Today’s nuclear technology depends on critical 
facilities for information in four important areas: 
1) reactor design, 2) weapons-systems design, 3) 
criticality limits on processing operations and 4) 
basic physics data. 
By far the largest number of the critical facili- 
ties in the U.S. are devoted to reactor design. In 
these the experimenter seeks to achieve a delicate 
balance between neutron production and loss over 
the full range of reactor operating conditions. 
For processing-operation critical experiments, 
the emphasis is on setting safety limits to ensure 
that batches of fissionable materials will never sup- 
port a self-sustaining reaction in operations such 
as storage and shipping, isotope separation, 
chemical processing of irradiated fuel and fertile 
The Role of 
Critical Facilities 
material, and fabrication of fuel elements. 
Through critical experiments, safety factors are 
established for systems of varying geometry, com- 
position, fuel concentration and degree of coupling. 
In contrast, weapons criticals are concerned 
with the understanding of supercritical assemblies 
with very short neutron lifetimes. Hence such 
experiments concentrate on determining the static 
and kinetic behavior of unmoderated nuclear 
systems. 
Critical facilities that perform any of these first 
three functions can also be used to obtain basic 
physics information. In addition much useful 
data has come from “‘pure-geometry”’ critical 
assemblies constructed especially for physics 
experiments. 
1. Critical Facilities for Reactor Design 
By W. C. REDMAN, Argonne National Laboratory, Lemont, Illinois 
In many ways the critical facility bears 
the same relation to reactor design as 
the wind tunnel does to aircraft design. 
Just as the aeronautical engineer at a 
certain point in the design process looks 
to the wind tunnel for crucial informa- 
tion so the nuclear engineer must check 
out his proposed reactor design in a 
critical facility before he can confi- 
dently begin construction of the full- 
scale reactor. 
In principle, a complete knowledge 
of all of the pertinent neutron cross sec- 
tions and yields as a function of energy 
would enable the reactor designer to 
calculate the neutron distribution in 
space and energy and thereby specify 
the reactor configuration in all detail. 
In practice, the large gaps in our knowl- 
edge of the nuclear interaction parame- 
ters and the limited accuracy of many 
of the existing data make the use of 
refined theories and improved calcula- 
tional techniques of questionable value 
in reactor design. 
108 
For example, theoretical calculations 
predicted a reference design loading for 
EBWR of 42 fuel assemblies with a 
probable error of +3% ink. As such 
calculations go this one would appear 
reasonably accurate; however, in terms 
of fuel loading +3% corresponds to 
almost a factor of two in fuel load— 
some 40 extra fuel assemblies. For 
EBWR the observed loading fell just 
within the probable error, requiring a 
total of 81 assemblies. A much more 
accurate estimate for the loading could 
have been given by a critical experi- 
ment—and in fact this is what the first 
EBWR runs were used to do (NU, 
July ’57, 60). 
A critical experiment can be com- 
pared to a computer in which a Monte 
Carlo type of calculation is performed. 
The appropriate neutron cross sections 
are inherent in the system being stud- 
ied, and the life histories of a tremen- 
dous quantity of neutrons are “‘calcu- 
lated” simultaneously. Thus critical 
experiments afford a means of obtain- 
ing required answers without a knowl- 
edge of the neutron cross sections. 
Analysis of the results of critical experi- 
ments enables the reactor theorist to 
arrive at better parameter values for 
use in reactor-design calculations. 
Experimental Objectives 
Critical experiments serve primarily 
to yield information on the nuclear per- 
formance of the heart of the reactor 
system—namely, the reactor core and 
reflector. 
Working within the limits set by 
heat-transfer, metallurgical and nu- 
clear considerations, the critical experi- 
menter seeks an optimum design. 
Variables typically at his disposal in- 
clude quantity and enrichment of fuel, 
amount of moderator, dimensions of 
the reactor, lattice spacing of fuel chan- 
nels, arrangement of fuel in the chan- 
nel, area of coolant channel around fuel, 
physical properties and pressure of 
