Slope stability analysis.
Practical reference for the engineer running slope stability analyses in Malaysia. Limit equilibrium methods (Bishop's Simplified, Janbu's Simplified, Spencer, Morgenstern-Price, Sarma, ordinary method of slices) and finite element strength reduction (SRM) in PLAXIS, FLAC, RS2. Slip surface searching (circular, non-circular, entry-exit, grid, optimization). Pore pressure modeling (piezometric line, Ru coefficient, transient seepage). Software comparison (Slope/W, Slide, PLAXIS LE, FLAC, RS2). FoS interpretation against JKR / Eurocode 7 / BS 6031 targets. Common pitfalls. By Infraconcrete - CIDB G7 specialist geotechnical contractor.
Jump to a topic.
What LEM does, and what it doesn't.
Limit equilibrium method (LEM) divides the soil mass above a trial slip surface into vertical slices, then solves force and/or moment equilibrium of each slice for the factor of safety (FoS). It assumes rigid-perfectly-plastic soil at failure, no progressive failure, and a pre-defined slip surface shape (circular or non-circular). The methods differ in how they handle interslice forces - the unknowns that make the problem statically indeterminate.
| Method | Equilibrium satisfied | Shape | Speed | Use case |
|---|---|---|---|---|
| Ordinary (Fellenius) | Moment only | Circular | Fastest | Hand calc / sanity check only - conservative |
| Bishop's Simplified | Vertical force + moment | Circular | Fast | Production work for circular slips - industry default |
| Janbu's Simplified | Horizontal + vertical force | Non-circular | Fast | Translational / wedge - needs correction factor |
| Spencer | Force + moment (constant interslice angle) | Any | Iterative | Rigorous, handles any shape |
| Morgenstern-Price | Force + moment (variable interslice function) | Any | Iterative | Most rigorous, default for non-circular |
| Sarma | Force + moment + acceleration | Any | Iterative | Seismic / pseudo-static |
The industry workhorse for circular slips.
Basis
Published by Alan W. Bishop in 1955. Satisfies vertical force equilibrium of each slice and overall moment equilibrium of the failing mass about the centre of the circular slip surface. Assumes interslice shear forces sum to zero (horizontal interslice equilibrium not enforced).
Iterative: FoS appears on both sides of the equation. Initial guess (typically 1.0), iterate to convergence (4-8 iterations typical).
When to use
- Circular slip surface, homogeneous or layered slope without strong directionality
- Routine cut and fill design
- Embankment on uniform foundation
- Preliminary analysis before detailed Spencer / MP run
Avoid for: non-circular slips (weak seam, bedding plane, soil-rock interface), wedge failures, wall-soil composite analyses.
FoS = (1 / sum[W*sin(alpha)]) * sum[ (c'*b + (W - u*b)*tan(phi')) / m_alpha ]
where m_alpha = cos(alpha) + sin(alpha)*tan(phi')/FoS
(W = slice weight, alpha = base angle, b = slice width, c' / phi' = effective parameters, u = pore pressure)
For non-circular slips - with a known limitation.
Basis
Published by Nilmar Janbu (1954, 1973). Satisfies horizontal and vertical force equilibrium of each slice. Does NOT enforce moment equilibrium. Assumes interslice shear forces are zero. Applicable to any slip surface shape.
Janbu published a correction factor (f_o) to compensate for the missing moment equilibrium - applied to FoS based on slip surface shape (depth-to-length ratio) and soil type. Without the correction, Janbu under-estimates FoS by 5-12 percent.
When to use
- Translational failure on a known weak plane (bedding, soil-rock interface, weathered seam)
- Wedge failure in jointed rock or stratified soil
- Quick scan of multiple non-circular geometries
Best practice. Use Janbu's Simplified with the correction factor for preliminary work, then verify with Spencer or MP for the final design FoS.
Rigorous - both force and moment satisfied.
Basis
Published by E. Spencer (1967). Satisfies both force and moment equilibrium. Assumes a constant ratio between interslice normal and shear forces (interslice force inclination is constant for the entire slip surface, but unknown - solved iteratively).
Two unknowns solved simultaneously: FoS and the interslice angle (theta). Convergence iteration searches for the (FoS, theta) pair that satisfies both equilibrium conditions.
When to use
- Non-circular slip surfaces (translational, wedge, layered)
- Heterogeneous slopes with strong stratigraphic contrasts
- Cases where Bishop and Janbu disagree by more than 5 percent
- Final design check for permanent slopes
The default for non-circular and complex geometry.
Basis
Published by N.R. Morgenstern and V.E. Price (1965). Generalization of Spencer's method - the interslice force inclination is allowed to vary along the slip surface according to a user-specified function f(x).
Common functions: half-sine (default), constant (reduces to Spencer), trapezoidal. Lambda parameter scales the function. Solution iterates on FoS and lambda.
When to use
- Strongly non-circular geometry (toe wedge + circular middle + sliding back)
- Multi-layer slopes with stratum-controlled slip path
- Walls and reinforced slopes (MSE, soil nail, anchors)
- Final design FoS for high-consequence slopes (highway, rail, public infrastructure)
When LEM isn't enough.
How it works
Strength reduction method (SRM): the soil shear strength (c, phi) is progressively reduced by a strength reduction factor (SRF). At each SRF level, the FEM solves for stress/strain. The slope is "stable" while the analysis converges; when convergence fails (large unbalanced forces, runaway displacements), the slope has reached limit equilibrium. FoS = the SRF at which collapse occurs.
No pre-defined slip surface - failure surface emerges from the strain field where plastic shear strain concentrates.
When FEM is mandatory
- Strain-softening / progressive failure (sensitive clay, residual soil)
- Soil-structure interaction (anchors, soil nails, MSE reinforcement, piles)
- Deformation criterion (limit on slope movement, not just FoS)
- Tunnel portal slopes
- Deep excavations adjacent to existing structures
- JKR Class III / IV slopes (high consequence)
Finding the critical slip - the search method matters.
The critical slip surface is the one with the lowest FoS. Modern software automates the search but the engineer must specify the search method, the search domain, and the slip surface shape. A poorly-set search misses the critical surface and over-reports FoS.
Circular search methods
- Grid and radius: classic - grid of centres, range of radii, brute force. Reliable but slow. Use for verification.
- Auto-locate: intelligent search using moving grids and radius optimization. Default in most software.
- Entry-exit: specify entry and exit ranges along the slope surface; software searches all geometrically valid circles. Good for cuts where the slip exits at toe.
Non-circular search methods
- Block search: series of polylines through user-defined zones. Use when stratum boundary is known.
- Path search: piecewise non-circular - automated optimization.
- Optimization: starts from a candidate (often Bishop circular result), perturbs vertices to lower FoS. The most rigorous - but vulnerable to local minima.
Get the water wrong, get the FoS wrong.
Pore pressure has the largest single impact on slope FoS short of the soil shear parameters themselves. A 10 percent error in pore pressure can move FoS by 5-15 percent. The pore pressure model must be defensible.
| Method | How it works | When to use | Limitation |
|---|---|---|---|
| Piezometric line | User-drawn phreatic surface; u = gamma_w * (h_pl - h_slice base) | Phreatic level well-defined from boreholes / piezometers; steady state | Cannot capture transient rainfall infiltration; assumes hydrostatic below line |
| Ru coefficient | u = Ru * gamma * h, applied to all slices in zone | Preliminary work; rough approximation; Ru = 0.25-0.35 for residual soil under steady seepage | Over-simplified; not defensible for permanent design |
| B-bar (Skempton) | u = B * (sigma_v - sigma_h_0); for undrained loading | Embankment construction over soft clay; rapid loading | Requires triaxial test data; rarely used in routine slope work |
| Steady-state seepage FE | Solve Laplace equation; import u into stability analysis | Drainage system design; complex boundary conditions | Steady state only - misses rainfall transients |
| Transient seepage FE | Solve unsaturated flow with rainfall boundary; import u(t) | Rainfall-induced failure analysis; rapid drawdown; antecedent rainfall effects | Computationally heavy; needs SWCC and k(theta) curves |
What's used in Malaysia.
| Software | Vendor | Type | Strengths | Weaknesses |
|---|---|---|---|---|
| Slope/W (GeoStudio) | Bentley / Seequent | LEM (2D) | Industry default in Malaysia. All LEM methods. Probabilistic. Easy to learn. Coupled with Seep/W for pore pressure. | 2D only. LEM only - no FEM SRM directly (use Sigma/W). |
| Slide2 / Slide3 | Rocscience | LEM (2D / 3D) | Best slip search algorithms. 3D capability. Probabilistic. Excellent post-processing. | Higher learning curve. Subscription cost. |
| PLAXIS 2D / 3D | Bentley | FEM (continuum) | Premier soil-structure FEM in Malaysia. Excellent for walls, anchors, deep excavations, embankments on soft soil. Includes SRM for slopes. Advanced soil models (HS, HSsmall, SS, MC, NGI-ADP). | Steep learning curve. Long compute. Overkill for routine cut/fill slopes. |
| RS2 / RS3 | Rocscience | FEM (continuum) | Strong for rock slopes and tunnel-slope interaction. SSR analysis. Coupled stress-flow. | Less common in Malaysia than PLAXIS. |
| FLAC / FLAC3D | Itasca | FDM (continuum) | Best for large strain, progressive failure, dynamic. Used for sensitive analysis (rapid drawdown, earthquake, debris flow initiation). | Specialist tool. Programming-style input. Few users in Malaysia. |
| Plaxis LE (formerly SVSlope) | Bentley | LEM (2D / 3D) | 3D LEM. Coupled with Plaxis FEM. Solid for waste rock dumps, tailings dams, large complex 3D geometry. | Newer to Malaysia. |
Compliance ranges per Malaysian practice.
| Condition | FoS target | Source |
|---|---|---|
| Long-term, drained, permanent slope | 1.4 minimum | JKR SEM, BS 6031, EC7 |
| Long-term, public infrastructure (highway / rail / dam) | 1.5 - 1.6 | JKR SEM, AASHTO, BS 8006 |
| Long-term, JKR Class III / IV (high consequence, public exposure) | 1.5 - 1.6 | JKR Slope Hazard Class |
| Short-term, undrained (end-of-construction embankment) | 1.3 | JKR SEM, BS 6031 |
| Temporary works | 1.2 - 1.3 | BS 5975, BS 6031 |
| Seismic (pseudo-static) | 1.1 - 1.2 | EC8, BS 6031 |
| Rapid drawdown (dam, retention pond) | 1.2 - 1.3 | USACE, BS 6031 |
| Eurocode 7 partial factor (DA1-2 equivalent global) | ~1.4 - 1.5 | BS EN 1997-1 |
What goes wrong in slope analysis.
Geometry & geology pitfalls
- Missed weak seam. Stratigraphy from boreholes interpolated as continuous - actual weak seam not captured. Ground-truth with trial pits, geophysics, additional boreholes.
- Bedrock interface ignored. Tropical residual soil over weathered rock can slide on the soil-rock interface. Always model the interface as a potential slip plane.
- Joint orientation in rock slopes. LEM in soil mode misses kinematic failure (plane sliding, wedge, toppling). Use stereonet and rock-mass methods first.
- Search domain too narrow. Critical slip can be deeper than searched. Always extend search beyond the obvious zone.
Parameter pitfalls
- Peak vs residual. Brittle soils mobilize peak first, then drop to residual along the slip surface. For pre-existing slip surfaces (reactivated landslides), use residual phi-r. For first-time failure, peak parameters with progressive failure consideration.
- Su vs c-prime / phi-prime. Mixing total and effective stress parameters in one analysis is invalid. Pick one framework per slice / per zone.
- Cementation overestimated. Tropical residual soil with cementation is often given high c-prime - but cementation breaks down under strain. Run with and without cementation; design for the lower FoS.
- Unit weight wrong sign. Submerged weight vs total weight confusion in below-water-table zones. Effective stress requires (gamma - gamma_w) below WT.
Pore pressure pitfalls
- Steady-state when transient applies. Wet-season failure under-states FoS by 10-20 percent in residual soil if you assume steady state.
- Phreatic surface flat across the section. Real water tables follow topography roughly with attenuation. Use FE seepage or piezometer-calibrated surface.
- Negative pore pressure (matric suction) abused. Suction adds apparent cohesion in unsaturated soil but is lost on rainfall infiltration. Don't bank on suction for design FoS.
Method & reporting pitfalls
- Single method reported. Always run two methods (e.g. Bishop + Spencer or MP) and report the lower FoS.
- Only critical surface shown. Show the FoS contour or top-10 lowest surfaces - not just one. Multiple low-FoS surfaces clustered together indicate a weak zone, not a one-off.
- Software defaults accepted. Default search range, default tension crack settings, default interslice function - all may be inappropriate for your geometry. Verify each setting.
- FEM mesh too coarse. SRM with coarse mesh over-states FoS. Mesh-refinement check is mandatory.
Codes and references.
| Topic | Reference |
|---|---|
| Earthworks & slope stability (general) | BS 6031, BS EN 1997-1 (Eurocode 7), JKR Slope Engineering Manual |
| Site investigation | BS 5930, JKR/SPJ Section 1, Eurocode 7 Part 2 |
| Highway slope design | FHWA-NHI-14-007, AASHTO LRFD, JKR ATJ |
| Embankment over soft soil | BS 8006-1, FHWA-NHI-12-024, JKR SEM |
| Reinforced soil structures | BS 8006-1, BS 8006-2, FHWA-NHI-10-024 |
| Seismic slope design | BS EN 1998 (Eurocode 8), Malaysian National Annex |
| Probabilistic / reliability | JCSS Probabilistic Model Code, ISO 2394 |
| FE modeling guidance | NAFEMS guides, PLAXIS material model manual, PIANC |
Analysis questions.
Bishop, Janbu, Spencer, or Morgenstern-Price - which one? +
When do I need FEM strength reduction instead of LEM? +
What FoS should I target for Malaysian slopes? +
How do I model rainfall infiltration? +
What's the difference between c-prime / phi-prime and Su analysis? +
Need slope analysis support?
Send the borehole logs, geometry, and constraints. Same-day response from the engineering team. We run Slope/W, Slide, PLAXIS, and RS2 in-house under design-build, or as the specialist contractor under your appointed consulting engineer.