Polaris stainless steel reflector group data preparation

[Pavel Suk]

Hello, 

   I have doubts about the results of the homogenized data prepared for the stainless steel reflector in Polaris. I am preparing data via Triton also and when I perform a calculation of a model with almost a full reflector made from stainless steel, the differences from Triton are large. Especially when the multiplication factor of the homogenized data preparation model is compared, the discrepancies can be up to 1150 pcm. 

 

I am attaching the files here: 

 

Rad2 models consist of water and stainless steel block and the discrepancy between Triton and Polaris is approximately 175 pcm in multiplication.

 

Rad3 models are mainly made from stainless steel and the discrepancy between Triton and Polaris is more than 1300 pcm in multiplication of the model. 

 

This discrepancy produces macro data that affects full core calculation up to 30 ppm of boron (small core). 

 

I am not able to find errors in the Rad3 cases (when you compare the files with Rad2, you will see that there are just different labels and dimensions of the stainless steel). 

I am also writing this confusion because a few years ago I found similar behavior in the CASL Vera code that uses also the MoC. The CASL Vera calculated the multiplication factor of the core with the stainless steel reflector wrongly (in comparison to the Serpent code). When the amount of stainless steel decreased. The results were in good consistency. 

 

Do you have any idea how to set up the Polaris to obtain proper XS data?

 

The XS data obtained from Triton works in the full core model correctly and the multiplication and power distribution are consistent with the reference full core Serpent model. 

 

Thanks, Pavel 

 

[Matt Jessee]

Pavel,

Thanks for your inquiry.  Attaching 3 files that I ran with SCALE 6.3.1

1) Polaris model, k-eff = 1.15151

2) TNEWT model, k-eff = 1.15172

3) CSAS6 model, k-eff = 1.15530[140]

All are fairly consistent with one another, so I am not concerned about a fundamental flaw or deficiency in MOC/NEWT or MG XS self-shielding differences.Key things to note:

1) Polaris reflector models mirror BCs on all 4 sides, just like a single-assembly model. We will add vacuum BC in future release as part of are enhancements to reflector modeling. But note for this case, the refl BC assumption should have a minor effect on the homogenized cross sections.

2) In your NEWT model, I modified quadrature (s6 to s10) and scattering options (all p2, just like Polaris), but I don't think it would have much of a difference.

3) In the Polaris model, there appears to be something wrong with the code handling a defined "coolant" composition using the CONC option, coupled with a state card. As you see in my model, I just commented out the state card and it works fine. I did not test, but I expect if you use the default composition for COOL.1 and COOL.2 (which is borated water), then the state card should work as expected. 

Thanks, Matt

[Pavel Suk]

Matt, 

  thank you for your prompt reply. I think the problem will be in the boundary condition. I will try to calculate my tasks and I will see the influence of the boundary condition into the full core calculation. Regarding point 3. What is the problem thar you are talking about? I used the CONC definiton with state parameter and it worked correctly. 

Pavel 

Dne středa 17. ledna 2024 v 18:34:53 UTC+1 uživatel Matt Jessee napsal:

[Matt Jessee]

Pavel,

Good deal, for point 3), disregard my prev comments. I ran several combinations of your model, and cannot reproduce the results anymore.

[Pavel Suk]

Hi Matthew,

I hope you are doing well. We have resolved the issue related to boundary conditions in the SCALE Polaris calculation two years ago.

As I understand it, Polaris currently uses mirror boundary conditions in all directions for both geometries (fuel-only and fuel + reflector). Is there any possibility to modify the boundary conditions in the fuel + reflector model? If not, is this feature planned to be implemented in a future SCALE Polaris release?

Thank you in advance.

Best regards,

Dne neděle 21. ledna 2024 v 14:46:41 UTC+1 uživatel Matt Jessee napsal: