Shielding calculation with MCNP5

In summary, the original input file for gamma shielding calculation did not work well. There were problems with three different shielding materials (concrete, iron, and lead) and the input files did not work.
  • #1
damyoro
12
0
Greetings to all

I had some problems using MCNP5 for gamma shielding calculation. The original input file is from a document by George E. Chabot, Jr., PhD, CHP Shielding of Gamma Radiation. The description of the problems are following:

1) Three different shielding materials, namely concrete, iron and lead are used to shield plane mono-energetic source particles of energy 0.511 MeV.
2) The said shielding materials are sub-divided into 10 cylindrical cells with equal radius (50 cm) but with different thickness from one material to another: i.e. 5 cm thick (concrete), 2 cm thick (iron) and 1 cm (lead).
3) The doses or kermas are tallied for different thicknesses of a given shielding material and also one tally in the absence of any shielding is used as reference.
4) For each case we wrote 11 input files and tried to run them at least to obtain doses for different geometrical configurations. Had it been successful we could have determined corresponding transmission factors.
5) Unfortunately, despite of several run trials, the input files did not work well as expected.

Comment on results

A: Concrete
• The input file for the concrete shield did not work satisfactorily. The messages generated are kindly reported but not limited to : -
o inp = cc4.txt outp = run runtpe = runtal
o starting mcnp execution
o comment. 22 surfaces were deleted for being the same as others.
o warning. surface 85 is not used for anything.
o imcn is done
o xact is done
The message ‘surface 85 is not used for anything’ is reported. This surface is where the source particles are generated
B: Iron
The input files for iron seemed to work, based on the message generated in the output file
‘run terminated when 10000000 particle histories were done’. Yet, all the 2 tallies are zero. Other messages are issued such as: ‘2 of the tallies did not pass all the statistic checks’, ’the importance may be poor’.
C: Lead

• The typical messages after the run are : -
o inp = pb3.txt outp = pb3o runtpe = run mctal = runtal
o starting mcnp execution
o comment. 22 surfaces were deleted for being the same as others.
o warning. surface 85 is not used for anything.
o imcn is done
o warning. material 1 has been set to a conductor.
o xact is done
o warning. importance function may be poor. see print table 120.
o warning. 2 of 2 tallies did not pass all 10 statistical checks.
o warning. 2 of 2 tallies were all zeros.
o mcrun is done

I attached 2 input and 1 outpout files for reference
Thank in advance for your help
 

Attachments

  • fe2.txt
    1.3 KB · Views: 991
  • pb3.txt
    1.3 KB · Views: 883
  • fe2o.txt
    37.9 KB · Views: 1,058
Last edited:
Physics news on Phys.org
  • #2
The warning about the surface not being used means that the surface
was not used by that name. I have not looked at your file, but the source
definition probably didn't use the surface but rather it specified the coordinates
in the sdef statement.

The warnings about various tallies being zero or not passing statistical checks
means that few (zero) particles made it to the tally location. This is because,
as usual in shielding calcs, the number of particles behind the shield is very
small compared to the number at the source. That's what shielding is for.

But the stats in a tally depend on the number of particles that contribute.
Very roughly speaking, the statistical error is proportional to one over the
square root of the number of particles.

There are two generic methods to deal with this: brute force or variance
reduction.

Brute force:If you have plenty of time and a big CPU then crank up the number
of particles you track. Just keep in mind that you may have to use a completely
unreasonable amount of particles to get a reasonable statistic.

Variance reduction: This is the generic name for a variety of methods where
you decrease the time spent in uninteresting parts of the calculation and
increase it for interesting parts.

For shielding calcs, the usual approach is to adjust weights in cells.
You want to divide the shielding into layers.
The layers closer to the source will naturally have more particles, and
those farther will have less. You want to increase the weight in cells that
are closer to your detector. This will mean that, when a particle comes
to the boundary between cells of different weights MCNP does some form
of the following. Suppose the weight changes from 1 to 2. Then the
particle goes from a single weight one particle, to two weight 0.5
particles. Each particle is then independently tracked with diffrent
random numbers. The result is that this is statistically unbiased.
Two weight 0.5 particles have the same result as one weight 1.
But now there are two of them in the "more interesting" area closer
to your detector.

So with several layers, each with twice the weight of the neighbor,
you get layers like so.

Source | 1 | 2 | 4 | 8 | 16 | 32 | 64 | etc. | detector

At each boundary, the weight of the particle goes down but the
number goes up. This acts in the other direction to particles
getting absorbed by the shielding. So by the time you get to
your detector you have a large number of very low weight
particles. Each gives a fractional count at the detector, thus
giving you good stats.

The usual advice from MCNP developers in this situation is to
adjust the weights and cell widths to keep the number of
particles the same in each cell. Also, try not to change the
weight much more than about a factor of 2 in adjacent cells.

Read up in the MCNP manual about weights, Russian roulette,
weight windows, and other forms of variance reduction.

Hope this helps.
Dan
 
  • #3
Thank you very much for your helpful reply. I will certainly follow your advice and come back to you soon.
 
  • #4
Input file MCNP5

Greetings to all

After some corrections, the MCNP input files for shielding calculations for iron and lead run without problem, but the input file for the concrete shielding calculations did not work, although the files are similar , I have only replaced in the material definition the compositions of the material concerned.

Please could you check the input file attached (cc4.txt) for the concrete to help me find out what the problem is. I have also attached an input file which run properly for your reference (fe3.txt) and thr result (fe3o.txt)

Thank you in advance
 

Attachments

  • cc4.txt
    1.5 KB · Views: 812
  • fe3.txt
    1.3 KB · Views: 747
  • fe3o.txt
    46.6 KB · Views: 777
  • #5
damyoro said:
After some corrections, the MCNP input files for shielding calculations for iron and lead run without problem, but the input file for the concrete shielding calculations did not work, although the files are similar , I have only replaced in the material definition the compositions of the material concerned.

In what way "did not work?"

I changed your files name to cc4 from cc4.txt. MCNP has an 8 character limit on files names. Then I ran it using the following command.

mcnp5 i=cc4

It ran for about 8 minutes, and finished. You need to clean up the weighing values you use. The code gives you some guidance on that. And you can read some more in the manual about how to choose good weight values.

How did it "not work" for you?
Dan
 
  • #6
I use both MCNP5 visual Editor and command line. When I run the file the nps count doesn't even start.
1. This is the message generated when I run with Visual

inp = cc4.txt outp = CC4O runtpe = run mctal = runtal
starting mcnp execution
comment. 22 surfaces were deleted for being the same as others.
warning. surface 85 is not used for anything.
imcn is done
xact is done

And there is no results (cf. CC4O.txt) I added .txt in order to attach the output CC4O.

2. When I run with the command line here the message is below. Unfortunately I could not even see the output I don't there it may be located.

C:\Program Files (x86)\LANL\MCNP5>mcnp5 n=cc4
mcnp ver=5 , ld=11042003 05/08/13 18:08:15
Thread Name & Version = MCNP5_RSICC, 1.20
Copyright LANL/UC/DOE - see output file
_
._ _ _ ._ ._ |_
| | | (_ | | |_) _)
|
comment. 22 surfaces were deleted for being the same as others.
warning. surface 85 is not used for anything.
comment. using random number generator 1, initial seed = 19073486328125
imcn is done
dump 1 on file cc4r nps = 0 coll = 0
ctm = 0.00 nrn = 0
xact is done
cp0 = 0.01
run terminated when 10000000 particle histories were done.
warning. importance function may be poor. see print table 120.
warning. tally 6 tfc bin did not pass 1 of 10 statistical checks.
warning. 1 of 2 tallies did not pass all 10 statistical checks.
dump 2 on file cc4r nps = 10000000 coll = 81751568
ctm = 22.97 nrn = 1292251531
mcrun is done

C:\Program Files (x86)\LANL\MCNP5>


Could explain how 'to clean up the weighing values ...'.

As a beginner I need your help to progress.
Thank for all your effort.
 

Attachments

  • CC4P.txt
    17.8 KB · Views: 723
  • #7
DEvens ,may I ask about mcnp code?
 
  • #8
If I got it right, here is the code
MNCP Input for Concrete Attenuation calculation
c
c Cell cards:
1 1 -2.3 -1
5 1 -2.3 -5
10 1 -2.3 -10
15 1 -2.3 -15
20 2 -0.001205 -20
25 2 -0.001205 -25
30 2 -0.001205 -30
35 2 -0.001205 -35
40 2 -0.001205 -40
45 2 -0.001205 -45
c Tally...
60 2 -0.001205 -60 $ Tally
70 2 -0.001205 -70 60 $ Air surroundings
c Environnment...
80 0 -80 $ Sources region
90 0 1 5 10 15 20 25 30 35 40 45 70 80 -90 $ Surroundings
99 0 90 $ Void

C Surface cards:
c concrete
1 RCC 0 0 0 0 0 5 50
5 RCC 0 0 5 0 0 5 50
10 RCC 0 0 10 0 0 5 50
15 RCC 0 0 15 0 0 5 50
20 RCC 0 0 20 0 0 5 50
25 RCC 0 0 25 0 0 5 50
30 RCC 0 0 30 0 0 5 50
35 RCC 0 0 35 0 0 5 50
40 RCC 0 0 40 0 0 5 50
45 RCC 0 0 45 0 0 5 50
c Tally:
60 RCC 0 0 50 0 0 0.1 1 $ Tally
70 RCC 0 0 50 0 0 5 59 $ Air surrounding Tally
c Environment:
80 RCC 0 0 -10 0 0 10 9 $ Source vacuum
85 PZ -5 $ Source
90 RCC 0 0 -10 0 0 80 60 $ Surrounds

mode p
c Materials:
M1 1000 -0.0221 $ oncrete
6000 -0.002484
8000 -0.574930
11000 -0.015208
12000 -0.001267
13000 -0.019953
14000 -0.304627
19000 -0.010045
20000 -0.042951
26000 -0.006435
M2 7000 -0.755267 $ AIR
8000 -0.231781
18000 -0.012827
6000 -0.000124
c
c Impotrances:
IMP:P 1 1 2 2 4 $ 1, 20
8 8 16 32 64 $ 25,70
64 64 1 1 0 $ 80, 99
c
c Sources:
SDEF POS = 0 0 -5 SUR = 85 RAD = d1 VEC = 0 0 1 DIR = 1 Par = 2 ERG = 0.511
SI1 50
c
c Tally cards:
F4:P 60
F6:P 60
c
c Outpouts:
PRINT 110 120 130 160 162
c
c Cutoffs:
phys:p 1 0 0
NPS 10000000
 
  • #9
Sorry, I was out of town the last few days.

damyoro said:
[snips]
When I run the file the nps count doesn't even start.
[snips]
run terminated when 10000000 particle histories were done.

Nope. You got the number of particle histories you asked for, and it stopped. This is why I asked, in what way did it not work? It terminated normally.

damyoro said:
Could explain how 'to clean up the weighing values ...'.

The program gives hints. What you are looking for is the number of particles (histories, tracks, whichever you find more useful) to be roughly the same in each layer. So you want the weights to change by a factor of about 2 between layers (not a lot more). You can make layers thinner or thicker, make more or fewer of them, change the weights, etc., to try to get this to happen. You may need to iterate a bit. Make a guess, run a case, look at the table that gives cell information including histories, readjust.

Like I said, read more in the user manual about weights and selecting good values. Also, there are other methods of reducing variance, also good reading in the manual.

It really isn't practical to teach the program in the forum. If you have specific questions I will try to help, but I can't spend more than a few minutes at a time.
Dan
 
  • #10
crnjak said:
DEvens ,may I ask about mcnp code?

Feel free to post questions here. I will try to answer if I can.
 
  • #11
Thank you for your help and your availability. I really appreciated. I will focus on reading the manual.
 
  • #12
Hi,I ve run the mcnp code and get : fatal error cell 2 have no fill.

1 1 -1 -1 fill=1 imp:n=1
2 1 -1 -2 lat=1 u=1 imp:n=1
3 0 1 imp:n=0

1 box -5 -5 -10 10 0 0 0 10 0 0 0 20
c 1 REC 0 0 0 0 0 70 20 0 0 0 10 0
2 box -5 -5 -10 1 0 0 0 1 0 0 0 1
c box -1 -1 -1 1 0 0 0 1 0 0 0 1

m1 94239.66c 1.0
c tally
f4:n 1 (2<2[0:9 0:9 0:19])
sdef pos .1 .1 .1
print

any tips?

Cheers, Zoki
 
  • #13
Greetings to all
I need some some help to proceed the calculations of gamma shielding using MCNP.
The details of the problem are as follows:
- A square irradiation room/cell of maximum of 2 m of side
- A cobalt source with an activity of 5,000 Ci placed on a table inside the room,
- A glass window to enable to see inside.
- material to be used: lead, concrete, iron, lead glass (for the window)
Essentially, I need guidance on:
- geometry configurations,
- source configurations (if Co source is considered as a istropic point source) to be able to focus on a region of interest ( one side of the room) in order to limit the running time,
- types of tallies that may be used for this problem to relate them to the limits dose of 20 mSV/year or the dose rate of 10 μSV/hr.
- How to take account of the buildup.

Thank you in advance
 
Last edited:
  • #14
Help for ?CNP input

Greetings
I am using MCNP to estimate the shielding of Cobalt source placed at the center of cube of 2 meter of side. I wrote an input file using macroby. I used successive cubes and filled them with either air or lead (or other shielding material: tungsten, steal, etc) to vary the thickness. I use F6 tally to record the dose outside the shield in a small cylindrical cell. The limit dose outside is not to exceed the annual dose limit that is 20mSV/year or 10 microSV/hr if the consider 2000 hr/year.When I run my input there are some fatal errors. I will appreciate if you can have a look at it and help me to find the problem. I attached the input and the output files.
Thank you very much in advance. Your suggestions are most welcome, especially for the choice of the tally and the geometry
 

Attachments

  • HC1.txt
    1.2 KB · Views: 734
  • HC1p.txt
    84.9 KB · Views: 764
  • #15
Comparison of Dose rates throgh MCNP and MICROSHIELD

I calculated dose rate at five different points of a radioactive waste tank, containing four radioisotopes having radioactivity as under
Radioactivity of Cs-137= 4.184E+2
Radioactivity of Cs-134=4.498E+2
Radioactivity of Pr-144=9.826E+3
Radioactivity of Nb-95=9.826E+3
these radioistopes are homogeniously mix in 100 liters water
Dimension of the tank i use is H=108 cm, r=54cm, stainless steal tank (density 7.92g/cc)
%difference b/w the MCNP and MICROSHIELD calculated dose rate is from 25% to 95% at five different points of tank surface, is this difference acceptable, if not then please tell me the possible reason of error
 
  • #16
damyoro said:
Greetings to all

I had some problems using MCNP5 for gamma shielding calculation. The original input file is from a document by George E. Chabot, Jr., PhD, CHP Shielding of Gamma Radiation. The description of the problems are following:

1) Three different shielding materials, namely concrete, iron and lead are used to shield plane mono-energetic source particles of energy 0.511 MeV.
2) The said shielding materials are sub-divided into 10 cylindrical cells with equal radius (50 cm) but with different thickness from one material to another: i.e. 5 cm thick (concrete), 2 cm thick (iron) and 1 cm (lead).
3) The doses or kermas are tallied for different thicknesses of a given shielding material and also one tally in the absence of any shielding is used as reference.
4) For each case we wrote 11 input files and tried to run them at least to obtain doses for different geometrical configurations. Had it been successful we could have determined corresponding transmission factors.
5) Unfortunately, despite of several run trials, the input files did not work well as expected.

Comment on results

A: Concrete
• The input file for the concrete shield did not work satisfactorily. The messages generated are kindly reported but not limited to : -
o inp = cc4.txt outp = run runtpe = runtal
o starting mcnp execution
o comment. 22 surfaces were deleted for being the same as others.
o warning. surface 85 is not used for anything.
o imcn is done
o xact is done
The message ‘surface 85 is not used for anything’ is reported. This surface is where the source particles are generated
B: Iron
The input files for iron seemed to work, based on the message generated in the output file
‘run terminated when 10000000 particle histories were done’. Yet, all the 2 tallies are zero. Other messages are issued such as: ‘2 of the tallies did not pass all the statistic checks’, ’the importance may be poor’.
C: Lead

• The typical messages after the run are : -
o inp = pb3.txt outp = pb3o runtpe = run mctal = runtal
o starting mcnp execution
o comment. 22 surfaces were deleted for being the same as others.
o warning. surface 85 is not used for anything.
o imcn is done
o warning. material 1 has been set to a conductor.
o xact is done
o warning. importance function may be poor. see print table 120.
o warning. 2 of 2 tallies did not pass all 10 statistical checks.
o warning. 2 of 2 tallies were all zeros.
o mcrun is done

I attached 2 input and 1 outpout files for reference
Thank in advance for your help
damyoro said:
Greetings to all

I had some problems using MCNP5 for gamma shielding calculation. The original input file is from a document by George E. Chabot, Jr., PhD, CHP Shielding of Gamma Radiation. The description of the problems are following:

1) Three different shielding materials, namely concrete, iron and lead are used to shield plane mono-energetic source particles of energy 0.511 MeV.
2) The said shielding materials are sub-divided into 10 cylindrical cells with equal radius (50 cm) but with different thickness from one material to another: i.e. 5 cm thick (concrete), 2 cm thick (iron) and 1 cm (lead).
3) The doses or kermas are tallied for different thicknesses of a given shielding material and also one tally in the absence of any shielding is used as reference.
4) For each case we wrote 11 input files and tried to run them at least to obtain doses for different geometrical configurations. Had it been successful we could have determined corresponding transmission factors.
5) Unfortunately, despite of several run trials, the input files did not work well as expected.

Comment on results

A: Concrete
• The input file for the concrete shield did not work satisfactorily. The messages generated are kindly reported but not limited to : -
o inp = cc4.txt outp = run runtpe = runtal
o starting mcnp execution
o comment. 22 surfaces were deleted for being the same as others.
o warning. surface 85 is not used for anything.
o imcn is done
o xact is done
The message ‘surface 85 is not used for anything’ is reported. This surface is where the source particles are generated
B: Iron
The input files for iron seemed to work, based on the message generated in the output file
‘run terminated when 10000000 particle histories were done’. Yet, all the 2 tallies are zero. Other messages are issued such as: ‘2 of the tallies did not pass all the statistic checks’, ’the importance may be poor’.
C: Lead

• The typical messages after the run are : -
o inp = pb3.txt outp = pb3o runtpe = run mctal = runtal
o starting mcnp execution
o comment. 22 surfaces were deleted for being the same as others.
o warning. surface 85 is not used for anything.
o imcn is done
o warning. material 1 has been set to a conductor.
o xact is done
o warning. importance function may be poor. see print table 120.
o warning. 2 of 2 tallies did not pass all 10 statistical checks.
o warning. 2 of 2 tallies were all zeros.
o mcrun is done

I attached 2 input and 1 outpout files for reference
Thank in advance for your help

I have read your files fe2.txt and pb3.txt, I think the discription of cell 70 is not appropriate, you can try "70 2 -0.001205 -70 60 "
Hope it's helpful for you
 
  • #17
damyoro said:
If I got it right, here is the code
MNCP Input for Concrete Attenuation calculation
c
c Cell cards:
1 1 -2.3 -1
5 1 -2.3 -5
10 1 -2.3 -10
15 1 -2.3 -15
20 2 -0.001205 -20
25 2 -0.001205 -25
30 2 -0.001205 -30
35 2 -0.001205 -35
40 2 -0.001205 -40
45 2 -0.001205 -45
c Tally...
60 2 -0.001205 -60 $ Tally
70 2 -0.001205 -70 60 $ Air surroundings
c Environnment...
80 0 -80 $ Sources region
90 0 1 5 10 15 20 25 30 35 40 45 70 80 -90 $ Surroundings
99 0 90 $ Void

C Surface cards:
c concrete
1 RCC 0 0 0 0 0 5 50
5 RCC 0 0 5 0 0 5 50
10 RCC 0 0 10 0 0 5 50
15 RCC 0 0 15 0 0 5 50
20 RCC 0 0 20 0 0 5 50
25 RCC 0 0 25 0 0 5 50
30 RCC 0 0 30 0 0 5 50
35 RCC 0 0 35 0 0 5 50
40 RCC 0 0 40 0 0 5 50
45 RCC 0 0 45 0 0 5 50
c Tally:
60 RCC 0 0 50 0 0 0.1 1 $ Tally
70 RCC 0 0 50 0 0 5 59 $ Air surrounding Tally
c Environment:
80 RCC 0 0 -10 0 0 10 9 $ Source vacuum
85 PZ -5 $ Source
90 RCC 0 0 -10 0 0 80 60 $ Surrounds

mode p
c Materials:
M1 1000 -0.0221 $ oncrete
6000 -0.002484
8000 -0.574930
11000 -0.015208
12000 -0.001267
13000 -0.019953
14000 -0.304627
19000 -0.010045
20000 -0.042951
26000 -0.006435
M2 7000 -0.755267 $ AIR
8000 -0.231781
18000 -0.012827
6000 -0.000124
c
c Impotrances:
IMP:P 1 1 2 2 4 $ 1, 20
8 8 16 32 64 $ 25,70
64 64 1 1 0 $ 80, 99
c
c Sources:
SDEF POS = 0 0 -5 SUR = 85 RAD = d1 VEC = 0 0 1 DIR = 1 Par = 2 ERG = 0.511
SI1 50
c
c Tally cards:
F4:P 60
F6:P 60
c
c Outpouts:
PRINT 110 120 130 160 162
c
c Cutoffs:
phys:p 1 0 0
NPS 10000000
Can you explain IMP: and PHYS:P? I don't really understand it
 
  • #18
I don't really understand PHYS:. Can you explain? Thank you
 
  • #19
If I have a point isotropic source into a Cone of Directions with angle 30 degree. How to specify SIn, SPn and special SBn card. And if possible, please explain more explicit SBn card.Thanks a lot.
 
  • #20
how to do gamma ray attenuation using attenuation law with mcnp modelling
detector- NaI
source - cs137
absorber-lead
please help me...
 
  • #21
This thread is more than three years old. Please open new threads for new questions.
 

1. How does MCNP5 calculate shielding?

MCNP5 uses the Monte Carlo method to simulate the transport of particles through materials. It tracks individual particles as they interact with the material and calculates the probability of various outcomes, such as scattering or absorption. This information is used to determine the shielding effectiveness of the material.

2. What types of radiation can MCNP5 calculate for?

MCNP5 can calculate for a wide range of radiation types, including neutrons, photons, and electrons. It can also simulate other types of particles, such as alpha and beta particles, as well as heavy ions.

3. How accurate are the results from MCNP5?

The accuracy of the results from MCNP5 depends on various factors, such as the complexity of the geometry, the number of particles simulated, and the accuracy of the input parameters. In general, MCNP5 is considered a highly accurate tool for shielding calculations, but it is always important to carefully review and validate the results.

4. Is MCNP5 difficult to use?

MCNP5 is a complex software that requires some level of expertise to use effectively. It has a steep learning curve and requires knowledge of nuclear physics, radiation transport, and computer programming. However, with proper training and practice, it can be a powerful tool for shielding calculations.

5. Can MCNP5 simulate different materials?

Yes, MCNP5 can simulate the transport of particles through various materials, including metals, concrete, water, and biological tissues. Users can specify the composition and density of the material in the input file, and MCNP5 will calculate the shielding effectiveness accordingly.

Similar threads

  • Nuclear Engineering
Replies
2
Views
2K
Replies
6
Views
1K
  • Nuclear Engineering
Replies
5
Views
1K
  • Nuclear Engineering
Replies
2
Views
2K
Replies
6
Views
1K
Replies
2
Views
3K
  • Nuclear Engineering
Replies
1
Views
2K
  • General Engineering
Replies
5
Views
5K
  • Nuclear Engineering
Replies
9
Views
2K
  • Nuclear Engineering
Replies
2
Views
2K
Back
Top