In this paper we propose an algorithm that computes a low cardinality PI-type (Positive weights and Interior nodes) algebraic cubature rule of degree n, with at most (n +1)(n + 2)/2 nodes, over curvilinear polygons defined by piecewise rational functions. Typical examples are domains whose boundary is defined piecewise by NURBS curves or by composite Bezier curves. The key tools are a relevant but overlooked theorem of 1976 by Wilhelmsen on Tchakaloff sets, a specific in-domain algorithm for such curvilinear polygons and the sparse nonnegative solution of underdetermined moment matching systems by the Lawson-Hanson Nonnegative least squares solver. Many numerical tests are performed to show the flexibility of this approach and the implemented MATLAB toolbox is freely available to the users, in particular for possible applications within FEM/VEM with NURBS-shaped curvilinear elements.
Low cardinality positive interior cubature on NURBS-shaped domains
Sommariva, A;Vianello, M
2023
Abstract
In this paper we propose an algorithm that computes a low cardinality PI-type (Positive weights and Interior nodes) algebraic cubature rule of degree n, with at most (n +1)(n + 2)/2 nodes, over curvilinear polygons defined by piecewise rational functions. Typical examples are domains whose boundary is defined piecewise by NURBS curves or by composite Bezier curves. The key tools are a relevant but overlooked theorem of 1976 by Wilhelmsen on Tchakaloff sets, a specific in-domain algorithm for such curvilinear polygons and the sparse nonnegative solution of underdetermined moment matching systems by the Lawson-Hanson Nonnegative least squares solver. Many numerical tests are performed to show the flexibility of this approach and the implemented MATLAB toolbox is freely available to the users, in particular for possible applications within FEM/VEM with NURBS-shaped curvilinear elements.File | Dimensione | Formato | |
---|---|---|---|
nurbscatch.pdf
accesso aperto
Descrizione: preprint
Tipologia:
Preprint (submitted version)
Licenza:
Accesso libero
Dimensione
1.28 MB
Formato
Adobe PDF
|
1.28 MB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.