Time dependence in external source

Report all good and bad behavior here
EdsonHenrice
Posts: 10
Joined: Sun Mar 09, 2014 5:18 pm
Security question 1: No
Security question 2: 92

Time dependence in external source

Post by EdsonHenrice » Thu May 15, 2014 9:56 pm

I saw in a post something I believe to be about time dependence in "Dynamic simulation mode".
then I'm trying to use this:

set nps 20000 100 intervalos
tme intervalos 2 10000 0 0.01

when I use the code above, Detector time binning show me external neutrons behavior like only one pulse in the very begining. In a few seconds neutrons die ( I simulate a sub-critical system)

I was looking at the User's Manual in section 9.2.4. But I'm not sure how to use the time in source files.

Could you explain more about the inputs for the time dependence in external source calculation mode?
For example, I would like the external neutron source to produce 2E+8 neutrons/s during 100 seconds.

User avatar
Jaakko Leppänen
Site Admin
Posts: 2377
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: Time dependence in external source

Post by Jaakko Leppänen » Fri May 16, 2014 9:08 am

At the moment the only options to apply time dependence in external source is to read the source points from a file or to use a user-specified subroutine. These options are described in a related post. Writing the points in file using a Matlab script that sets the coordinates and energy and samples the direction and time randomly is probably the easiest way to go.
- Jaakko

EdsonHenrice
Posts: 10
Joined: Sun Mar 09, 2014 5:18 pm
Security question 1: No
Security question 2: 92

Re: Time dependence in external source

Post by EdsonHenrice » Thu May 29, 2014 5:09 pm

I think I don’t exactly understand how to write a source file or there is something wrong in my Serpent input.

I need to measurement first 100 seconds of nuclear subcritical operation for my phd thesis. When I use time binning counter, I only saw neutrons in first output or second one raw (my tme card is: tme intervalos 2 100 0.0 1). This occurs when I don’t use delayed neutron in my input, even if I had changed the line like you told me to do in other post.

When I set to use delayed neutrons in Serpent input, the output file show me large numbers of neutrons in the first row and the delayed neutrons disappear in a few seconds in the time binning counter (in this case I use tme card as : tme intervalos 2 1000 0.0 10 for example). In this case, the raws below the first raw show me neutrons 3 or 4 order of ten smaller than the neutrons numbers in the first row. The neutron number decreasis quickly.

I guess the neutron source appears only in the beginning of my simulation or every time I use “dep daystep” function (for this porpoise I did 1day/ 86400second, I believe this is not the better way to use this function, bus it is only a test)

In the field set nps, I was increased <Nsrc> field and the number of neutrons showed in the output file is the same (there is a few fluctuation in this number, I believe it occurs in function of random numbers only). This occurs even if I change the number of rows in the file fonte.c, it doesn’t matter if I put 100 or 100 hundred rows, the results in the time binning output are the same.

In the example below, tme field goes 0 to 1 second. In the future I will use 100 seconds, for example.

How can I do to the external neutron source creates neutrons for every time I am analyzing? The external neutron source is pulsed? Is it possible to maintain external neutron source continually? How can I do it for 100 seconds?

Below it follows how I use time binning option.

set nps 4000 100 intervalos

src testen sf fonte.c 1
set nbuf 1e3

det contador dm fuelinp1r1 di intervalos
tme intervalos 2 100 0.0 1

Same liner of the fonte.c file:

0.000000E+00 0.000000E+00 0.000000E+00 -0.580422E+00 -0.256661E+00 -0.772810E+00 0.100000E+02 1.0 0.856802E-04
0.000000E+00 0.000000E+00 0.000000E+00 -0.823497E+00 0.566800E+00 0.243044E-01 0.105000E+02 1.0 0.327188E-05
0.000000E+00 0.000000E+00 0.000000E+00 0.421503E+00 0.350987E+00 -0.836148E+00 0.110000E+02 1.0 0.723138E-04

Obs: I consider the external neutron source in the x=0 y=0 and z=0 from reactor.

User avatar
Jaakko Leppänen
Site Admin
Posts: 2377
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: Time dependence in external source

Post by Jaakko Leppänen » Thu May 29, 2014 6:21 pm

Could you send me the input and source file by e-mail?
- Jaakko

User avatar
Jaakko Leppänen
Site Admin
Posts: 2377
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: Time dependence in external source

Post by Jaakko Leppänen » Fri May 30, 2014 1:53 pm

This is what I get for detector "contador":

Image

when this recently discovered bug:

http://ttuki.vtt.fi/serpent/viewtopic.p ... 4b92#p5091

is fixed. To me this looks like what can be expected of a sub-critical system. I don't quite understand what the purpose of the burnup parameters is? This is simply a time-dependent transport simulation that doesn't involve burnup.

It also seems that the source is sampled correctly from the file. The code selects random points from the list and sets the variables accordingly. Time is the last value on each row.
- Jaakko

User avatar
Jaakko Leppänen
Site Admin
Posts: 2377
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: Time dependence in external source

Post by Jaakko Leppänen » Fri May 30, 2014 2:12 pm

So what happens in the simulation is that all neutrons in the source file are emitted between 4E-8 and 1E-2 seconds, which means that the prompt fission chains initiated by these neutrons die out within the first time bin. This explains the initial peak. The tail of the distribution results from fission chains initiated by delayed neutrons.

Is this the behaviour you expect, or do you intend to simulate a source that is constant in time?
- Jaakko

EdsonHenrice
Posts: 10
Joined: Sun Mar 09, 2014 5:18 pm
Security question 1: No
Security question 2: 92

Re: Time dependence in external source

Post by EdsonHenrice » Tue Jun 03, 2014 5:47 am

I'm interested on studying the begining of operation of subcritical systems.
I was writing the source file in a wrong way. That was not what I wanted.

Now I guess I understand. for the external source to operate for 100 seconds my source file randomily peak a time between 0 and 100 second (the last value on each row). One of the cases I'm working on is to simutate a source that is constant in time. Is that way I can do it?

User avatar
Jaakko Leppänen
Site Admin
Posts: 2377
Joined: Thu Mar 18, 2010 10:43 pm
Security question 2: 0
Location: Espoo, Finland
Contact:

Re: Time dependence in external source

Post by Jaakko Leppänen » Tue Jun 03, 2014 10:59 am

Each entry in the source file represents an individual source neutron. So if you want a constant source over the time interval, you need to sample the initial time (the last value of each row) uniformly between the beginning and end.
- Jaakko

MojoJojo
Posts: 18
Joined: Fri Jul 06, 2018 9:27 pm
Security question 1: No
Security question 2: 93
Location: Canada

Re: Time dependence in external source

Post by MojoJojo » Sat Feb 22, 2020 1:33 am

Jaakko Leppänen wrote:
Tue Jun 03, 2014 10:59 am
Each entry in the source file represents an individual source neutron. So if you want a constant source over the time interval, you need to sample the initial time (the last value of each row) uniformly between the beginning and end.
Hey Jaakko,

I was trying to simulate a beamed external neutron source. So I made a Matlab code to do the following (as you instructed):
1- Generate 'n' random locations on a circular plane: x0 y z (i.e. the beam port).
Image
2- Assign the same direction to each point source: -1.0 0.0 0.0
3- Assign the same energy to each point source: 1 MeV
4- Assign the same weight to each point source: 1.0
5- Assign a time to point source groups (i.e. first 'n' point source at t = 0, second 'n' point sources at t = 1, etc.): 0 to 100 seconds

Code: Select all

96.000000   6.277595 135.251644  -1.0   0.0   0.0  1.0000e+00  1.00000  0.0
96.000000   3.545154 140.054213  -1.0   0.0   0.0  1.0000e+00  1.00000  0.0
.
.
.
96.000000   1.007386 139.781269  -1.0   0.0   0.0  1.0000e+00  1.00000  6.0
96.000000   0.659124 139.036558  -1.0   0.0   0.0  1.0000e+00  1.00000  6.0
.
.
.
96.000000  -0.468055 136.086386  -1.0   0.0   0.0  1.0000e+00  1.00000 100.0
96.000000   3.953465 141.458997  -1.0   0.0   0.0  1.0000e+00  1.00000 100.0
I still get one pulse at t = 0!
Image

In the serpent input code, I have:

Code: Select all

%%% Time structure of neutron source emission (type 2: uniform binning)
tme srctime 2 100 0.000 100.00

%%% External Neutron Source Mode Enabled via nps input card
set nps 10000 100 srctime

%%% Source Definition (type 1: ASCII format)
src 1 sf "source.txt" 1
What am I doing wrong??
The generated source file works well in steady state external neutron source mode (all times set to zero).
Name: Mohammed Mahdi
Occupation: MASc. Student
Research Interest: Liquid Fuel Molten Salt Reactor
Institution: McMaster University
Email Address: Mahdim4@mcmaster.ca

Ville Valtavirta
Posts: 457
Joined: Fri Sep 07, 2012 1:43 pm
Security question 1: No
Security question 2: 92

Re: Time dependence in external source

Post by Ville Valtavirta » Mon Feb 24, 2020 11:31 am

Hi,

could you try re-running the simulation with just a single time interval of 100 seconds in the srctime definition?

Serpent is probably sampling points from the source file only for the first time interval (and rejecting all points that are after the end of that first interval) and the source for the further time intervals comes from the neutrons that reached the end of the previous interval.

-Ville

Post Reply