Burn

Burn is a model for simulating the burning of materials. It uses tools from the FEniCS project to automate mesh generation, discretization, and solution of the problem. The goal of developing Burn is to enable predictions of burning rate and flame spread across common flammable items. Such a capability will enable materials scientists and engineers to design safer products.

Overview

The physical model underlying Burn is expressed as a system of partial differential equations describing conservation of energy and component mass fraction. Diffusional heat and mass transfer are allowed, and state dependent thermophysical properties may be specified by the user. The current version of Burn assumes constant density, and so it is most appropriate for non-charring and non-intumescent materials. Both Dirichlet and Neumann-type boundary conditions may be specified by the user. One of the unique features of Burn as compared to similar models of burning materials is that 3d material deformation is modeled. This deformation is achieved through an arbitrary Lagrangian-Eulerian (ALE) formulation of the governing equations.

The governing equations of Burn are discretized using linear finite elements, and time-stepping is by the implicit Euler method. Since the discretized equations are nonlinear, Newton's method is used to update the solution at each time step.

Burn is implemented as a set of Python modules. These modules are found in the scripts/ directory. A simulation is executed using the burn.py script along with an input file specified as a command line argument. Input files for Burn take the form of Python modules, but require little to no knowledge of Python to create or modify. For examples of input file modules look in demos/, validation/, or verification/. Several meshes generated using Gmsh may be found in meshes/. Gmsh *.msh files may be converted to XML files as used by Burn using the dolfin-convert.py script found here.

Setup and Usage

Burn requires an installation of FEniCS. The easiest way to use FEniCS is through the FEniCS Docker images. Installers for Docker may be found here. Note that FEniCS requires that Windows users install the Docker Toolbox, and not Docker for Windows. A nice discussion of how to use FEniCS with Docker may be found here.

Once you have Docker installed, it is straightforward to start running Burn. The process is outlined as follows:

1. Start a Docker Quickstart terminal. Your installation should create an executable for launching such a terminal.
2. Change directories into your Burn repository. The Burn repository may be obtained by clicking the green "Clone or download" button above.
3. Start a FEniCS session with access to your working directory by executing the following command:
   $ docker run -ti -v $(pwd):/home/fenics/shared quay.io/fenicsproject/stable
4. Change directories to the location of your input file (e.g., input.py). An example of this is given below.
5. Run Burn: $ ~/shared/burn.py input.py

The output of the simulation will be stored locally in an automatically generated folder. Burn writes temperature and mass fraction fields to vtk files, which may be visualized using ParaView.

Example Simulation

To run the plate demo, begin by starting a Docker Quickstart Terminal. Change directories to your Burn repository. Start a FEniCS session (see Step 3 above). Change directories to the plate verification case:

$ cd verification/conduction/plate/

Then run the simulation using

$ ~/shared/burn.py plate.py

This example is steady state, and so it won't take long to finish. Once the simulation is complete, start ParaView. From the File menu, click on open and find and open .../burn/verification/conduction/plate/output_plate/paraview/T_data.pvd. Click the Apply button on the left of the ParaView window, and a colored plot of the temperature field will appear.

All of the other input files in the demos/, validation/, and verification/ folders can be run similarly. Be aware that some of the 3d simulations can take a while to run.