🎓 Lesson 13
D5
Interpreting Fracture Density Maps from Blast Face Imagery
Fracture density maps show how many cracks or breaks are packed into a given area of a rock face after blasting, helping engineers see how well the blast broke the rock.
🎯 Learning Objectives
- ✓ Analyze drone-captured blast face imagery to identify and classify fractures using standardized protocols
- ✓ Calculate planar fracture density (fractures/m²) from annotated image segments with sub-pixel accuracy
- ✓ Explain how local fracture density variations correlate with blast design parameters (e.g., burden, spacing, delay timing)
- ✓ Apply fracture density thresholds to classify zones of under-break, optimal break, and over-break on a blast face map
📖 Why This Matters
In modern open-pit mines, 70–80% of operational cost inefficiencies stem from poor fragmentation — leading to crusher damage, secondary breaking, and haul truck downtime. Fracture density maps generated from post-blast drone surveys provide the first objective, quantifiable feedback on blast performance — not just 'did it break?', but 'how uniformly and finely did it break?'. This lesson bridges visual inspection with engineering analytics, turning pixels into actionable geotechnical insight.
📘 Core Principles
Fracture density mapping rests on three interdependent layers: (1) Image acquisition fidelity — resolution ≥ 2 mm/pixel at 30 m standoff ensures detectable fractures ≥5 mm wide; (2) Discontinuity delineation — using edge-detection (Canny), texture analysis (GLCM), and deep learning (U-Net) to segment fractures from intact rock; (3) Spatial normalization — converting pixel counts to real-world area via ground sampling distance (GSD) calibration and orthorectification. Critically, fracture density is not a scalar rock property — it reflects *blast-induced* discontinuities, not pre-existing joint sets alone; thus, interpretation must distinguish natural from induced fractures using timing context, orientation clustering, and energy deposition modeling.
📐 Planar Fracture Density Calculation
Planar fracture density (ρ_f) expresses the number of identifiable fractures per square meter of blast face area. It requires accurate GSD calibration and manual or semi-automated fracture tracing. The formula converts counted fractures in an image region to physical area using the drone’s calibrated pixel-to-meter ratio.
Planar Fracture Density (ρ_f)
ρ_f = N_f / A_faceQuantifies the number of identifiable fractures per unit area of exposed blast face.
Variables:
| Symbol | Name | Unit | Description |
|---|---|---|---|
| ρ_f | Planar fracture density | fractures/m² | Number of fractures intersecting a unit area of the blast face surface. |
| N_f | Total counted fractures | dimensionless count | Number of fracture traces ≥0.3 m long identified in the analyzed region. |
| A_face | Real-world face area | m² | Orthorectified area corresponding to the image region, calculated as (width_px × GSD) × (height_px × GSD). |
Typical Ranges:
Well-fragmented hard rock (e.g., granite): 12 – 20 fractures/m²
Moderately fractured sedimentary rock: 6 – 12 fractures/m²
Under-break zone (poor energy coupling): 0 – 4 fractures/m²
💡 Worked Example
Problem: A 4000 × 3000 pixel orthomosaic covers a 12 m × 9 m blast face section. Automated fracture detection identifies 1,842 fracture segments within that region. GSD = 3.2 mm/pixel (0.0032 m/pixel). Calculate ρ_f.
1.
Step 1: Compute real-world area: width = 4000 px × 0.0032 m/px = 12.8 m; height = 3000 px × 0.0032 m/px = 9.6 m → area = 12.8 × 9.6 = 122.88 m²
2.
Step 2: Apply formula: ρ_f = total_fractures / area = 1842 / 122.88
3.
Step 3: Round to two decimals: 1842 ÷ 122.88 ≈ 14.99 fractures/m²
Answer:
The planar fracture density is 15.0 fractures/m², which falls within the typical range for well-fragmented hard rock (12–20 fractures/m²).
🏗️ Real-World Application
At Newmont’s Boddington Mine (Western Australia), drone-based fracture density mapping revealed a 35% lower ρ_f zone (7.2 fractures/m²) in the toe region of a 15-m bench blast — correlating with excessive burden (5.2 m vs. design 4.0 m) and delayed bottom-hole initiation. Engineers revised the delay pattern and reduced burden by 0.8 m in the next round, increasing toe ρ_f to 16.3 fractures/m² and reducing >300 mm oversize by 62%. The map directly informed a 12% reduction in secondary breaking hours per blast cycle.