Lennart J.
Schulze

Dr.-Ing. in Computational Sciences at the MPI-CBG and TU Dresden

About Me

Lennart J. Schulze

Hey! I am Lennart, a mechanical engineer by training and a computational scientist by practice. I have recently finished my Ph.D. thesis at the TU Dresden and the Max-Planck Institute of Molecular Cell Biology and Genetics and am transitioning into industry. The past four years, I have been part of the group of Ivo Sbalzarini that focuses on scientific machine learning and numerical simulations. My research here has focused on the description and adaptive discretization of dynamic surfaces, which are ubiquitous in nature and engineering. Before, I have studied Mechanical Engineering for my bachelor's and master's degrees at the TU München, with stays in Tokyo and as a visiting researcher at the Northwestern University, where I have been working on the simulation of water waves in the group of Robert Dalrymple.

My work sits right at the intersection of physics, numerical methods and high-performance computing.

For more details, see my CV . You can reach me at .

Research

Research

Simulating morphogenetic models

Self-organizing shapes emerge from the interaction of physics and geometry. Here, we solve the non-linear and second-order reaction-diffusion equations within the continuously deforming surface, and allow for local growth depending on the concentration of the surface chemical. The PDE and the observed Turing patterns in the surface are influenced by the local emerging curvature, and the local curvature is determined by the PDE dynamics in the surface: A full feedback loop, from which organic shapes arise.

Manuscript in preparation.

Adaptive sampling of surfaces

Sampling curved surfaces with local regularity and global adaptivity is a long-standing problem in computer graphics and numerical methods. In this work, we use a gradient descent constrained to the curved surface to regularize point distributions for analytical surfaces and obtained from 3D scanning devices such as the 3D Stanford bunny.
The computed point discretizations of the curved surface allow for straightforward rendering without artifacts and robustness and accuracy of numerical methods during simulations.

Currently under revision. Preprint is available at arXiv:2605.03803 .

...
...

Describing and quantifying dynamic surfaces in numerical simulations

Real-world problems involve complex, and often dynamic surfaces. Either as boundary conditions, or as the domain of interest. An illustrative example of a dynamic surface is a hydrodynamic, incompressible droplet under surface tension, governed by the multi-phase Navier-Stokes equations.
To track the surface, we propose a method based on local high-order polynomial regression, enabling the computation of derivative geometric quantities such as surface normals and mean and Gaussian curvatures.

The paper is available at the Journal of Computational Physics .

Saving computational resources in water wave simulations

Adaptive resolution methods that place more computational nodes where required are a classic approach to reduce computational effort and accordingly speed up simulations, while retaining high accuracy. In this work, we proposed a different approach: While the point discretization of the domain of interest is homogeneous, we use a higher-order variant of Smoothed Particle Hydrodynamics (SPH) in regions with demand for high accuracy, and a standard, more cost-efficient variant in the other regions. Especially for deep water waves, this enables a much faster simulation of water wave propagation.

This work was presented at the SPHERIC 2022 International Workshop and its proceedings paper is available at arXiv:2511.10064 .

Water wave simulation

Publications

2026

L. J. Schulze, and I. F. Sbalzarini: "Globally adaptive and locally regular point discretization of curved surfaces." Under review, available at arXiv:2605.03803

2024

L. J. Schulze, S. K. T. Veettil, and I. F. Sbalzarini. "A high-order fully Lagrangian particle level-set method for dynamic surfaces." Journal of Computational Physics , 515, 113262 (2024)

2022

L. J. Schulze, V. Zago, G. Bilotta, and R. A. Dalrymple. "Localized Kernel Gradient Correction for SPH Simulations of Water Wave Propagation." Proceedings paper at the SPHERIC 2022 International Workshop. Available at arXiv:2511.10064

2021

V. Zago, L. J. Schulze, G. Bilotta, N. Al Mashaan, and R. A. Dalrymple. "Overcoming excessive numerical dissipation in SPH modeling of water waves." Coastal Engineering , 170, 104018 (2021)



Ph.D. Thesis

L. J. Schulze. "Mesh-free Methods for the Adaptive Discretization and High-order Approximation of Dynamic Surfaces." Ph.D. thesis, TU Dresden, October 2025. Available at Sächsische Landesbibliothek - Staats- und Universitätsbibliothek Dresden .

More stuff

Knowing a city means knowing its public transport, how hallmark places connect and how to traverse them quickly. Inspired by the Field Guide To Clouds App from the UCAR Center for Science Education , I created a similar app for public transport in Berlin and Munich. Click here to open it in a new tab, otherwise find it embedded here. Happy guessing :)

Contact Me