adtzlr / contique Goto Github PK

Numeric continuation of equilibrium equations

License: GNU General Public License v3.0

Python 100.00%

arc-length scientific-computing python mathematics solid-mechanics numeric-continuation

contique's Introduction

Hi there 🖐️,

this is Andreas, a mechanical engineer graduated from Graz University of Technology, based in 🏰⛰️ Graz, Austria 🇦🇹. In my free time, I like running 🏃‍, skiing ⛷️ and snowboarding 🏂 while I also enjoy family times at home 👨‍👩‍👧‍👦.

Currently, I'm an engineer in industry (during the day) and a PhD student at Graz University of Technology at the Institute of Structural Durability and Railway Technology (well, at night... 📚 🕯️). All the tools related to my scientific work are available here on my GitHub profile.

I'm the author of 🔍 FElupe, an open-source finite element analysis package focussing on the formulation and numerical solution of nonlinear problems in continuum mechanics of solid bodies. Most of the open source finite element packages I found are either super-difficult to install, needs to be compiled or are great but slow (or at least too slow for my needs).

With FElupe, I try to fill a gap in between.

I'm convinced that static input files 🖨️ which are passed to a standalone fea solver 🖩 are a thing of the last decades 💾. Instead, scripts are input files: easy to adopt scripts with access to third-party libraries 🛒, written in common scripting languages are the way to go. With common languages I mean something easy-to-learn for engineers, like Python, Matlab/Octave or Julia, not another proprietary simulation file format. FElupe is just another one of many open-source finite element analysis packages using this approach. Well defined and public available scripting interfaces hopefully accelerate the introduction of flexible natural language-processing for simulations.

Projects

🔧 Mechanics: felupe, hyperelastic, matadi, trusspy, contique, cubrium, fiberreinforcedrubber
🔢 Math: tensortrax, ttb, neinsum

Code Snippets

contique's People

Contributors

Stargazers

Watchers

Forkers

kinow

contique's Issues

Use a generator instead of a list for collecting the results

Improve memory usage

Add user-defined pre-modification of unknowns as well as update and check functions in Newton-Rhapson iterations

This should involve

preprocess the unknowns y = pre(x) and by consequence, calculate f as f(y, *args); this defaults to pre = lambda x: x
update the solution x = update(x, dx); this defaults to lambda x, dx: x + dx
check if success success = check(dx, x, f); this defaults to check = lambda dx, x, f, tol=np.finfo(float).eps: np.linalg.norm(dx) < tol

With this modifications it should be possible to combine contique and felupe. The resulting newtonrhapson function should be very abstract and hence, the code should have good readability.

Use a Generator instead of a list of Results

[Enhancement] Add constant part of external forces

For now (v0.1.7) it is only possible to prescribe external quantities and apply a lpf-value to a certain type of these quantities (e.g. normal forces). It would be great to seperate constant and lpf-associated properties. This would especially be of interest for the kinematic-type of lpf, where some gridvector components should be fixed.