Rock mechanics and ground control

In short

Rock mechanics is the study of how rock and rock masses deform and fail under stress; ground control is the engineering use of that knowledge to keep mine roofs, pillars, ribs, benches and dumps standing. Every underground gallery, every pillar left for support, every opencast bench angle and every dump height in India is sized from rock-mass strength, in-situ stress and a factor of safety. It is examined in GATE (Mining Engineering) under "Rock mechanics / ground control" with numerical questions on pillar stress, RMR and slope stability, and in university papers under strata control. Employers — Coal India subsidiaries (ECL, BCCL, CCL, WCL, SECL, MCL, NCL), SCCL, NLC, Hindustan Zinc, Tata Steel, Vedanta, NMDC, Hindustan Copper — test it in interviews for mining engineer, strata-control officer and geotechnical roles, and DGMS requires support design to be backed by it.

Explained

Rock material properties and laboratory tests

• Intact rock vs rock mass: a laboratory specimen is intact rock; the rock *mass* includes joints, bedding, faults and water, and is far weaker and more deformable. Classification systems exist to bridge the two.
• Uniaxial compressive strength (UCS, σc): cylinder of length/diameter 2–3 loaded to failure; Indian coal 10–40 MPa, shale/sandstone roof 20–80 MPa, quartzite/basalt 100–300 MPa. Tested per IS 9143 (and ISRM Suggested Methods); specimen preparation per IS 9179.
• Tensile strength: Brazilian (indirect) test, σt ≈ UCS/10 to UCS/20 — rock is weak in tension, which is why roof beams fail by sagging tension at the mid-span and corners.
• Point load strength index (Is50): quick field/core test (IS 8764); UCS ≈ 20–25 × Is50. Useful when cores are too short for UCS.
• Triaxial test: gives cohesion c and friction angle φ from a σ1–σ3 plot; essential for Mohr-Coulomb design.
• Elastic constants: Young's modulus E and Poisson's ratio ν from stress-strain curves (IS 9221); also dynamic E from sonic velocity. Rock-mass modulus is 10–50% of intact E.
• Other indices: Protodyakonov, slake-durability (weatherability of shales), swelling index, Brazilian disc, shear box on joints (JRC-JCS joint model).
• Stress-strain behaviour: elastic → yield → peak → post-peak. Brittle rock sheds strength suddenly (bump/burst risk); ductile/strain-softening rock (coal at depth, soft shale) yields gradually. A soft loading system relative to pillar stiffness makes violent failure more likely.

In-situ stress and stress around openings

• Vertical stress σv = γH, about 0.025 MPa per metre of depth (25 kN/m³ overburden) — 5 MPa at 200 m, 15 MPa at 600 m.
• Horizontal stress σh = k σv. Elastic (gravity-only) theory gives k = ν/(1−ν) ≈ 0.25–0.43, but measurements almost always give more. Sheorey's thermo-elastic model, calibrated on Indian data, adds a geothermal term and predicts k falling with depth — high ratios (1.5–2.5) near surface, tending towards 1 at great depth. Indian coalfields commonly show horizontal stress exceeding vertical above ~300 m, which causes roof guttering and cutter roof along one direction of galleries.
• Stress measurement methods: flat jack (slot and repressurise), hydraulic fracturing in boreholes, overcoring with USBM borehole deformation gauge or CSIRO hollow inclusion cell, borehole slotter, and back-analysis of convergence.
• Kirsch solution (circular opening, elastic, 2-D) gives stress concentration at the opening wall: roof/floor tangential stress = σv(3k − 1), sidewall = σv(3 − k). If k < 1/3 the roof goes into tension and delaminates — the classic argument for bolting a flat-bedded roof.
• Rectangular galleries concentrate stress at corners (hence rounded or arched profiles); openings closer than about 2–3 diameters interact and stress-superpose.
• Pressure arch / abutment: above a wide extracted area the load transfers to solid abutments; partial extraction relies on this, and the arch collapses once the span exceeds roughly the depth.

Rock-mass classification

• RQD (Rock Quality Designation, Deere): sum of intact core pieces ≥100 mm divided by total core run length, in %. 0–25 very poor, 25–50 poor, 50–75 fair, 75–90 good, 90–100 excellent.
• RMR (Bieniawski, 1989) — six parameters summed to 100: UCS, RQD, discontinuity spacing, discontinuity condition, groundwater, and an adjustment for discontinuity orientation. Classes I (81–100 very good) to V (<21 very poor); gives stand-up time, support pressure and rock-mass modulus.
• CMRI-ISM RMR (CIMFR, Dhanbad — the classification actually used for Indian coal-mine roofs): five parameters, each 0–20, totalling 100 — layer thickness, structural features, weatherability, roof-rock compressive strength and groundwater seepage; then adjusted for depth, stress and induced effects. Classes: very poor (0–20), poor (20–40), fair (40–60), good (60–80), very good (80–100). Rock load and bolt pattern follow from the adjusted RMR.
• Q-system (Barton, NGI): Q = (RQD/Jn)(Jr/Ja)(Jw/SRF) — block size, inter-block shear strength and active stress. Used with ESR/span to read support off the Q-chart; widespread in tunnelling and metal mines.
• GSI (Geological Strength Index): visual estimate of blockiness and joint-surface condition, feeding the generalised Hoek-Brown criterion for numerical modelling.
• SMR (Romana) adapts RMR to slopes with factors for joint-vs-slope orientation and excavation method.
• Limitations: all are empirical, derived from a particular database; they cannot capture a single adverse feature (a fault, a slickensided plane, a water-charged sandstone channel) that governs the actual failure.

Failure criteria

• Mohr-Coulomb: τ = c + σn tanφ — simple, linear, used for joints, slopes and pillars; overestimates strength at high confinement and predicts too much tensile strength.
• Hoek-Brown: non-linear, curve-fitted to rock masses via GSI, mi and σci; the standard for modern slope and cavern analysis.
• Griffith (tensile crack theory) and the Mohr envelope with tension cut-off are the exam answers for why rock fails in tension at a fraction of its compressive strength.

Underground ground control: pillars, support and caving

• Bord-and-pillar development: galleries (statutory width not exceeding 4.8 m in Indian coal mines) form pillars. The Coal Mines Regulations prescribe a minimum pillar centre-to-centre distance as a table of depth against gallery width; the strength-based design must satisfy that statutory minimum as well, so always check the current schedule rather than relying on the formula alone.
• Pillar design steps: (1) estimate pillar stress by tributary area; (2) estimate pillar strength (CMRI/Sheorey, Obert-Duvall, Bieniawski, Salamon-Munro); (3) FoS = strength/stress — typically ≥1.5 for development pillars meant to stand long-term, 1.0–1.3 for pillars to be extracted or short-lived, 2.0+ under surface structures or water bodies; (4) check extraction ratio, roof span and floor strength (punching/heaving in soft fireclay floors).
• Failure modes: pillar spalling and necking, progressive "domino" collapse of a panel, floor heave, roof fall (local fall, massive fall), rib spalling, sidewall slabbing, coal bump/rock burst in strong brittle strata at depth.
• Support types: timber/steel props, chocks and cogs, roof bolts (full-column resin/cement grouted, mechanically anchored slot-and-wedge, friction-anchored), W-straps, wire mesh, cable bolts, roof stitching, steel arches and girders, shotcrete/fibrecrete (metal mines), hydraulic props, and powered roof supports (chock shields) in longwalls.
• Roof-bolting mechanisms: suspension (hang a weak immediate roof from a competent bed above), beam building (clamp thin laminations into one thick beam — stiffness rises with the cube of thickness), keying of blocky roof, and friction/shear reinforcement across bedding.
• Systematic Support Rules (SSR): the Coal Mines Regulations, 2017 require an approved SSR for each district — specifying bolt type, length, capacity, grid spacing, maximum unsupported width/length after a blast, and support of junctions, which carry the largest span and therefore the heaviest rock load.
• Caving and goaf: after extraction the immediate roof caves, bulks and fills the void. Caving height ≈ extracted thickness/(bulking factor − 1); strong massive sandstone/ sill roofs hang up, over-run and cause air blasts — controlled by induced blasting or by adopting a different method (stowing).
• Depillaring against a line of goaf, with splitting and slicing, is where most Indian strata-control accidents occur; hydraulic sand stowing is used where surface features must be protected.

Subsidence and surface protection

• Mechanism: extraction → caving → bending of strata → surface trough, with vertical subsidence, tilt, curvature, horizontal displacement and strain; tensile strain at the trough edge cracks buildings, compressive strain at the centre buckles rails and pipelines.
• Maximum subsidence S = a·m, where m is extracted thickness and a the subsidence factor — in Indian coalfields roughly 0.3–0.8 (often taken ~0.6) for caving, and far less for stowing (0.05–0.25). It varies with strata, so use measured local values where available.
• Critical width: full subsidence develops only when panel width ≳ 1.2–1.4 × depth; narrower (subcritical) panels subside less. Angle of draw 20–35° defines the surface area affected.
• Protection: leaving barriers/support pillars under rivers, railways, buildings and against waterlogged workings; harmonic/partial extraction; stowing; structural stiffening. DGMS permission is required to work under or near surface features and water bodies.

Opencast slope and dump stability

• Geometry terms: bench height, bench width, bench face angle, overall (interramp) slope angle, berm, haul-road width.
• Failure modes: planar (plane) sliding on a daylighting joint, wedge on two intersecting planes, circular/rotational in weak or soil-like material and dumps, toppling of steep adversely dipping beds, and rockfall.
• Kinematic admissibility (stereonet): a plane slides only if it dips less than the face but more than its friction angle, and strikes within about ±20° of the face.
• Water is the chief destabiliser: pore pressure reduces effective normal stress; horizontal drains, perimeter drains and depressurisation wells are the cheapest stabilisation. A tension crack filled with water adds a large driving thrust.
• Design FoS: commonly 1.2–1.3 for temporary/individual benches, 1.3–1.5 for overall pit slopes, and higher where a haul road, crusher or township lies above. DGMS circulars require slope-stability study and monitoring for deep opencast mines; dump height, dump slope and the foundation on which a dump is raised are specifically regulated.
• Blast damage behind the final wall is limited by pre-split/trim (cushion) blasting and reduced charge per delay — it enters the Hoek-Brown disturbance factor D.

Instrumentation and monitoring

• Roof/strata: telltale (dual-height roof extensometer — the mandated low-cost warning device in bolted coal roofs), multi-point borehole extensometer, convergence indicator/tape, sag/strata movement recorder, bore-scope for bed separation.
• Load and stress: bolt load cells, hydraulic/vibrating-wire stress cells, prop load cells, pressure cells in powered supports.
• Slope: total-station prism monitoring, extensometers/crackmeters, inclinometers, piezometers for water level, GNSS, slope-stability radar and InSAR.
• Seismic: microseismic/acoustic-emission monitoring to detect pre-failure cracking and bump-prone zones.
• Rule of practice: instruments give *rates*. An accelerating convergence or sag rate, not an absolute number, is the evacuation trigger. Readings must be plotted and reviewed, with threshold/alarm levels fixed in advance.
• Statutory frame: the Occupational Safety, Health and Working Conditions Code, 2020 (in force 21 November 2025), which repealed the Mines Act, 1952, and the Coal Mines Regulations, 2017 / Metalliferous Mines Regulations, 1961, continued under its savings clause (section 143(3)) and administered by DGMS, govern support, pillar sizes and slope/dump design, with DGMS technical circulars on bolting, SSR and opencast slopes. Draft successor regulations under the Code were published in 2026 — confirm the current regulation numbering from DGMS before quoting one.

Key points

• Intact rock vs rock mass - joints, not the specimen, decide how a mine opening behaves.
• RQD - % of core in pieces ≥100 mm; a jointing index only, blind to joint condition and water.
• RMR (Bieniawski) - six parameters to 100, gives class, stand-up time and support pressure.
• CMRI-ISM RMR - five parameters (layer thickness, structure, weatherability, strength, water), the Indian coal-roof standard, used to fix bolt length and grid.
• Q-system - block size × joint shear strength × active stress; Q-chart gives support for a given span.
• GSI - visual index feeding generalised Hoek-Brown for numerical models.
• σv = γH ≈ 0.025 MPa/m; horizontal stress often exceeds vertical at shallow depth in India, causing cutter roof.
• Kirsch - roof stress = σv(3k − 1): tension in the roof when k < 1/3.
• Tributary area - pillar stress rises as the square of (pillar + gallery)/pillar width.
• FoS - ≥1.5 for standing development pillars, 1.0–1.3 for pillars to be extracted, higher under surface features.
• Roof bolting - works by suspension, beam building and keying; full-column grouted bolts are standard in Indian coal.
• SSR - DGMS-approved Systematic Support Rules are mandatory; junctions need extra support.
• Caving height ≈ m/(bulking factor − 1); massive hanging roof causes air blast.
• Subsidence S = a·m, a ≈ 0.3–0.8 for caving in Indian coalfields; full trough needs panel width ≳1.2 × depth.
• Slope failure modes - planar, wedge, circular, toppling; water pressure is the usual trigger.
• Mohr-Coulomb vs Hoek-Brown - linear joint/slope criterion vs non-linear rock-mass criterion.
• Telltale - the simplest and most effective roof-fall warning; watch the rate of sag, not the value.
• An automated or empirical "safe" answer is not a conclusion - confirm with instrumentation and geology.

Formulas and examples

Example 1 — RQD and Q from a core run. A 2.0 m run yields intact pieces of 250, 180, 420, 120, 310 and 150 mm (rest fragmented).

• Σ(≥100 mm) = 250+180+420+120+310+150 = 1,430 mm; RQD = 1,430/2,000 = 71.5% → "fair".
• With three joint sets (Jn = 9), rough planar joints (Jr = 1.5), slightly altered coatings (Ja = 2), dry (Jw = 1), moderate stress (SRF = 1):

Q = (71.5/9) × (1.5/2) × (1/1) = 7.94 × 0.75 = 5.96 ≈ 6 → "fair" rock mass.

Example 2 — point load to UCS. Diametral test on NX core, De = 54 mm, failure load P = 8.5 kN.

• Is = 8,500/54² = 8,500/2,916 = 2.915 MPa.
• Size correction F = (54/50)^0.45 = 1.08^0.45 = 1.035 → Is50 = 2.915 × 1.035 = 3.02 MPa.
• UCS ≈ 22 × 3.02 ≈ 66 MPa — a competent sandstone roof.

Example 3 — triaxial data to c and φ. Triaxial tests give σ1 = 45 MPa at σ3 = 5 MPa and σ1 = 75 MPa at σ3 = 15 MPa.

• Slope = (75 − 45)/(15 − 5) = 3 = (1+sinφ)/(1−sinφ) → sinφ = 0.5, φ = 30°.
• Intercept (σ3 = 0): UCS = 45 − 3×5 = 30 MPa.
• c = UCS(1 − sinφ)/(2cosφ) = 30 × 0.5/(2 × 0.866) = 8.66 MPa.

Example 4 — pillar design (bord-and-pillar). Depth H = 200 m, seam height h = 3 m, gallery B = 4.8 m, γ = 0.025 MN/m³, coal σc = 25 MPa. Trial pillar w = 20 m (square).

• σp = 0.025 × 200 × (24.8)²/(20)² = 5 × 1.5376 = 7.69 MPa.
• S = 0.27 × 25 × 3^−0.36 + (200/250 + 1)(20/3 − 1) = 6.75 × 0.6733 + 1.8 × 5.667 = 4.55 + 10.20 = 14.75 MPa.
• FoS = 14.75/7.69 = 1.92 — comfortably above 1.5, so the pillar can be slimmed to win back extraction.
• Retry w = 16 m: σp = 5 × (20.8/16)² = 5 × 1.690 = 8.45 MPa; S = 4.55 + 1.8 × (16/3 − 1) = 4.55 + 7.80 = 12.35 MPa; FoS = 1.46 — just short. At w = 17 m: σp = 8.22 MPa, S = 12.95 MPa, FoS = 1.58 — acceptable.
• Extraction ratio rises from 1 − 1/1.5376 = 35% (w = 20 m) to 1 − 1/1.644 = 39% (w = 17 m) — the gain from a slimmer pillar. Check against the statutory minimum pillar centre distance for 200 m depth before adopting.

Example 5 — roof-bolt pattern from CMRI RMR. Gallery B = 4.8 m, adjusted RMR = 45, roof γ = 2.5 t/m³.

• Bracket: 1.7 − 0.037(45) + 0.0002(45²) = 1.7 − 1.665 + 0.405 = 0.44.
• Rock load P = 4.8 × 2.5 × 0.44 = 5.28 t/m²; rock-load height = 5.28/2.5 ≈ 2.11 m.
• Bolt of 11 t anchorage capacity with FoS 2 → safe load 5.5 t. Area per bolt = 5.5/5.28 = 1.04 m² → spacing √1.04 ≈ 1.0 m grid.
• Bolt length L = 1.4 + 0.184 × 4.8 = 2.28 m → adopt 2.4 m full-column grouted bolts at 1.0 m × 1.0 m, closer at junctions.

Example 6 — planar slope failure. Bench H = 12 m, face angle ψf = 60°, joint dipping ψp = 35° and daylighting, c = 25 kPa, φ = 30°, γ = 25 kN/m³, dry (U = 0).

• L = 12/sin35° = 12/0.5736 = 20.92 m.
• W = ½ × 25 × 144 × (cot35° − cot60°) = 1,800 × (1.4281 − 0.5774) = 1,800 × 0.8507 = 1,531 kN/m.
• Resisting = 25 × 20.92 + 1,531 × cos35° × tan30° = 523 + 1,531 × 0.8192 × 0.5774 = 523 + 724 = 1,247 kN/m.
• Driving = 1,531 × 0.5736 = 878 kN/m → FoS = 1,247/878 = 1.42. Acceptable for a working bench, but if the joint fills with water the FoS drops below 1.1 — so drainage is mandatory.

Example 7 — caving and subsidence. Seam m = 3 m, bulking factor k = 1.3, depth 150 m, panel width 200 m, subsidence factor a = 0.6.

• Caving height = 3/(1.3 − 1) = 10 m of immediate roof must cave to fill the goaf.
• Panel width/depth = 200/150 = 1.33 ≥ 1.2 → essentially full (critical) subsidence.
• Smax = 0.6 × 3 = 1.8 m at the surface; with 25° angle of draw the affected zone extends 150 × tan25° ≈ 70 m beyond the panel edge.

Where it is used

• Underground coal (CIL subsidiaries, SCCL): pillar sizing for development, SSR and bolt design, depillaring with caving or sand stowing, longwall face and gate-road support at Adriyala, Jhanjra and Moonidih.
• Underground metal mines (Hindustan Zinc Rampura Agucha/Sindesar Khurd, HCL Malanjkhand, Tata Steel): stope and sill-pillar design, cable bolting, shotcrete, backfill (paste/hydraulic) strength design, cavity monitoring.
• Deep opencast (NCL Jayant/Nigahi, MCL Basundhara, NMDC Bailadila, Vedanta): bench and interramp slope angles, overburden dump stability, slope radar and piezometer monitoring, high-wall mining feasibility.
• Research and consultancy: CIMFR Dhanbad, IIT (ISM) Dhanbad, IIT Kharagpur/Bombay, NIRM Bengaluru carry out statutory strata-control and slope-stability studies; DGMS requires such a study for most permissions.
• Job roles: strata-control officer, geotechnical/rock-mechanics engineer, mine planning engineer, mine safety officer, tunnel/geotechnical engineer in civil projects (metro, hydel caverns), instrumentation engineer.
• Adjacent sectors: tunnelling and underground metro works, hydro-power caverns, quarry slope design, and rock-support product suppliers (resin capsules, bolts, mesh, shotcrete).

Interview questions

• Q: What is the difference between RQD and RMR? A: RQD measures only the degree of core fracturing (% pieces ≥100 mm); RMR combines RQD with strength, joint spacing, joint condition, water and orientation to give a design class and support pressure.
• Q: Why is CMRI-ISM RMR preferred to Bieniawski's RMR in Indian coal mines? A: It was developed on Indian coal-measure roofs and uses parameters that govern there — bedding layer thickness, weatherability of shale and seepage — and leads directly to a rock load and bolt pattern.
• Q: Why does the roof of a gallery crack in tension when the rock is strong in compression? A: The Kirsch solution gives roof tangential stress σv(3k − 1), which becomes tensile when the horizontal-to-vertical stress ratio k < 1/3; rock's tensile strength is only about a tenth of its compressive strength.
• Q: How does a roof bolt actually support the roof? A: By suspending a weak immediate roof from a competent bed, by clamping thin laminations into a single thicker, stiffer beam, by keying blocky rock, and by adding shear resistance across bedding — not by "holding up" the full dead weight.
• Q: Why does pillar stress rise so fast when galleries are widened? A: Tributary stress varies as ((w+B)/w)², so the load transferred to the remaining pillar grows with the square of the ratio; a 4.8 m gallery around a 20 m pillar already adds 54% to the cover load.
• Q: What FoS would you adopt for pillars and why? A: About 1.5 or more for development pillars that must stand for years, 1.0–1.3 for pillars about to be extracted, and higher under a river, railway or township where failure is unacceptable.
• Q: What is a telltale and what reading worries you? A: A dual-height roof extensometer giving bed separation above and within the bolted horizon; an accelerating rate of sag, or movement above the bolt anchorage, is the danger signal — not any single absolute figure.
• Q: Why is a massive sandstone roof a hazard during depillaring? A: It does not cave with the extraction, hangs over a large goaf and then collapses suddenly, compressing the air into an air blast; it is handled by induced blasting or by stowing.
• Q: What governs whether a slope fails on a plane? A: Kinematics — the plane must daylight in the face, dip less than the face angle and more than its friction angle, and strike roughly parallel to the face — plus water pressure, which is usually the trigger.
• Q: Mohr-Coulomb or Hoek-Brown? A: Mohr-Coulomb for joints, soils, dumps and simple pillar/slope work; Hoek-Brown with GSI for jointed rock masses and numerical modelling, because rock-mass strength is non-linear with confinement.
• Q: How is surface subsidence limited over a working panel? A: By stowing instead of caving, by partial or harmonic extraction, by leaving support/barrier pillars under the feature, and by keeping panel width well below the critical width — all under DGMS permission.
• Q: Where does a classification system let you down? A: When a single feature governs — a fault, a slickensided plane, a water-bearing channel or high horizontal stress — a good RMR can coexist with a real failure risk; instrumentation and geological mapping must confirm the design.

Common mistakes

• Using cover load γH as the pillar stress and forgetting the tributary-area multiplier (w+B)²/w².
• Mixing units: γ as 25 kN/m³ with strength in MPa, or t/m² with MPa — convert once at the start (1 MPa ≈ 100 t/m², 2.5 t/m³ ≈ 0.025 MN/m³).
• Applying the CMRI pillar formula with a large-specimen UCS instead of the 25 mm cube strength it was calibrated on, or ignoring the statutory minimum pillar size once the formula gives a smaller one.
• Quoting an RQD or RMR class as the design answer without the orientation adjustment, water condition or depth/stress adjustment.
• Omitting water pressure (and tension-crack water thrust) in slope FoS, which overstates stability by 20–40%.
• Treating rock mass strength as equal to intact specimen strength, or comparing a lab UCS directly with field behaviour.
• Reporting an isolated instrument reading instead of the trend and rate, and treating a stable number as proof of safety.

Related topics

Drilling and blasting · Mine planning and design software · Mine ventilation · Mine surveying · Mining methods and pillar extraction · Mine safety legislation and DGMS practice