This module is used for a small semester projet @EPFL. It is free of use, at your own risk.
The code is intended for the following setup:
- Two PMMA plates are sheared against each other
- Strain gauges are placed along the plate, close to the sheared frontier
- The strains are recorded in files readable with pandas/polars with one column for time and triplets of strain gauge columns. In short, the files should ressemble the following:
| Relative time | Ch01 | Ch02 | Ch03 | … | Ch18 | Ch19 | Ch20 |
|---|---|---|---|---|---|---|---|
| -0.005 | -0.177943 | -2.0 | -2.0 | … | -0.367605 | -2.0 | 0.017335 |
| -0.004998 | -0.263445 | -2.0 | -2.0 | … | -0.43433 | -2.0 | 0.013566 |
| … | … | … | … | … | … | … | … |
| 0.0049975 | -0.20223 | -2.0 | -2.0 | … | -0.367605 | -2.0 | 0.000502 |
| 0.0049995 | -0.3772 | -2.0 | -2.0 | … | -0.354459 | -2.0 | -0.006029 |
read()(for reading thecsvfiles)lowfilter()for reducing noisestraindf()for computing the strains from the signalstressdf()for computing the stresses from the strainsselect_files()is a simple GUI tool for choosing the files to work with
In order to pass from the rosette signals to strains, an amplification factor and a rotation matrix (3x3) are used.
- elongation
$\gets$ amplification factor * signal - strains
$\gets$ rotation matrix @ elongation
The rotation_matrix function is implemented for any orientations but it is simply used for the standard (45, 90, 135) holy trinity of angles for the measure of shear. The broadcasting of the matrix to all measures uses numpy.einsum.
Here is a gif to simulate an interactive plot
Is a duplicate of the BigBrother script with polars instead of pandas.