Empirical Design Charts: Barton-Bandis, Hoek-Brown Failure Criterion - Mining Engineering Knowledge & Tools Platform

Industry Applications

Hard-rock tunneling (e.g., Gotthard Base Tunnel), open-pit slope design, underground mine pillar stability, cavern storage

Key Standards

ISRM Suggested Methods (1995, 2007), ASTM D3148, Eurocode 7 Annex A.3

Typical Scale

Applied from meter-scale drifts to 20+ km tunnel alignments; parameters validated up to 1000 m depth

⚠️ Why It Matters

1

Inaccurate rock mass strength estimation

↓

2

Over- or under-designed support systems

↓

3

Unexpected roof collapse or wall spalling

↓

4

Costly remediation and schedule delays

↓

5

Increased risk to personnel and equipment

↓

6

Regulatory non-compliance and project stoppages

📘 Definition

Empirical design charts—specifically the Barton-Bandis Joint Roughness–Joint Wall Strength–Joint Condition (JRC–JCS–J a) system and the Hoek-Brown failure criterion—are semi-empirical, field-calibrated frameworks for estimating the strength and deformability of rock masses. They bridge intact rock properties with discontinuity characteristics (e.g., joint spacing, roughness, weathering) using dimensionless parameters derived from extensive in situ and laboratory testing. These charts enable rapid, practical estimation of rock mass strength parameters (e.g., σ_cm, m_b, s, a) without requiring complex numerical modeling at early design stages.

🎨 Concept Diagram

AI-generated illustration for visual understanding

💡 Engineering Insight

Never treat Hoek-Brown or Barton-Bandis as plug-and-play equations—these are *boundary-conditioned tools*. Their accuracy collapses when applied outside their calibration domain: Hoek-Brown fails for highly anisotropic or heavily fractured rock masses where GSI < 15, and Barton-Bandis overpredicts shear strength if JCS is estimated from hammer rebound rather than direct joint wall testing. Always verify with at least one back-analysis of a similar excavation.

📖 Detailed Explanation

Empirical design charts originated from the need to move beyond idealized intact rock mechanics into the messy reality of natural rock masses—where strength isn’t uniform but governed by fractures, weathering, and structure. Early practitioners like Terzaghi and Wickham recognized that rock mass behavior couldn’t be captured by lab-derived UCS alone; they introduced descriptive indices (e.g., RQD) that correlated strongly with field performance.

The Hoek-Brown criterion (1980, revised 1997, 2002, 2018) formalized this intuition into a power-law failure envelope: σ₁ = σ₃ + σ_ci (m_b σ₃/σ_ci + s)^a. Its brilliance lies in its scalability: m_i reflects intact rock quality, while m_b, s, and a degrade systematically with GSI and disturbance factor D. Meanwhile, Barton-Bandis (1974) addressed the dominant role of discontinuities, defining shear strength not by cohesion and friction angle, but by joint geometry (JRC) and wall strength (JCS)—parameters measurable in the field with simple tools.

Advanced application now integrates these charts with digital workflows: GSI is assigned via photogrammetric 3D joint mapping; Hoek-Brown parameters feed into discrete fracture network (DFN) models in UDEC or RS2; and time-dependent degradation (e.g., stress corrosion cracking in quartz-rich joints) is approximated using modified JRC decay functions. The 2023 Hoek-Brown update explicitly links D-factor to TBM advance rate and vibration spectra—bridging empirical charts with modern excavation metrics.

🔄 Engineering Workflow

Step 1

Step 1: Regional geologic mapping & structural trend analysis

→

Step 2

Step 2: Core drilling (HQ/NQ), detailed core logging (RQD, joint count, orientation, infill, weathering)

→

Step 3

Step 3: Lab testing (UCS, tensile strength, JCS, Schmidt hammer rebound, point load index)

→

Step 4

Step 4: Field GSI assessment (using ISRM GSI chart) and RMR calculation

→

Step 5

Step 5: Derive Hoek-Brown parameters (m_b, s, a) and Barton-Bandis τ–σ_n envelope

→

Step 6

Step 6: Calibrate against back-analyzed failures or monitoring data (convergence, bolt load cells)

→

Step 7

Step 7: Update design iteratively using digital twin feedback loop

📋 Decision Guide

Rock/Field Condition	Recommended Design Action
High GSI (>75), low joint frequency (<1/m), UCS > 100 MPa	Use Hoek-Brown with m_i = 17–25; apply empirical support charts (e.g., NGI Tunnel Support Guidelines); minimal systematic bolting required.
Low GSI (25–45), high joint density (>3/m), moderate groundwater inflow	Apply reduced m_b and s values per Hoek-Brown ‘disturbed’ rock mass; install patterned rebar dowels + wire mesh + shotcrete (100–150 mm).
Persistent planar joint set dipping toward excavation face (JRC < 4, JCS < 10 MPa)	Switch to Barton-Bandis shear analysis; install dowel-type rock bolts across joint strike; consider pre-splitting or directional drilling to avoid wedge failure.
RMR < 30, swelling clay seams present, high horizontal stress	Avoid empirical charts alone; require FLAC2D/Phase2 modeling with plastic softening; specify steel ribs + invert closure + stress-relief slots.

📊 Key Properties & Parameters

UCS

5–300 MPa (basalt ~250 MPa; chalk ~5 MPa)

Uniaxial Compressive Strength: peak axial stress a cylindrical rock specimen sustains under unconfined compression.

⚡ Engineering Impact:

Primary input for Hoek-Brown constant m_i and governs allowable stress ratios in tunnel crown design.

RMR

0–100 (poor: 0–20; fair: 40–60; good: 70–85; excellent: 90–100)

Rock Mass Rating: an index-based classification (0–100) integrating UCS, RQD, joint spacing, condition, and groundwater.

⚡ Engineering Impact:

Directly maps to Hoek-Brown m_b and s values via conversion tables; used to select initial support types and spacings.

JRC

0–20 (smooth slickensided = 0–2; very rough undulating = 16–20)

Joint Roughness Coefficient: a dimensionless number (0–20) quantifying surface irregularity of natural rock joints based on visual comparison or profile measurement.

⚡ Engineering Impact:

Controls Barton-Bandis shear strength (τ = σ_n tan[JRC log₁₀(JCS/σ_n) + JCS]); critical for slope stability along persistent bedding planes.

JCS

1–200 MPa (weathered shale ~1 MPa; fresh granite ~120 MPa)

Joint Wall Compressive Strength: uniaxial compressive strength of intact joint wall material, measured perpendicular to the joint plane.

⚡ Engineering Impact:

Determines dilational behavior and peak shear resistance in Barton-Bandis; low JCS increases potential for asperity degradation during sliding.

GSI

5–85 (disintegrated volcanic tuff ~5; massive quartzite ~85)

Geological Strength Index: a qualitative index (0–100) describing rock mass structure and surface condition, derived from field mapping and core logging.

⚡ Engineering Impact:

Primary input for Hoek-Brown m_b, s, and a exponents; dictates whether rock mass behaves as a continuum or discontinuum in numerical models.

📐 Key Formulas

Hoek-Brown Failure Criterion (2018)

σ₁ = σ₃ + σ_ci (m_b σ₃/σ_ci + s)^a

Estimates major principal stress at failure given minor principal stress, intact rock strength, and rock mass quality parameters.

Variables:

Symbol	Name	Unit	Description
σ₁	Major Principal Stress	MPa	Maximum principal stress at failure
σ₃	Minor Principal Stress	MPa	Minimum principal stress
σ_ci	Uniaxial Compressive Strength of Intact Rock	MPa	Intact rock strength
m_b	Modified Hoek-Brown Constant	dimensionless	Rock mass quality parameter related to Geological Strength Index and disturbance factor
s	Hoek-Brown Constant s	dimensionless	Rock mass quality parameter dependent on rock type and disturbance
a	Hoek-Brown Constant a	dimensionless	Exponent reflecting rock mass heterogeneity and stress dependence

Typical Ranges:

Hard massive rock (GSI=80)

m_b = 12–20, s = 0.001–0.005, a = 0.5–0.65

Weak foliated schist (GSI=25)

m_b = 0.02–0.08, s = 1e−6–5e−5, a = 0.45–0.55

⚠️ For GSI < 15, use alternative criteria (e.g., Mohr-Coulomb with reduced φ); avoid extrapolating s < 1e−7

🔧 Open Calculator

Barton-Bandis Shear Strength

τ = σ_n tan[JRC · log₁₀(JCS/σ_n) + JCS]

Predicts peak shear strength of a rock discontinuity under normal stress σ_n.

Variables:

Symbol	Name	Unit	Description
τ	Shear Strength	MPa or Pa	Peak shear strength of the rock discontinuity
σ_n	Normal Stress	MPa or Pa	Effective normal stress acting on the discontinuity
JRC	Joint Roughness Coefficient	dimensionless	Empirical parameter quantifying surface roughness of the discontinuity
JCS	Joint Wall Compressive Strength	MPa or Pa	Uniaxial compressive strength of the discontinuity wall rock

Typical Ranges:

Dry, tight joint (σ_n = 1–5 MPa)

τ = 1.2–8.5 MPa

Water-saturated, weathered joint (σ_n = 0.1–0.5 MPa)

τ = 0.15–1.1 MPa

⚠️ Not valid for σ_n > JCS; avoid if JCS/σ_n < 1 or > 1000 (use limit-state alternatives)

🔧 Open Calculator

🏭 Engineering Example

Cullinan Diamond Mine, South Africa

Kimberlite (hypabyssal volcanic pipe)

GSI

38

JCS

18 MPa

JRC

6.5

RMR

41

UCS

32 MPa

Hoek_Brown_s

0.00025

Hoek_Brown_m_b

0.12

🏗️ Applications

Tunnel face stability assessment
Pillar design in block caving
Slope reinforcement layout optimization
TBM disc cutter wear prediction

🔧 Try It: Interactive Calculator

🔧

Mine Ground Control & Rock Mechanics Calculator

Multi-tab calculator for Mine Ground Control & Roc...

Open →

📋 Real Project Case

Deep-Level Gold Mine Rockburst Mitigation

Mponeng Mine, South Africa — 4.2 km depth expansion

Challenge: Frequent high-energy rockbursts causing fatalities and equipment damage

Read full case study →

Limestone Quarry Slope Stabilization Underground Copper Mine Pillar Recovery Optimization Coal Mine Longwall Gate Road Support Upgrade

❓ Frequently Asked Questions

What is the difference between the Barton-Bandis and Hoek-Brown empirical design charts?

The Barton-Bandis system focuses specifically on rock joint behavior, using three dimensionless parameters—Joint Roughness Coefficient (JRC), Joint Wall Strength (JCS), and joint alteration condition (Ja)—to estimate peak shear strength along discontinuities. In contrast, the Hoek-Brown failure criterion is a rock mass-scale model that relates the uniaxial compressive strength of intact rock (σ_ci) to the equivalent compressive strength of the rock mass (σ_cm) via empirical constants (m_b, s, a), incorporating geological strength index (GSI), rock material constant (mi), and disturbance factor (D). While Barton-Bandis targets discontinuity shear response, Hoek-Brown estimates overall rock mass failure under compressive stress.

How are the JRC, JCS, and Ja values determined in practice?

JRC is estimated visually or quantitatively (e.g., profilometry) by comparing joint surface profiles to standard reference profiles (0–20 scale). JCS is measured via uniaxial compression tests on intact joint wall samples (typically 50 mm diameter cores), reported in MPa. Ja accounts for joint condition factors—including aperture, infilling, weathering, and water presence—and is assessed qualitatively using field classification tables (e.g., Barton’s 1974 chart), with typical values ranging from 0.75 (very tight, fresh, rough joints) to 4.0 (wide, clay-filled, weathered joints).

What role does the Geological Strength Index (GSI) play in the Hoek-Brown criterion?

GSI is a key input parameter in the Hoek-Brown criterion that quantifies the structural integrity and surface condition of a rock mass—integrating both joint geometry (e.g., spacing, continuity) and joint surface characteristics (e.g., roughness, alteration). It ranges from 0 (exceptionally poor, crushed rock) to 100 (intact, unjointed rock). GSI directly influences the reduced constants m_b and s, which govern the shape and position of the rock mass failure envelope; lower GSI values indicate higher discontinuity density or poorer joint conditions, resulting in significantly reduced rock mass strength.

Can these empirical charts replace numerical modeling in rock engineering design?

No—they are not substitutes for numerical modeling but rather complementary tools for rapid preliminary assessment and screening. Empirical charts provide first-order estimates of rock mass strength and deformability early in design (e.g., feasibility studies, conceptual tunnel support design), where site data is limited and computational resources are constrained. However, for detailed analysis—especially involving complex stress paths, time-dependent behavior, or non-linear interactions—numerical methods (e.g., distinct element or finite element modeling) remain essential, often calibrated *using* parameters derived from these empirical frameworks.

Why are Barton-Bandis and Hoek-Brown considered 'semi-empirical' rather than purely empirical or theoretical?

They are termed 'semi-empirical' because they combine physical mechanics principles (e.g., Mohr-Coulomb theory for Barton-Bandis; generalized failure envelopes inspired by Griffith and Coulomb theories for Hoek-Brown) with extensive field and laboratory calibration. Their functional forms include theoretically motivated exponents and scaling relationships, but the coefficients and correlations (e.g., JRC–JCS–Ja shear strength equation; GSI–m_b–s mapping) are derived from statistical fitting to real-world rock test data—not derived from first-principles physics alone. This blend ensures practical applicability while retaining mechanistic interpretability.

🎨 Technical Diagrams

📚 References

[1]

Hoek-Brown Failure Criterion – 2018 Edition — Canadian Geotechnical Society

[2]

The Shear Strength of Rock Joints in Theory and Practice — Rock Mechanics and Rock Engineering

[3]

ISRM Suggested Methods for the Quantitative Description of Discontinuities in Rock Masses — International Society for Rock Mechanics