Skip to content

Latest commit

 

History

History
200 lines (156 loc) · 13.6 KB

README.md

File metadata and controls

200 lines (156 loc) · 13.6 KB

OSTR-CONSTELATION

Coupled model using STAR-CCM+ and Serpent 2 for a HENRI-based experiment in the Oregon State TRIGA Reactor. Based on CONSTELATION for HENRI in TREAT (CONSTELATION GitHub).

This project utilizes the same software, structure, and coupling method as CONSTELATION for TREAT, but applies it to the OSTR with a custom designed experiment. The purpose is to quantify the extent to which the coupled model CONSTELATION is accurate in predicting the density evolution of the helium-3 during an injection for a TREAT transient pulse and its effects on the reactor. The OSTR and the experiment designed for it will provide the data for this quantification. This project contains the code, scripts, and instructions to execute the coupled STAR-CCM+ and Serpent 2 simulation for the OSTR.

An example transient coupling script can be found at the Serpent 2 wiki: MSCS for coupled transients.

Requirements

  • Python 3 (tested using Python 3.9 from the Anaconda distribution)
  • STAR-CCM+ (used version 2020.2.1 on INL HPC)
  • Serpent 2 (used version 2.1.31 on INL HPC)

File Structure

The file structure required to begin the simulation is displayed below. This example is for a simple model using only one STAR-CCM+ simulation and no other coupled physics, like fuel feedback.

|-- Archive/
|-- ExtractedData/
|   |-- He3Data_table.csv
|-- CONSTELATION.py
|-- functions.py
|-- STAR_coupled.sim
|-- STAR_load_data.java
|-- Serpent_input
|-- STAR_job.sh
|-- Serpent_job.sh

Any additional STAR-CCM+ simulations will require their own _job.sh and _load_data.java files. The user can also add fuel feedback and may need to place an initial fuel.ifc file in the directory and enable or write fuel functions in CONSTELATION as necessary.

Usage

The beginning of the ostr_CONSTELATION_3.py file contains the inputs required to run the simulation. The inputs have comments in the script that describe what they do, but we will also define them here.

Main Input File

Serpent_file = 'TRIGA'

This is the name of the Serpent 2 input file being used for the coupled simulation. This should be in its base form, the script will add the necessary lines for the simulation. See the documentation on the Serpent 2 model for more information.

Time Step Inputs

timestep = 5E-6

This is the time step in seconds for the Serpent 2 simulation.

STAR_STEP = 100

This is the number of steps the STAR-CCM+ simulation performs before sharing its data with the Serpent 2 simulation. It must be the ratio of the Serpent 2 time step to the STAR-CCM+ time step. For example if the Serpent 2 time step is 2E-6 and the STAR-CCM+ time step is 5E-8, then STAR_STEP = 40.

step_length = 100

This is the same as STAR_STEP, except that it does not change through the simulation and is used to increment STAR_STEP.

Cluster Job Inputs

run_SERP = "qsub SERPENT_job.sh"
run_STAR1 = "qsub STARTop_Job.sh"

These are the commands to run the simulations on the cluster.

Serpent 2 COM Files

comin_name = 'com.in'
comout_name = 'com.out'

These are the names of the com files used by Serpent to communicate. It is nice to define them here and reference the variable in the code. However, it is not expected that these names change.

Serpent 2 Mesh Inputs

helium_mesh = '1 -100 100 1 0 100 500 -33.02 30.78\n'

This is the mesh used in Serpent 2 to keep track of the helium-3 of interest. The meshes are defined in the form (NX XMIN XMAX NY YMIN YMAX NZ ZMIN ZMAX) If there are multiple STAR-CCM+ simulations modeled, there will need to be a mesh for each one. Additionally, if the user wishes to model fuel temperature feedback, a fuel mesh must be define here.

Cross-code Communication Inputs

reference_conversion = [20.405, 40.605, 226.7903]
unit_conversion = [0, -1/100, -1/100]

These are the values used to convert between reference frames. Both inputs are of the form [x, y, z], where the first value is for the x-direction, etc. The description of the position_Serpent_to_STAR function later provides more detail on how to use these inputs. The reference_conversion input is the difference in translation between the two models' reference frames; and the unit_conversion input accounts for any unit change, direction change, or any other applicable multiplier.

Heat_csv_outfile = 'STAR_HeatTop.csv'
Heat_csv_Title = ['X(m)','Y(m)','Z(m)','VolumetricHeat(W/m^3)']
Heat_csv = STAR_csv(Heat_csv_outfile,Heat_csv_Title)

This set of inputs provides the name and header for the .csv file Serpent 2 writes out for STAR-CCM+ to read in regarding the volumetric heating of the helium-3.

Serpent_ifc_top = Serpent_ifc('HE3.ifc','2 helium3 0\n','1\n',helium_mesh)

This input defines the Serpent_ifc class object for the .ifc file used to pass information with Serpent 2. The first entry is the file name, then the header, then the mesh type, then the mesh itself. Further information is provided in the description of the class later in this document.

STARHeat_table = './ExtractedData/He3Data_table.csv'
columns = ['Position in Cartesian 1[X] (cm)', 'Density(g/cm^3) (kg/m^3)', 'Temperature (K)']
STAR_csv_top = STAR_csv(STARHeat_table,columns)

This defines the STAR_csv class object used for the .csv file written by STAR-CCM+ to pass information. The first input is the file path and the second input is the header (or names of the columns). Further information about the class is provided later in this document.

Serpent_det_heat = 'Serpent2STop'

This input provides the name of the detector that will be read later by the script. Any additional detectors should also be defined here, either for additional STAR-CCM+ simulations or for fuel temperature.

Serpent_done = './SerpentDone.txt'
STARTop_Done = './STARTopDone.txt'
STAR_read = './ReadTop.txt'

These inputs provide the file names to communicate that each simulation is done with its respective time steps for the cycle. The STAR_read input was created to ensure the STAR-CCM+ simulation read the Serpent_done file before the file was deleted.

Constants and Conversions

cm3_to_m3 = 1E-6
time_to_wait_default = 3600
sig_notdigit = 42

These are inputs that provide values used later in the script to avoid numbers floating in the script with no definition. The first is the conversion multiplier for going from cubic centimeters to cubic meters. The time_to_wait_default constant was selected to be 1 hour in seconds as a standard for most of the wait times in the script. Finally, sig_notdigit is the number printed to the terminal when the com.out file is empty or its contents cannot be converted into an integer. These inputs are used in functions described later in this document.

Classes

The following section describes the classes defined and used in CONSTELATION.

Serpent_ifc(object)

This class is used to store information about the interface files used by Serpent 2. It is initialized with 4 attributes:

Top_ifc = Serpent_ifc(name,header,mesh_type,mesh)

These 4 attributes represent the first 4 lines written into the .ifc files that tell Serpent 2 what the file is. An additional attribute that is not required upon initialization is the data attribute, which is a place to store data associated with the .ifc file if necessary. The data does not need to be stored in the object when used in the csv_to_ifc function as the data is written out to the file.

STAR_csv(object)

The STAR_csv class is similar to the Serpent_ifc class and contains information relevant to the STAR-CCM+ .csv files used to pass information to and from STAR-CCM+. It is initialized with 2 attributes:

Top_csv = STAR_csv(name,header)

The name attribute is the file name used in the python script and the header attribute is the first row of the .csv file. This class also has an optional data attribute that can be defined later.

Functions

The following section describes the functions defined and used in CONSTELATION.

wait_for_file(file,wait)

This function takes in the file path and wait time in seconds and waits until the file exists or the wait time is exceeded. It is used a few times throughout CONSTELATION to make sure a file is there to be read or to wait for a file to be created during the set up. If the wait time is exceeded, there is an error that is raised to stop the program.

position_Serpent_to_STAR(data,reference_conversion,unit_conversion)

This function converts position values from the Serpent 2 reference frame to the STAR-CCM+ reference frame. It takes 3 inputs: data, reference_conversion, and unit_conversion and is defined as:

data = (data-reference_conversion)*unit_conversion
return data

The reference_conversion value is the difference between the two model's reference frames. Normally, the STAR-CCM+ model is centered on the HENRI cartridge on the $x$- and $y$- axes. In the Serpent model, the HENRI cartridge has some translation from the origin as the origin is at the center of the reactor. The unit_conversion value addresses any discrepancy between the units of the two codes. For example, the STAR-CCM+ model may be in meters, but Serpent uses centimeters. For 2-D CFD simulations, the unit_conversion value can be set to zero for the dimension that does not exist in the STAR-CCM+ model.

SerpentHeat_to_Star_csv(detector,outfile,title)

This function writes out Serpent 2 detector position and tally values to a .csv file for STAR-CCM+ to read, specifically for the helium-3 heating detector. The 3 inputs for this function are the serpentTools Detector (serpentTools and serpentTools DetectorReader) object of interest, the .csv file to write to, and the header row of the output file. The function includes the reference frame and unit conversion and calls the position_Serpent_to_STAR function as needed for the position values.

There are 12 columns in the Serpent 2 detector (Serpent manual):

  1. Value index
  2. Energy bin index
  3. Universe bin index
  4. Cell bin index
  5. Material bin index
  6. Lattice bin index
  7. Reaction bin index
  8. Z-mesh bin index
  9. Y-mesh bin index
  10. X-mesh bin index
  11. Mean value
  12. Relative statistical error

The serpentTools package allows for any part of the detector to be accessed. We are specifically interested with the .tallies portion, which is the mean value reported by Serpent. There are also $x$, $y$, and $z$ grid meshes that are printed in the detector file that the DetectorReader reads. These are accessed using .grids['X'], .grids['Y'], and .grids['Z'].

The detectors defined for passing the Serpent 2 data to STAR-CCM+ are only binned with respect to the $x$, $y$, and $z$ meshes. The serpentTools.Detector objects are reshaped such that the resultant array has an axis for each bin. Because the $x$ direction only has one bin, the shape of the detector array is ($z$-bins,$y$-bins).

min_temp_fix(array)

The cross section values used for helium-3 that have the KERMA data do not have data for helium-3 below 300K. This function fixes any temperature that is below 300K to be 300K. The input is an array of values in the form [position,density,temperature]. The function checks every value in the third column and makes sure it is at or above 300.0.

read_to_numpy(STAR_csv)

This function reads data from a .csv file written by STAR-CCM+ and converts it to a numpy array. It uses a pandas function to read only the specified columns, then converts the data frame to a numpy array. The required input is a STAR_csv object with the file name in the name attribute and the desired columns in the header attribute.

csv_to_ifc(STAR_csv,Serpent_ifc)

This function handles writing the data from a STAR-CCM+ .csv file to a Serpent 2 .ifc file. It calls both min_temp_fix and read_to_numpy within it. The inputs are the STAR_csv object for the STAR-CCM+ generated file being read and the Serpent_ifc object for the file being written to. While the incoming data has position, density, and temperature, the .ifc requires only the density and temperature data, so the function truncates the position values off when it writes to the .ifc file.

wait_for_file(file,wait)

There are many places in CONSTELATION where the script is waiting for a file to be created so it can be read, or be used as communication about the status of the simulation. This function takes in the file name and the maximum wait time in seconds and waits until the file exists. If the file does not exist at the end of the wait time, and error is raised.

com_check_digit(line,sig_notdigit)

For a reason that may never be determined, sometimes the contents of the com.out file are removed. This may be caused by the server, the simulation, both, or neither. This function checks that the contents of the file are able to be converted into an integer so that the simulation is not ended by an error trying to convert nothing to an integer. The inputs are the line that was read in from the com.out file and the false signal to give the output variable if the contents cannot be read properly. This false signal has its own place in the if loop that interprets the signal, which continues the simulation as if the signal was to resume current iteration, but it prints out the number provided in the sig_notdigit input.