Re: CHI_D data missing -> assume delayed neutrons born in top group for materialfuel

[Alexander Lindsay]

Hi Olga,

Unfortunately it may be a while until I can dig through this. It would be great if a current ARFC (the University of Illinois research group in charge of the code) member looked into this. I've cc'd the moltres-users group for that purpose.

Alex

On Wed, Mar 13, 2019 at 7:02 AM Olga Negri <olga....@gmail.com> wrote:

Hi Alex!

Thank you very much for your reply. I have updated the file: generated new cross-sections using SERPENT (the keff was raised from 1 to about 1.6 now) and increased the size of the reactor.

The reason I was referring to the MSRE cross sections is that they were giving some response in terms of temperature change. The response that I am getting with the fast reactor cross sections is the same every time, even if I change cross section data and the reactor size.

I am not sure what I did wrong...

I have attached the input and the data library files as well as the start of the postprocessor output.

Olga

+----------------+----------------+----------------+----------------+----------------+

| time           | group1_current | group1_old     | multiplication | temp_fuel      |

+----------------+----------------+----------------+----------------+----------------+

|   0.000000e+00 |   0.000000e+00 |   0.000000e+00 |   0.000000e+00 |   0.000000e+00 |

|   1.000000e-04 |   2.940000e+02 |   3.125000e+02 |   9.408000e-01 |   9.000000e+02 |

|   2.200000e-04 |   2.940000e+02 |   2.940000e+02 |   1.000000e+00 |   9.000000e+02 |

|   3.640000e-04 |   2.940000e+02 |   2.940000e+02 |   1.000000e+00 |   9.000000e+02 |

+----------------+----------------+----------------+----------------+----------------+

On 8 Mar 2019, at 22:43, Alexander Lindsay <alexlin...@gmail.com> wrote:

Looking at the results from the input file you sent me, it looks like the reactor is simply sub-critical for the given geometry and the material cross sections at 900 K. There is the initial jump in neutrons in the first time step, but that is simply due to the arbitrary initial condition. For every step afterwards, the multiplication is < 1. With no neutrons to heat the reactor, the temperature will just stay at 900 K. After a long enough time, the initial residual falls below the nonlinear absolute tolerance (set through the nl_abs_tol parameter in the Executioner block) and so the solver does nothing. That's why after a while the variable values stop changing.

To get a reactor closer to critical, you could think about a couple of things:

1. Are you sure you have the correct cross sections for your reactor design? You said you want to do a fast reactor, but it looks like you're using the cross section data that we generated for the thermal MSR.

2. Are your reactor dimensions correct? If you increase the reactor size, then of course you will have less leakage.

On Thu, Mar 7, 2019 at 6:29 AM Alexander Lindsay <alexlin...@gmail.com> wrote:

Hoping I can take a look at it at the end of the day today. Sorry for the delay. 

On Mar 7, 2019, at 2:54 AM, Olga Negri <olga....@gmail.com> wrote:

Hi Alex,

Just wondering if you had a chance to have a look?

Best wishes,

Olga

Begin forwarded message:

From: Olga Negri <olga....@gmail.com>

Subject: Re: CHI_D data missing -> assume delayed neutrons born in top group for materialfuel

Date: 1 March 2019 at 15:08:57 GMT

To: Alexander Lindsay <alexlin...@gmail.com>

Alex,

Thank you for your response!

I have tried to change the time-step, but it seems like the behaviour is still the same. The input file is attached.

Looking forwards to hearing from you,

Best wishes,

Olga

<gen-mesh-one-material-1.i>

On 28 Feb 2019, at 18:06, Alexander Lindsay <alexlin...@gmail.com> wrote:

On Thu, Feb 28, 2019 at 3:57 AM Olga Negri <olga....@gmail.com> wrote:

Hi Alex,

Thank you for your reply, your suggestions have fixed the problem.

I am trying to model a MSFR (so far in 2D only), a fast spectra MSR design.

One of the problems that I have faced now is that once I get the file to run, the variables would start changing, in this case increasing and then decreasing for both the groups and the temperature, and at a certain moment temperature would go back to it original value and any other parameters would not change with time as well. I am not sure what I am doing wrong there. Is that a common thing or am I doing something completely wrong? I have tried to do this with models containing moderator material and this has never happened.

I don't know how I dropped our mailing list in my first reply, but I'm adding it back here... So this is behavior that other users and I have encountered. I know that having too large of a time-step can cause this problem, but I believe you can also run into it in other situations. Could you share the input file that's generating this behavior? I can take a look at what might be happening.

Alex

Looking forward to hearing from you,

Best regards,

Olga

On 26 Feb 2019, at 05:41, Alexander Lindsay <alexlin...@gmail.com> wrote:

On Mon, Feb 25, 2019 at 4:52 AM Olga Negri <olga....@gmail.com> wrote:

Hello!

I am very new to MOLTRES, I was trying to modify 'gen-mesh-one-material' tutorial to account for fuel velocity and precursor drift. I have received 

CHI_D data missing -> assume delayed neutrons born in top group for materialfuel

warning

This is a warning and is not fatal. Presumably you are running in tests/nts and there are a bunch of files in there that look like msr2g_enrU*.txt. The warning is essentially saying there are no msr2g_enrU*CHID.txt files, e.g. there is no data about what energy groups delayed neutrons are born into. Since it doesn't have that data, moltres is just going to assume all delayed neutrons are going to be born into the highest energy group. If you don't want that to be the case, then you will need to provide the *CHID* data files.

and a segmentation fault. I am not sure what I should do to fix this I would appreciate any advice and comments.

The reason for the segmentation fault is that you have an EigenKernel in your kernels block (CoupledFissionEigenKernel) but you are not using an Eigen Executioner like NonlinearEigen or InversePowerMethod. You are using the Transient executioner. Either you can't use the CoupledFissionEigenKernel or you need to switch to using an Eigen executioner.

What is it that you're trying to ultimately accomplish? Maybe we can help direct you. Thanks for investigating Moltres!

Sincerely,

Alex

My input file is attached.

Looking forward to hearing from you!

Best regards,

Olga

--
You received this message because you are subscribed to the Google Groups "moltres-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to moltres-user...@googlegroups.com.
To post to this group, send email to moltre...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/moltres-users/50df3417-962a-491a-a4ee-8d04f43ce2fb%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Katy Huff]

Sun Myung - is this related to the issues you've been seeing?

To view this discussion on the web visit https://groups.google.com/d/msgid/moltres-users/CANFcJrFD8ENZOTxoZDg9PbUnSMKDOu-4UHY3b92%3Di0W_2hKB3Q%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

--

Kathryn Huff

Assistant Professor

Nuclear, Plasma, and Radiological Engineering

The University of Illinois at Urbana-Champaign

Email: kdh...@illinois.edu 

Preferred Pronoun: she/her

Individual site: http://kdhuff.npre.illinois.edu/

Research group site: http://arfc.npre.illinois.edu/ 

Office: 118 Talbot Laboratory, 104 S. Wright St. 

Book a meeting with me: https://katyhuff.youcanbook.me/