Convergent FEM for a membrane model of liquid crystal polymer networks

Lucas Bouck,Ricardo H. Nochetto,Shuo Yang

2023-07-25

Abstract:We design a finite element method (FEM) for a membrane model of liquid crystal polymer networks (LCNs). This model consists of a minimization problem of a non-convex stretching energy. We discuss properties of this energy functional such as lack of weak lower semicontinuity. We devise a discretization with regularization, propose a novel iterative scheme to solve the non-convex discrete minimization problem, and prove stability of the scheme and convergence of discrete minimizers. We present numerical simulations to illustrate convergence properties of our algorithm and features of the model.

Numerical Analysis

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is to develop a finite element method (FEM) to simulate the non - convex stretching energy minimization problem in the liquid crystal polymer network (LCNs) membrane model. Specifically, the author focuses on:

1. **Discretization of non - convex stretching energy**: Design a discretization scheme with regularization to solve the problem of the lack of weak lower - semicontinuity of the energy functional due to non - convexity.
2. **Iterative solution algorithm**: Propose a new iterative scheme to solve the discretized non - convex minimization problem and prove the stability of this scheme and the convergence of the discrete minima.
3. **Numerical simulation**: Demonstrate the convergence characteristics of the algorithm and the characteristics of the model through numerical simulation.

### Detailed Interpretation

#### 1. Background and Problem Description
Liquid crystal polymer networks (LCNs) are materials that can spontaneously deform under temperature or optical stimulation. These materials combine the properties of rubber and nematic liquid crystals, where the direction of the nematic molecules (i.e., the direction of the mesogenic groups) affects the deformation of the material under stimulation. This property makes LCNs important in applications such as microrobots and biomedical devices.

The paper focuses specifically on the modeling of thin - layer LCNs membranes. Such materials are usually modeled as three - dimensional hyperelastic bodies \( B:=\Omega\times(-t / 2,t / 2)\), where \( \Omega\subset\mathbb{R}^2\) is a bounded Lipschitz domain and \( t\) is a small thickness parameter. Classical elasticity theory uses dimension - reduction techniques to derive two - dimensional models from three - dimensional models to describe the deformation \( y(\Omega)\) of the mid - plane.

#### 2. Model Characteristics
- **Nematic direction field and order parameter**: Due to the nematic - elastic coupling in LCNs, the unit - length vector field \( m:\Omega\rightarrow S^1\) describes the direction of LC molecules and plays a key role in the deformation of the material.
- **Energy density**: The three - dimensional elastic energy density of LCNs is given by the classical energy density formula proposed by Bladon, Warner, and Terentjev:
\[
W^{3D}(x,z),F = \left\| L^{-1/2}_n F L^{1/2}_m\right\|^2 - 3,
\]
where \( F\in\mathbb{R}^{3\times3}\) satisfies \( \det F = 1\) (assuming incompressibility), and \( L_m\) and \( L_n\) are the step - length tensors in the reference configuration and the deformed configuration, respectively.

#### 3. Numerical Methods and Contributions
- **Discretization**: The author proposes a continuous \( P_1\) Lagrange finite - element approximation \( y_h\) to discretize the deformation and introduces a regularization term \( R_h[y_h]\) to handle non - convexity.
- **Non - linear gradient - flow scheme**: To solve the discrete minimization problem, a non - linear gradient - flow scheme with embedded Newton sub - iterations is designed to efficiently solve the non - linear discrete equations.
- **Convergence analysis**: Prove the convergence of the discrete minima and show useful tools and techniques, such as the construction of recovery sequences and compactness arguments.

In summary, this paper solves the non - convexity and weak lower - semicontinuity problems that are difficult to handle with traditional methods by developing an effective numerical method to handle the non - convex energy minimization problem in the LCNs membrane model.