There is a Better Way to Automate and Manage Your (Fluid) Simulations

DVC (Data Version Control) is an open-source command-line tool that helps manage simulation data pipelines and version large files that Git cannot handle effectively

Key workflow components:

• Use Python as a control/glue language to automate simulation software
• Track and manage simulations with DVC
• Create rich comparisons in JupyterLab
• Sync PowerPoint presentations with simulation data

DVC advantages:

• Agnostic regarding simulation software vendors
• Handles large files efficiently
• Tracks dependencies between pipeline stages
• Only re-executes necessary stages when changes occur
• Enables experiment tracking and comparison

Pipeline automation features:

• Declarative approach using DVC.yml files
• Can specify software versions
• Manages input parameters and output metrics
• Supports parallel execution
• Integrates with existing Git workflows

PowerPoint integration capabilities:

• Maintains unique shape names for reference
• Allows manual updates while keeping automation
• Creates updatable custom templates
• Syncs presentations with new simulation data
• Preserves existing team workflows

Storage and hosting:

• Uses S3, Azure blob storage, SSH servers or local folders
• No dedicated hosting required
• Integrates with existing Git infrastructure
• Manages large simulation files separately from Git

The approach aims to:

• Improve iteration speed
• Increase transparency and traceability
• Reduce effort per simulation
• Standardize reporting while maintaining flexibility
• Enable better parameter studies and optimization