Work

Assessment of Hybrid NURBS-Lagrange Discretization for Finite Deformation Hyperelastic Problems

Year: 2026
Type: article
Abstract: This study extends a hybrid NURBS-Lagrange discretization to finite-deformation hyperelastic analysis and evaluates its accuracy and computational efficiency. Different discretization strategies are implemented within a unified finite-deformation hyperelastic framework. Five benchmark problems involving compressible and nearly incompressible materials are considered, with comparisons against analytical solutions, conventional linear finite element analysis (FEA), and quadratic isogeometric analysis (IGA). The hybrid models accurately reproduce the analytical solutions for uniaxial stretching and a thick-walled cylinder subjected to internal pressure. For nearly incompressible block compression and large-deformation bending of a curved beam, the linear hybrid model generally achieves higher accuracy than linear FEA at comparable mesh resolutions, while higher-order models provide further improvements at increased computational cost. Excluding preprocessing time, the R2L1L1 hybrid model incurs a modest computational overhead relative to linear FEA but remains substantially less expensive than quadratic IGA and higher-order hybrid models at comparable element counts. Comparisons of error against computational time further demonstrate the favorable accuracy–cost balance of the linear hybrid model. In addition, the rubber connecting sheet benchmark further demonstrates the applicability of the hybrid formulation to geometries involving holes and curved boundaries. Overall, the R2L1L1 hybrid discretization provides a practical alternative to conventional linear FEA and IGA for finite-deformation hyperelastic problems, balancing numerical accuracy, computational cost, and geometric flexibility.
Language: English
Open Access status: gold
Estimated APC paid: $2,896
OpenAlex ID: W7235948334
FWCI:
Cites:
Cited by:
Related to: 10