This interactive workshop was held at the Exploring the Status of Data-Driven Engineering Online Seminar, held on 25-26th of February 2026.

T​he Colab notebook is accessible only for the attendees of the event until September 2026, after that this will automatically become accessible to all NAFEMS Members.

Interactive Workshop from the Engineering Data Science Working Group

How can we identify a fatigue material card when properties change based on the 3D printing process? This interactive notebook provides a hands-on workflow for translating Additive Manufacturing (AM) process parameters into predictive fatigue life models.

What’s Inside?

This notebook follows the CRISP-DM framework to move from raw experimental data to a deployed model:

• Data Understanding: Visualizing S-N curves and surface condition correlations.

• Classical Modeling: Comparing Linear Regression, Random Forest, and XGBoost.

• Deep Learning: Building Neural Networks to capture non-linear material behavior.

• Physics-Informed ML (PINNs): A look at advanced models that enforce monotonic physics constraints (S-N behavior) during training.

How to Access

To view the code and run the simulations yourself without installing any software, use the link.

• When the page opens, click "Open in Playground" (top left) or "File > Save a copy in Drive" to run the code cells.

• No Python installation is required; the simulations run entirely on Google’s cloud server.