MCNP code for Neutron Spectroscopy

[Hamidul]

Hello everyone , in my mcnp coding for finding neutron spectroscopy I used F2 tally across a surface. Is it correct or I should use f4 tally? Morever I need to transform the flux data into neutron fluence. How can I do that. Here I uploaded my code. Though my data from codes is way more different from my experimental data.

 

[PSRB191921]

I would have used a F4:n 6

 

[Alex A]

It might be better if this was all kept in one thread. I am not seeing the excel data or the runs attached here or in the other thread. Are you saying you used a neutron spectrometer to get your results?

 

[Hamidul]

It might be better if this was all kept in one thread. I am not seeing the excel data or the runs attached here or in the other thread. Are you saying you used a neutron spectrometer to get your results?

Hello Alex, I uploaded both the file, but due to some issues of the network it did not worked.
Yes, I used NNS (Nested Neutron Spectrometer) to get my results. My professor said that my experimental results are good and I have to simulate the same things to get a spectra.

Moreover, should I stop this thread and go back to the previous thread. ? If this is convenient, then I will upload my simulated output file and the experimental spectra in the previous thread.

 

[Alex A]

I don't think it matters which thread. Yes, please post your results.

 

[Hamidul]

Hello ALEX , here are my both results,

 

[Alex A]

The main problem is that you have a small detector one meter away from the source. This means most neutrons in the problem do not hit the detector and leave the problem with the CPU time wasted. Your spectrum is basic because only one neutron hit it.

I would consider calculating a slightly different problem that might be expected to have a similar result that can be calculated with less CPU time. Such as making the detector a ring, or even a spherical shell around the problem.

You could just run the problem for a long time. With 10^5 neutrons the spectrum has one or two neutrons. 10^7 a spectrum starts to appear. 10^9 would probably take 2 core days with MCNP5, and maybe twice that with MCNPX.

Or you could do both and get the spectrum 'right' with a bigger detector and then do a long run for the final result.

I note the X manual section H gives different constants for the watt spectrum of Cf-242, I do not know if the difference matters. I would also consider setting every bin of your energy range to match your real results.

 

[Hamidul]

The main problem is that you have a small detector one meter away from the source. This means most neutrons in the problem do not hit the detector and leave the problem with the CPU time wasted. Your spectrum is basic because only one neutron hit it.

I would consider calculating a slightly different problem that might be expected to have a similar result that can be calculated with less CPU time. Such as making the detector a ring, or even a spherical shell around the problem.

You could just run the problem for a long time. With 10^5 neutrons the spectrum has one or two neutrons. 10^7 a spectrum starts to appear. 10^9 would probably take 2 core days with MCNP5, and maybe twice that with MCNPX.

Or you could do both and get the spectrum 'right' with a bigger detector and then do a long run for the final result.

I note the X manual section H gives different constants for the watt spectrum of Cf-242, I do not know if the difference matters. I would also consider setting every bin of your energy range to match your real results.

Thank you Alex, I will keep updating my outcomes.

 

[Hamidul]

The main problem is that you have a small detector one meter away from the source. This means most neutrons in the problem do not hit the detector and leave the problem with the CPU time wasted. Your spectrum is basic because only one neutron hit it.

I would consider calculating a slightly different problem that might be expected to have a similar result that can be calculated with less CPU time. Such as making the detector a ring, or even a spherical shell around the problem.

You could just run the problem for a long time. With 10^5 neutrons the spectrum has one or two neutrons. 10^7 a spectrum starts to appear. 10^9 would probably take 2 core days with MCNP5, and maybe twice that with MCNPX.

Or you could do both and get the spectrum 'right' with a bigger detector and then do a long run for the final result.

I note the X manual section H gives different constants for the watt spectrum of Cf-242, I do not know if the difference matters. I would also consider setting every bin of your energy range to match your real results.

Thanks a lot Alex, By following your instruction I was able to find out all of my spectra which matched with my real result beautifully. Without your and others help in this forum, may be I would not been able to finish that. Long live the PHYSICSFORUM. Sorry for late update.

 

[Hamidul]

Hello Alex, are you here? I want to measure the dose rate in my same simulation geometry surface. I did also write a code for that, got a single data. But I am struggling to interpret my data, I need to convert it dose rate microsievert per hour.

 

[Alex A]

You are specifying a dose function rather than using a built in function and I need to spend a bit more time reading the manual. Should your coefficients be in pico Sv/Hr? That might explain why your result is so high.

Do you know the activity of your Cf-252 source?

 

[PSRB191921]

your dose calculation is strange, because:
- you do the calculation through a sphere while the notion of dose is punctual. In principle with mcnp we calculate the fluence at a point (for example with a type 5 tally) and we apply a DE/DF to it.
- Cf-252 is a spectrum not monoenergetic at 2.26 MeV you must apply a watt spectrum.

I think you DF is in pSv.cm2 so you must multiply by the neutrons flux in n/s*3600*1E-12 to have it in Sv/h

 

[Hamidul]

your dose calculation is strange, because:
- you do the calculation through a sphere while the notion of dose is punctual. In principle with mcnp we calculate the fluence at a point (for example with a type 5 tally) and we apply a DE/DF to it.
- Cf-252 is a spectrum not monoenergetic at 2.26 MeV you must apply a watt spectrum.

I think you DF is in pSv.cm2 so you must multiply by the neutrons flux in n/s*3600*1E-12 to have it in Sv/h

If I do so, I will get dose against various energy. Right? But, I need also measure the dose at various distances like 30cm, 40cm,... 100cm from the source? For getting that which
tally should I use? F6?
the results in my input file is huge.

 

[PSRB191921]

with F5:n x y z .1 you calculate the fluence at the coordinate (x,y,z). With DE/DF the fluence is transformed into equivalent dose.
In my input file I put :
F5:n 0 0 10 .1 for a distance of 10 cm from the source
F15:n 0 0 50 .1 for a distance of 50 cm from the source
F25:n 0 0 100 .1 for a distance of 100 cm from the source
I changed your DE/DF to calculate the ambiant dose equivalent (H*(10) from ICRP 74)
I also changed your watt spectrum data (from ICRP 107)
for F25 (dose equivalent at 100 cm) I obtaine 2.7280E-03 and the unit is pSv for one neutron.
You know that for Cf-252 you have 0,1164 n/s/Bq so at 1 meter your obtain :
2.7280E-03*0.1164*3600*1E-12=1.14E-12 Sv/h

 

[Hamidul]

That is awesome. This results agrees with my calculated ambient dose equivalent( actually free field dose equivalent ). though the values are little bit smaller than the calculated FFDE. I do not know how to interpret this. In addition , I have also measured FFDE with survey meter, the values are comparable, though the survey meter gives some larger values . By the way, can I know your real name and county? You all helped me a lot.

 

[Hamidul]

with F5:n x y z .1 you calculate the fluence at the coordinate (x,y,z). With DE/DF the fluence is transformed into equivalent dose.
In my input file I put :
F5:n 0 0 10 .1 for a distance of 10 cm from the source
F15:n 0 0 50 .1 for a distance of 50 cm from the source
F25:n 0 0 100 .1 for a distance of 100 cm from the source
I changed your DE/DF to calculate the ambiant dose equivalent (H*(10) from ICRP 74)
I also changed your watt spectrum data (from ICRP 107)
for F25 (dose equivalent at 100 cm) I obtaine 2.7280E-03 and the unit is pSv for one neutron.
You know that for Cf-252 you have 0,1164 n/s/Bq so at 1 meter your obtain :
2.7280E-03*0.1164*3600*1E-12=1.14E-12 Sv/h

One more query, please suggest me a Watt energy spectrum for Am-Be neutron source. I have another code with Am-Be with same geometry. My existing function is -3 0.933020 3.46195 for Am-Be

 

[PSRB191921]

Am-Be is not a fission source, so it is not a watt spectrum. You must input the spectrum by bin.

 

[Hamidul]

Thank you so much