Skip to content

Fix IC in open boundaries and add simple_advection_2d.jl#852

Closed
LasNikas wants to merge 6 commits intotrixi-framework:mainfrom
LasNikas:simple-advection
Closed

Fix IC in open boundaries and add simple_advection_2d.jl#852
LasNikas wants to merge 6 commits intotrixi-framework:mainfrom
LasNikas:simple-advection

Conversation

@LasNikas
Copy link
Copy Markdown
Collaborator

@LasNikas LasNikas commented Jul 1, 2025

The additional example simple_advection_2d.jl is to quickly test (new) open boundaries and also to benchmark potential new open boundary implementations.

@LasNikas LasNikas requested a review from efaulhaber July 1, 2025 13:37
@LasNikas LasNikas added the bug Something isn't working label Jul 1, 2025
@LasNikas LasNikas self-assigned this Jul 1, 2025
@codecov
Copy link
Copy Markdown

codecov bot commented Jul 1, 2025

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 70.50%. Comparing base (e72c597) to head (7f74e25).
⚠️ Report is 9 commits behind head on main.

Additional details and impacted files
@@            Coverage Diff             @@
##             main     #852      +/-   ##
==========================================
+ Coverage   70.47%   70.50%   +0.03%     
==========================================
  Files         106      106              
  Lines        6893     6897       +4     
==========================================
+ Hits         4858     4863       +5     
+ Misses       2035     2034       -1     
Flag Coverage Δ
unit 70.50% <100.00%> (+0.03%) ⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:
  • ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

@LasNikas
Copy link
Copy Markdown
Collaborator Author

LasNikas commented Jul 1, 2025

/run-gpu-tests

Comment thread examples/fluid/simple_advection_2d.jl
Comment on lines +3 to +144

# ==========================================================================================
# ==== Resolution
particle_spacing = 0.01

# Make sure that the kernel support of fluid particles at a boundary is always fully sampled
boundary_layers = 4

# Make sure that the kernel support of fluid particles at an open boundary is always
# fully sampled.
# Note: Due to the dynamics at the inlets and outlets of open boundaries,
# it is recommended to use `open_boundary_layers > boundary_layers`
open_boundary_layers = 6

# ==========================================================================================
# ==== Experiment Setup
tspan = (0.0, 2.0)

# Boundary geometry and initial fluid particle positions
domain_size = (1.0, 0.4)

flow_direction = [1.0, 0.0]
reynolds_number = 100
const prescribed_velocity = 1.0

boundary_size = (domain_size[1] + 2 * particle_spacing * open_boundary_layers,
domain_size[2])

fluid_density = 1000.0

sound_speed = 10 * prescribed_velocity

state_equation = nothing

pipe = RectangularTank(particle_spacing, domain_size, boundary_size, fluid_density,
velocity=[prescribed_velocity, 0.0],
n_layers=boundary_layers, faces=(false, false, true, true))

# Shift pipe walls in negative x-direction for the inflow
pipe.boundary.coordinates[1, :] .-= particle_spacing * open_boundary_layers

NDIMS = ndims(pipe.fluid)

n_buffer_particles = pipe.n_particles_per_dimension[2]^(NDIMS - 1)

# ==========================================================================================
# ==== Fluid
wcsph = true

smoothing_length = 1.2 * particle_spacing
smoothing_kernel = SchoenbergQuinticSplineKernel{NDIMS}()

fluid_density_calculator = ContinuityDensity()

kinematic_viscosity = prescribed_velocity * domain_size[2] / reynolds_number

viscosity = ViscosityAdami(nu=kinematic_viscosity)

# Alternatively the WCSPH scheme can be used
if wcsph
state_equation = StateEquationCole(; sound_speed, reference_density=fluid_density,
exponent=1)

density_diffusion = DensityDiffusionMolteniColagrossi(delta=0.1)

fluid_system = WeaklyCompressibleSPHSystem(pipe.fluid, fluid_density_calculator,
state_equation, smoothing_kernel,
density_diffusion=density_diffusion,
smoothing_length, viscosity=viscosity,
buffer_size=n_buffer_particles)
else
fluid_system = EntropicallyDampedSPHSystem(pipe.fluid, smoothing_kernel,
smoothing_length,
sound_speed, viscosity=viscosity,
density_calculator=fluid_density_calculator,
buffer_size=n_buffer_particles)
end

# ==========================================================================================
# ==== Open Boundary

velocity_function2d(pos, t) = SVector(prescribed_velocity, 0.0)

open_boundary_model = BoundaryModelTafuni()

boundary_type_in = InFlow()
plane_in = ([0.0, 0.0], [0.0, domain_size[2]])
inflow = BoundaryZone(; plane=plane_in, plane_normal=flow_direction, open_boundary_layers,
density=fluid_density, particle_spacing,
boundary_type=boundary_type_in)

open_boundary_in = OpenBoundarySPHSystem(inflow; fluid_system,
boundary_model=open_boundary_model,
buffer_size=n_buffer_particles,
reference_velocity=velocity_function2d)

boundary_type_out = OutFlow()
plane_out = ([domain_size[1], 0.0], [domain_size[1], domain_size[2]])
outflow = BoundaryZone(; plane=plane_out, plane_normal=(-flow_direction),
open_boundary_layers, density=fluid_density, particle_spacing,
boundary_type=boundary_type_out)
open_boundary_out = OpenBoundarySPHSystem(outflow; fluid_system,
boundary_model=open_boundary_model,
buffer_size=n_buffer_particles,
reference_velocity=velocity_function2d)

# ==========================================================================================
# ==== Boundary
boundary_model = BoundaryModelDummyParticles(pipe.boundary.density, pipe.boundary.mass,
AdamiPressureExtrapolation(),
state_equation=state_equation,
smoothing_kernel, smoothing_length)

boundary_system = BoundarySPHSystem(pipe.boundary, boundary_model)

# ==========================================================================================
# ==== Simulation
min_corner = minimum(pipe.boundary.coordinates .- particle_spacing, dims=2)
max_corner = maximum(pipe.boundary.coordinates .+ particle_spacing, dims=2)

nhs = GridNeighborhoodSearch{NDIMS}(; cell_list=FullGridCellList(; min_corner, max_corner),
update_strategy=ParallelUpdate())

semi = Semidiscretization(fluid_system, open_boundary_in, open_boundary_out,
boundary_system, neighborhood_search=nhs,
parallelization_backend=PolyesterBackend())

ode = semidiscretize(semi, tspan)

info_callback = InfoCallback(interval=100)
saving_callback = SolutionSavingCallback(dt=0.02, prefix="")

particle_shifting = nothing

extra_callback = nothing

callbacks = CallbackSet(info_callback, saving_callback, UpdateCallback(),
particle_shifting, extra_callback)

sol = solve(ode, RDPK3SpFSAL35(),
abstol=1e-5, # Default abstol is 1e-6 (may need to be tuned to prevent boundary penetration)
reltol=1e-3, # Default reltol is 1e-3 (may need to be tuned to prevent boundary penetration)
dtmax=1e-2, # Limit stepsize to prevent crashing
save_everystep=false, callback=callbacks);
Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Isn't this just another pipe? Can't you base it on pipe_flow_2d.jl?

Copy link
Copy Markdown
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I want a clean, simple example for the open boundaries. I've run into issues with trixi_include so many times and ended up with unnecessarily frustrating bugs: you simply forget to enable or disable something. For complexity reasons, it's also better for beginners to have a simple example.

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this is more a problem with the structure of our setup files. But anyway I think it would be better to further simplify pipe_flow_2d and base your setup on that.

Comment on lines +1 to +2
# This is a simple advection example with open boundaries. The setup is similar to
# `periodic_channel_2d.jl`, but uses open boundaries instead of periodic ones.
Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Not the same style.

@LasNikas LasNikas mentioned this pull request Jul 28, 2025
4 tasks
LasNikas pushed a commit to LasNikas/TrixiParticles.jl that referenced this pull request Jul 29, 2025
@LasNikas
Copy link
Copy Markdown
Collaborator Author

superseded by #866

@LasNikas LasNikas closed this Jul 29, 2025
efaulhaber added a commit that referenced this pull request Sep 8, 2025
* swap extrapolation

* swap density pressure

* adapt docs

* rm example

* fix missing nhs update

* implement different mirror methods

* add docs

* fix tests

* fix typos

* add docs

* add tests

* apply formatter

* add `average_velocity!` for `BoundaryModelLastiwka`

* fix

* add interpolation functions

* restructure again

* fix gpu bugs

* remove unnecessary argument

* fix gpu again

* fix gpu

* implement suggestions

* fix

* add comments

* implement suggestions

* first prototype: NOT VALIDATED YET

* fix avg vel in LastiwkaModel

* fix avg in mirroring

* fix tests

* adapt example file

* fix allocations

* fix gpu

* first GPU working prototype

* improve API

* fix tests

* rm export

* fix

* fix type stability

* implement suggestions

* rm ref

* format

* Rename 'tlsph' to 'place_on_shell' (#814)

* rename

* format

* forgot some

* format

* naming

* forgot some more

* fix test

* incorporate review comments

* format

---------

Co-authored-by: Niklas Neher <73897120+LasNikas@users.noreply.github.com>

* fix merge bugs

* supersede #852

* supersede #850

* fix

* rename boundary models

* fix?

* fix tests

* Rename 'tlsph' to 'place_on_shell' (#814)

* rename

* format

* forgot some

* format

* naming

* forgot some more

* fix test

* incorporate review comments

* format

---------

Co-authored-by: Niklas Neher <73897120+LasNikas@users.noreply.github.com>

* fix characteristics

* revise pipe flow example

* fix include bug

* Rename 'tlsph' to 'place_on_shell' (#814)

* rename

* format

* forgot some

* format

* naming

* forgot some more

* fix test

* incorporate review comments

* format

---------

Co-authored-by: Niklas Neher <73897120+LasNikas@users.noreply.github.com>

* Combine PST and TVF into a unified framework (#884)

* Combine PST and TVF into a unified framework

* Require update callback for PST

* Fix WCSPH

* Update PST only in callback

* Fix EDAC

* Update docs

* Fix alle example files

* Fix tests

* Fix periodic channel

* Fix docs

* Update news

* Fix tests

* Fix example file

* add doc strings

* implement suggestions

* remove duplicated system

* rm unnecessary variable

* apply formatter

* set pressure for WCSPH

* setter for open boundary

* rename reference values

* fix gpu tests

* fix sign

* implement suggestions

* fix unit tests

* revise doc strings

* add NEWS entry

* update NEWS

---------

Co-authored-by: LasNikas <niklas.nehe@web.de>
Co-authored-by: Sven Berger <berger.sven@gmail.com>
Co-authored-by: Erik Faulhaber <44124897+efaulhaber@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

bug Something isn't working

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants