Add HDF5 + XDMF output module

CMakeLists.txt

@@ -460,8 +460,10 @@ function(MFC_SETUP_TARGET)
         endif()
 
         if (ARGS_HDF5)
-            find_package(HDF5 REQUIRED)
+            find_package(HDF5 REQUIRED COMPONENTS Fortran)
             target_link_libraries(${a_target} PRIVATE HDF5::HDF5)
+            target_include_directories(${a_target} PRIVATE ${HDF5_Fortran_INCLUDE_DIRS})
+            target_link_libraries(${a_target} PRIVATE ${HDF5_Fortran_LIBRARIES})
         endif()
 
         if (ARGS_FFTW)

docs/documentation/case.md

@@ -638,7 +638,10 @@ To restart the simulation from $k$-th time step, see [Restarting Cases](running.
 
 The table lists formatted database output parameters. The parameters define variables that are outputted from simulation and file types and formats of data as well as options for post-processing.
 
-- `format` specifies the choice of the file format of data file outputted by MFC by an integer of 1 and 2. `format = 1` and `2` correspond to Silo-HDF5 format and binary format, respectively.
+- `format` specifies the choice of the file format of data file outputted by MFC by an integer of 1, 2, or 3:
+  - `format = 1`: Silo-HDF5 format (default). Works with VisIt and ParaView 5.11 and earlier.
+  - `format = 2`: Binary format. Lightweight output for custom analysis pipelines.
+  - `format = 3`: Native HDF5 + XDMF format. Recommended for ParaView 5.12 and later compatibility. Produces portable, self-describing output without external library dependencies beyond HDF5.
 
 - `precision` specifies the choice of the floating-point format of the data file outputted by MFC by an integer of 1 and 2. `precision = 1` and `2` correspond to single-precision and double-precision formats, respectively.
 

docs/documentation/testing.md

@@ -97,8 +97,21 @@ To test the post-processing code, append the `-a` or `--test-all` option:
 ./mfc.sh test -a -j 8
 ```
 
-This argument will re-run the test stack with `parallel_io='T'`, which generates silo_hdf5 files.
+This argument will re-run the test stack with `parallel_io='T'`, which generates output files depending on the `format` parameter.
 It will also turn most write parameters (`*_wrt`) on.
-Then, it searches through the silo files using `h5dump` to ensure that there are no `NaN`s or `Infinity`s.
-Although adding this option does not guarantee that accurate `.silo` files are generated, it does ensure that the post-process code does not fail or produce malformed data.
+
+The test framework validates post-process output for the following formats:
+
+| Format | Output Directory | Validation |
+|--------|------------------|------------|
+| `format=1` | `silo_hdf5/` | Checks HDF5 files for NaN/Infinity values |
+| `format=3` | `hdf5_xdmf/` | Checks HDF5 files for NaN/Infinity, then validates XDMF metadata structure |
+
+For Silo-HDF5 output (`format=1`), the test searches through the `.silo` files using `h5dump` to ensure that there are no `NaN`s or `Infinity`s.
+
+For native HDF5+XDMF output (`format=3`), the test:
+1. Validates all `.h5` files contain no `NaN` or `Infinity` values
+2. Validates `.xdmf` metadata files have correct XML structure with required elements (`Domain`, `Grid`, `Topology`, `Geometry`, `Attribute`)
+
+Although adding this option does not guarantee that fully accurate output files are generated, it does ensure that the post-process code does not fail or produce malformed data.
 

docs/documentation/visualization.md

@@ -1,5 +1,70 @@
 # Flow visualization
 
+MFC supports three output formats for post-processed simulation data. You may choose the format that best matches your visualization workflow:
+
+| Format | Parameter | Output Directory | Best For |
+|--------|-----------|------------------|----------|
+| Silo-HDF5 | `format = 1` | `silo_hdf5/` | VisIt, ParaView ≤ 5.11 |
+| Binary | `format = 2` | `D/` | Custom analysis scripts |
+| HDF5 + XDMF | `format = 3` | `hdf5_xdmf/` | ParaView 5.12-6.0+, portable workflows |
+
+## Output Format Details
+
+### Format 1: Silo-HDF5 (Default)
+
+Silo-HDF5 is the traditional MFC output format. It requires the Silo library and produces databases compatible with VisIt and older versions of ParaView.
+
+**Directory structure:**
+```
+case_dir/silo_hdf5/
+├── root/           # Index files for visualization
+│   └── *.silo      # One per time step
+└── p0/, p1/, ...   # Per-processor data files
+```
+
+**Pros:** Well-established, works with VisIt  
+**Cons:** Requires separate Silo library installation; ParaView 5.12-6.0.1 Silo reader is broken
+
+### Format 2: Binary
+
+Binary format writes raw floating-point data for custom analysis pipelines.
+
+**Directory structure:**
+```
+case_dir/D/
+└── *.dat           # Raw data files
+```
+
+**Pros:** Lightweight, easy to parse programmatically  
+**Cons:** No built-in visualization support; requires custom readers
+
+### Format 3: HDF5 + XDMF (Recommended for ParaView 5.12+)
+
+Native HDF5 output with XDMF metadata provides a portable, self-describing format that works reliably with modern visualization tools without external library dependencies beyond HDF5.
+
+**Directory structure:**
+```
+case_dir/hdf5_xdmf/
+├── root/                    # XDMF metadata files
+│   ├── timestep_0.xdmf      # Open these in ParaView
+│   ├── timestep_100.xdmf
+│   └── ...
+└── p0/, p1/, ...            # Per-processor HDF5 data
+    └── *.h5
+```
+
+**Pros:**
+- Works with ParaView 5.12+ and later without Silo dependencies
+- Fully portable across systems and software versions
+- Self-describing via XDMF metadata
+- Supports all MFC features: grid data, MHD fields, and Lagrangian bubble particles
+
+**Cons:** Larger file sizes than Silo-HDF5 (internal compression not implemented yet)
+
+**Usage:** Set `format = 3` in your post_process input file, then open the `.xdmf` files from `hdf5_xdmf/root/` in ParaView.
+
+---
+
 A post-processed database in Silo-HDF5 format can be visualized and analyzed using Paraview and VisIt.
 After the post-processing of simulation data (see section [Running](running.md)), a directory named `silo_hdf5` contains a silo-HDF5 database.
 Here, `silo_hdf5/` includes a directory named `root/` that contains index files for flow field data at each saved time step.
@@ -17,6 +82,27 @@ The figure below shows the iso-contour of the liquid void fraction (`alpha1`) in
 
 ![](../res/paraview.png)
 
+### Visualizing HDF5+XDMF with ParaView 5.12+
+
+ParaView 5.12-6.0.1 have issues with their ViSiT Silo plugin. Use `format = 3` (HDF5+XDMF) for post-processing:
+
+1. Run post-processing with `format = 3` in your input file
+2. Navigate to the `hdf5_xdmf/root/` directory
+3. Open the `.xdmf` file for your desired time step (e.g., `timestep_0.xdmf`), or open the master time series file for aggregated time steps
+4. ParaView will automatically load the grid and all available flow variables
+
+**Opening Multiple Time Steps:**  
+Each XDMF file represents a single time step. To animate through time:
+- Open the master time series XDMF file if generated
+- Or, use the XDMF reader's time support if your ParaView version supports it
+
+**Lagrangian Bubble Visualization:**  
+When simulations include Lagrangian bubbles (`lagrange_bubbles = .true.`), the XDMF file contains both:
+- **Grid data:** Eulerian field variables on the computational mesh
+- **Particle data:** Bubble positions and properties as point clouds
+
+In ParaView, the particle data appears as a separate "Particles" grid that can be visualized using the Glyph filter to render spheres at bubble locations.
+
 ### Visualizing data in cylindrical coordinates
 
 Visualizing data in cylindrical coordinates requires a coordinate transformation of the raw data in the database file.