Pybaqus is a python library to import the output files generated by Abaqus in the ASCII *.fil format, and create a VTK object from the data.
The results can then be analyzed in pure python (i.e. no Abaqus licence needed) with the great tools provided by PyVista.
Pybaqus is in a very early development stage. Therefore, there are still many unimplemented functionalities. However, basic operations like importing the mesh and the nodal and element results is implemeneted and can be used for some analysis.
The following feeatures are either already implemented or planned:
- Import 2D meshes
- Import 3D meshes
- Import nodal results
- Import element results
- Element and node sets
- Extrapolate element results from Gaussian points to nodes (implemented for some elements)
- Interpolate element/nodal results to any point within the element considering the shape functions of the element
- Import history output
- Compute stresses along paths
- Compute section forces and moments
- Implement functions to easily create animations
pip install git+https://github.com/cristobaltapia/pybaqus
The first thing you need is to tell Abaqus that you want am ASCII *.fil result file.
To get that you need to write the following lines in your *.inp file, within the step definition (before the *End Step command) e.g.:
...
*FILE FORMAT, ASCII
*EL FILE, DIRECTIONS=YES
S, E, COORD
*NODE FILE
COORD, U
*End Step
...
You can specify different output variables (as long as they are available for the elements you are using, of course).
After submitting your model you will get a *.fil file.
This is the file you need to import it with Pybaqus.
Import the *.fil file like this:
from pybaqus import open_fil
res = open_fil("your_result.fil")Great! That was it. :)
Now you have your results as a VTK object, wrapped by PyVista, and there's nothing that can get in your way to analyze your results with pure python.
import pyvista as pv
mesh = res.get_mesh()
plot = pv.Plotter()
plot.add_mesh(mesh, show_edges=True, color="white")
plot.view_xy()
plot.show()
Cool! But something's missing there. Colors! We can plot some of our results like this:
mesh = res.get_deformed_mesh(step=1, inc=1, scale=3)
s2 = res.get_nodal_result(var="S2", step=1, inc=1)
mesh.point_arrays["S2"] = s2
plot = pv.Plotter()
plot.add_mesh(mesh, show_edges=True, color="white",
scalars="S2", show_scalar_bar=True)
plot.view_xy()
plot.show()
That's it! Since the API is still under development, some of these functions might change.
And now you can de whatever you want with your results in python. Have fun!