Anyone test in 2.1.30 yet?

Separate section for discussion on gamma transport
Post Reply
User avatar
gavin.ridley.utk
Posts: 95
Joined: Wed Mar 16, 2016 7:07 am
Security question 1: No
Security question 2: 40

Anyone test in 2.1.30 yet?

Post by gavin.ridley.utk » Tue Mar 27, 2018 12:03 am

Hey folks,

I'm having some trouble running coupled n, gamma calculations in version 2.1.30. When not using MPI, stuff seems to work in 2.1.29. I can present an offending input file if whichever developer who decides to check this out can't reproduce the issue with whatever simple examples you all have.

Have a nice day, and thanks for this awesome code!
Gavin Ridley

Yellowstone Energy

Ville Valtavirta
Posts: 468
Joined: Fri Sep 07, 2012 1:43 pm
Security question 1: No
Security question 2: 92

Re: Anyone test in 2.1.30 yet?

Post by Ville Valtavirta » Wed Mar 28, 2018 11:44 am

Hi Gavin,

did you apply the fixes for the bugs listed in http://ttuki.vtt.fi/serpent/viewtopic.php?f=25&t=2839 already?

If not, that would be a good place to start.

-Ville

kotlyar
Posts: 4
Joined: Thu Sep 08, 2011 2:36 pm

Re: Anyone test in 2.1.30 yet?

Post by kotlyar » Wed Apr 04, 2018 4:55 pm

Hi all,

I am trying to obtain the gamma heat deposition using ver. 2.1.30
I have included the following cards in the input file:
1) Library
set acelib "/DATA/sss_endfb7u.xsdata"
set pdatadir "/DATA/PhotonData"
2) Transport option
set ngamma 2
3) Detector
det P1 p dm ZrHx dr -26 void

I am getting the following warnings and error:
Adding photon interaction data...
***** Warning: Photon interaction data not available for hydrogen (Z = 1)
***** Warning: Photon interaction data not available for oxygen (Z = 8)
***** Warning: Photon interaction data not available for zirconium (Z = 40)
***** Warning: Photon interaction data not available for tungsten (Z = 74)
........
Input error:
No photon data in photon transport problem

I've doubled checked and the files for the 'p' mode are in the "/DATA/PhotonData" path.
Any help will be greatly appreciated.

kotlyar
Posts: 4
Joined: Thu Sep 08, 2011 2:36 pm

Re: Anyone test in 2.1.30 yet?

Post by kotlyar » Wed Apr 04, 2018 7:16 pm

I have just realized that the photon ace datafile was not available in my directory.

Toni Kaltiaisenaho
Posts: 11
Joined: Sat Jun 27, 2015 3:31 pm
Security question 1: No
Security question 2: 92

Re: Anyone test in 2.1.30 yet?

Post by Toni Kaltiaisenaho » Thu Apr 05, 2018 3:46 pm

Hi,

I would suggest that you use the analog energy deposition detector dr -12 which gives the correct result within the limits of the accuracy of the physics models used in Serpent. The dr -26 gives the collision flux estimate of the energy deposition (or track-length flux estimate with the dtl detector parameter), and it uses the average energy deposition data given in the photon ACE library. You may get incorrect results with the dr -26 if the data is not in agreement with the photon physics models or simulation parameters. For example, if you use a high cut-off energy for photons, the dr -26 will underestimate the energy deposition. However, the dr -26 is currently the only choice if you need to improve the statistics e.g. in low density materials by using the dtl parameter or the "set cfe" card.

User avatar
gavin.ridley.utk
Posts: 95
Joined: Wed Mar 16, 2016 7:07 am
Security question 1: No
Security question 2: 40

Re: Anyone test in 2.1.30 yet?

Post by gavin.ridley.utk » Thu Jun 28, 2018 6:56 pm

Hi Toni,

Thanks for this tip. Serpent tells me that the material must be void in order to use the -12 MT for photons. Why is that? The heating should be zero in void, so this confuses me.

One more question: it appears that in using the photon transport mode, my "fuel" material seems to be renamed internally to "fuelp7r1", if I have my fuel set to deplete using "burn 1". As such, later attempting to use the old "fuel" material in power normalization "e.g., set powdens .03 fuel" or in a detector like "det fuelPowerGamma p dr -26 fuel" fails to work. It took quite some time to figure out what was going on here.

What really ends causing the failure is that the material called "fuelp7r1" has its macroscopic cross sections allocated in allocmacroxs.c, whereas "fuel" has no macroscopic XS allocated. Then, in trying to use the aforementioned photon heating detector, linkreactions.c ends up seeing "fuel", and attempts to address "rea" where it expects the macroscopic heating cross section to be. Since the macroscopic heating cross section only got allocated for "fuelp7r1" rather than fuel, rea stays less than VALID_PTR and this raises the error at line 327 at linkreactions.c to get risen. That error does not fit the issue at hand that the fuel material was renamed. Instead, the error says "MT = -4 not found", which caused a lot of confusion for me.

Presumably the fuel got renamed internally for depletion purposes. Maybe some more helpful error messages could be added here?
Gavin Ridley

Yellowstone Energy

Ville Valtavirta
Posts: 468
Joined: Fri Sep 07, 2012 1:43 pm
Security question 1: No
Security question 2: 92

Re: Anyone test in 2.1.30 yet?

Post by Ville Valtavirta » Fri Jun 29, 2018 2:00 pm

Hi,

the <mat> part of a "dr <resp> <mat>" simply tells Serpent which material to use as a response any time the detector is scored. It does not tell Serpent to limit the scoring to collisions in that material (see also my reply to your other question).

Setting that <mat> to void simply tells Serpent to use whichever material is located in the collision point.

In order to limit scoring the detector in just a single material you'll want to use the "dm <mat>" option in the detector card instead.

That said, using

Code: Select all

det fuelPowerGamma p dr -26 void dm fuel
instead of

Code: Select all

det fuelPowerGamma p dr -26 fuel
should give you the photon KERMA in fuel material whether or not the fuel material is divided to burnable zones.

I also did not have any problems using the "fuel" material in power normalization. The error from linkreactions.c seems to be solely related to using a material that is not part of the transport calculation (e.g. due to being divided into multiple pins) as a detector response material in the "dr" option of a detector. As you noted, the error itself is not very informative in this case.

The following input ran with Serpent 2.1.29 and 2.1.30:

Code: Select all

pin p1
fuel   0.46950
void   0.4791
clad   0.5464
cool

pin p2
fuel   0.46950
void   0.4791
clad   0.5464
cool

pin p3
fuel2  0.46950
void   0.4791
clad   0.5464
cool

lat l1  1  0.0 0.0 15 15 1.4427
p2 p3 p3 p3 p3 p3 p3 p3 p3 p3 p3 p3 p3 p3 p3
p3 p3 p3 p3 p3 p3 p3 p3 p3 p3 p3 p3 p3 p3 p3
p3 p3 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1
p3 p3 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1
p3 p3 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1
p3 p3 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1
p3 p3 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1
p3 p3 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1
p3 p3 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1
p3 p3 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1
p3 p3 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1
p3 p3 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1
p3 p3 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1
p3 p3 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1
p3 p3 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1 p1

surf s3 sqc 0.0 0.0 10.82025

% --- Cell definitions:

cell c3  0  fill l1  -s3
cell c9  0  outside   s3     % Outside world

mat fuel    -10.283 tmp 551.0 burn 1
92235.03c       -0.04275144
92238.03c       -0.83872152
8016.03c        -0.11852704

mat fuel2    -10.283 tmp 551.0 burn 1
92235.03c       -0.04275144
92238.03c       -0.83872152
8016.03c        -0.11852704

mat clad     -6.55 tmp 551
40090.03c   -0.98135
24052.03c   -0.00100
26056.03c   -0.00135
28058.03c   -0.00055
50120.03c   -0.01450
8016.03c    -0.00125

mat cool     -0.766 tmp 551 moder lwtr 1001 rgb 50 50 255
1001.03c    0.6666666667
8016.03c    0.3333333333

therm lwtr lwj3.11t

set pop 4000 50 50 1 5

set bc 3 3 3

% --- Test detector

det myTest dr -8 void dm fuel

% --- Set power normalization

set power 100 fuel

% --- Set depletion step

dep daystep 1
Running -checkvolumes yields

Code: Select all

Volumes (2D problem, the values are in cm2) :

Material fuel       : 1.1773E+02 1.2031E+02 (0.01613) :  0.02195   
Material fuelpp2r1  : 6.9250E-01 7.0247E-01 (0.25733) :  0.01439   
Material fuelpp1r1  : 1.1703E+02 1.1961E+02 (0.01613) :  0.02199   
Material fuel2      : 3.8088E+01 3.7324E+01 (0.02878) : -0.02004   
Material clad       : 4.8785E+01 4.8704E+01 (0.02847) : -0.00165   
Material cool       :        N/A 2.5546E+02 (0.00923) :      N/A * 
showing that the material "fuel" is indeed divided into two materials (p1r1 an p2r1).

The power in material "fuel" (from _det0.m) is 100 Watts as requested and the total power of the input is approximately 132 Watts as some power is generated in fuel2.

-Ville

User avatar
gavin.ridley.utk
Posts: 95
Joined: Wed Mar 16, 2016 7:07 am
Security question 1: No
Security question 2: 40

Re: Anyone test in 2.1.30 yet?

Post by gavin.ridley.utk » Fri Jun 29, 2018 5:05 pm

Ville, I tremendously appreciate your guidance! Thank you kindly for the prompt reply.
Gavin Ridley

Yellowstone Energy

Afigue
Posts: 5
Joined: Mon May 07, 2018 6:49 pm
Security question 1: No
Security question 2: 93

Re: Anyone test in 2.1.30 yet?

Post by Afigue » Fri Nov 02, 2018 5:26 pm

Hi! where do i find the photon ace files? I haven't been able to run any simulations using photon transport.

Best.

Antonio

Ville Valtavirta
Posts: 468
Joined: Fri Sep 07, 2012 1:43 pm
Security question 1: No
Security question 2: 92

Re: Anyone test in 2.1.30 yet?

Post by Ville Valtavirta » Mon Nov 05, 2018 10:46 am

Hi,

I think people have been generally using photon ACE-files from the MCNP distribution. The other possibility would be to process your own photoatomic interaction data. I don't think there are yet processed photon ACE-data distributed for Serpent.

-Ville

Post Reply