PENGoLINS introduction

PENGoLINS (PENalty-based GLuing of Isogeometric Non-matching Shells) is an open-source Python library for penalty of non-matching Kirchhoff–Love shells using a FEniCS-based implementation of isogeometric analysis (IGA) called tIGAr. Users can directly perform shell structure analysis on CAD models, in STEP or IGES format, consisting of multiple B-spline/NURBS patches with non-matching parametrizations at their intersections. The pythonOCC, a Python interface of OpenCASCADE, is leveraged for computation of surface-surface intersections, where displacement and rotational continuities are maintained using penalty method. Furthermore, this library can be extended to fluid-structure interaction (FSI) and nonlocal contact analysis by integrating with existing open-source Python frameworks, viz., VarMINT, ShNAPr and CouDALFISh. Coupling between separately-parametrized patches uses a modification of the penalty formulation proposed by Herrema et al., which is verified using several benchmark problems. Applications to aerospace structures and prosthetic heart values are demonstrated in this documentation.

Penalty formulation and design of PENGoLINS are illustrated at length in the following article

@article{Zhao2022,
title = "An open-source framework for coupling non-matching isogeometric shells with application to aerospace structures",
journal = "Computers \& Mathematics with Applications",
volume = "111",
pages = "109--123",
year = "2022",
issn = "0898-1221",
doi = "https://doi.org/10.1016/j.camwa.2022.02.007",
author = "H. Zhao and X. Liu and A. H. Fletcher and R. Xiang and J. T. Hwang and D. Kamensky"
}

A preprint of the above article can be found here. We have since updated the penalty formulation to be more similar to that of Herrema et al., but with a numerical approximation of the tangent vectors of intersection curves. This updated formulation is more robust in nonlinear analyses of certain geometries. Differences from the benchmark results for linear problems documented in the paper are negligible.