Dear Jyoti,
the main issue with your analysis that you have the order of numbers wrong. The ASCII files of USRBIN results are read line by line continuously, and the order of number are the following:
(r_1, \phi_1), (r_2,\phi_1), ... (r_{17},\phi_1), (r_1\phi_2), ...
Also, the big errors in the “boundary” and “detector” region need to be addressed. They are caused by the unbiased photons leaving your geometry on the sides, when they are back scattered in air. They have significantly higher weight compared to the photons going through the shielding, so they contribute to the dose much more in value. But this happens very rarely, thus the statistics will be bad.
These two contribution is best to simulate separately:
- For the contribution through the shielding it is the easiest to set the “Void” and “detector” to VACUUM.
- For the contribution due to the back scattered photons, the biasing should be turned off, and to speed up the simulations, the shielding material could be changed to BLCKHOLE instead of LEAD.
Cheers,
David