bjam-toolbox 2.2.0

Tools for Binder Jet Additive Manufacturing process analysis

BJAM Toolbox

Open-source tools for Binder Jet Additive Manufacturing (BJAM) process development.

This repository provides two complementary analysis tools designed to help researchers and operators assess and improve print quality:

1. Dimensional Accuracy Tool - Measure placement accuracy, feature size, and shape fidelity of printed calibration patterns
2. Ink Concentration Tool - Analyze ink deposition properties through region-of-interest measurements

Both tools use high-resolution flatbed scanners as the primary imaging device, providing micron-scale measurement capability at commodity cost.

---

Installation

Quick Install (all platforms)

pip install bjam-toolbox

From Source

git clone https://github.com/mattlmccoy/rfam-toolbox.git
cd rfam-toolbox
pip install .

Using the Install Scripts

Linux / macOS:

bash scripts/install.sh

Windows (PowerShell):

.\scripts\install.ps1

The install scripts check for Python 3.9+, tkinter, and install the package automatically.

Prerequisites

• Python 3.9 or higher (download)
• tkinter (usually bundled with Python)

If tkinter is missing:

Platform	Command
macOS	brew install python-tk
Ubuntu / Debian	sudo apt-get install python3-tk
Fedora / RHEL	sudo dnf install python3-tkinter
Windows	Reinstall Python and check "tcl/tk and IDLE"

Optional: PDF Support

To import scanned PDFs directly, install poppler:

# macOS
brew install poppler

# Ubuntu/Debian
sudo apt-get install poppler-utils

# Windows — download from https://github.com/osber/pdf2image

Then install the PDF extra: pip install bjam-toolbox[pdf]

For detailed platform-specific instructions, see INSTALL.md.

---

Quick Start

Launch the tool

bjam-toolbox

Or equivalently:

python -m bjam_toolbox

A dialog will ask which workflow you want to use:

• Yes → Dimensional Accuracy Analysis
• No → Ink Concentration Analysis

Verify Installation

bjam-toolbox --version

---

Features

Dimensional Accuracy Analysis

• Analyze four calibration feature types:
  • Dot arrays: Measure placement accuracy, dot diameter, circularity, and eccentricity
  • Checkerboards: Extract rotation (yaw) and scale with edge-based detection
  • Concentric rings: Stress-test edge fidelity and linewidth consistency
  • Pitch rulers: Characterize minimum resolvable feature size
• Generate compensation recommendations (scale factors, rotation corrections)
• Produce health indicators that flag when process tuning is needed
• Export standardized CSV data products for tracking calibration over time

Ink Concentration Analysis

• Interactive region-of-interest (ROI) selection with polygon, circle, and ruler tools
• Compute intensity metrics (mean, median, std, IQR, skewness, kurtosis, entropy)
• Compute shape metrics (circularity, perimeter, convexity, inertia ratio)
• Halo/spreading analysis via eccentricity measurements
• Generate statistical plots and heatmaps

---

Workflows

Dimensional Accuracy Workflow

1. Select image: Load a scanned calibration pattern (PNG, JPG, TIF, or PDF)
2. Enter scanner DPI: The tool calculates pixel-to-mm conversion automatically
3. Enter metadata: Sample ID, ink type, feathering percentage, gantry speed
4. Select features: Check which features to analyze (dots, checkerboard, rings, pitch rulers)
5. Draw ROIs: For each feature, draw a bounding box around the region to analyze
6. Review results: The tool generates:
   • results.csv: Summary metrics
   • raw.csv: Per-instance measurements
   • Debug overlay images
   • Dimensional health report with compensation recommendations

Ink Concentration Workflow

1. Select image: Load a scanned ink deposition sample
2. Draw ROIs: Use keyboard shortcuts to draw regions:
   • P: Polygon mode (click vertices, right-click to close)
   • C: Circle mode (hold and drag)
   • L: Ruler mode (calibrate pixel-to-mm scale)
3. Label ROIs: Press 1-4 to assign ink type labels
4. Toggle metrics: Press I/S/H for intensity/shape/halo analysis
5. Generate results: Press Q to finish and export

---

Repository Structure

rfam-toolbox/
├── bjam_toolbox/              # Main package
│   ├── __main__.py            # python -m bjam_toolbox entry point
│   ├── launcher.py            # Unified entry point
│   ├── dimensional/           # Dimensional accuracy analysis
│   │   ├── gui.py             # GUI for dimensional workflow
│   │   ├── analysis.py        # Feature analysis algorithms
│   │   └── geometry.py        # Test pattern generation
│   ├── ink_concentration/     # Ink analysis
│   │   ├── main.py            # Workflow orchestration
│   │   ├── gui.py             # ROI selection interface
│   │   ├── analyzer.py        # Metric computation
│   │   └── plots.py           # Visualization
│   └── common/                # Shared utilities
│       ├── dataio.py          # CSV export
│       └── utils.py           # Helper functions
├── geometries/                # Reference calibration patterns
│   └── gold_standard_patterns/
├── scripts/                   # Install helper scripts
│   ├── install.sh             # Linux / macOS installer
│   └── install.ps1            # Windows installer
├── .github/workflows/         # CI/CD
│   ├── ci.yml                 # Test matrix (Linux, macOS, Windows)
│   └── publish.yml            # Auto-publish to PyPI on release
├── pyproject.toml             # Package configuration
├── requirements.txt
├── Makefile
├── LICENSE
├── README.md
├── INSTALL.md                 # Detailed platform install guide
└── CONTRIBUTING.md            # Developer guide

---

Output Data Products

Each analysis run creates a timestamped folder under session_data/ containing:

File	Description
results.csv	Summary metrics (one row per feature type)
raw.csv	Per-instance measurements (e.g., each dot's diameter)
dimensional_summary.txt	Human-readable summary
dimensional_compensation.txt	Recommended calibration adjustments
*_overlay.png	Visualization of detected features
*_roi.png	Cropped region of interest
source_image.png	Copy of input image

---

Calibration Pattern Design

The tool is designed to work with calibration patterns containing:

• Dot array: 5x5 grid of 2mm diameter dots at 6mm pitch
• Checkerboard: 8x8 pattern of 2mm squares
• Concentric rings: 20 rings with 0.5mm line width and 0.5mm spacing
• Pitch rulers: Bars with widths from 0.1mm to 6mm

You can generate custom patterns using:

from bjam_toolbox.dimensional.geometry import generate_pattern
generate_pattern(dpi=4800, output_path="my_pattern.png")

Pre-generated gold standard patterns at various DPI values are included in geometries/gold_standard_patterns/.

---

Understanding the Health Report

The dimensional health report separates two types of outputs:

Compensation Recommendations

• Scale factors: Multiply X/Y coordinates to correct placement errors
• Rotation correction: Angular adjustment to correct yaw misalignment
• These correct motion-system calibration issues

Health Indicators

• Normalized circularity: Values below ~0.9 indicate spreading/coalescence
• Eccentricity: Non-zero values indicate directional stretching
• Ring variability: High values indicate edge quality issues
• These flag when process tuning (not calibration) is needed

---

Citation

If you use this tool in your research, please cite:

@article{mccoy2025scanner,
  title={A Low-Cost Scanner-Based Diagnostic Pipeline for Dimensional Metrology
         of Jetting Fidelity in Binder Jet Additive Manufacturing},
  author={McCoy, Matthew L. and Salda{\~n}a, Christopher J. and Seepersad, Carolyn C.},
  journal={Additive Manufacturing},
  year={2025}
}

---

Contributing

Contributions are welcome! See CONTRIBUTING.md for development setup and guidelines.

---

License

This project is licensed under the MIT License - see the LICENSE file for details.

---

Acknowledgments

This work was supported by the National Science Foundation under Grant No. DGE-2039655.

Contact

Matthew L. McCoy - [e-mail]

George W. Woodruff School of Mechanical Engineering Georgia Institute of Technology