solution Solution Sphere A 465.011.2 462.611.6 465.411.1 466.011.2 463.312.3 465.311.0 462.311.1 462.311.1 462.311.1 B 3.9611.4 6.7711.7 2.6511.5 8.4112.1 10.7312.8 9.9711.9 5.0711.7 7.2411.3 5.9211.4 F 309.810.8 308.514.2 312.612.1 311.011.6 311.312.3 310.512.6 307.213.6 307.213.6 307.2+3.6 G 0.0 0.0 0.0 4.0212.2 4.0212.2 4.0212.2 0.0 0.0 0.0 H 769.512.3 768.911.7 768.210.8 769.811.1 768.911.1 769.511.5 767.811.2 767.811.2 767.811.2 I 62.612.3 62.611.4 61.711.3 61.912.0 62.912.2 62.712.8 69.511.9 69.511.9 69.511.9 J 768.212.5 768.311.7 768.011.4 769.810.7 769.311.1 769.110.9 767.811.2 767.811.2 767.811.2 K 65.812.2 65.111.4 65.711.1 60.511.3 61.111.2 60.810.8 66.612.9 66.612.9 66.612.9 L 765.211.4 766.511.7 767.311.4 775.010.7 775.010.7 775.010.7 768.813.7 768.813.7 768.813.7 M 768.612.4 767.112.4 768.411.4 769.313.6 769.3+3.6 769.313.6 769.812.5 769.812.5 769.812.5 N 9.9411.1 11.6514.3 6.2312.3 11.0112.4 10.3312.9 10.7712.4 6.1813.8 8.3513.7 7.0313.7 O 0.0 0.0 0.0 4.2812.9 4.2812.9 4.2812.9 0.0 0.0 0.0 SZ! a Refer to Figures 19 through 21 for identification of dimensions. See Table XVI for core separation (D). ta o 5d\ o p *0 w trq I I (D to O 00 o> 00 50 ^ to cn NUREG/CR-0674 RFP-2895 Page 90 FIGURE 21 Configuration of core cuboid and reflector showing the elevation and plan views. Dashed lines indicate core boundaries, and solid lines represent reflector boundaries. NUREG/CR-0674 RFP-2895 Page 91 Top O K ] ■ H ELEVATION VIEW N G- -D Bottom East PLAN VIEW West NORTH SOUTH CORE CORE FIGURE 21 NUREG/CR-0674 RFP-2895 Page 93 CRITICAL RESULTS Critical Parameters Experimental parameters for the ten critical arrange ments are presented in Table XIX. The table separations are measured for one positive reactor period (slightly into delayed supercritical) and for one negative reactor period (slightly subcritical). The critical table separation is determined by a linear interpolation between these two points. In these experiments, the table separation means core and reflector separations. Any irregularity of the core reflector face is included in the uncertainty for the core spearation given. As discussed in the Experimental Procedure section, the correlation between the differential transformer readout and the separation of the core sections was obtained with small balls of a putty-like substance. Typically, five small mounds were located over the face of the core area and eight mounds were distributed around the reflector. When the table closed to some known differential transformer reading, the pliable mounds compressed to a thickness which could be measured later. This correlation between the core separation and the differential transformer reading was established within ~ 10 mm of the critical separation in all cases. *0 W 3 P» h rj S TO W (D I M to O to 00\ (D O TABLE XIX Oi to I Principal Critical Parameters of Ten 5x5x5 Arrays o of Damp, Compacted, Low-Enriched Uranium Oxide 0) METAL DIAMETER CRITICAL PERIODDATA TOTAL OXIDE NUMBER TYPE CORE DRIVER OR Negative Positive REFLECTOR WEIGHT OF SOLUTION HEIGHT^ SEPARATION Separation^ Separation^ MASS Period Period (kg) DRIVER (mm) (mm) (mm) CANS° (kg) (mm) (seo) (sec) 93.12% 29.870 0.0 (ID), 146.68(OD) 12.60 ± 1.19 1080 12.69 ± 1.19 468 12.39 ±1.19 concrete 1864.6 120 Enriched 29.870 0.0 (ID), 146.68(0D) 8.01± 1.67 1716 8.08 ± 1.67 492 7.74 ± 1.67 plastic 1864.6 and Uranium 4(S) Metal 33.543 40.02(ID), 153.32(0D) 14.64 ± 2.26 1134 14.73 ± 2.26 624 14.50 ± 2.26 minimal 1864.6 105.47 (S), 106.22 (N) ± 1.79 1218 6.7 1 ± 1.79 462 6.60 ±1.79 concrete 1824.9 High 14.066 6.68 119 Concentration 14.844 111.30 (S), 112.10 (N) 9.42 ± 3.01 882 9.50 ± 3.01 510 9.27 ±3.01 plastic 1824.9 and Solution 2(S) (351.2 kg U/m3) 16.143 121.04 (S), 121.91 (N) 9.04 ± 1.44 486 9.20 ± 1.44 408 8.85 ± 1.44 minimal 1824.9 12.446 124.08 (S), 124.42 (N) 8.64 ± 1.31 930 8.76 ± 1.31 516 8.4 1 ± 1.31 concrete 1824.9 Low 119 Concentration 12.875 128.28 (S), 128.74 (N) 9.12 ± 1.42 474 9.33 ± 1.42 744 8.98 ± 1.42 plastic 1824.9 and Solution 13.001 129.54 (S), 129.99 (N) 11.17 ± 1.42 492 11.28 ± 1.42 462 11.05 ± 1.42 plastic 1824.9 (86.4 kg U/m^) 2(S) 13.999 139.46 (S), 139.99 (N) 6.30 ± 1.66 522 6.41 ±1.66 498 6.18 ± 1.66 minimal 1824.9 inner diameter; CD = outer diameter; S = south; N = north. special oxide can. 'The ± is on the absolute value. The relative accuracy between the two period data points is ± 0.08 mm. NUREG/CR-0674 RFP-2895 Page 95 Mathematically summarizing the above descriptions, the core separation, S^, is expressed as ^ ^P “ 2.294 (Dp - D^,) in mm * where is a core separation measured by the putty ball. Dp the differential transformer reading at the point of putty ball measurement, D^ a differential transformer reading based on the critical core separation, and 2.294 a conver sion factor between the table separation and the differential transformer reading. The critical transformer reading, D^., is calculated by a linear interpolation between two differentia] transformer readings for one positive and one negative reactor period. The uncertainty, 6 8^,, for S^. is given by 6Sc = r (ASp)^ + 5.262 [(ADp)^ + (ADc)^]]^ where ASp, ADp, and AD^, are the standard deviations resulting from the measurement uncertainties. Concrete-Reflected Experiments The concrete-reflected oxide experiments were performed with three different drivers. The first criticality was attained with the high concentration (351.2 kg U/m^) solution driver in the 5x5x5 array when the core separation between the core parts was 6.68 mm. The array consists of 119 oxide cans, two special oxide cans, and two solution driver cans equivalent to four oxide cans as listed in NUREG/CR-0674 RFP-2895 Page 96 Table XX. For the solution drivers, the south can contained 7.034 kg solution, and the north 7.032 kg, for a total of 14.066 kg. When the low concentration (86.4 kg U/m^) solution driver was used, the critical core separation was 8.64 mm. In this critical experiment, the south driver can had 6.227 kg and the north can had 6.221 kg, for a total of 12.446 kg of solu tion. When criticality was achieved, the spherical uranium metal used in the 5x5x5 array of oxide as a driver was a 146.7-mm-diameter metal sphere weighing 29.87 kg, whose part numbers are 1 through 34 as listed in Table VIII. The array consisted of 120 oxide cans, four special oxide cans, and a uranium metal driver whose diameter is almost equal to the width of an oxide can. See Table XX for the array configuration. The core separation was 12.60 mm, and the metal sphere was almost centered in the array with 2.85 mm gaps at five sides (top, bottom, east, west, and north) except for south when the criticality was attained. The gap at the south side was 15.45 mm. Plastic-Reflected Experiments The low-enriched oxide critical experiments reflected by the plastic cubical shell were performed with three drivers: a high concentration solution, a low concentration solution, and a uranium metal sphere. NUREG/CR-0674 RFP-2895 Page 97 TABLE XX Weight(in grams) and Location of Oxide Cans as Placed into the Low-Enriched Oxide Array with H/U 0.77 Row Column West Second Third Fourth East 16049 16037 16034 16075 16048 15999 15989 15996 15989 16000 North 16076 15996 16027 16035 16019 0) iH 15970 15968 15966 15950 15950 XI 15935 15925 15961 15993 15991 a E-i 16090 15985 15979 15974 15950 ■p 16014 16016 16020 16015 16005 O Second 15985 a b 16020 16020 15966 15923 15951 16027 15989 15976 15989 15982 16010 15982 16011 16016 h 16025 16001 16006 16015 g 15996 16009 Third 16000 c d 15954 15969 16040 16034 f 16023 16023 16015 16031 e 16021 16000 Q) rH 16048 16013 16030 16019 16019 X a 15999 16020 16005 16025 16018 Eh Fourth 16016 16000 16013 15976 16023 xi 15957 15944 15965 15975 15985 -p 15971 15995 16015 16010 15993 0 15954 15945 15913 15950 16024 16021 16002 15995 15935 16001 South 15945 16000 16025 16030 16021 15971 15990 15954 16005 16014 16000 15955 15980 15976 16016 ^Weight includes can + oxide + water + plastic. ®The five weights in any box of this table correspond to cans placed into the array from bottom layer to top in the associated column and row. These weights were measured just before the experiments were performed (March 1, 1978). For the solution driver case, a, b, c, and d are occupied by two solution driver cans. In this case, e = 16031, f = 16028, g = 13150, and h = 13171. For the metal driver case, d is occupied by a metal driver. In this case, a = 16034, b = 16056, c = 15971, e = 13150, f = 13171, g = 13100, and h = 13225. NUREG/CR-0674 RFP-2895 . Page 98 For the low concentration driver, two critical data points were obtained. For both, the array consisted of 119 oxide cans, two special oxide cans, and two solution driver cans. When the solution driver contained a total of 13.001 kg (6.500 kg in the north and 6.501 kg in the south), the critical separation was 11.17 mm. When the driver was changed to a total of 12.875 kg (6.437 kg in the north and 6.438 kg in the south), the critical separation was 9.12 mm. With the high concentration driver, the oxide system was critical at the core separation of 9.42 mm. In this experiment, a total of 14.844 kg solution was contained in the south driver can (7.423 kg) and the north can (7.421 kg). A 146.7-mm-diameter metal sphere weighing 29.87 kg (part numbers are 1 through 34 as shown in Table VIII) was used in the 5x5x5 array of oxide. The array consisted of 120 oxide cans, four special oxide cans, and a uraniiun metal driver. See Table XX for the array configuration. The core separation was 8.01 mm when the array was critical. The position of the sphere was the same as for the concrete- reflected case except for south side. The gap at the south side was 10.86 mm. Minimally-Reflected Experiments The minimally-reflected oxide experiments were conducted with the same three driver types. NUREG/CR-0674 RFP-2895 Page 99 With the high concentration solution driver, the 5x5x5 oxide array was critical when the core separation was 9.04 nun. The array configuration (see Table XX) was the same as for the concrete- and plastic-reflected cases with the solution drivers. For the solution driver, the south can contained 8.072 kg solution and the north can 8.071 kg, a total of 16.143 kg. Criticality was achieved with the low concentration driver in the 5x5x5 oxide array when the core separation was 6.30 mm. In this critical experiment, a total of 13.999 kg solution was used, 6.999 kg in the south driver can and 7.000 kg in the north can. A 153,32-mm-diameter uranivim metal sphere with a 40.02- mm-diameter spherical void at the center was used as a driver in the same oxide array as before. The metal driver weighed 33.543 kg and consisted of part nvimbers 3 through 36 as listed in Table VIII. The array configuration was the same as for the concrete- and plastic-reflected cases with the metal drivers. The critical core separation was 14.64 mm. Figure 22 shows the experimental cans loaded inside the steel reflector on the horizontal table. There were no gaps between the sphere and cans at five sides (top, bottom, east, west, and north) except for south when the table was closed. The gap at the south side was the same as the core separation. NUREG/CR-0674 RFP-2895 Page 100 FIGURE 22 Minimally-reflected oxide cans with the uranium metal driver. Both tables are open. NUREG/CR-0674 RFP-2895 Page 101 y . ^ 't . '• > . t '!?v ' s. m ...... NUREG/CR-0674 RFP-2895 Page 103 DISCUSSION OF UNCERTAINTIES Measured uncertainties in parameters given in the tables are the standard deviation of the measurements and are defined by the formula N CTf = N - 1 ^"^i “ ^ ^ i=l where is the value of a particular measurement and X is the average value. The only exception is the core separa tion. The uncertainty is proportionally propagated from the standard errors of differential transformer reading and putty ball measurements. The plastic-reflected experiment with the metal driver was repeated six months after the original experiment, to account for the uncertainties due to the oxide weight gain and statistical error of measurement. The critical core separation of the repeated and original experiments was 8.26 ± 1.26 and 8.01 ± 1.67 mm, respectively, under the same conditions except for the oxide weight gain and a small perturbation of the core resulting from the reassembly of the oxide cans. The main oxide weight gain was due to: [l] using one different can in the repeated experiment, increasing the total weight by 35 grams; and [2] the absorp tion of oxygen from the air, previously discussed. The change NUREG/CR-0674 RFP-2895 Page 104 in reactivity in terms of change of the core separation near critical, is 10.2 ^/mm for the original experiment and 17.7 ^/mm for the repeated experiment, averaging 14.0 ^/mm. This indicates that the repeated experiment was in good agreement with the original, i.e. Ak ^ 0.0003. Similarly, the concrete-reflected experiment with the metal driver was repeated ten months after the original experiment. The critical core separation of the repeated and original experiments was 18.16 ± 1.23 and 12.60 ± 1.19 mm, respec tively. The difference (5.56 mm) of the two core separa tions corresponds to Ak ~ 0.005, which is approximately one sigma of the k^^f value calculated by the KENO-IV code using 100 batches of 300 neutrons per batch. Therefore, the repeated experiments imply that variations of the oxide weight gain and small perturbations of the can positions have little effect on the reactivity of the system. One oxide can was disassembled on January 19, 1979, to find out water distribution throughout the can. The water distribution was found to be almost 96%) uniform through out the can. This also indicates that any change in water distribution during the experimental period had little effect on the reactivity. NUREG/CR-0674 RFP-2895 Page 105 REFERENCES 1. S. R. Bierman and E. D. Clayton, "Critical Separation Between Subcritical Clusters of 2.35 Wt-% 235u Enriched UO2 Rods in Water with Fixed Neutron Poisons", Trans.Am. Nuc.Soc.. 27, 422 (1977). 2. B. M. Durst, S. R. Bierman, and E. D. Clayton, "Bench mark Calculations for Subcritical Clusters of 2.35 Wt-® 235u Enriched UOg Rods in Water with Fixed Neutron Poisons", Trans.Am.Nuc.Soc.. 27. 424 (1977). 3. W. B. Rogers, Jr., and F. E. Kinard, "Material Buckling and Critical Masses of Uranium Rods Containing 3 Wt-% 235u in H2 O", Nucl.Sci.Eng.. 20. 266 (1964). 4. D. W. Magnuson, Critical Experiments with Enriched Uranium Dioxide. Y-DR-120, Oak Ridge Y-12 Plant, Oak Ridge, Tennessee (1973). 5. R. E. Rothe and I. Oh, "Benchmark Critical Experiments on High-Enriched Uranyl Nitrate Solution Systems", Nuc1.Tech.. 41. 207 (1978). NUREG/CR-0674 RFP-2895 Page 106 6. I. Oh and R. E. Rothe, "A Calculational Study of Bench mark Critical Experiments on High-Enriched Uranyl Nitrate Solution Systems", Nuc1.Tech.. 4 ^ 226 (1978). 7. C. L. Schuske, D. Dickinson, D. C. Hunt, R. E. Rothe, and G. Tuck, Reference Critical Experiments Progress Report for Period July 1. 1975. through December 31. 1975. RFP- NUREG-2481, Rocky Flats Plant (January, 1976). 8. K, D. Lathrop, DTF-IV. A Fortran-IV Program for Solving the Multigroup Transport Equation with Anisotropic Scattering. LA-3373, Los Alamos Scientific Laboratory (1965). 9. G. E. Hansen and V. H. Roach, Six and Sixteen Group Cross Sections for Fast and Intermediate Critical Assemblies. LAMS-2543, Los Alamos Scientific Laboratory (1961). 10. L. M. Petrie and N. F. Cross, KENO-IV - An Improved Monte Carlo Criticality Program. ORNL-4938, Oak Ridge National Laboratory (1975). 11. D. C. Coonfield, G. Tuck, H. E. Clark, and B. B. Ernst, "Critical Mass Irregularity of Steel-Moderated Enriched Uranium Metal Assemblies with Composite Steel-Oil Reflectors", Nucl.Sci.Eng.. 39. 320 (1970). NUREG/CR-0674 RFP-2895 Page 107 12. Reactor Physics Constants. ANL-5800, Argonne National Laboratory (1963),