In order to compute the transmission coefficient as measured experimentally, the primary particle is considered as transmitted if there is at least one particle of the same kind escaping the absorber forward. See the pictures for e- or e+ primary. The electrons/positrons are visualised as red/blue tracks and there are 10 events superimposed in the pictures.
The transmission (or backscattering) of low energy particles is a very sensitive test of the low energy simulation code. In order to get realistic results one has to use small production thresholds. The results below has been obtained with 1 micron production threshold (both for e+/e- and photons).
960 keV electron/positron transmission through lead foils : Geant4/Geant3/data comparison
There are 10 000 simulated events in every bin. the experimental data are from H.H. Seliger,Phys.Rev.100(1955)1029, the relative error of the data is about 10 %.
One can see on the plots, that Geant4 reproduces the experimental data better than Geant3. It should be noticed that Geant4 distinguishes e+ from e-, while Geant3 does not.
It is worth to mention that the low energy Monte Carlo code ETRAN and Geant4 gives similar results: Geant4/ETRAN/data comparison. ETRAN results have been taken from M.J.Berger Appl.Radiat.Isot.42 (1991) 905.
The code of this example is in geant4/examples/extended/electromagnetic/TestEm5