Working with VTK Files

The Visualization Toolkit (VTK) establishes a data format which is designed for use in high performance computing (HPC) applications. Detailed documentation of data format can be found on the website at VTK File Formats. This page provides a quick start guide to working with VTK files generated by FDS.

Compilation with HDF5 support for VTKHDF output

This section describes the process to compile FDS with HDF5 support with the Ubuntu linux distribution.

Intel compilers (impi_intel_linux*)

Install prerequisites
sudo apt install zlib1g-dev cmake libhwloc-dev
Download HDF5 library and extract to build directory
mkdir HDF5 && cd HDF5

wget https://github.com/HDFGroup/hdf5/releases/download/hdf5_1.14.4.3/hdf5-1.14.4-3.tar.gz

wget https://github.com/madler/zlib/releases/download/v1.3/zlib-1.3.tar.gz

wget https://github.com/zlib-ng/zlib-ng/archive/refs/tags/2.1.6.tar.gz

wget https://github.com/MathisRosenhauer/libaec/releases/download/v1.1.3/libaec-1.1.3.tar.gz

tar -zvxf hdf5-1.14.4-3.tar.gz
Navigate to build directory
mkdir hdf5-1.14.4-3/build && cd hdf5-1.14.4-3/build
Configure the HDF5 library with autotools
../configure --enable-fortran --enable-parallel CXX="$(which mpiicpc) -cc=$(which icpx)" CC="$(which mpiicc) -cc=$(which icx)" FC="$(which mpiifort) -fc=$(which ifort)" LDFLAGS="-L/opt/intel/oneapi/mpi/2021.11/lib -L/opt/intel/oneapi/mpi/2021.11/lib/release/" --prefix=/path/to/install/directory
Build the library
make -j 8
install library
make install
set HDF5_HOME directory in your .bashrc
export HDF5_HOME=/path/to/install/directory
add runtime library directory to LD_LIBRARY_PATH in your .bashrc
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$HDF5_HOME/lib
build FDS with supported target (impi_intel_linux*) target. The build should include the -DWITH_HDF5 flag similar to
mpiifort -c -m64 -fc=ifort -O2 -ipo -no-wrap-margin -fpp -DGITHASH_PP=\"FDS-6.9.1-1053-ga2ec224-vtkhdf\" -DGITDATE_PP=\""Sun Oct 13 17:33:39 2024 -0400\"" -DBUILDDATE_PP=\""Oct 14, 2024 11:56:22\"" -DWITH_MKL -I"/data/opt/intel/oneapi/mkl/2024.0/include" -DWITH_HDF5 -I"/raid/jhodges/HDF5/hdf5-1.14.4-3/build/install/mod/static" -I"/raid/jhodges/HDF5/hdf5-1.14.4-3/build/install/include" -L"/raid/jhodges/HDF5/hdf5-1.14.4-3/build/install/lib" -lhdf5_fortran -DUSE_IFPORT -diag-disable=10448 ../../Source/vtkf.f90

Openmpi compilers (ompi_gnu_linux*)

Install prerequisites
sudo apt install zlib1g-dev cmake libhwloc-dev
Download HDF5 library and extract to build directory
mkdir HDF5 && cd HDF5

wget https://github.com/HDFGroup/hdf5/releases/download/hdf5_1.14.4.3/hdf5-1.14.4-3.tar.gz

wget https://github.com/madler/zlib/releases/download/v1.3/zlib-1.3.tar.gz

wget https://github.com/zlib-ng/zlib-ng/archive/refs/tags/2.1.6.tar.gz

wget https://github.com/MathisRosenhauer/libaec/releases/download/v1.1.3/libaec-1.1.3.tar.gz

tar -zvxf hdf5-1.14.4-3.tar.gz
Navigate to build directory
mkdir hdf5-1.14.4-3/build && cd hdf5-1.14.4-3/build
Configure the HDF5 library with autotools
../configure --enable-fortran --enable-parallel CXX="$(which mpicxx)" CC="$(which mpicc)" FC="$(which mpifort)" LDFLAGS="-L$OMPI_HOME/lib" --prefix=/path/to/install/directory
Build the library
make -j 8
install library
make install
set HDF5_HOME directory in your .bashrc
export HDF5_HOME=/path/to/install/directory
add runtime library directory to LD_LIBRARY_PATH in your .bashrc
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$HDF5_HOME/lib
build FDS with supported target (ompi_gnu_linux*) target. The build should include the -DWITH_HDF5 flag similar to
mpifort -c -m64 -O2 -std=f2018 -frecursive -ffpe-summary=none -fall-intrinsics -cpp -DGITHASH_PP=\"FDS-6.9.1-1132-g766c7f6-vtkhdf\" -DGITDATE_PP=\""Wed Oct 16 20:09:09 2024 -0400\"" -DBUILDDATE_PP=\""Oct 17, 2024 11:10:52\"" -DWITH_HDF5 -I"/raid/jhodges/HDF5_ompi/hdf5-1.14.4-3/build/hdf5/mod/static" -I"/raid/jhodges/HDF5_ompi/hdf5-1.14.4-3/build/hdf5/include" -L"/raid/jhodges/HDF5_ompi/hdf5-1.14.4-3/build/hdf5/lib" -lhdf5_fortran ../../Source/vtkf.f90

Quick-start

This section will demonstrate a basic example to generate VTK data with FDS and visualize the results in Paraview.

Open the storage example

Installed FDS: Examples/VTK/storage.fds
Source FDS: Verification/VTK/storage.fds

Edit the input file

Add "WRITE_VTK=T" to the &DUMP namelist. This tells FDS to generate VTK data.
Add "RESULTS_DIR='storage_results'" to the &DUMP namelist. This tells FDS to locate all the VTK and binary data in the subdirectory, "storage_results".
The namelist should now be:
&DUMP NFRAMES=100, DT_CTRL=1., DT_DEVC=10., DT_HRR=10., SIG_FIGS=4, WRITE_VTK=T, RESULTS_DIR='storage_results' /

Run the example with fds

Navigate to the directory with the input file
Run fds with the command: mpiexec -np 8 fds storage.fds

Open Paraview and import the state file

Open Paraview
Click File->Load State
Select the "storage_paraview.py" file and click ok

Visualize fire and smoke results

Manipulate the viewport by using the preset views or clicking in the window and dragging the mouse.
Click play to start the animation of the fire, smoke, and sprinkler droplets.
After sprinklers activate, the view should be similar to the following.

Visualize slice data

Hide the fire, smoke, and particles by clicking the open "eye" icon next to these two datasets in the pipeline browser
Show the slice at a z-axis height of 7.4m by clicking the closed "eye" icon in the pipeline browser.
Click on the z-axis slice in the pipeline browser and the properties window will appear.
Click the drop down icon under "Coloring" and change it to "TEMPERATURE".
Use a similar approach to activate the "3D Slice Extraction" in the pipeline browser. Click on "3D Slice Extraction" in the pipeline browser and a plane visualization will appear in the viewport which can be used to translate and rotate extracted plane from the 3D data.
Move the plane to Origin: -3, 0, 4.5 with Normal 1, 0, 0 and click apply.
Click on the drop-down icon under "Coloring" and change it to "TEMPERATURE".
Deselect the "3D Slice Extraction" from the pipeline browser by clicking any of the other datasets or the "builtin" root tree at the top. This will hide the plane which was used to move the slice extraction plane.
Click the run icon to see the animation of slice data.

Visualize boundary data

Hide the slice data from the previous step by clicking the "eye" icon for each of the slice datasets in the pipeline browser.
Click the "eye" icon next to the boundary data and select coloring by "HRRPUA". You may see overlap between the geometry and boundary data. This can be removed by hiding the geometry data.

Frequently Asked Questions

What is an attribute not found error in Paraview and how do I fix it?

The python state file interface in Paraview has syntax changes over different releases of Paraview. This error typically occurs when the configuration file generated by FDS tries to modify a parameter that does not exist in the version of Paraview used. If this occurs you can comment out the line in the Paraview state file and reload it. Please report these observations in the issue tracker with the tag "Paraview compatibility" and provide your FDS input file and Paraview version.

How do I use Paraview on a remote system?

Using Paraview on a remote system depends on how the server is configured. Oak Ridge provides information a typical configuration OLCF Paraview Guide. See the user guides for existing HPC configurations and discuss with the system administrator for configuration of new systems.

How do I output human-readable / ASCII VTK data with FDS?

Add the flag VTK_BINARY=F to the &MISC to change from binary to ASCII output in VTK files.

What processing operations exist in Paraview?

See the Paraview User's Guide.

How do I do [processing operation] in Paraview?