💻 computer science

The cross-sectional warping problem for hyperelastic beams: An efficient formulation in Voigt notation

This paper presents an efficient, fully material formulation for the cross-sectional warping problem of hyperelastic beams using Voigt notation to overcome traditional limitations of small strains and rigid cross-sections, validated through numerical examples and accompanied by an open-source isogeometric finite element implementation.

Original authors: Juan C. Alzate Cobo, Tobias Henkels, Oliver Weeger

Published 2026-04-15

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Juan C. Alzate Cobo, Tobias Henkels, Oliver Weeger

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are designing a flexible robot arm, a soft medical device, or a high-tech fabric. To make these things work, engineers need to know exactly how they will bend, twist, and stretch under pressure.

Traditionally, engineers have used a "shortcut" to calculate this. They pretend the beam (the rod-like part of the structure) is a rigid stick that doesn't change its shape when it bends, and they only calculate how it stretches a tiny bit. This is like trying to predict how a wet noodle will flop around by pretending it's a stiff wooden dowel. It works for simple things, but fails miserably when the material is soft, stretchy, or deforms wildly.

To get the real answer, you usually have to simulate the entire 3D object in a computer. But this is like trying to count every single grain of sand on a beach to predict how the tide moves—it's incredibly slow and requires massive computing power.

The "Cross-Sectional Warping" Solution
A few years ago, researchers came up with a clever middle ground called the Cross-Sectional Warping Problem (CSWP).

Think of a beam not as a solid stick, but as a stack of pancakes (the cross-sections). When you twist or bend a beam, those "pancakes" don't just stay flat; they warp, bulge, and twist like a wet sponge. The CSWP method says: "Let's ignore the long length of the beam for a moment and just solve the puzzle of how one single pancake deforms."

Once you know how that one slice deforms, you can mathematically figure out how the whole beam behaves. It's much faster than simulating the whole 3D object, but it captures the complex, squishy reality that the old "rigid stick" models miss.

The Problem with the Old Math
The original version of this method (developed by Arora et al.) was brilliant, but the math behind it was like trying to drive a car with a steering wheel that had too many gears. It used a complex, asymmetrical way of calculating stress and strain (called the "PK1 formulation"). This made the computer code messy, slow, and hard for other scientists to understand or build upon. It was like using a sledgehammer to crack a nut.

The New "PK2" Formulation
This paper introduces a new, streamlined version of that method. The authors, Juan C. Alzate Cobo and his team, decided to rewrite the math using a different set of tools (called the "PK2 formulation" and "Voigt notation").

Here is the analogy:

The Old Way (PK1): Imagine trying to organize a messy room by moving every single item individually, one by one, in a chaotic order. It gets the job done, but it's exhausting and prone to errors.
The New Way (PK2): Imagine you have a smart sorting system that groups items by shape and size automatically. You use a standard, symmetrical grid (Voigt notation) to organize the data. Suddenly, the room is tidy, the process is twice as fast, and anyone else can easily see how you did it.

Why This Matters

Speed & Efficiency: By using this new "symmetrical" math, the computer calculations become much faster. It's like upgrading from a dial-up modem to fiber optic internet.
Accuracy: It allows engineers to simulate beams made of "hyperelastic" materials (like rubber, soft robots, or biological tissues) that can stretch and twist wildly without breaking.
Open Source: The authors didn't just write the math; they built the actual software and put it on GitHub for free. They are saying, "Here is the code, here is the data, go ahead and use it to build your own amazing things."

In a Nutshell
This paper is about taking a powerful but clunky tool for simulating flexible beams and giving it a sleek, high-performance engine. It makes it easier and faster for scientists and engineers to design the next generation of soft robots, medical implants, and smart materials, ensuring they work exactly as intended before they are ever built in the real world.

1. Problem Statement

Classical beam theory is traditionally limited to small elastic strains, linear elastic materials, and the assumption that cross-sections remain planar (no warping). While 3D continuum models can capture complex behaviors like large strains, hyperelasticity, and cross-sectional warping, they are computationally expensive for structural analysis.

To bridge this gap, the Cross-Sectional Warping Problem (CSWP) was introduced (notably by Arora et al.) to derive general, nonlinear constitutive relations for beam models. However, existing CSWP formulations rely on the First Piola-Kirchhoff (PK1) stress and the deformation gradient. This approach has significant drawbacks:

It requires manipulating full, non-symmetric second-order tensors.
It involves complex linearization with fourth-order constitutive tensors.
It is less efficient for numerical implementation compared to standard finite strain continuum mechanics formulations.

The authors aim to reformulate the CSWP using a fully material framework based on symmetric tensors to improve computational efficiency and simplify implementation.

2. Methodology

A. Theoretical Reformulation (PK2 & GL Strain)

The authors reinterpret the CSWP using the Second Piola-Kirchhoff (PK2) stress tensor ( $\mathbf{S}$ ) and the Green-Lagrange (GL) strain tensor ( $\mathbf{E}$ ).

Kinematics: The deformation map includes a warping function $\mathbf{u}$ defined on the cross-section. The deformation gradient $\mathbf{F}$ is decomposed into a rotation $\mathbf{R}$ and a cross-section stretch $\mathbf{F}_{\{cs\}}$ .
Energy Potential: The beam's internal energy potential $\Psi_b$ is defined as the minimum of the total strain energy over the cross-section, subject to constraints that prevent rigid body motions (translation and rotation of the cross-section).
Constraints: Rigid body motions are constrained using Lagrange multipliers ( $\lambda$ for translation, $\mu$ for rotation), formulated as a constrained minimization problem.
Voigt Notation: Because $\mathbf{S}$ and $\mathbf{E}$ are symmetric, the formulation is expressed in Voigt notation. This reduces second-order tensors to 6-component vectors and fourth-order tensors to $6 \times 6$ matrices, significantly simplifying the algebraic structure.

B. Numerical Implementation (Isogeometric Analysis)

The problem is discretized using Isogeometric Analysis (IGA) with B-spline basis functions.

Discretization: The cross-section and the warping displacement are approximated using NURBS/B-splines.
Strain-Displacement Operators: The authors derive specific differential operators ( $\mathbf{B}$ and $\tilde{\mathbf{B}}$ ) that map nodal displacements to the GL strain vector in Voigt notation. These operators account for the beam's strain measures (axial strain $\varepsilon$ and curvature/twist $\kappa$ ) and the warping function.
Solution Scheme: The resulting nonlinear system is solved using Newton's method. The tangent stiffness matrix is derived analytically, leveraging the symmetry of the Voigt formulation.

C. Sensitivity Analysis for Beam Stiffness

A critical requirement for beam models is the calculation of the tangent stiffness matrix ( $\mathbf{C}^b$ ), which requires the sensitivity of the warping function $\mathbf{u}$ with respect to the beam strain measures ( $\mathbf{q} = \{\varepsilon, \kappa\}$ ).

Adjoint Method: Instead of deriving complex additional weak forms for sensitivities, the authors employ an adjoint equation approach. By differentiating the discretized residual equations with respect to the strain measures, they solve a linear system involving the tangent stiffness matrix to obtain the sensitivities ( $\mathbf{u}_{,q}$ ).
Efficiency: This allows for the direct computation of the beam stiffness coefficients without re-solving the full nonlinear problem for every strain increment.

3. Key Contributions

Fully Material PK2 Formulation: The paper presents a novel reformulation of the CSWP using PK2 stress and GL strain, enabling the use of symmetric tensor properties.
Voigt Notation Efficiency: By utilizing Voigt notation, the authors reduce the complexity of the constitutive tensors, leading to a more concise theoretical description and a computationally efficient numerical implementation.
Derivation of Operators: Detailed derivation of the strain-displacement operators ( $\mathbf{B}$ and $\tilde{\mathbf{B}}$ ) specific to the cross-sectional warping problem in Voigt notation.
Adjoint Sensitivity Calculation: An efficient method to compute the sensitivities of the warping function required for the beam stiffness matrix, avoiding complex linearizations.
Open-Source Reproducibility: The authors provide a complete open-source implementation (MATLAB/NLIGA library) on GitHub, including both the new PK2 formulation and the reference PK1 formulation, along with all numerical examples.

4. Results and Verification

The proposed PK2 formulation was rigorously validated against:

Literature: Results were compared with established data from Arora et al. (2019) and Herrnböck et al. (2023).
Reference Implementation: A custom implementation of the original PK1 formulation was created to serve as a direct benchmark.

Key Findings:

Equivalence: The PK2 formulation yields identical results to the PK1 formulation in terms of residual norms, warping displacements, stress resultants (forces and moments), and tangent stiffness matrices.
Validation Cases:
- Pure Torsion: Normalized torsional stiffness and torsional moments matched literature data perfectly for rectangular and square cross-sections.
- Multi-axial Loading: Under combined axial strain and curvature, the out-of-plane warping displacements and force/stiffness responses were identical for both formulations across three hyperelastic material models (Saint Venant-Kirchhoff, Neo-Hooke, and Mooney-Rivlin).
- Pure Shear: Visualizations of displacement and von Mises stress confirmed qualitative agreement with existing literature.

5. Significance

Computational Efficiency: The shift to a symmetric, material-based formulation (PK2/GL) in Voigt notation simplifies the code structure and reduces computational overhead associated with handling non-symmetric tensors and complex fourth-order linearizations.
Generalizability: The approach is not limited to Isogeometric Analysis; the authors note it can be easily adapted to classical Finite Element Analysis (FEA).
Foundation for Advanced Models: This formulation provides a robust, unified framework for developing beam constitutive models that include hyperelasticity, inelastic effects (plasticity, swelling), and multiphysics coupling, while maintaining kinematic consistency with 3D continuum mechanics.
Reproducibility: The release of the code and data sets a high standard for reproducibility in computational mechanics, allowing other researchers to build upon this efficient formulation for complex nonlinear beam problems.