Comparison of Bending-Energy Discretization Methods for Anisotropic Meshes in Morphogenetic Simulations

Tomohiro Mimura, Yasuhiro Inoue · 2025

Read Comparison of Bending-Energy Discretization Methods for Anisotropic Meshes in Morphogenetic Simulations by Tomohiro Mimura, Yasuhiro Inoue (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-00302.txt.

Introduction

Tomohiro Mimura and Yasuhiro Inoue compared this 2025 set of math recipes for bending energy in simulations of growing tissue. Such simulations track cells whose positions carry real meaning, so the mesh cannot simply be rebuilt when it stretches unevenly. Picking the right bending recipe matters, yet guidance was scarce. The authors tested rival recipes in two stages.

First, they checked each recipe's energy and force against exact values on a wrinkled sheet and a smooth sphere. Then they grew wrinkles in a sheet under two cell-division patterns and read the spacing of the folds. One model, Hamann's, proved most exact for hard numbers. Another, Julicher's, ran cheaper and sufficed when only shape trends matter. The study gives modelers a clear selection rule.

Catalog record

CREATORTomohiro Mimura, Yasuhiro Inoue
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE58 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-00302

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Comparison of Bending-Energy Discretization Methods for Anisotropic Meshes in Morphogenetic Simulations Tomohiro Mimura, Yasuhiro Inoue Abstract Accurately modeling bending energy in morphogenetic simulations is crucial, especially when dealing with anisotropic meshes where remeshing is infeasible due to the biologically meaningful entities of vertex positions (e.g., cells). This study addresses the underexplored question of which bending-energy discretization methods are most accurate and suitable for such simulations. The evaluation consists of two stages: First, the accuracy of each method is tested by comparing predicted bending energy and force against theoretical values for two benchmark cases–a wrinkled planar sheet and a smooth spherical sheet. Second, we simulate the formation of wrinkles in a planar sheet caused by anisotropic cell division, analyzing the resulting wavenumber patterns for two division orientations: uniaxial and random. The results highlight that the choice of the optimal discretization method depends on the application. For simulations requiring precise quantitative predictions, the Hamann model demonstrates superior accuracy. Conversely, for simulations where qualitative trends in morphology are of primary interest, the Jülicher model provides a computationally efficient alternative. These findings provide guidance for selecting appropriate bending-energy discretization methods in morphogenetic simulations, ultimately leading to more accurate and efficient modeling of complex biological forms. 1 Introduction Thin shell deformation has been applied to solve various problems across diverse fields, including the study of vesicle and red blood cell (RBC) shapes in fluids [ 1 , 2 , 3 , 4 , 5 ] , buckling of thin films placed on fluid surfaces [ 6 , 7 , 8 , 9 , 10 ] , and cloth simulations [ 11 , 12 , 13 , 14 , 15 ] . Numerical solution methods are indispensable for analyzing these complex phenomena. A widely used method involves representing continuous surfaces through triangular element decomposition. To address mechanical problems using this discrete framework, it is necessary to account for both in-plane elastic and bending deformations through a suitable bending-energy discretization method. Discretization of bending energy is based on Canham’s formula using principal curvatures [ 16 ] and Helfrich’s formula using mean and Gaussian curvatures [ 17 ] . However, these methods require the second- and fourth-order derivatives of surface coordinates to calculate the bending energy and resulting force, respectively [ 18 , 19 ] . Various methods exist to approximate these parameters. For example, Kantor and Nelson used the angle between two adjacent triangles to express the bending energy and simulated the phase transition of a shape in polymer films [ 20 ] . Later, Jülicher [ 21 ] and Gommper and Kroll [ 22 ] constructed models to represent the discrete curvatures of vesicles. Since then, various discretization methods have been developed, and compared [ 18 , 19 , 23 , 24 ] . Guckenberger et al. [ 18 ] calculated the mean curvature and force of red blood cells, evaluated the errors against theoretical values, and notes that the errors increase for non-uniform triangular meshes. Bian et al. [ 19 ] compared equilibrium shapes calculations for lipid bilayers and highlights that different discretization methods produce varying results. Therefore, systems that can be represented by isotropic triangular meshes, such as vesicles and RBCs, have been extensively studied. However, not all phenomena can be accurately represented using isotropic triangular meshes. Biological processes often involve complex, dynamic shapes that require anisotropic modeling approaches. In biology, the cell-center model [ 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 ] is widely employed to study cell movement and tissue formation by representing individual cells as vertices within a mesh. This approach allows simulating diverse processes, such as cell movement in intestinal crypts [ 25 , 27 ] and cell migration [ 28 , 30 , 31 , 32 , 33 ] in two dimensions, by defining cell–cell interactions based on vertex connectivity. We have previously extended this model to 3D to simulate the morphogenesis of sheet-like epithelial tissues [ 34 ] . This model allowed us to explore how cells utilize mechanisms, such as cell division [ 35 , 36 ] and apical constriction [ 37 , 38 ] , to control global tissue deformation and buckling [ 39 , 40 ] . Cell-division axis orientation plays a crucial role in shaping the tissue morphology during morphogenesis [ 41 , 42 , 43 ] . However, this process often leads to anisotropic cell arrangements and, consequently, anisotropic meshes. Because the vertices in these meshes represent individual cells, remeshing to maintain isotropy during the simulations is not feasible. Therefore, research on the accuracy of different-bending energy discretization methods for anisotropic meshes is limited. This study addresses this gap by comprehensively evaluating the accuracy and applicability of various bending-energy discretization methods for simulating anisotropic meshes using a cell-center model. Specifically, we compared 13 existing discretization methods by incorporating them into the cell-center model and simulating the 3D deformation of the monolayer epithelial sheet tissue. Two sets of simulations were conducted to assess the accuracy of each method. First, to investigate the bending-energy error in anisotropic meshes, we prepared anisotropic sheets with simple wrinkle patterns, allowing for theoretical bending-energy calculations, and compared the accuracies of each method across different cell configurations. Next, we compared them under the dynamic conditions by simulating sheet deformation owing to cell division with two axis orientations: uniaxial, which leads to anisotropic meshes, and random, which maintains mesh isotropy. We simulated epithelial folding caused by cell proliferation to identify the most suitable bending-energy discretization method for accurately capturing morphogenetic events in the cell-center model. 2 Cell-center model In this study, the cell-center model was used to construct a triangular mesh by connecting adjacent cell centers in the epithelial sheet. As illustrated in Fig. 1 (a), triangular elements are formed by connecting the centers of adjacent cells, assuming three-cell adjacency. The cell shape is then defined by connecting the centers of mass of the triangles that share a common cell, as shown in Fig. 1 (b). This approach eliminates the need to explicitly maintain the Delaunay property and simplifies mesh construction. The model simulates the tissue dynamics by performing mechanical and geometrical calculations at each time step. 2.1 Mechanical calculation The cell-center coordinates \bm{x_{i}} evolve over time based on the equations of motion at the cell center i . \eta\frac{d\bm{x_{i}}}{dt}=-\nabla_{\bm{x_{i}}}E+\bm{F^{B}_{i}} (1) where \eta is the viscosity coefficient, E is the energy function of the system excluding the bending stiffness, and \bm{F^{B}_{i}} is the force applied to cell i by the bending energy. A detailed explanation is provided in Section 3. The energy function E comprises the mechanical energy E_{\rm{Mech}} excluding the bending stiffness and the constraint energy E_{\rm{Center}} to satisfy the cell-center constraint [ 34 ] . E=E_{\rm{Mech}}+E_{\rm{Center}} (2) The mechanical energy E_{\rm{Mech}} is composed of the edge elastic energy between adjacent cells E_{L} , area elastic energy of the triangle connecting adjacent cells E_{S} , excluded volume energy of the cell E_{\rm{Rep}} , and the constraint energy from the surrounding environment when deformed out-of-plane E_{Z} . E_{\rm{Mech}}=E_{L}+E_{S}+E_{\rm{Rep}}+E_{Z} (3) The edge elastic energy E_{L} is the sum of the energies required to adjust the edge length L_{ij} between adjacent cells i and j to the equilibrium length L_{\rm{eq}} . This relationship

Read the complete text