Research Tools
1. Data-Driven Modelling
Scikit-rom
scikit-rom is a lightweight, Python-based platform designed for projection-based model reduction of finite element models with moderate to large problem sizes (up to approximately 600,000 degrees of freedom). Built on top of scikit-fem, an easy-to-use finite element assembly library, this repository demonstrates reduced-order modeling (ROM) techniques for a range of linear and nonlinear problems, including time-dependent ones. Current examples primarily focus on thermal and mechanical systems, with extensions to fluid dynamics planned for future releases. Beyond its application to real-world finite element problems, the library also serves as a valuable testing ground for new model reduction algorithms and as a tutorial resource for those new to the field.
2. Mechanics of meta-materials
Curved Beam Element
This web application provides an interactive tool for computing the (12x12) stiffness and consistent mass matrices of open three-dimensional Timoshenko beams with straight, analytically defined, or point-based curved centerlines. Point-defined geometries can be reconstructed using Chebyshev or cubic-spline interpolation and inspected through an interactive 3D visualization. For validation, the stiffness matrix is compared with the analytical straight-beam solution or, for curved geometries, with a refined assembly of conventional straight Timoshenko beam elements whose internal degrees of freedom are statically condensed to the two beam endpoints.
3. Writing tool
Pub Assist
A workflow tool for preparing LaTeX manuscripts for journal submission, revision, and reviewer response. Pub Assist provides a local web app for browser-based access and a set of Jupyter notebooks for cell-by-cell control.
Texreader
This is a lightweight, web-based accessibility tool designed to make LaTeX documents more accessible by reading them aloud, effectively bridging the gap between complex mathematical typesetting and text-to-speech technology. The site allows users to paste or load LaTeX source code to instantly view the rendered document or compiled PDF. Its core feature is a built-in voice narration system with adjustable playback speeds, specifically tailored for academic content. Users can customize how mathematical expressions are vocalized by choosing between natural or literal reading styles, or they can opt to skip equations entirely for a smoother listening experience. Additionally, the platform includes a narration queue that provides step-by-step control over the reading sequence, making it an invaluable resource for visually impaired individuals or anyone who prefers auditory learning when reviewing scientific and mathematical documents.
PubQ Studio
PubQ Studio is a browser-based publication-quality scientific figure editor built with Python and Jupyter. It allows researchers and engineers to import data or create mathematical equations, then build highly customizable 1D, 2D, and 3D plots, multi-panel figures, scalar/vector/tensor field visualizations, annotations, curve fits, smoothing, and animations. The platform emphasizes publication-ready output, offering precise control over typography, axes, layouts, colors, legends, camera views, figure dimensions, DPI, and export settings. Figures can be exported in formats such as PDF, SVG, EPS, PNG, MP4, and GIF, with tools such as publication preflight, templates, and style controls to help produce consistent scientific graphics.