Learning Resources
Case Study

Integrating Rocscience Tools and Modelling DFN's Can Lead to Better Rock Slope Analysis

Published on: Apr 28, 2025 Updated on: Apr 28, 2025 Read: 8 minutes

Traditional approaches to rock slope stability analysis can help determine a factor of safety — but they don’t always reveal the full story. When slopes are heavily jointed or structurally complex, conventional methods may underestimate potential failure volumes or misrepresent the actual failure mechanisms at play.

Researchers Elvis Karikari Mensah, Reginald Hammah, Hassan Basahel and Hani Mitri recently investigated these limitations by comparing conventional kinematic and limit equilibrium methods with 3D discrete fracture network modelling across variable joint persistence and spacing.

In the case study below, you can see how using a combination of Rocscience tools — RS3, RocSlope3 and Dips — provides the perfect workflow to detect, define and model discrete fracture networks for complex slopes. Using two real-world road cuts, the research shows how this integrated approach delivers a more realistic view of slope behavior.

For a more detailed overview, read the full publication here.

The Geological Setting

The research focuses on two road-cut slopes situated along a steep, mountainous transportation corridors in the Jazan region of southwest Saudi Arabia. These road cuts are part of critical infrastructure connecting remote villages. The geology is part of the Arabian Shield, composed of fractured Precambrian syenite and metamorphosed rocks of the Proterozoic Sabya Formation. The area is characterized by steep slope angles, multiple discontinuity sets, and a high degree of structural deformation, including folds and faults that trend northwest.

Figure 1. A photograph of Site 1 (left), Site 2 (middle) and the study area’s simplified geological map (right). Reproduced from Mensah et al. (2024) under CC BY 4.0.
Figure 1. A photograph of Site 1 (left), Site 2 (middle) and the study area’s simplified geological map (right). Reproduced from Mensah et al. (2024) under CC BY 4.0.
Figure 2. The joint sets of the area generated using Dips. Reproduced from Mensah et al. (2024) under CC BY 4.0.
Figure 2. The joint sets of the area generated using Dips. Reproduced from Mensah et al. (2024) under CC BY 4.0.

The Research Challenge

Initial slope stability assessment using conventional kinematic and limit equilibrium analysis identified possible planar and wedge failures. However, the calculated factor of safety (FoS) did not fully align with field observations — particularly in areas where the slope appeared stable despite unfavorable joint orientations. What these methods lacked was the ability to account for joint persistence, spacing, and the volumetric nature of potential block failures — factors critical to accurate slope failure analysis.

Kinematic analysis using Dips revealed multiple intersecting joint sets with poles falling inside critical zones for planar and wedge sliding, indicating susceptibility to complex mechanisms Limit equilibrium models in SWedge under dry, wet, and seismic loading (see Figure 3 below) returned FoS values below design thresholds:

  • Dry: Design FoS 1.5, actual FoS 0.8737
  • Wet: Design FoS 1.1, actual FoS 0.7871
  • Seismic: Design FoS 1.05, actual FoS 0.8504

However, no significant instability had occurred historically — suggesting that key physical controls such as intact rock bridges and discontinuity trace length variability were being overlooked by traditional models.

Figure 3. The planar failure analysis for Site 1, conducted in Dips. Reproduced from Mensah et al. (2024) under CC BY 4.0.
Figure 3. The planar failure analysis for Site 1, conducted in Dips. Reproduced from Mensah et al. (2024) under CC BY 4.0.
Figure 4. Wedge failure analysis conducted for dry conditions (left) wet conditions (middle) and seismic conditions (right). Reproduced from Mensah et al. (2024) under CC BY 4.0.
Figure 4. Wedge failure analysis conducted for dry conditions (left) wet conditions (middle) and seismic conditions (right). Reproduced from Mensah et al. (2024) under CC BY 4.0.

The Solution

To bridge the gap between analysis and reality, researchers applied a discrete fracture network modelling approach using RS3 to construct 3D deterministic fracture models, then RocSlope3 to calculate the intersection of joints to find valid blocks and evaluate stability. Discrete fracture networks (DFNs) are explicit models of fracture sets, constructed using statistical data on orientation, spacing, and persistence.

Instead of simulating all discontinuities, the DFN models focused on critical joint sets responsible for the observed failure modes. By varying joint persistence and spacing, the team generated 24 models for Site 1 and 6 for Site 2 under dry, wet, and seismic conditions.

The DFN workflow involved building the slope geometry in RS3 and modelling each joint set as parallel planes with defined statistical inputs for spacing and trace length. The resulting DFNs were then imported into RocSlope3, then computations occured in two stages: (1) identification of potential block geometries via joint intersections, and (2) stability evaluation based on driving and resisting forces. Randomized seed values were also used to test model sensitivity and eliminate bias due to deterministic trace alignment.

Figure 5. The DFN modelling workflow from the research using RS3 and RocSlope3, from geometry construction and joint definition to block computation and stability analysis. Reproduced from Mensah et al. (2024) under CC BY 4.0.
Figure 5. The DFN modelling workflow from the research using RS3 and RocSlope3, from geometry construction and joint definition to block computation and stability analysis. Reproduced from Mensah et al. (2024) under CC BY 4.0.

The Results

The DFN results offered a more detailed and realistic picture of slope behavior:

  • When joints were fully persistent, the models captured block failures with a minimum FoS as low as .853 and significant failure volumes, matching what traditional analysis had predicted but providing additional volumetric insight.
  • When persistence dropped below 1, even minor changes rendered the slope stable, eliminating failure volumes and increasing FoS — aligning more closely with field conditions.
  • Increasing spacing only affected results at full persistence, confirming that intact rock bridges play a key role in real-world stability.
  • The DFN approach revealed how joint trace intersections can form failure wedges even when persistence appears low — insights that conventional tools miss.
  • Slope behavior across different joint orientations and trace limits further confirmed the DFN model's sensitivity and robustness.
Figure 6. The DFN model geometry for model 4 (2m spacing, 0.7 persistence) of Site 1 (top left) and the generated DFN (top right); the DFN model geometry for model 2 (0.9 persistence) of Site 2 (bottom left) and the generated DFN (bottom right). Reproduced from Mensah et al. (2024) under CC BY 4.0.
Figure 6. The DFN model geometry for model 4 (2m spacing, 0.7 persistence) of Site 1 (top left) and the generated DFN (top right); the DFN model geometry for model 2 (0.9 persistence) of Site 2 (bottom left) and the generated DFN (bottom right). Reproduced from Mensah et al. (2024) under CC BY 4.0.

Site 1: Larger Failures, Closer Match to Field Conditions

Failure volumes ranged from 8.75 m³ to 39.49 m³, depending on joint spacing (0.75 m–2.0 m) and dip direction limits. These were only observed when joints were fully persistent. At lower persistence values, no failures were recorded, and FoS values rose above critical thresholds, closely matching real-world performance.

Figure 7. Effect of joint spacing on minimum factor of safety under dry, wet, and seismic conditions, showing that only fully persistent joints result in instability across all spacing values. Reproduced from Mensah et al. (2024) under CC BY 4.0.
Figure 7. Effect of joint spacing on minimum factor of safety under dry, wet, and seismic conditions, showing that only fully persistent joints result in instability across all spacing values. Reproduced from Mensah et al. (2024) under CC BY 4.0.
Figure 8. Minimum factor of safety results at 0.8 joint persistence, showing the effect of randomization under dry, wet, and seismic conditions.
Figure 8. Minimum factor of safety results at 0.8 joint persistence, showing the effect of randomization under dry, wet, and seismic conditions.

Site 2: Smaller Failures, Consistent Pattern

For joint pair J1 and J3, a failure volume of 0.08 m³ and FoS of 0.60 were observed at full persistence. For J3 and J4, the model returned a failure volume of 1.36 m³ and FoS of 1.304. When persistence was reduced, failure volumes dropped to zero and FoS increased — demonstrating a consistent response to fracture continuity like Site 1.

Figure 9. Failure volume results for joint pair J3–J4 under varying persistence levels, illustrating that significant failure occurs only when joints are fully persistent. Reproduced from Mensah et al. (2024) under CC BY 4.0.
Figure 9. Failure volume results for joint pair J3–J4 under varying persistence levels, illustrating that significant failure occurs only when joints are fully persistent. Reproduced from Mensah et al. (2024) under CC BY 4.0.

The Verdict

By capturing the true nature of discontinuities — how they connect, how far they extend, and how they interact — discrete fracture network modelling offers a more realistic assessment of slope stability than conventional methods alone. When used together, the Rocscience software suite —RS3, RocSlope3, RocSlope2 and Dips — empowers engineers to go beyond a single factor of safety and begin visualizing and quantifying the actual failure mechanism. In this case, DFN modelling aligned better with field observations and highlighted the significance of joint persistence in understanding slope performance.

Try DFN Modelling in Your Own Slope Projects

Experience the next generation of rock mechanics software with DFN capabilities tailored for geotechnical professionals. Start your free trial of RS3, RocSlope3, and Dips today.

Learn More About Discrete Fracture Network Modelling

When to Use DFN Modelling in Your Project

  • Complex slopes with multiple joint sets.
  • Discrepancies between analysis and field observations.
  • Projects requiring volumetric failure insight or probabilistic analysis.
  • Advanced slope failure repair method planning.
  • Environments where joint persistence, trace length, and spacing are difficult to generalize.
  • Situations where failure volume quantification is needed for hazard assessment or mitigation planning.

How Much Field Data Do I Need to Build a Reliable DFN?

At a minimum, you need orientation, spacing, trace length, and persistence data for the key joint sets. The more site-specific measurements you have, especially for persistence and spacing, the more accurate your DFN results will be. Digital mapping techniques like LiDAR or photogrammetry can significantly improve data quality.

Rocscience Tools That Make It Possible

  • RS3: A powerful 3D finite element analysis program for complex geotechnical problems.
  • RocSlope3: A 3D block stability risk assessment software tailored to structurally controlled rock slope failures.
  • RocSlope2: A 2D limit equilibrium tool for analyzing rock slopes susceptible to wedge, planar, and toppling failure.
  • Dips: A stereonet-based program used for the analysis of orientation-based geological data.

Resources to Build Your DFN Skills

Ready to model discrete fracture networks?

Start your free trial of RS3, RocSlope2, RocSlope3, and Dips today!

Start Your Free Trial
Back to top

More from Rocscience