- AutorIn
- Alejandra Maria Foggia Technische Universität Dresden#Center for Systems Biology Dresden#Max Planck Institute of Molecular Cell Biology and Genetics
- Titel
- Numerical solution of scalar and vector differential equations on surface point clouds
- Zitierfähige Url:
- https://nbn-resolving.org/urn:nbn:de:bsz:14-qucosa2-984574
- Erstveröffentlichung
- 2025
- Datum der Einreichung
- 11.11.2024
- Datum der Verteidigung
- 30.04.2025
- Abstract (EN)
- In this thesis, I present the Surface Discretization-Corrected Particle Strength Exchange (Surface DC-PSE) method for vector-valued partial differential equations (PDEs) on curved surfaces. Surface DC-PSE is a numerical meshfree collocation method for approximating surface differential operators. The method requires a surface point cloud and the surface normal at each point, and provides a mathematically embedding but computationally embedding-free approximation of the fields and their spatial derivatives at each point. The Surface DC-PSE method is an extension of the DC-PSE method, which is a mesh- free numerical method for approximating Euclidean differential operators. Surface DC-PSE is based on the idea of extending the field constant along the normal into the embedding space, constructing the kernels using both surface and “virtual” extended particles along the normal, and collapsing the resulting ”bulk” kernels into surface ones. This thesis presents a comprehensive study of the Surface DC-PSE method, including its mathematical formulation, numerical implementation, and verification tests. The results show that the method provides convergent and stable approximations of scalar and vector surface differential operators, and is applicable to a wide range of surfaces. I also discuss the limitations of the method, including the requirement of a homogeneous point cloud, the sensitivity to particle distributions, and the computational cost of constructing the kernels. Future work is proposed to address these limitations, including modifying the neighborhood construction algorithm, evaluating the accuracy and stability of vector Surface DC-PSE on deforming surfaces, and exploring applications of the method to non-analytical surfaces. The Surface DC-PSE method has the potential to be used in a wide range of applications, including simulating biological fluid surfaces, studying morphogenesis, and modeling complex surface phenomena. The method provides a robust and reliable framework for approximating surface differential operators, and it provides a basis for further research and development in these and other areas of scientific computing.
- Nachfolger
- Numerical solution of scalar and vector differential equations on surface point clouds
- Freie Schlagwörter (EN)
- partial differential equations, surfaces, point cloud
- Klassifikation (DDC)
- 004
- Klassifikation (RVK)
- SK 540
- ST 130
- GutachterIn
- Prof. Dr. Ivo Fabian Sbalzarini
- Prof. Dr. Jens Honore Walther
- Den akademischen Grad verleihende / prüfende Institution
- Technische Universität Dresden, Dresden
- Förder- / Projektangaben
- Deutsche Forschungsgemeinschaft (DFG)
Vector- and tensor-valued surface PDEs
ID: FOR-3013 - Sonstige beteiligte Institution
- Max Planck Institute of Molecular Cell Biology and Genetics, Dresden
- Version / Begutachtungsstatus
- publizierte Version / Verlagsversion
- URN Qucosa
- urn:nbn:de:bsz:14-qucosa2-984574
- Veröffentlichungsdatum Qucosa
- 15.08.2025
- Dokumenttyp
- Dissertation
- Sprache des Dokumentes
- Englisch
- Lizenz / Rechtehinweis
CC BY 4.0- Inhaltsverzeichnis
Abstract List of publications 1 Introduction 1.1 Morphogenesis 1.2 Active matter 1.3 Surfaces and curvature in the context of morphogenesis 1.4 Challenges related to surfaces 1.5 State of the art of numerical methods for the solution of surface PDEs 1.5.1 Finite Element methods (FEMs) 1.5.2 Collocation methods 1.5.3 Other methods 1.6 This thesis: an overview 2 Surface differential operators 2.1 Motivation 2.2 Hierarchy of manifolds 2.3 Vector fields: covariant and contravariant notation 2.4 Differentiating vector fields on (curved) manifolds 2.4.1 Parallel transport 2.4.2 Covariant derivative 2.4.3 Covariant derivative in a Riemannian manifold 2.4.4 Summary 2.5 Covariant derivative of hypersurfaces in the Euclidean space 2.6 Curvature 2.6.1 Summary 2.7 A reference case: the S2 sphere 2.7.1 Operator expressions 2.7.2 Eigenfunctions of the scalar and vector Laplacian 2.7.3 An orthonormal basis on the S2 sphere 3 Surface DC-PSE 33 3.1 Particle methods for the approximation of differential operators 3.1.1 DC-PSE: Two-dimensional case 3.1.2 DC-PSE: General case 3.2 Approximating surface differential operators using particle methods 3.3 Surface DC-PSE 3.3.1 Neighborhood construction 3.3.2 Normal extension 3.3.3 Surface kernel construction 3.4 Verification and tests of Surface DC-PSE for scalar fields 3.4.1 Convergence tests 3.4.2 Analysis of the influence of rc and the curvature on the error 3.4.3 Strong scaling of Surface DC-PSE 3.4.4 Comparison with the Surface Finite Element Method 3.5 Reaction-diffusion and patterning with Surface DC-PSE 4 Vector surface PDEs 4.1 Surface DC-PSE for vector fields 4.2 Verification tests of Surface DC-PSE for vector fields 4.2.1 Convergence of vector Laplacian on the unit sphere 4.2.2 Convergence of vector diffusion on unit sphere 4.2.3 Vector diffusion on other shapes 4.3 Flows on surfaces 4.3.1 Verification of combined Surface DC-PSE and EDAC on the unit sphere 4.3.2 Incompressible flow on tori of different curvature 5 Conclusion and Outlook 5.1 Summary 5.2 Limitations 5.3 Looking into the future Appendices A Numerical methods for function approximation and PDE solution A.1 Collocation methods A.2 Galerkin methods A.3 Spectral methods B Differential operators on the unit sphere S2 B.1 Gradient of a scalar B.2 Gradient of a contravariant vector B.3 Gradient of a covariant vector B.4 Divergence of a contravariant vector B.4.1 Divergence of a covariant vector B.5 Laplacian (Laplace-Beltrami) of a scalar B.6 (Connection) Laplacian of a contravariant vector B.7 Crossproduct C Scalar spherical harmonics D Vector spherical harmonics E An orthonormal basis on the unit S2 sphere F DC-PSE algorithm: Auxiliary equations G Simulation parameters G.1 Fibonacci distribution of points on a sphere G.2 Error metrics for vector quantities G.3 Convergence of the Laplace-Beltrami operator on the unit circle and the unit sphere G.4 Mean and Gauss curvature of ellipsoid and Stanford bunny G.5 Analysis of the influence of rc and the curvature on the error of the Laplace-Beltrami operator on the unit sphere G.6 Comparison with FEMs G.7 Reaction-diffusion patterns G.8 Convergence of the connection Laplacian on the unit sphere G.9 Convergence of the surface divergence operator on the unit sphere G.10 Convergence of the solution of the vector diffusion equation on the unit sphere G.11 Vector diffusion equation on the torus and a nonic surface G.12 Verification of combined Surface DC-PSE and EDAC for the solution of the incompressible surface Navier-Stokes equation on the unit sphere G.13 Test for the incompressible surface Navier-Stokes equation on tori G.14 Spherical harmonics for the solution of surface diffusion on the unit sphere G.15 Error computation using Richardson extrapolation Bibliography Index Acknowledgments