trixi-framework · LasNikas · Mar 19, 2026 · Apr 16, 2026 · Apr 17, 2026 · Apr 20, 2026
diff --git a/NEWS.md b/NEWS.md
@@ -8,6 +8,86 @@ used in the Julia ecosystem. Notable changes will be documented in this file for
 
 ### API Changes
 
+- Clipping of negative pressure values in the `BoundaryModelDummyParticles` is now disabled
+  by default and can be enabled with the keyword argument `clip_negative_pressure=true` (#1143).
+- The example files are now loading sub-packages of OrdinaryDiffEq.jl instead of
+  OrdinaryDiffEq.jl itself. For example, `using OrdinaryDiffEqLowStorageRK` instead of
+  `using OrdinaryDiffEq` (#1154).
+- `DensityDiffusionAntuono` now only has the kwarg `delta` and no positional
+  arguments (#1142).
+- Return type of `vtk2trixi` changed to `NamedTuple` including an optional
+  `:initial_condition` field if `create_initial_condition=true` is passed (#959).
+- Public system constructors now use keyword arguments for configuration values.
+  This affects `WeaklyCompressibleSPHSystem`, `EntropicallyDampedSPHSystem`,
+  `ImplicitIncompressibleSPHSystem`, `TotalLagrangianSPHSystem`, `DEMSystem`,
+  and `BoundaryDEMSystem`.
+
+#### Migration Guide for Keyword-Based System Constructors
+
+The initial condition remains the only positional argument for the affected system
+constructors. All method, material, and configuration arguments that were previously
+passed positionally now need to be passed as keywords.
+Existing optional keyword arguments remain available. For example, the
+`density_calculator` keyword of `EntropicallyDampedSPHSystem` still defaults to
+`SummationDensity()`.
+
+Here are examples, where `ic` is the initial condition, `kernel` is the
+smoothing kernel, and `h` is the smoothing length:
+
+```julia
+WeaklyCompressibleSPHSystem(ic, density_calculator, state_equation, kernel, h;
+                            kwargs...)
+# becomes
+WeaklyCompressibleSPHSystem(ic; smoothing_kernel=kernel, smoothing_length=h,
+                            density_calculator, state_equation, kwargs...)
+
+EntropicallyDampedSPHSystem(ic, kernel, h, sound_speed; kwargs...)
+# becomes
+EntropicallyDampedSPHSystem(ic; smoothing_kernel=kernel, smoothing_length=h,
+                            sound_speed, kwargs...)
+
+ImplicitIncompressibleSPHSystem(ic, kernel, h, reference_density; kwargs...)
+# becomes
+ImplicitIncompressibleSPHSystem(ic; smoothing_kernel=kernel, smoothing_length=h,
+                                reference_density, kwargs...)
+
+TotalLagrangianSPHSystem(ic, kernel, h, young_modulus, poisson_ratio; kwargs...)
+# becomes
+TotalLagrangianSPHSystem(ic; smoothing_kernel=kernel, smoothing_length=h,
+                         young_modulus, poisson_ratio, kwargs...)
+
+DEMSystem(ic, contact_model; kwargs...)
+# becomes
+DEMSystem(ic; contact_model, kwargs...)
+
+BoundaryDEMSystem(ic, normal_stiffness)
+# becomes
+BoundaryDEMSystem(ic; normal_stiffness)
+```
+
+### Performance
+
+- Greatly improved GPU performance of WCSPH and TLSPH
+  (#1128, #1117, #1124, #1125, #1130, #1116, #1139, #1149).
+  See [#1131](https://github.com/trixi-framework/TrixiParticles.jl/issues/1131)
+  for a detailed breakdown including benchmark results.
+
+### Features
+
+- Added a `SortingCallback` that can be used to sort particles by their spatial location
+  to improve performance (#1044).
+
+### Important Bugfixes
+
+- Fixed a bug where `DensityDiffusionAntuono` could not be used together with open
+  boundaries and `BoundaryModelDynamicalPressureZhang` (#1043).
+- Fixed a bug where no-slip boundary conditions were not applied correctly when not using
+  `ViscosityAdami` (#1089).
+
+## Version 0.4.4
+
+### API Changes
+
 - Custom quantities called in the `PostprocessCallback` are now passed CPU arrays when
   the simulation is run on a GPU (#1065).
 
@@ -21,13 +101,6 @@ used in the Julia ecosystem. Notable changes will be documented in this file for
 - Added new validation case hydrostatic water column (#724).
 - Added Carreau–Yasuda non-Newtonian viscosity model (#1010).
 
-### Performance
-
-- Greatly improved GPU performance of WCSPH and TLSPH
-  (#1128, #1117, #1124, #1125, #1130, #1116, #1139, #1149).
-  See [#1131](https://github.com/trixi-framework/TrixiParticles.jl/issues/1131)
-  for a detailed breakdown including benchmark results.
-
 ### Important Bugfixes
 
 - Fixed the periodic array of cylinders example file (#975).
@@ -62,7 +135,6 @@ used in the Julia ecosystem. Notable changes will be documented in this file for
 
 - Added GPU and FP32 support for DEM (#979).
 
-
 ### Performance
 - Improved GPU performance with shifting up to a factor of 10x (#974, #993).
 

diff --git a/Project.toml b/Project.toml
@@ -1,6 +1,6 @@
 name = "TrixiParticles"
 uuid = "66699cd8-9c01-4e9d-a059-b96c86d16b3a"
-version = "0.4.5-dev"
+version = "0.5"
 authors = ["erik.faulhaber <44124897+efaulhaber@users.noreply.github.com>"]
 
 [deps]

diff --git a/README.md b/README.md
@@ -75,13 +75,14 @@ with Julia v1.10 and newer. We recommend using the latest stable release of Juli
 ### For users
 TrixiParticles.jl is a registered Julia package.
 You can install TrixiParticles.jl,
-[OrdinaryDiffEq.jl](https://github.com/SciML/OrdinaryDiffEq.jl) (used for time integration)
-and [Plots.jl](https://github.com/JuliaPlots/Plots.jl) by executing the following commands
+time integration sub-packages of [OrdinaryDiffEq.jl](https://github.com/SciML/OrdinaryDiffEq.jl)
+(for example `OrdinaryDiffEqLowStorageRK`), and [Plots.jl](https://github.com/JuliaPlots/Plots.jl)
+by executing the following commands
 in the Julia REPL:
 ```julia
 julia> using Pkg
 
-julia> Pkg.add(["TrixiParticles", "OrdinaryDiffEq", "Plots"])
+julia> Pkg.add(["TrixiParticles", "OrdinaryDiffEqLowStorageRK", "OrdinaryDiffEqSymplecticRK", "Plots"])
 ```
 
 ### For developers
@@ -92,7 +93,7 @@ git clone git@github.com:trixi-framework/TrixiParticles.jl.git
 cd TrixiParticles.jl
 mkdir run
 julia --project=run -e 'using Pkg; Pkg.develop(PackageSpec(path="."))' # Add TrixiParticles.jl to `run` project
-julia --project=run -e 'using Pkg; Pkg.add(["OrdinaryDiffEq", "Plots"])' # Add additional packages
+julia --project=run -e 'using Pkg; Pkg.add(["OrdinaryDiffEqLowStorageRK", "OrdinaryDiffEqSymplecticRK", "Plots"])' # Add additional packages
 ```
 
 If you installed TrixiParticles.jl this way, you always have to start Julia with the
@@ -133,7 +134,7 @@ You can find the documentation for the latest release
   [![arXiv:2506.21206](https://img.shields.io/badge/arXiv-2506.21206-yellow)](https://arxiv.org/abs/2506.21206)
   [![doi:10.1016/j.cpc.2025.109898](https://zenodo.org/badge/doi/10.1016/j.cpc.2025.109898.svg)](https://doi.org/10.1016/j.cpc.2025.109898)
   [![reproduce me!](https://img.shields.io/badge/reproduce-me!-brightgreen)](https://github.com/trixi-framework/paper-2025-particle-based_preprocessing)
-  
+
 ## Cite Us
 
 If you use TrixiParticles.jl in your own research or write a paper using results obtained
@@ -190,4 +191,3 @@ or [create an issue](https://github.com/trixi-framework/TrixiParticles.jl/issues
 </p>
 
 The project has benefited from funding from [hereon](https://www.hereon.de/), [HiRSE](https://www.helmholtz-hirse.de/), and through [ScienceServe](https://www.helmholtz.de/en/research/current-calls-for-applications/article/scienceserve-boosting-research-software-at-helmholtz/) for the MATRIX project.
-
diff --git a/docs/Project.toml b/docs/Project.toml
@@ -5,7 +5,7 @@ Documenter = "e30172f5-a6a5-5a46-863b-614d45cd2de4"
 DocumenterCitations = "daee34ce-89f3-4625-b898-19384cb65244"
 DocumenterMermaid = "a078cd44-4d9c-4618-b545-3ab9d77f9177"
 Literate = "98b081ad-f1c9-55d3-8b20-4c87d4299306"
-OrdinaryDiffEq = "1dea7af3-3e70-54e6-95c3-0bf5283fa5ed"
+OrdinaryDiffEqLowStorageRK = "b0944070-b475-4768-8dec-fb6eb410534d"
 Plots = "91a5bcdd-55d7-5caf-9e0b-520d859cae80"
 
 [compat]
@@ -15,5 +15,5 @@ Documenter = "1"
 DocumenterCitations = "1"
 DocumenterMermaid = "0.2"
 Literate = "2"
-OrdinaryDiffEq = "6"
+OrdinaryDiffEqLowStorageRK = "1"
 Plots = "1"
diff --git a/docs/literate/src/tut_custom_kernel.jl b/docs/literate/src/tut_custom_kernel.jl
@@ -4,11 +4,12 @@
 # of TrixiParticles.jl by custom implementations from within a simulation file,
 # without ever cloning the repository.
 
-# First, we import TrixiParticles.jl and
+# First, we import TrixiParticles.jl and the time integration sub-package
+# `OrdinaryDiffEqLowStorageRK` of
 # [OrdinaryDiffEq.jl](https://github.com/SciML/OrdinaryDiffEq.jl), which we will
 # use at the very end for the time integration.
 using TrixiParticles
-using OrdinaryDiffEq
+using OrdinaryDiffEqLowStorageRK
 
 # ## Load a simulation file
 
@@ -117,13 +118,12 @@ nothing # hide
 # `examples/fluid/dam_break_2d.jl` with our custom kernel and the corresponding
 # smoothing length.
 # ```@cast @__NAME__; width=100, height=50, delay=0, loop=true, loop_delay=5
-# trixi_include(@__MODULE__, joinpath(examples_dir(), "fluid", "dam_break_2d.jl"),
+# trixi_include(@__MODULE__, joinpath(examples_dir(), "fluid", "dam_break_2d.jl");
 #               smoothing_kernel=MyGaussianKernel(),
-#               smoothing_length=smoothing_length);
+#               smoothing_length);
 # ```
-trixi_include(joinpath(examples_dir(), "fluid", "dam_break_2d.jl"), #!md
-              smoothing_kernel=MyGaussianKernel(), #!md
-              smoothing_length=smoothing_length) #!md
+trixi_include(joinpath(examples_dir(), "fluid", "dam_break_2d.jl"); #!md
+              smoothing_kernel=MyGaussianKernel(), smoothing_length) #!md
 
 # See [Visualization](@ref) for how to visualize the final solution.
 # For the simplest visualization, we can use [Plots.jl](https://docs.juliaplots.org/stable/):