The neutron activation system at JET is a pneumatic transfer system capable
of positioning activation samples close to the plasma. Its primary purpose
is to provide a calibration for the time-dependent neutron yield monitors
(fission chambers and solid state detectors). Various activation reactions
with different high energy thresholds were used including Fe-56(n,p) Mn-56,
Al-27(n,alpha) Na-24, Nb-93(n,2n) Nb-92m, and Si-28(n,p) Al-28 reactions.
The silicon reaction, with its short half life (2.25 min), provides a promp
t determination of the 14 MeV DT yield. The neutron induced gamma-ray activ
ity of the Si samples was measured using three sodium iodide scintillators,
while two high purity germanium detectors were used for other foils. It wa
s necessary to use a range of sample masses and different counting geometri
es in order to cover the wide range of neutron yields (10(15)-10(19) neutro
ns) while avoiding excessive count rates in the detectors. The absolute ful
l energy peak efficiency calibration of the detectors was measured taking i
nto account the source-detector geometry, the self-attenuation of the sampl
es and cross-talk effects. An error analysis of the neutron yield measureme
nt was performed including uncertainties in efficiency calibration, neutron
transport calculations, cross sections, and counting statistics. Cross cal
ibrations between the different irradiation ends were carried out in DD and
DT (with 1% and 10% tritium content) discharges. The effect of the plasma
vertical displacement was also experimentally studied. An agreement within
10% was found between the 14 MeV neutron yields measured from Si, Fe, Al, N
b samples in DT discharges. (C) 1999 American Institute of Physics. [S0034-
6748(99)59101-6].