Added PINN vs FEM structural dynamics example

.gitignore

@@ -90,4 +90,5 @@ forumat.for
 build
 bin
 __pycache__
-*.dat
+*.dat
+.DS_Store

EXAMPLES/PINN_vs_FEM_example/README.md

@@ -0,0 +1,26 @@
+# PINN vs FEM Example
+
+This example compares a classical Finite Element Method (FEM) structural dynamics solution using OpenSeesPy against a Physics-Informed Neural Network (PINN) approximation implemented in PyTorch.
+
+## Features
+
+- **SDOF dynamic system**: Modeling a single-degree-of-freedom oscillator.
+- **Harmonic excitation**: Application of sinusoidal force to observe frequency response.
+- **FEM transient response**: Numerical integration using the OpenSeesPy framework.
+- **PINN approximation**: Neural network training constrained by the governing differential equations.
+- **Response comparison plots**: Visual validation of the neural network's accuracy against the FEM baseline.
+
+## Requirements
+
+The following libraries are required to run the comparison:
+
+- `openseespy`
+- `torch` (PyTorch)
+- `numpy`
+- `matplotlib`
+
+## Project Structure
+
+- `main.py`: Primary script to run both simulations and generate plots.
+- `models/`: Contains the PINN architecture and training logic.
+- `fem/`: OpenSeesPy model definitions and solver settings.