Biasing - big error in last bin

Dear Experts,

In similar lines of my previous question - on using Biasing and scoring (“How to improvise”) - I have scored from within the shield to the outer surface of the shield in the present input. Scored quantity : DOSE-EQ; user routine for 12 cylindrical sources; Energy from BEAM card.

I have few questions on the results obtained:
a) In DEbel1X usrbin - the error suddenly increased from 8.7 to 47.58 - I couldn’t understand the reason. The same increase in error is observed when scored along Y-direction. The decrease in DOSE-EQ seems logical (shield IN to shield OUT) but then sudden increase in error (outermost bin) confuses.
b) Similarly in the usrbins - DEtopgap and DEscrew - sudden increase in the error for the last bin ( as the outer surface is reached) - what could be the reason ?
c) the .out file - energy deposition - shows (***) for EM showers and 0.000 for particles escaping the system - Is it because of VOID is filled with VACUUM, it is not quantifying this ? Usually it shows definite %age of EM showers and particles escaping the system.
d) Biasing is done only in a part of the Shield (top side) - is this use correct ?

Please find attached the required files.

Can you please explain where is the mistake to get such increase in errors ?

Thanking you in advance. With regards,
biascylin.flair (7.5 KB)
biascylin.lis (4.3 KB)
biascylin.out (160.1 KB)
cylin12.f (11.3 KB)

Dear Raksha,

  1. It is not surprising to me that the last bin has such an error, if you observe the difference between the previous bins, you see that those are also increasing with the same ratio.

  2. Same applies to the other scorings as well.

  3. The 0.0% for escaping particles is just coming from the rounding, you can see with the absolute value, that barely any particles leaving the shielding.

  4. Technically it is correct, but you need to think about if this make sense in regard what you want to do.


Dear David,
Thank you making me realize the ratio, I didn’t note that.
But I am not getting how to reduce the error In the outermost bin, as I am interested to score on the outer surface of the shield. If I increase the bin volume, I will be deviating from the near-point estimation, as the results are required on any point on the surface of the shield.
Can you please suggest some effective method to score efficiently on the surface of the shield.with statistical errors below 10 % ?
I am trying R-phi-Z binning now in the belly region.
Thank you again.

Dear Raksha,

you could split the lead regions into multiple layers and use biasing radially.


Good morning David,
I would surely attempt that. Thank you.