.. _about:
=====
About
=====
MFiX
----
MFiX is an open-source multiphase flow solver and is free to download
and use. A one-time, no-cost registration is required prior to
downloading the source code. To register, go to
https://mfix.netl.doe.gov/ and click on the "Register" button in the
upper right corner. Once you have read the notice, you can submit your
application by clicking on "REGISTER." After your application has been
reviewed and accepted, you will receive an email notification and
instructions on how to download the code. Please allow for 2-3 business
days for your registration to be processed.
Development state of MFiX models
--------------------------------
MFiX provides a suite of models that treat the carrier phase (typically
the gas phase) and disperse phase (typically the solids phase)
differently. Their current state of development is summarized in the
tables below.
Symbol description for the following tables:
.. |c| replace:: :math:`\circ`
.. |b| replace:: :math:`\bullet`
.. |s| replace:: :math:`\square`
+----------+----------------------------------------------------+
| Symbol |Description |
| | |
+==========+====================================================+
| |b| |Implemented and fully tested |
+----------+----------------------------------------------------+
| |c| |Implemented with limited testing |
+----------+----------------------------------------------------+
| |s| |Not tested or status unknown |
+----------+----------------------------------------------------+
| †|Models not extended to DMP-parallel are only |
| |available for serial runs |
+----------+----------------------------------------------------+
| ‡ |Models not extended to SMP-parallel are available |
| |for SMP runs but do not scale with thread count |
+----------+----------------------------------------------------+
.. _dvlpt-TFM:
MFiX-TFM (Two-fluid model)
~~~~~~~~~~~~~~~~~~~~~~~~~~
**MFiX-TFM (Two-fluid model)** is an Eulerian-Eulerian model, which
supports a broad range of capabilities for dense, reacting, multiphase
flows by representing the fluid and solids as interpenetrating continua.
This is the most mature MFiX model and is capable of modeling multiphase
reactors ranging in size from benchtop to industry-scale. Approximation
of the solid phase as a continuum typically allows for faster simulation
time than Lagrangian techniques; however, it also introduces the need
for accurate mathematical models to capture realistic solids phase
behavior. This includes transport properties, heterogeneous reaction
kinetics, and constitutive relations for interaction between fluid and
solid phases, e.g., solids phase drag and interphase heat transfer.
+----------------------+----------+--------+---------+
| Feature | Serial | †DMP | ‡SMP |
+======================+==========+========+=========+
| Momentum Equations | |b| | |b| | |b| |
+----------------------+----------+--------+---------+
| Energy Equations | |b| | |b| | |b| |
+----------------------+----------+--------+---------+
| Species Equations | |b| | |b| | |b| |
+----------------------+----------+--------+---------+
| Chemical Reactions | |b| | |b| | |
+----------------------+----------+--------+---------+
| Cartesian cut-cell | |b| | |b| | **â–¡** |
+----------------------+----------+--------+---------+
.. _dvlpt-DEM:
MFiX-DEM (Discrete element model)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**MFiX-DEM (Discrete element model)** is an Eulerian-Lagrangian model
that treats the fluid phase as a continuum and models the individual
particles of the solid phase. This is a relatively new variation on
MFiX. While the treatment of individual particles can provide higher
fidelity over a broad range of flow regimes (from dilute to packed), it
becomes very challenging (in terms of computational resources)
when dealing with very large numbers of particles
for large-scale simulations. These large-scale applications will require
high performance computing (HPC) resources and large amounts of computer
time. Code optimization and speed up are critical research fronts to
support industrial scale applications.
+----------------------+----------+--------+--------+
| Feature | Serial | †DMP | ‡SMP |
+======================+==========+========+========+
| Momentum Equations | |b| | |b| | |b| |
+----------------------+----------+--------+--------+
| Energy Equations | |b| | |b| | |
+----------------------+----------+--------+--------+
| Species Equations | |b| | |b| | |
+----------------------+----------+--------+--------+
| Chemical Reactions | |b| | |b| | |
+----------------------+----------+--------+--------+
| Cartesian cut-cell | |c| | |c| | |
+----------------------+----------+--------+--------+
.. _dvlpt-CGP:
MFiX-CGP (Coarse-grained particle)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**MFiX-CGP (Coarse-grained particle)** is an Eulerian-Lagrangian model
similar to DEM, where particles are grouped into larger coarse-grained
particles (CGP). The collisions between CGPs are resolved the same way
as in DEM. This modeling approach is more affordable than DEM since fewer
CGPs need to be tracked. This modeling approach falls between DEM and PIC
in terms of computational resources.
+----------------------+----------+--------+--------+
| Feature | Serial | †DMP | ‡SMP |
+======================+==========+========+========+
| Momentum Equations | |b| | |b| | |b| |
+----------------------+----------+--------+--------+
| Energy Equations | |b| | |b| | |
+----------------------+----------+--------+--------+
| Species Equations | |b| | |b| | |
+----------------------+----------+--------+--------+
| Chemical Reactions | |b| | |b| | |
+----------------------+----------+--------+--------+
| Cartesian cut-cell | |c| | |c| | |
+----------------------+----------+--------+--------+
.. _dvlpt-SQP:
MFiX-SQP (Superquadric particle)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**MFiX-SQP (Superquadric particle)** is an Eulerian-Lagrangian model
similar to DEM, where particles have non-spherical shapes. The particle
shape is defined with a superquadric surface. This allows representing a
wide variety of shapes with 5 parameters. This modeling approach is
much more computationlly intensive than MFiX-DEM due to the complexity
of contact detection for non-spherical shapes.
+----------------------+----------+--------+--------+
| Feature | Serial | †DMP | ‡SMP |
+======================+==========+========+========+
| Momentum Equations | |b| | |b| | |b| |
+----------------------+----------+--------+--------+
| Energy Equations | |b| | |b| | |
+----------------------+----------+--------+--------+
| Species Equations | |b| | |b| | |
+----------------------+----------+--------+--------+
| Chemical Reactions | |b| | |b| | |
+----------------------+----------+--------+--------+
| Cartesian cut-cell | |c| | |c| | |
+----------------------+----------+--------+--------+
.. _dvlpt-GSP:
MFiX-GSP (Glued-sphere particle)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**MFiX-GSP (Glued-sphere particle)** is an Eulerian-Lagrangian model
similar to DEM, where particles have non-spherical shapes. The particle
is an assembly of smaller spheres that are glued together.
This modeling approach is more computationlly intensive than MFiX-DEM due
to the larger particle count, but faster that MFiX-SQP model because the
contact detection is still based on sphere-sphere contact.
+----------------------+----------+--------+--------+
| Feature | Serial | †DMP | ‡SMP |
+======================+==========+========+========+
| Momentum Equations | |b| | |b| | |b| |
+----------------------+----------+--------+--------+
| Energy Equations | |b| | |b| | |
+----------------------+----------+--------+--------+
| Species Equations | |b| | |b| | |
+----------------------+----------+--------+--------+
| Chemical Reactions | |b| | |b| | |
+----------------------+----------+--------+--------+
| Cartesian cut-cell | |c| | |c| | |
+----------------------+----------+--------+--------+
.. _dvlpt-PIC:
MFiX-PIC (Multiphase particle-in-cell)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
**MFiX-PIC (Multiphase particle-in-cell)** is another
Eulerian-Lagrangian model that represents the fluid as a continuum while
using "parcels" to represent groups of real particles with similar
physical characteristics. The MFiX-PIC approach offers reduced
computational cost over MFiX-DEM as there are typically fewer parcels to
track and parcel collisions are not resolved. However, the added
modeling approximations influence the overall accuracy of the method.
Development, validation, and optimization of modeling approximations are
critical research fronts.
+----------------------+----------+--------+--------+
| Feature | Serial | †DMP | ‡SMP |
+======================+==========+========+========+
| Momentum Equations | |b| | |b| | |c| |
+----------------------+----------+--------+--------+
| Energy Equations | |b| | |b| | |
+----------------------+----------+--------+--------+
| Species Equations | |b| | |b| | |
+----------------------+----------+--------+--------+
| Chemical Reactions | |b| | |b| | |
+----------------------+----------+--------+--------+
| Cartesian cut-cell | |b| | |b| | |
+----------------------+----------+--------+--------+
GUI
---
The MFiX GUI is a front-end tool that allows users to quickly set-up MFiX
models, run the developed models, and provide post-processing tools
with the goal of making MFiX easy to use.
The GUI is written in Python using the cross-platform Qt framework. The Anaconda/
Miniforge platform is used for MFiX dependencies, which allows us to support MFiX on
Linux, Windows, and Mac. The `Visualization Toolkit (VTK)
<https://vtk.org/>`__ is used to visualize and manipulate the input geometry.
.. _support-forum:
Support forum
-------------
When your subscription to MFiX is accepted, you are automatically added to the
MFiX forum, where important announcements about MFiX are shared with
the MFiX community. If you are already an MFiX member, use you current credentials
to participate in the forum.
The Forum is a discussion platform for MFiX, Nodeworks and Tracker software.
Developers and users can post topics (discussion threads) to various categories
for each software.
Users are encouraged to participate in the discussion and share their work
(input files, udf's, images or small animations showing simulation results)
to benefit the entire MFiX community.
The support forum is located at https://mfix.netl.doe.gov/forum. Click
on the topics of interest to you.
Support forum etiquette:
1) Do not post offensive or inappropriate material. Stay courteous and respectful at all times.
2) Please allow sufficient time (say 2 to 3 business days) for MFiX developers and
users to reply before posting unanswered questions again.
3) Post questions in the appropriate forum category. Do not send requests or emails directly
to MFiX developers or other users unless a prior arrangement has been made.
This ensures questions and answers are archived, and it allows the entire MFiX community
to engage. Additionally, follow-up questions should occur in the same thread.
4) Do not ask for a copy of a reference, e.g., a journal article. Do not post copyrighted material.
5) Prior to posting questions regarding MFiX installation or compilation issues,
please search the forum to see if your question has already been answered.
6) When posting a question to the forum, provide a complete description of the issue you encountered.
It may be useful to provide the following information in your post:
a. MFiX version you are trying to install or run
b. Some details on your operating system environment (for Linux: copy and
paste the output of the command "uname –a", Linux distribution name and
version also)
c. Your compiler name and version number (e.g. ifort -v will give the version
number for Intel Fortran compiler)
d. Output for your $PATH environment (in csh type echo $PATH)
e. Your MPI library name and version number (if compilations problem with
DMP mode encountered but make sure you can compile and run a simple
hello world type MPI program with your current installation) Also please
provide hardware details such as number of cores per socket in your
system (or send the output for "cat /proc/cpuinfo" and how many cores
you are trying to utilize.
.. note::
Common reasons you may not receive an answer to your request
1. Your question has already been answered and is available in the archive.
2. You did not provide sufficient description of your problem
(saying "It doesn’t work" is not useful).
3. Your question is outside the scope of the MFiX forum.
User contribution
-----------------
If you wish to contribute to the development of MFiX, please contact the MFiX
team at admin@mfix.netl.doe.gov. We are looking for simulation results (figures,
animations, input files, user-defined subroutines), and new models that could
benefit the entire MFiX community. If you have written or know any publication
that uses MFiX, please let us know and we will post the citation on the website
(https://mfix.netl.doe.gov/publications). Proper credit
will be given to all contributors.