Slope Stability
Slope Stability tests where the geometry of a slope and of its discontinuities lets a block slide or topple. It checks four failure modes against the discontinuity sets you have measured — planar sliding, wedge sliding, flexural toppling and direct toppling — and reports each as the percentage of a set's planes, or of the lines where two sets' planes meet, that are critical. If you ask, it also works out limit-equilibrium factors of safety for planar sliding, wedge sliding and block toppling on each set's mean plane, and the probability of failure by Monte Carlo over the sets' spread, their strength and the water, judged against design targets for each scale.
It works two ways, with the same tests:
- On a mesh or point cloud, it fits the slope face around every point at a scale you choose — or at the bench, inter-ramp and overall scales at once — and writes a map for each mode, plus Kinematic Mode, the dominant mode at every point, the factors of safety and the probability of failure. Use it to find where along an outcrop or a cutting each set becomes a hazard.
- On the stereonet, it analyses one slope face you set, draws its construction and critical zones over your data, and lists the percentages, factors of safety and probabilities of failure set by set.
The kinematic analysis says where the geometry permits failure; a critical percentage is not a probability of failure. The factors of safety are worked out on each set's mean plane, with the height of the slope taken from the model and the strength you give each set; the probability of failure is only as good as the spreads you give it. Use the maps to decide where to look, then assess those places properly.
The tests, with the formulas, are in Slope Stability — Method.
The four failure modes
| Mode | Tested on | Critical when |
|---|---|---|
| Planar sliding | Each plane of a set | The plane dips out of the face more steeply than the friction angle but less steeply than the face, so it daylights, and its dip direction is within the lateral limit of the face's (default 20°). |
| Wedge sliding | The lines where planes of two sets meet | The line plunges out of the face more steeply than the friction angle and daylights (Markland's test). The block may then slide along the line or, by Hocking's test, down one of its planes; either way it counts. No lateral limit. |
| Flexural toppling | Each plane of a set | The plane dips steeply into the face — its pole lies below the slip-limit plane, which dips at the face dip minus the friction angle — within the lateral limit of straight in (default 30°). |
| Direct toppling | The lines where planes of two sets meet | The line plunges into the slope within the lateral limit of straight in (default 20°), and is no further from vertical than the face dip, so the columns it bounds lean out of the face. |
Wedge sliding and direct toppling need planes in at least two sets. Lines are only formed between planes of different sets, and two planes closer to parallel than the minimum wedge angle, 10° by default, form no line.
Before you start: discontinuity sets
A set can be either of two things, and both ways of running the analysis accept both:
- A project Set — the planes tagged with one Set, wherever they sit in the tree. Auto-cluster Selected Orientations on the Orientation branch tags its clusters this way, as J1, J2 and so on.
- An orientation folder — the planes directly in one folder under Orientations in the Interpretation tree. Planes in a subfolder belong to the subfolder, which is a set of its own. Every orientation folder counts, including one made with the stereonet's Group Selected.
Other points:
- Lineations are left out: they have no plane.
- A project Set can have a joint strength of its own. The kinematic tests then use its own friction angle, and the factors of safety its strength; a line of intersection takes the lower angle of its two sets. A folder, or a Set with no strength of its own, uses the friction angle of the analysis.
- On the stereonet, Sets From chooses Folders or Project Sets. Planes in no folder, or with no Set, form one more set, Ungrouped or No Set.
- A plane in two chosen sets — a folder and a Set, say — meets copies of its own set's planes in lines that no two sets form, which inflates wedge sliding and direct toppling. The messages panel warns how many planes are chosen twice; untick one of each pair.
Before sets were tags, Auto-cluster copied each cluster into a folder below the one it read and left the originals there. A folder that holds every plane of a folder below it is not taken as a set unless you choose it: the dialog leaves it unticked the first time, and the stereonet leaves its poles out and says how many in the key.
DFN fracture sets
The mesh and point cloud analysis also lists the fracture sets of every DFN in the project, as DFN: Rock/1: J1: the network's name, the set's number and its name. The stereonet's analysis reads plane interpretations only.
- It reads the active fractures of the current realisation, including those far from the outcrop and those of a hidden network or set. Inactive and clipped-away fractures do not count.
- A DFN set named like a project Set takes that Set's joint strength. Any other takes the friction angle of the analysis.
- A DFN is a sample of the rock volume, not of what one face shows, so Terzaghi (sampling bias) leaves its fractures at a weight of 1. Lean on the planes measured within the radius and Planes Within Radius use the fractures' centres as their positions.
- The first time on an object, DFN sets are ticked only when there are no measured planes, and then only those of the first DFN, since realisations are alternatives. Choose one representation of the rock: measured planes and the DFN built from them would meet each other in lines no two real sets form, and the warning about planes in two chosen sets does not catch this.
Mapping a mesh or point cloud
Typical workflow
- Sort the measured planes into sets: tag them with Auto-cluster or by hand, or put each set in a folder of its own.
- In the Data Tree, select one or more loaded meshes or point clouds, right-click, and choose Attributes & Analysis → Geotechnical → Slope Stability....
- In the dialog, tick the sets to use, the scales to fit the face at, the Friction angle and the modes to test. Tick Factors of safety (limit equilibrium) for factors of safety, and check its Settings..., where you can also ask for the probability of failure. Press OK.
- When the run finishes the object is coloured by Kinematic Mode, and the messages panel gives the share of the surface in each mode and, with factors of safety, the lowest factor and the share below 1.
- Look at the percentage layer of each mode to see how strongly a place is affected, at Factor of Safety for how close it is to failing, and at Face Dip and Face Dip Direction to check the faces were fitted at a sensible scale.
Prerequisites
- A loaded triangulated mesh or point cloud. A tile of a tiled model cannot be analysed.
- At least one set: planes tagged with a Set, a folder under Orientations with planes in it, or a DFN with fractures. Without one the command stops with a warning telling you to make the sets first.
- To tell a face from its back — an overhang from a slope facing the other way — the analysis
needs to know which side of the surface is out of the rock. A mesh has this from its
triangle winding. A point cloud has it only if it carries camera-oriented normals, the
Nx,NyandNzlayers of an SfM dense cloud; without them every face is taken to look upward.
The dialog
The line at the top names the object, its number of points and where the outward side of each face comes from. With several objects selected, one dialog is filled from the first and its options apply to all of them.
Discontinuity sets - tick orientation folders, project Sets or DFN sets: lists every folder under Orientations that holds planes, with its path, every project Set that tags planes, as Set: J1, and every DFN fracture set, as DFN: Rock/1: J1, each with its number of planes. The first time on an object, the project Sets are ticked when any planes are tagged; otherwise every folder is ticked, except one an older Auto-cluster copied its planes out of. With no measured planes at all, the sets of the first DFN are ticked. After that, the sets you ticked last time.
Slope face - fitted over a radius (model units) at each ticked scale
| Field | Description |
|---|---|
| Single radius | Fit the face over this radius and write the layers under their plain names. The note beside it gives the point spacing and a suggestion: 2% of the model's diagonal but no less than twenty point spacings. Ticked by default. |
| Bench, Inter-ramp, Overall | Fit the face at each of these scales too, each writing its own layers, named with the scale: Planar Sliding (Bench), Factor of Safety (Overall). They start at the suggested radius, four times it and sixteen times it. |
| Its normals point into the rock (flip them) | Reverse the outward side taken from the mesh winding or the point normals. Tick it for a mesh wound inside out — the messages panel warns when more than half of a mesh's faces look downward. Greyed when the object has no oriented normals. |
A radius under four point spacings is raised to four, and the messages panel says so.
Weighting
| Field | Description |
|---|---|
| Terzaghi (sampling bias) | None (default), Planes seen on the face or Traces across the face: weight each plane for how likely it was to be measured from the face at its position. See Weighting. |
| cap | The largest weight a plane can take, 1–100. Default 7. |
| Lean on the planes measured within the radius | At each point, blend each set's percentage for the planes measured near that point with its percentage over the whole project, so a sparsely mapped place does not read as fully critical. Off by default. |
| half say at … planes | The number of nearby planes at which the local share has half the say. Default 10. |
Kinematic tests
| Field | Description |
|---|---|
| Friction angle (deg) | Friction angle of the discontinuities, 0–89, for every set without a joint strength of its own. The first time, it is the angle whose tangent is the Friction Coefficient of the project's active stress tensor: about 31° (30.96°) for the default 0.6. |
| Classify at (% critical) | A point is given a mode in Kinematic Mode only when that mode's percentage reaches this. Default 10. |
| Planar sliding · Lateral limit (deg) | Test planar sliding, with how far a sliding plane's dip direction may swing from the face's. Default 20. |
| Wedge sliding | Test wedge sliding on the lines where planes of two different sets meet. |
| Flexural toppling · Lateral limit (deg) | Test flexural toppling, with how far a toppling plane may swing from dipping straight into the face. Default 30. |
| Direct toppling · Lateral limit (deg) | Test direct toppling on the lines of two sets, with how far a line may swing from plunging straight into the slope. Default 20. |
| Planes within … deg of parallel form no line of intersection | The minimum wedge angle of wedge sliding and direct toppling. Default 10, from 0.1 to 90. |
The lateral limits run from 0 to 90. At least one set, one scale and one mode must be ticked.
Factors of safety (limit equilibrium) — tick it to work out factors of safety as well, and press Settings... to set them, and the probability of failure (below). Pressing OK in the settings ticks the box.
Factors of safety: the settings
The Factors of Safety (Limit Equilibrium) dialog sets how the factors of safety are worked out. At every point they come from each set's mean plane and joint strength, the fitted face, and the Local Relief at the scale's radius as the slope height; block toppling takes each set's friction angle. A set without a joint strength of its own is taken as Mohr-Coulomb at the friction angle, with no cohesion. Model units are taken as metres.
Mechanisms
| Field | Default | Description |
|---|---|---|
| Planar sliding (Hoek & Bray) | Ticked | Planar sliding on each set's mean plane that daylights within the planar lateral limit. |
| Tension crack | Critical | None: the sliding plane runs out through the upper slope. Critical: a vertical crack behind the crest, at the position that gives the lowest factor of safety. Where the upper slope dips as steeply as the plane, the critical crack is always used. |
| Wedge sliding | Ticked | Wedge sliding on the mean planes of each pair of sets. |
| Longest trace (0: no limit) | 0 | The longest trace either plane of a wedge may leave on the face, m. A bigger wedge is scaled down to it. |
| Block toppling (Goodman & Bray) | Unticked | Block toppling of columns between a set's mean planes dipping into the face within the toppling lateral limit. It needs a block Width: OK refuses toppling without one. |
| Width | 0 | The block width, m: the toppling planes' spacing. |
| base step | 10 | How much steeper, in degrees, the stepped base through the blocks' feet runs than the block bases. |
Loading
| Field | Default | Description |
|---|---|---|
| Rock unit weight (kN/m3) | 26 | Unit weight of the rock. |
| Water, 0 dry to 1 full | 0 | How full of water a planar block's tension crack is, with the matching uplift on the plane, and how saturated a wedge is (Hoek & Bray's drained wedge). With Crack set to none the planar case has nowhere for the water to act, so its factor of safety is the dry one; the run says so in the messages panel. |
| Seismic coefficient kh | 0 | A horizontal pseudo-static load, kh times the weight, acting out of the slope. |
| Upper slope dip (deg) | 0 | The slope above the crest, dipping the same way as the face; negative dips back into the hill. |
Probability of failure (Monte Carlo)
| Field | Default | Description |
|---|---|---|
| Work it out | Off | Also work out the probability of failure at every point. |
| Samples | 200 | How many Monte Carlo samples, 10–100000. |
| Seed | 0 | The seed of the random draws, a whole number. 0 draws a new seed each run; the run records the seed it used, so it can be repeated. The same seed gives the same result. |
| Draw planes from each set's Fisher spread | Ticked | Draw each set's plane in every sample from a Fisher distribution about its mean plane. Off, every sample uses the mean plane. |
| Friction s.d. (deg) | 2 | Standard deviation of each set's friction angle — its residual friction angle for Barton-Bandis. |
| Cohesion CoV | 0.25 | Coefficient of variation of each set's Mohr-Coulomb cohesion. |
| Water from … to | 0 to 0 | The water, 0 dry to 1 full, drawn evenly from this range. Both 0 uses the Water above. |
| Targets, FoS at least / PoF % at most | see below | For each scale — Single, Bench, Inter-ramp and Overall — the factor of safety a slope should reach and the probability of failure it should stay under. |
The targets start from Read & Stacey's (2009) acceptance criteria for a moderate consequence of failure:
| Scale | FoS at least | PoF at most |
|---|---|---|
| Single | 1.3 | 10% |
| Bench | 1.1 | 25% |
| Inter-ramp | 1.2 | 20% |
| Overall | 1.3 | 10% |
Set your own when the consequence of failure, or your design standard, calls for it.
Running and cancelling
The run shows as a slopestability progress bar: sampling the surface, fitting the slope faces,
forming the intersections, then the kinematic tests and the factors of safety. Stop it with the
bar's terminate button or the tree's Cancel Operation on the object; with several objects or
scales, cancelling stops the rest too. A run fits the faces once and then evaluates each face
orientation once, so its time depends on the model's area and the radius far more than on its
number of points.
What it writes
Every run replaces these point (vertex) layers on the object. At the Bench, Inter-ramp and Overall scales each name ends with the scale in brackets, as Kinematic Mode (Bench); the single radius keeps the plain names.
| Layer | Values |
|---|---|
| Face Dip | Dip of the fitted slope face, 0–180°. Above 90 the face overhangs. |
| Face Dip Direction | The direction the face looks, 0–360°, on a cyclic colour map so 359 and 1 look alike. |
| Planar Sliding | Percentage critical, 0–100, of the most critical set. |
| Wedge Sliding | Percentage of the lines of the most critical pair of sets that pass Markland's test, whether the block slides along the line or down one of its planes. |
| Flexural Toppling | Percentage critical of the most critical set. |
| Direct Toppling | Percentage critical of the lines of the most critical pair of sets. |
| Kinematic Mode | The dominant mode: 0 none (grey), 1 planar sliding (red), 2 wedge sliding (orange), 3 flexural toppling (purple), 4 direct toppling (blue). |
| Local Relief | The height range of the surface within the radius: the highest point less the lowest, in model units. It is the slope height for the factors of safety. |
| Planes Within Radius | How many planes of the chosen sets lie within the radius of the point, DFN fractures by their centres. A plane in two chosen sets counts twice; a plane with no position counts nowhere. |
With Factors of safety ticked it also writes:
| Layer | Values |
|---|---|
| Factor of Safety | The lowest factor of safety over the mechanisms and sets at the point. |
| FoS Planar Sliding | The lowest planar-sliding factor over the sets. |
| FoS Wedge Sliding | The lowest wedge factor over the pairs of sets. |
| FoS Toppling | The lowest block-toppling factor over the sets; only when a block Width is given. |
| FoS Mode | The mechanism with the lowest factor: 0 none (grey), 1 planar sliding (red), 2 wedge sliding (orange), 3 toppling (purple). On a tie: planar, wedge, then toppling. |
With Probability of failure ticked in the settings it also writes:
| Layer | Values |
|---|---|
| Probability of Failure | The percentage of the samples in which the lowest factor of safety at the point is below 1, 0–100, green to red. |
| PoF Planar Sliding, PoF Wedge Sliding, PoF Toppling | The percentage of the samples in which that mechanism's factor is below 1. |
| Acceptance | The point against the scale's targets: 0 meets both (green), 1 its factor of safety is short of the target, 2 its probability of failure is too high, 3 both (dark red). A point where no block forms meets the factor-of-safety target. |
- A percentage layer holds the highest percentage over the sets, or pairs of sets, so one critical set is not diluted by others that are not. The stereonet reports each set separately.
- Kinematic Mode takes the tested mode with the highest percentage (on a tie: planar, wedge, flexural, then direct), or none when that is below Classify at.
- A mode you did not tick has no layer: one left by an earlier run is removed. So is a mechanism's FoS layer when the mechanism is not ticked.
- Points with too few neighbours within the radius for a face get no value in any layer.
- The percentage layers have a fixed 0–100 range, so colours compare between runs and objects.
- Factor of Safety layers are coloured on a fixed 0–3 range — reds below 1, yellow about 1.5, greens above. A factor above 10 is written as 10. Where no block forms — no set's mean plane daylights, say — the layer has no value and FoS Mode is 0. A wedge that lifts off both planes, and toppling that cannot stand at any friction, read 0.
- Kinematic Mode uses the Kinematic Failure Mode colour map, and FoS Mode the Factor of Safety Mode map, kept with the project's colour maps; change their colours there. A layer made before direct toppling existed keeps the older, four-colour Kinematic Mode map.
The messages panel ends the run with the share of the surface in each mode, the radius, the friction angle and the numbers of sets and planes, and, with factors of safety, how much of the surface has a factor, the lowest, and the share of those points below 1. With the probability of failure it gives the samples and seed, the highest probability, and the share of the surface that meets both targets, falls short on the factor, on the probability, or on both. It adds a notice when points got no face, when faces could not be told from their backs, when wedge sliding and direct toppling found no pair of sets to form lines from, and when no mean plane formed a block.
Each object and each scale is recorded under Analysis Runs in the Collections tree, with the settings and the sets it read. When a set changes afterwards — a plane added, removed or moved, its friction angle changed, or a DFN it read regenerated — the run is marked stale, so you know the layers are out of date. A run with the probability of failure records its seed.
Remembered settings
The settings of the last run — scales and radii, flip, weighting, friction angle, limits, minimum wedge angle, modes, threshold, the factor-of-safety settings and the sets used — are kept on the object and saved with the project, so the dialog reopens with them.
Probability of failure
The factor of safety of a set's mean plane says nothing about how much of the set could fail. The probability of failure draws the uncertain inputs many times and counts how often the slope fails:
- Planes. In every sample each set's plane is drawn from a Fisher distribution about its mean plane, with the set's own concentration, K = (N − 1) / (N − R), from its N measured planes.
- Strength. Each set's friction angle, and its cohesion, are drawn from normal distributions about their values, with the standard deviation and coefficient of variation you give.
- Water is drawn evenly from its range.
The probability of failure is the share of the samples in which the lowest factor of safety at a point is below 1. A sample in which no block forms does not fail. The same draws are used at every point, so the map is smooth rather than speckled by sampling noise. Faces are grouped to 2° of dip and dip direction, and heights to 5%, so the cost does not grow with the number of points.
The Acceptance layer compares each point with the targets of its scale: the deterministic Factor of Safety against the FoS target, and the probability against the PoF target.
The slope height and the face come from the model and are not drawn. Block width, unit weight and the seismic coefficient are fixed. A Barton-Bandis set draws its residual friction angle; its JRC and JCS are fixed.
Weighting
Two options in the dialog's Weighting group make the maps fairer to the way the planes were measured.
Terzaghi (sampling bias)
Planes measured on an outcrop are not a fair sample of the rock mass: how many of a set you see depends on its angle to the face. Each plane is weighted against the face fitted where it was measured — the face of the nearest point within the radius of its position. A plane with no position, or no face near it, counts once, and so does every fracture of a DFN set, which was not measured from a face.
- Planes seen on the face — weight 1/|cos| of the angle between the plane's pole and the face normal. This boosts the planes the face shows edge-on, which are under-represented when planes are picked as surfaces on the face (Lato et al. 2010).
- Traces across the face — weight 1/sin of the angle between the plane and the face. This boosts planes nearly parallel to the face, which leave few traces across it (Terzaghi 1965). Use it when the planes come from traces.
Either weight is capped at cap. The messages panel says how many planes were weighted and how many had no position or no face near them.
Lean on the planes measured within the radius
A map read from the whole project's sets says a place is critical if the sets could fail there, even where nobody measured a plane. With this option, each set's percentage at a point becomes
(n · local + k · project) / (n + k)
where n is the number of the set's planes measured within the radius of the point, local the critical share of those planes for the point's face, project the share over all the set's planes, and k the half say at value. With k planes nearby the local share has half the say; with none the point keeps the project-wide percentage. One or two critical planes cannot make a point read 100%. For a pair of sets, n is the geometric mean of the two sets' counts. Planes Within Radius shows how many planes each point leaned on.
Minimum wedge angle
Two nearly parallel planes meet in a line that a degree or two of measuring error can swing a long way, so the line says little about any block they cut. Wedge sliding and direct toppling leave out every pair of planes closer to parallel than the minimum wedge angle, 10° by default. It is Planes within … deg of parallel form no line of intersection in the dialog, and Minimum Wedge Angle on the stereonet, which offers it only in those two modes.
- A pair left out is on neither side of the percentage: it is neither critical nor among the lines the percentage is a share of. Raising the angle lowers N, and can move a wedge or direct-toppling percentage up or down. Sets close in orientation feel it most.
- Planar sliding and flexural toppling test planes, not lines, so it does not touch them. Nor does it touch direct toppling's Base column.
- The factors of safety work on each pair's mean planes, under a fixed limit of 1°, so it does not move them.
- A run records the value it used, under Analysis Runs. On the stereonet the key prints it when it is not the default, as Minimum wedge angle 15°, so an exported figure records it too.
On the stereonet
The stereonet analyses one slope face: it draws the construction for the chosen failure mode, shades the critical zones, and lists the percentage of each set's planes — or of the lines where two sets' planes meet — that are critical, and if you ask, the factor of safety of each set's mean plane.
Typical workflow
- Open a stereonet window on your orientation data: Home → Windows → Stereonet Window.
- In its property sheet, under Analysis Display Properties, set Sets From to Folders or Project Sets, and tick Kinematic Analysis.
- Pick the Failure Mode, and enter the Slope Dip and Slope Dip Direction of the face.
- Check the Friction Angle — the first time it is taken from the active stress tensor's friction coefficient — and the Lateral Limit.
- Read the zones on the net and the percentages in the key. Tick Factor of Safety for the factors as well.
Properties
Stereonet window only. These settings belong to the window and are not saved with the project.
| Property | Description |
|---|---|
| Kinematic Analysis | Draw the analysis. |
| Failure Mode | Planar Sliding, Wedge Sliding, Flexural Toppling or Direct Toppling. The rows below follow the mode. |
| Slope Dip | Dip of the slope face, 0–180°; above 90 it overhangs. Default 60. |
| Slope Dip Direction | The direction the face looks, in degrees clockwise from north. Default 180. |
| Friction Angle | Friction angle of the discontinuities, for sets without a joint strength of their own. It starts from the active stress tensor's Friction Coefficient the first time the analysis is shown in the window, as the mesh analysis does (about 31° for 0.6). |
| Lateral Limit | Degrees either side of straight out (planar sliding) or straight in (toppling) that a plane or line may swing and still count: 20 for planar sliding, 30 for flexural toppling, 20 for direct toppling, each kept separately. Wedge sliding has none, and the row is not shown. |
| Minimum Wedge Angle | Wedge Sliding and Direct Toppling only. Two planes closer to parallel than this form no line. Default 10. See Minimum wedge angle. |
| Intersection Scatter | Wedge Sliding and Direct Toppling only. Plot every line of intersection as a dot. Off by default. See The intersection scatter. |
| Factor of Safety | List the limit-equilibrium factor of safety of the chosen failure mode on each set's mean plane, or each pair of sets' for wedges, for this face. |
| FoS Slope Height | Height of the slope from toe to crest, in metres. Default 20. |
| FoS Unit Weight | Unit weight of the rock, kN/m³. Default 26. |
| FoS Water | 0 dry to 1 full: how full of water a planar block's tension crack is, and how saturated a wedge. Default 0. |
| FoS Block Width | Spacing of the toppling planes, in metres, for block toppling. 0 (default) leaves toppling out. |
| FoS Probability | Beside each factor of safety, the probability of failure from 1000 Monte Carlo samples, and below them the inputs that move the lowest factor most (see below). |
Sets From and Terzaghi Correction, in the same group, choose the sets and weight the percentages; see the Stereonet User Guide.
What is drawn
| Mode | The net shows | Curves | Zones |
|---|---|---|---|
| Planar Sliding | Poles | Slope face, friction cone, daylight envelope, lateral limits | Red: poles of the planes critical for planar sliding. |
| Wedge Sliding | Lines of intersection | Slope face, friction cone, friction plane | Red: lines that slide (Markland's test). Yellow: lines that daylight between the slope face and the friction plane, where a wedge can slide only on one of its planes. |
| Flexural Toppling | Poles | Slope face, slip limit plane, lateral limits | Red: poles of the planes critical for flexural toppling. |
| Direct Toppling | Lines, and the poles of base planes | Slope face, friction cone, slope angle cone, lateral limits | Red: direct toppling (Dips' zones 1 and 2). Yellow: oblique toppling (zone 3). |
The zones sit over the density contours and under the poles and planes; the curves are drawn over the planes, under the interpretation poles. Both work in equal-angle and equal-area. The zones are drawn for the Friction Angle of the analysis; a set with a friction angle of its own is tested at it, so its percentages may differ from what the zones suggest.
In wedge sliding and direct toppling the zones apply to lines of intersection, but the net plots poles. Tick Intersection Scatter to see the lines themselves, or turn on Show Planes and read the zones where the great circles cross; the percentages in the key count the lines.
The intersection scatter
Intersection Scatter, offered in Wedge Sliding and Direct Toppling, plots every pairing of one set's planes with another's as a small dot where the two planes meet, raw and unbinned, in the colour of that pair of sets:
- Solid — the line is critical for the mode at the pair's own friction angle, the lower of its two sets' angles: a Wedge line by Markland's test, or a Direct line.
- Pale — it is not. A line in the yellow zone alone, sliding on one plane or toppling obliquely, is pale.
While it is on, each row of the Lines table starts with a dot in its pair's colour, and a note under the table says what the dots are: Dots: intersections, solid where critical at the pair's friction. Past 3000 pairings an even sample is drawn, the same on every redraw, and the note gives it: Every 4-th pairing: 3000 dots. Below that it says how many pairings closer to parallel than the minimum wedge angle made no dot, if any did: 12 pairings too near parallel. The table's percentages and N are never sampled.
A dot is a measurement, so the Terzaghi Correction does not change it. The dots go into the SVG export, under the wedge markers, and are not drawn with Remove Regional Dip on.
The zones are drawn for the one Friction Angle of the analysis, but the dots are judged as the table judges each pair. In wedge sliding, a pair whose two sets both have a stronger joint strength can show pale dots inside the red zone, and a pair with a weaker set can show solid dots outside it. Trust the dots: they agree with the table. Direct toppling's test takes no friction angle, so its solid dots are exactly those in the red zone.
Wedge markers
With Factor of Safety on in Wedge Sliding, a larger dot marks where each pair of sets' mean planes meet: the line that pair's wedge factor of safety was worked out on. It is in the pair's colour, and a dot of the same colour starts that pair's row in the Wedges table, so the net and the table can be read against each other.
The marker is filled where the pair has a factor, and hollow where no wedge forms and the row reads -: the line does not daylight, say. A pair whose mean planes are within 1° of parallel gets no marker. The key says Dots: where mean planes meet., and Hollow: no wedge forms. when any are hollow. The markers go into the SVG export.
The key
The kinematic key is the first of the keys that share columns to the right of the net, below the stereonet key when that is shown; a long key continues in further columns rather than off the bottom of the view. It names the mode, the slope face, the friction angle and the lateral limits, gives a swatch for every curve and zone drawn, and then a table:
| Mode | Table |
|---|---|
| Planar Sliding, Flexural Toppling | Planes: the percentage Critical and the number of planes N, for All planes together and for each set. |
| Wedge Sliding | Lines: Wedge (passing Markland's test, as the map counts), 1 plane (blocks that slide on one of their planes rather than along the line: Hocking's cases among the Wedge lines, and lines in the secondary zone) and N, the number of pairs of planes that meet in a line, for All and for each pair of sets, A × B. |
| Direct Toppling | Lines: Direct and Oblique, and N; then Planes: Base, the share of each set's planes that can be the base of a toppling column. |
A set's percentage is what the map layers hold for a face of this orientation when that set is the most critical one. All counts every plane, or every line between two sets, together. A set with its own friction angle shows it beside its name, as J1 (34°). Long set names are shortened with an ellipsis in the middle, so they stay clear of the columns and names that start alike stay apart. When a folder's poles were left out as an older Auto-cluster's originals, the key ends with how many.
With Terzaghi Correction on, the key says the percentages are weighted and gives the sampling line, as plunge/trend, and the cap: Line 00/090, cap 7. N still counts what was measured: the percentages are shares of the weighted total, but a pair of planes each weighted 7 weighs 49 and adds 1 to N.
Factors of safety on the stereonet
With Factor of Safety ticked, the key adds a table headed Factor of safety, H 20 m (the FoS Slope Height):
- Planar Sliding — Mean planes and FoS: the planar factor of each set's mean plane.
- Wedge Sliding — Wedges, FoS and Slides: the factor of each pair of sets' wedge, and how it slides: both planes, the 1st or 2nd plane alone, or lifts off. Each row starts with the colour of its marker on the net.
- Flexural Toppling and Direct Toppling — Mean planes and FoS: the block-toppling factor of each set. It needs a FoS Block Width; the key asks for one.
With FoS Probability also ticked, each table gains a PoF column, the probability of failure of each set or pair, and below it the overall probability, PoF 12% (1000 samples). A Sensitivity table then lists the four inputs that move the lowest factor of safety most — a set's dip, dip direction, friction or cohesion, or the water — with the factor at -1 s.d. and +1 s.d. of each: one standard deviation either side of its value, a plane being moved by its set's Fisher spread. The stereonet draws 1000 samples with a fixed seed, so the numbers do not change between redraws; it uses a friction standard deviation of 2° and a cohesion coefficient of variation of 0.25, and draws each set's plane from its Fisher spread.
A factor of 10 or more reads >10, and - means the set forms no block on this face. The stereonet always uses the critical tension crack, a level upper slope, no seismic load and no trace limit; the mesh analysis takes those from its settings. The factors are worked out as on the maps, described in the method.
The probability of failure figure
Figure → Probability of Failure in the analysis report's editor adds the probability of failure of the Charts panel's stereonet to the report, for its kinematic face and failure mode with Factor of Safety on. The figure is a histogram of the lowest factor of safety over 1000 samples, with the limit at 1 and the probability of failure marked, and the tornado of the inputs that move it. Like the stereonet's FoS Probability, it uses a friction standard deviation of 2°, a cohesion coefficient of variation of 0.25 and each set's Fisher spread, not the spreads set for the maps. Open a report first; the figure needs the Geomechanics module. Athos can insert it too.
Exporting
Export → SVG writes the zones, the curves, the dots and the key with the rest of the net. The canvas widens to make room for the key and lengthens when the keys run below the net, so nothing is cropped; a stereonet figure in an analysis report is drawn the same way.
Choosing the parameters
- Radius and scale. Match the radius to the failures you are assessing: a few metres for bench-scale blocks, tens of metres for a whole cutting. Too small a radius follows rubble and noise, and leaves points on sparse parts of the surface without a face; too large a radius smooths benches and gullies into the overall slope. In a pit or a tall cutting, tick Bench, Inter-ramp and Overall and compare the maps: each answers a different design question. Check Face Dip on the model before trusting the percentages.
- Friction angle and joint strength. The friction angle of the discontinuities themselves. Give each set a joint strength of its own when you can: the friction angle is then the set's, and the factors of safety use its full strength.
- Lateral limits. 20° for planar sliding and 30° for flexural toppling are the usual values (Hoek & Bray; Goodman & Bray). Widening a limit is the conservative choice.
- Classify at. Raise it to show only places where a large share of a set is critical.
- Weighting. Use Terzaghi when one face orientation dominates the mapping and some sets are under-represented. Use Lean on the planes measured within the radius when the mapping is patchy, so unmapped stretches show the project-wide hazard rather than the worst case.
- Factors of safety. The slope height is the Local Relief at the scale's radius, so the radius sets the height as well as the face: at the bench scale it is roughly a bench, at the overall scale the whole slope. Give a block Width from the spacing of the toppling set for block toppling. Water and the seismic coefficient apply to every point alike.
- Probability of failure. 200 samples give a map to read; raise them for a design figure. Give a seed when you need to reproduce a run exactly, or re-run it from Analysis Runs, which uses the seed it recorded. A set with few planes has a poorly known K, so its spread is poorly known too.
Tips and troubleshooting
- "Slope stability needs discontinuity sets". No planes are tagged with a Set, no folder under Orientations holds planes, and no DFN has fractures. Tag each set's planes with a Set (Auto-cluster does), put them in a folder of their own, or generate a DFN, and run it again.
- Wedge Sliding and Direct Toppling are 0 everywhere. They need planes in two sets that are not parallel; the messages panel says so when there are none. On the stereonet the key says Lines need planes in two sets. Check the minimum wedge angle too.
- Pale dots sit inside the red zone. The pair's sets have joint strengths of their own; see Why pale dots can sit inside the zone.
- A warning says planes are in more than one chosen set. A folder and a Set, or an old Auto-cluster folder and the folders below it, hold the same planes. Untick one of each pair.
- Overhangs read as slopes facing the other way. The faces could not be told from their backs.
On a point cloud without
Nx,Ny,Nzlayers this is always so; analyse a mesh of the same surface instead. On a mesh, a surface that faces both ways within the radius does the same: use a smaller radius. - More than half of a mesh's faces look downward. Its normals probably point into the rock. Run it again with Its normals point into the rock ticked.
- Many points have no value. The radius is too small for the point spacing or the gaps in the surface. Raise it.
- "Block toppling needs the toppling planes' spacing". Give a block Width in the factor-of-safety settings, or untick block toppling.
- The factor-of-safety layers hold no values. No set's mean plane forms a planar block, a wedge or a toppling column on the faces. Check the sets' mean planes against the faces on the stereonet.
- An analysis run says stale. A set it read has changed since. Run Slope Stability again from the object to bring the layers up to date; Re-run on the run repeats it with the planes it recorded.
- No zones are drawn on the stereonet. Remove Regional Dip is on: the zones are drawn for the orientations as measured, and the key asks you to turn it off.
- Part of the key is off the right of the window. A long key continues in further columns to the right. Widen the window or pan the view; the SVG export always holds the whole key.
- The 3D stereosphere shows no zones. The kinematic overlay is drawn on the 2D stereonet only.
References
- Read, J., & Stacey, P. (eds.) (2009). Guidelines for Open Pit Slope Design. CSIRO Publishing, Melbourne.
- Lato, M. J., Diederichs, M. S., & Hutchinson, D. J. (2010). Bias correction for view-limited lidar scanning of rock outcrops for structural characterization. Rock Mechanics and Rock Engineering, 43, 615–628.
- Terzaghi, R. D. (1965). Sources of error in joint surveys. Géotechnique, 15(3), 287–304.
The references for the kinematic tests and the factors of safety are on the method page.
See also
- Slope Stability — Method — the tests, the percentages, the face fitting, the weighting and the factors of safety, with the formulas and references.
- Geomechanics — how slope stability fits with the other tools.
- Stereonet User Guide — plotting, grouping and contouring the orientations the analysis reads.
- Sets — the project Sets and their joint strength.
- Orientation — Auto-cluster Selected Orientations.
- Analysis Runs — the record of each run.
- Bench As-built — bench heights, face angles and berm widths from the scan.
- Attributes — displaying, filtering and colouring the layers the analysis writes.