Appearance

Introduction to the Computational Simulation Workflow ​

At Avkalan Labs, computational engineering forms the backbone of how we design, validate, and optimize ideas. Every simulation we run — whether it’s a simple structural model or a full-scale multi-physics analysis — follows a structured workflow. This workflow ensures that our results are reliable, reproducible, and scalable.

The Typical Process We Follow ​

A complete computational study usually goes through three major stages:

  1. Pre-processing – creating the geometry and mesh
  2. Solving – defining the physics and running the computation
  3. Post-processing – visualizing and analyzing the results
  4. Communication and documentation – recording your observations, sharing updates, and maintaining clarity in your work
  5. Collaboration for research output – working together across teams, reviewing results, and combining efforts to achieve a higher research impact

Let’s go through each one briefly.

1. Geometry and Meshing – Salome ​

We start by building the geometry and the mesh using Salome. Salome is a versatile pre-processing tool designed for CAD modeling and high-quality mesh generation. It allows us to create simple parts, complex assemblies, and parametric models, all within the same environment.

From defining geometric entities to assigning region markers and generating 2D or 3D meshes, Salome ensures that the model is simulation-ready. The output meshes are then exported (typically in .med or .xdmf format) for use inside FEniCS.

2. Simulation and Solving – FEniCS ​

Once the mesh is ready, we move to FEniCS, our main solver engine. FEniCS is a powerful open-source platform that allows us to define and solve partial differential equations (PDEs) with high efficiency.

With FEniCS, we can scale our simulations to handle systems with millions or even billions of degrees of freedom, depending on the problem. It also supports a wide range of physics:

  • Structural and mechanical analysis
  • Dynamic and modal analysis
  • Heat transfer and thermo-mechanical coupling
  • Electromagnetic and electro-thermal problems
  • Magneto-mechanical and other multi-physics systems

This flexibility makes FEniCS the computational core of most of our projects.

3. Post-Processing – ParaView ​

After the solution is complete, we use ParaView for post-processing and visualization. ParaView provides a rich set of tools for slicing, plotting, animating, and analyzing simulation data. It can handle both 2D and 3D results efficiently, and integrates well with our .xdmf output files.

From quick visual checks to detailed quantitative analysis, ParaView gives us the clarity we need to interpret the results and make design decisions.

4. Communication and Documentation – Slack, Obsidian, and GitHub ​

Once the technical work is done, we move to what holds it all together — communication and documentation. At Avkalan Labs, every study is documented, shared, and discussed openly. We use Slack for real-time communication, Obsidian for structured note-keeping, and GitHub for version control and collaborative documentation.

Good documentation ensures that anyone on the team can trace the logic behind a study, reproduce results, and build upon your work. It’s not just about writing notes — it’s about capturing decisions, reasoning, and outcomes so that our collective learning compounds over time.

5. Collaboration for Research Output – Working Together Across Teams ​

Research thrives on collaboration. Once individual studies are well-documented and validated, they’re shared across teams to connect ideas and accelerate progress. We often combine simulation insights, experimental data, and design inputs from different groups to arrive at stronger, more comprehensive conclusions.

Through shared repositories, cross-team reviews, and regular project discussions, we ensure that each piece of work contributes meaningfully to the larger research narrative. This collaborative approach is what allows Avkalan Labs to consistently turn isolated studies into high-impact research outcomes.

The Flow of the Course ​

In this brief training course, we’ll go through each part of this workflow step by step. By the end, you’ll have a complete understanding of how to set up, run, and analyze simulations — the same way we do at Avkalan Labs.

Here’s the flow we’ll follow:

  1. Getting started with the environment setup — installing WSL, Ubuntu, and setting up FEniCS using Conda or Docker
  2. Understanding Docker and containerization — building reproducible simulation environments and managing containers
  3. Setting up the essential tools — configuring VS Code, GitHub, and Obsidian for your daily workflow and documentation
  4. Building good communication and documentation habits — sharing progress, writing commit messages, and maintaining clear records of your work
  5. Learning terminal and Linux essentials — navigating the file system, managing packages, and automating tasks using bash scripts
  6. Pre-processing with Salome — creating geometries, performing meshing, and preparing simulation-ready models
  7. Post-processing with ParaView — loading, visualizing, and analyzing simulation results effectively
  8. Bringing it all together through workflow integration — connecting all tools into a seamless end-to-end process and automating repetitive steps
  9. Working on cloud and remote servers — setting up SSH access, running remote jobs, and visualizing results from server environments
  10. Applying everything through exercises and projects — completing guided tasks, simulations, and a capstone project to put your learning into practice

Each module will include step-by-step instructions, reference commands, and small practice tasks. By the time you finish this course, you’ll have a complete system setup — ready to build, run, and document your own simulations confidently.