The reflector elastically scatters or reflects neutrons when they escape from the fissioning
core, in order that they are not lost and can go on to cause further fissions. Thus the
material choice for a strong reflector is determined by its mean free path (mfp) for
scattering; good reflectors have a short mfp for neutrons. To reduce the neutron travel time
it is also important for the neutron reflector to be near the fissile core, preferably in direct
contact with it.

Because of the extremely short timescales on which the device is operating, the neutron
population in the fissioning core approximately doubles during the time when a neutron
travels one mfp - thus the importance of an average reflected neutron to the fission chain
reaction is reduced by this "time absorption" effect. Because of time absorption and the
effects of geometry, the effectiveness of a reflector thus drops very rapidly when the mfp

A usual mistake is to assume that a thicker reflector that has the same mfp is as good or
better than a thin reflector. However most of the benefit of a reduction of the critical mass
of a core is obtained in the first couple of mfps. When the thickness of a reflector increases
to be more than a few mfps, the law of diminishing returns applies, and no large additional
neutron reflection benefit is obtained.

For a constant mfp, an increase of reflector thickness also has a point of diminishing returns.
Most of the benefit of critical mass reduction occurs where there is reflector thickness of
one 1 mfp. With 2 mfp of reflector, usually the critical mass reduces to a point in a few
percent of its value for an infinitely thick reflector. Time absorption also forces the benefits
of a reflector to reduce rapidly for thicknesses more than approximately one mfp. A very
thick reflector gives only few benefits in comparison with a relatively thin reflector.

Suitable materials which have a sufficiently low mfp:

  • Graphite
  • Beryllium
  • Lead
  • Steel
  • Tungsten

This leads to the possibility of a singular component "tamper/reflector", which can give
tamping and neutron reflection - again, they both work the best when they are found as
near as possible to the fission core.

The "Demon core" accidents that occurred at Los Alamos in the middle of the 1940s are
examples in history of the effectiveness of a neutron reflector.

On August 21, 1945 a 6.2-kilogram subcritical mass of plutonium produced a burst of
neutron radiation which killed Harry Daghlian, a physicist at Los Alamos. Daghlian was
carrying out neutron reflection experiments on the core. The core was put in a stack of
neutron-reflective tungsten carbide bricks, and the addition of each brick moved the
assembly closer to criticality.

When he tried to place another brick around the assembly Daghlian accidentally dropped it
onto the core and made the core go critical. Despite quick action as a result of moving the
brick from the assembly, Daghlian received a fatal dose of radiation. He died after 25 days
following acute radiation poisoning.

On May 21, 1946, a physicist (Louis Slotin) and seven other scientists were in a laboratory in
Los Alamos. They were conducting an experiment to verify the exact point when a subcritical
mass of fissile material can become critical by means of the positioning of neutron reflectors.
The operator needed to place two half-spheres of beryllium (a neutron reflector) around the
core which the scientists wanted to test, and manually lower the top reflector over the core
through use of a thumb hole at the top. As the operator manually moved the reflectors
closer and farther away one from the other, scintillation counters were measuring the
relative activity from the core. To permit them to close fully would mean the immediate
formation of a critical mass and a lethal power excursion - the only thing that was preventing
this was a regular flathead screwdriver which the scientist operated in his other hand. The
test was called "tickling the dragon’s tail" because of its extreme risk, and notoriously did not
permit even the smallest mistake. A lot of scientists refused to carry out the test, but Slotin
(a man of bravado) became the local expert and carried out the test almost 10 times. He did
these often in his bluejeans and cowboy boots before a room of observers. They say that
Enrico Fermi told Slotin and the others they will be "dead within a year" if they continue to
carry out the test.

When he was lowering the top reflector, Slotin’s screwdriver slipped by a small fraction -
this made the top reflector fall into its place around the core. Immediately there was a flash
of blue light and a wave of heat on Slotin’s skin - the core became supercritical, and it
released an enormous burst of neutron radiation. Quickly he pushed the two halves apart,
stopped the chain reaction and likely saved the lives of the other men in the laboratory.
Slotin’s body over the apparatus also shielded the others from a lot of the neutron radiation.
He received an enormous lethal dose in a second and died after nine days of acute radiation