A new method called neutron signal transfer analysis has been developed for
quantitative determination of hydrogenous distributions from neutron radio
graphic measurements. The technique is based on a model which describes the
detector signal obtained in the measurement as a result of the action of t
hree different mechanisms expressed by signal transfer functions. The expli
cit forms of the signal transfer functions are determined by Monte Carlo co
mputer simulations and contain only the distribution as a variable. Therefo
re an unknown distribution can be determined from the detector signal by re
cursive iteration. This technique provides a simple and efficient tool for
analysis of this type while also taking into account complex effects due to
the energy dependency of neutron interaction and single and multiple scatt
ering. Therefore this method provides an efficient tool for precise quantit
ative analysis using neutron radiography, as for example quantitative deter
mination of moisture distributions in porous building materials. A comparis
on between neutron signal transfer analysis and conventional methods has be
en made by simulating an experiment with a sample containing a known moistu
re distribution by Monte Carlo calculation. By using the neutron signal tra
nsfer method the error is less than 3%. Whereas the calibration curve metho
d overestimates the moisture content by a factor in the range around 10% an
d the exponential attenuation method underestimates the moisture content by
a factor in the range of 2. (C) 1999 Published by Elsevier Science B.V. Al
l rights reserved.