gcs

gcs is a Python library for creating generalized cylindrical shells (GCS).

The GCS design space uses eleven continuous parameters to generate trillions of geometrically diverse, manufacturable cylindrical-shell structures. This compact parameterization enables large-scale optimization while capturing the nonlinear buckling, plasticity, and self-contact mechanisms that govern energy absorption. This library is used in our paper:

Kelsey L. Snapp, Benjamin Verdier, Aldair E. Gongora, Samuel Silverman, Adedire D. Adesiji, Elise F. Morgan, Timothy J. Lawton, Emily Whiting, Keith A. Brown
Superlative mechanical energy absorbing efficiency discovered through self-driving lab-human partnership
Nature Communications (2024)

Installation

The package requires Python 3.10 or later and is available on PyPI. Install it with:

pip install gcs-shape

You can then import the package in Python:

import gcs

GCS Design Parameters

GCS are parameterized by 11 values:

c4 and c8 Relationship

The geometry of each top or bottom face is described by a polar equation whose shape is controlled by the parameters c4 and c8. This parameterization was adopted from Overvelde and Bertoldi (2014).

twist_linear, twist_amplitude, and twist_cycles Relationship

The parameters twist_linear, twist_amplitude, and twist_cycles control how the cross-section rotates as it is interpolated from the base face to the top face. twist_linear sets the linear rotation, while twist_amplitude and twist_cycles control the amplitude and frequency of the oscillatory rotation along the height.

Quickstart

import gcs

# Create a generalized cylindrical shell
shape = gcs.GCS(c4_base=0.3,
                c8_base=-0.2,
                c4_top=0.4,
                c8_top=-0.3,
                twist_linear=2,
                twist_amplitude=0.05,
                twist_cycles=3,
                perimeter_ratio=1.5,
                height=20,
                mass=2.1,
                thickness=0.48)

# Assert the shape is valid
print(shape.valid)

# Save shape to STL file
gcs.io.save_mesh(file='shape.stl', shape=shape)

Documentation

Documentation is provided in NumPy-style docstrings throughout the codebase. Docstrings can be viewed interactively with Python's built-in help() function. For example:

import gcs
help(gcs.GCS)

Contributing

The contribution guidelines can be found in CONTRIBUTING.md.

Maintainers

• Sam Silverman - sssilver@bu.edu

Citation

@article{Snapp:2024:SuperlativeMechanicalEnergy,
author={Snapp, Kelsey L. and Verdier, Benjamin and Gongora, Aldair E. and Silverman, Samuel and Adesiji, Adedire D. and Morgan, Elise F. and Lawton, Timothy J. and Whiting, Emily and Brown, Keith A.},
title={Superlative mechanical energy absorbing efficiency discovered through self-driving lab-human partnership},
journal={Nature Communications},
year={2024},
month={May},
day={21},
volume={15},
number={1},
pages={4290},
issn={2041-1723},
doi={10.1038/s41467-024-48534-4},
url={https://doi.org/10.1038/s41467-024-48534-4}
}