FlexiblePatternGenerator with boundary_type=CIRCLE Error

[willschlueter]

I am using FlexiblePatternGenerator to create a mesh that has fuel pins arranged in a large hexagonal pattern but with each vertex pin removed. For some reason, no matter how large I make the boundary_size parameter, I keep getting the same error that I believe implies my boundary is too small to contain all of the pins despite having an arbitrarily large boundary size. When I run the mesh with all the same parameters except with boundary_type set to HEXAGON, it works fine. Something about the circle seems to be going wrong.

Code:

Error:

[GiudGiud]

Hello

please avoid screenshots for text and use triple backquotes to format it instead
I cant copy paste your input rn

[willschlueter]

Apologies--did not realize you'd want to copy/paste the input yourself and thought it would be easier to view as a png. I am no longer in the office which has the computer these files are stored on so I'll have to get back to you. Thanks anyway.

[willschlueter]

@GiudGiud

n_layers = 20    # number of axial extrusion layers; for the converged case,
                  # we set this to 20 to get a finer mesh
height = '${units 35.56 cm -> m}' # SNAP height
[GlobalParams]
  quad_center_elements = true
[]
[Mesh]
  [fuel_pin]
    type = PolygonConcentricCircleMeshGenerator
    num_sides = 6
    polygon_size = ${fparse 0.01448         / 2.0}
    ring_radii = '0.0067564 0.00681 0.00685 0.007134' 
    ring_intervals = '5 2 2 4'
    num_sectors_per_side = '4 4 4 4 4 4'
    ring_block_ids = '10 11 12 13'
    ring_block_names = 'Fuel Ceramic GAPHE Clad'
    background_block_ids = '1'
    background_block_names = 'coolant'
    background_intervals = 2
  []
   [core]
    type = FlexiblePatternGenerator
    inputs = fuel_pin
    hex_patterns = '1 0 0 0 0 0 0 0 1;
                            0 0 0 0 0 0 0 0 0 0;
                           0 0 0 0 0 0 0 0 0 0 0;
                         0 0 0 0 0 0 0 0 0 0 0 0;
                       0 0 0 0 0 0 0 0 0 0 0 0 0;
                      0 0 0 0 0 0 0 0 0 0 0 0 0 0;
                     0 0 0 0 0 0 0 0 0 0 0 0 0 0 0;
                    0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0;
                   1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1;
                     0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0;
                      0 0 0 0 0 0 0 0 0 0 0 0 0 0 0;
                       0 0 0 0 0 0 0 0 0 0 0 0 0 0;
                        0 0 0 0 0 0 0 0 0 0 0 0 0;
                          0 0 0 0 0 0 0 0 0 0 0 0;
                            0 0 0 0 0 0 0 0 0 0 0;
                             0 0 0 0 0 0 0 0 0 0;
                               1 0 0 0 0 0 0 0 1'
    hex_pitches = 0.014481
    boundary_type = CIRCLE
    boundary_sectors = 10
    boundary_size = 1
  []
  [extrude]
    type = AdvancedExtruderGenerator
    input = core
    heights = ${height}
    num_layers = ${n_layers}
    direction = '0 0 1'
  []
[]
# The following content is adding postprocessor(s) to check sideset areas.
# The reactor module is unfortunately quite brittle in its assignment of sideset
# IDs, so we want to be extra sure that any changes to sideset numbering are detected
# in our test suite.
[Problem]
  type = FEProblem
  solve = false
[]
[Postprocessors]
  [area_walls]
    type = AreaPostprocessor
    boundary = 'fluid_solid_interface'
  []
[]
[Executioner]
  type = Steady
[]
[Outputs]
  csv = false
  exodus = true
[]

[GiudGiud]

@willschlueter can you increase the number of sectors?
The number of sectors helps make a balanced circular mesh. I used 128 to get a result.

I also think a boundary_size of 1 is excessive here? It seems like the circle is huge compared to the assembly ?

[willschlueter]

Looks like that worked. I believe I tried boundary_sectors = 100 earlier...any reason 128 is better other than being slightly larger? Also, I had boundary_size of 1 simply because I though the error was related to the boundary being too small and 1 was just an arbitrary large number.

If my end goal is to be shrinking this outer boundary as small as I can (closest to touching the outer pins as possible) should I keep increasing the number of sectors to avoid this error?

[willschlueter]

Actually, it looks like 128 sectors works for the actual radius I'm looking for. Thank you so much!!