The NRCC accelerator is a Vickers research accelerator with no head assembly.
The geometry of the experimental set up (fig.1)
includes
a titanium exit vacuum window (0.127 mm thick),
a tungsten scattering foil (0.092 mm thick),
a square collimator (steel, 20.3 cm thick)
and a monitoring ion chamber (very thin mylar).
The water phantom container is made of PMMA (9 mm thick).
The beams irradiate the water phantom horizontally.
The field size is 8
8 cm
on the phantom surface and SSD = 96 cm.
The energy of the electron beam exiting the vacuum window
is known independently with an accuracy of 
1%[9].
The two beams (10.0 and 20.0 MeV) have an identical experimental
set up except for the incident electron energy at the exit vacuum window.
The dose measurements have been done with
a small diode detector, a parallel plate ion chamber and a cylindrical
farmer chamber. After correction of the ion chamber results for stopping
power ratios, all three measurement techniques gave the same results
within 0.5%.
The measured data are from Carl Ross of National Research Council of
Canada.
The results for these two beams are shown in
figure 9 and 10 respectively.
Because of the geometry of the experimental set up, the electron's
energy spread is very small and there are few contaminant photons in the beams.
The two beams have narrow angular spread although the electrons from the jaws
lead to a small tail right out to 80
.
Note that in order to match the experimental values of R
in
the central-axis depth-dose curves the incident electron energies
at the exit vacuum window used in the
simulation were increased by 1.5% compared to those
determined independently (viz. 10 and 20 MeV).
This slight discrepancy may be caused by the measured energy or
by the stopping power used in the calculation
(to be resolved in future investigations).
The remarkable agreement between calculated and measured dose distributions
shown in fig.10 marked a milestone in benchmarking the
simulation code BEAM[10].