Slope User Manual

Slope signs in with your hetGE account. After signing in you land on the projects screen: a project groups related designs, and each design is one slope model with its own geometry, materials, units and analysis settings. Create a project, add a design, and it opens in the editor. Cloud designs save your changes to your account automatically as you work; you can also export any design as a JSON project file and load it again later, so nothing is locked in.

The switch at the top of the editor moves between Model, where you build the cross-section, Analyze, where you choose an analysis method and search strategy and solve, and Report, where you assemble and export a PDF calculation package. The drawing canvas is shared: the same section stays in front of you, and each workspace changes which tools and overlays surround it. Your model, results and settings persist as you move between them.

Two layouts are offered from the user menu's settings. Classic keeps the header, the sidebar and its panels around the canvas. Modern maximises the canvas to fill the window, moving the tools onto a rail beside it and the workspace switch into a single top bar; the sidebar becomes a scrolling column that does not take the drawing with it. Both show the same model and the same results — it is a matter of how much chrome you want around the section. The colour theme (light or dark) is set in the same place.

In the Model workspace, pick the region tool and click out the outline of your slope — each click places a vertex, and closing the outline completes the region. If you already have coordinates in a spreadsheet, paste them instead: the coordinate entry accepts one x y pair per line. Add a region for each soil layer, then create materials with unit weight, cohesion and friction angle and assign one to each region. Slope derives the ground surface automatically from the top edges of your regions. If groundwater matters, draw the water table as a polyline across the section.

Switch to Analyze. The default configuration — simplified Bishop, a grid–radius search at normal resolution, plain factor of safety — is a sensible starting point, so you can simply press Solve. Slope sweeps the search space of trial circular surfaces, discards the ones that are not kinematically admissible, and reports the critical surface: the one with the minimum factor of safety. On a model like the example below this takes well under a second.

The critical surface is drawn on the canvas with its factor of safety, and a spectral heat map shows how the minimum FoS varies across the section — red is where it fails. Open the slice details to see the full ledger behind the number: a free-body diagram of any slice with every force annotated, the interslice forces, and base normal, shear and pore-force distributions along the surface. If the model changes after a solve, the results are flagged as stale until you solve again.

Switch to Report, choose what the calculation package should contain, and generate it: a real PDF with a cover page, table of contents, true-to-scale figures, slice-by-slice tables and reinforcement summaries, built in seconds in the browser. This is the document trail for review — the same numbers you saw in Analyze, laid out for checking.

Open the Help menu (the ? button in the header) and choose Load example project: a two-material bench cut with a groundwater table, a crest surcharge and a soil nail that solves instantly with default settings. The Help menu also replays the guided tour, opens this manual and lists every keyboard shortcut.

A region is a closed polygon representing one soil unit. Draw it vertex by vertex with the region tool — snap-to-grid keeps coordinates tidy, Enter (or clicking the first vertex) closes the outline, and Esc cancels a draft. Vertices can be dragged afterwards, and each region has a coordinate table for exact numeric editing. The exposed ground surface is derived automatically from the top edges of the regions, so slip surfaces always daylight on the geometry you drew.

Regions can also be created without drawing: paste a column of x y pairs — one vertex per line, straight from a spreadsheet — and Slope builds the polygon for you. This is the quickest route when the section comes from a survey or a CAD export. The same coordinate table is available for every existing region, with a live vertex and area summary.

A region can be divided after the fact with the split tool: click a line of points across the section and finish it, and every region the line passes through is cut along it. There is no selection to make first — the stroke decides what it divides, which is the point. A boundary drawn through a layered slope has to divide each layer along the same vertices; cutting one layer at a time re-derives that boundary each time, and any difference leaves a sliver the analysis then refuses to run on.

The line has to genuinely cross a region — enter it and leave it — so a stroke that stops inside cuts nothing, and one that leaves and comes back does not either. Anything drawn outside a region is ignored, so you can start and finish well clear of the geometry. A line that zigzags across a region divides it into as many pieces as it crosses, and every piece inherits the original's material. The whole stroke is one undo step, however many layers it cut.

Materials carry unit weight above and below the water table (γ, γ_sat), a display colour, and a constitutive strength model. Mohr–Coulomb (c′, φ′) is the default; Slope also supports undrained strength (constant or increasing with depth), Hoek–Brown for rock masses, anisotropic strength, a hyperbolic envelope, SHANSEP, and the special no-strength and infinite-strength models for water bodies and bedrock. Assign a material to each region from the materials panel; a legend overlay on the canvas keeps the colour coding in view.

A material's soil condition (controlled fill, uncontrolled fill, in situ) is shown only when a standard in use reads it (AS 4678), or when the material already has a condition other than in situ. For a no-strength or infinite-strength material it is shown only where pull-out reads it.

Materials can start from the built-in soil presets (BS 8002:2015 and Hong Kong Geoguide 1) through From library or Built-in soils. Each preset material shows where it came from and whether it has been edited, and the report cites the source. Presets are typical values for use in the absence of test results, not a substitute for site investigation. See the built-in soil presets reference.

The water table is a polyline across the section. Below it, pore pressure is computed hydrostatically (u = γ_w·h) and soil weight uses the saturated unit weight; the drained / undrained treatment of that pore pressure is chosen in the Analyze workspace. A tension crack is modelled as a crack-bottom line near the crest with a water-fill fraction from 0 (dry) to 1 (full), which adds the corresponding hydrostatic thrust where a trial surface intersects the crack.

Two load types cover most surcharge situations: distributed loads (a pressure in kPa over an x-range of the ground surface) and line loads (a force per metre of run at a point on the surface), both with an optional inclination. Each load is categorised as permanent or variable, which matters when a partial-factor design standard is active — the two categories receive different load factors. Load endpoints can be dragged directly on the canvas.

Four reinforcement types are available: ground anchors, soil nails, geosynthetic layers and piles. Linear elements are drawn head-to-tip on the section (the head snaps to the ground surface); geosynthetics are horizontal sheets at an elevation. Each element carries its capacities — tensile or characteristic capacity (the design standard's factors give the design value), pullout via a specified bond stress or one calculated from effective normal stress, plate (stripping) capacity, pile shear — and an out-of-plane spacing that converts element forces to per-metre values. The engine resolves each element's available force where it crosses a trial surface and reports the governing failure mode, so the Analyze results show which parameter to fix, not just a smaller number.

Capacities are labelled characteristic (T_k, V_s,k, τ_k, T_p,k), because the standard's factors turn them into design values. An element's product and model factor fields appear only where a standard in use reads them for that element, or once you enter a value. The note beside them reads Required by … and names every standard that reads the factor, … — enter it to run when a required value is missing, or Not used by … when none of the standards in use applies it, so a run is never blocked by a factor you could not see.

A layer under BS 8006-1 or AS 4678 also has direct-sliding coefficients: a′, which multiplies tan φ′ (greater than 0, at most 1), and a′bc, which multiplies c′ or c_u (between 0 and 1), both from the product's direct-shear tests. They set the strength of the soil sliding across the layer's surface in the Direct sliding check. The Auto search and the report need them for a layer that can slide; manual surfaces and the probabilistic analysis never read them. The note beside them reads Read by … for direct sliding in the Auto search.

Repeating elements are placed as an array rather than one at a time: anchor the array on the ground surface and give it a pitch, a count and a direction, and its members are derived from those parameters — edit the array and every member follows. Two spacings are in play and mean different things: the array pitch is the in-plane step along the face (the lift height for geosynthetics), while the out-of-plane spacing is what converts each element's capacity to a per-metre value for the two-dimensional analysis.

Each project is either metric (m, kN/m³, kPa) or US customary (ft, pcf, psf), chosen when the project is created and fixed for its lifetime — values are stored natively, never converted behind your back. Defaults for new materials, loads and reinforcement follow the project's unit system.

Seven methods of slices are available: Ordinary (Fellenius), simplified Bishop, simplified Janbu with an f₀ correction, Corps of Engineers, Lowe–Karafiath, Spencer, and Morgenstern–Price — a GLE formulation with a half-sine interslice force function. Bishop and Ordinary satisfy moment (or force) equilibrium on circular surfaces only and are auto-gated off polylines rather than report a silently wrong number; Janbu and the GLE family — Corps of Engineers, Lowe–Karafiath, Spencer and Morgenstern–Price — run on both circular and non-circular surfaces. Morgenstern–Price's force polygon visibly closes in the slice inspector when it converges. The engine break-aligns slice edges to ground-surface vertices, so no slice straddles a slope break.

Four strategies locate the critical surface. Grid–radius sweeps a slope-aligned grid of circle centres that expands and refines itself around the minimum. Entry–exit sweeps daylight points instead: two ranges on the ground surface — where surfaces enter and where they exit — that you can drag directly on the canvas. Block searches for a critical block or wedge mechanism through a thin weak seam, bracketing the seam with two windows and fanning active/passive wedges from each end. Manual lets you draw individual trial surfaces yourself, which is useful for checking a known or suspected mechanism. A surface can be a circular arc, a non-circular polyline drawn point by point or loaded from file, or a block mechanism — not every method runs on every surface kind (see above). All automatic searches respect a maximum-depth cap, and surfaces that leave the soil mass are rejected automatically.

The coarse / normal / fine presets trade run time for search density — fine sweeps thousands of trial surfaces. The number of slices per surface is set separately; the default suits most sections, and the report can re-state the count used for each result. Solve progress is shown live while a search runs.

By default Slope reports a plain (unfactored) factor of safety. Alternatively, pick one of 17 built-in partial-factor design standards — Eurocode 7 design approaches, BS, AASHTO LRFD and CSA families among them — or define a custom factor set. Factors are applied rigorously inside the engine: material strengths (tan φ′, c′, c_u), soil weight, permanent and variable loads, reinforcement capacities and seismic actions are factored where the code puts them, not divided out of a single result afterwards. The verdict compares the factored result against the standard's target.

Some standards ask for a choice about the project rather than a number: a consequence class (Eurocode 7 2024), an f_n category (BS 8006-1), a structure class (AS 4678) and, for BS 8006-2 soil nails, a geotechnical category of 1, 2 or 3. Each picker appears only when the standard you chose uses it. Category 3 runs Table 5 with every non-unity factor raised by 10 % (note A), and the report states the category. Projects saved under the former separate “Geotechnical Category 3” standard open on BS 8006-2 at category 3 and give the same results.

The Model workspace's material and reinforcement editors follow the same rule, showing a standard's inputs only where a standard in use reads them; see those sections.

BS 8006-1 (both editions) and AS 4678 also run a Direct sliding check: a search for a wedge whose base runs along a layer, with the interface strength the standard prescribes (fs = 1.3; or Φu 0.8 × Φn in controlled fill). It is a check of its own in the results table and can govern; the Auto search runs it, manual surfaces do not.

The partial-factors window has a Values | Sources switch. Sources shows, for each factor, the clause, table and page it comes from, how Slope interprets it and its validation status. A custom copy cites its base preset for the factors you have not changed and says which ones you changed. A custom set can be saved to your engineering library (Save to library) and reused in other designs.

Pseudo-static seismic loading applies horizontal and vertical coefficients k_h and k_v to every slice's weight. Enable the toggle, set the coefficients, and they participate in the equilibrium equations of whichever method you run; saved projects remember the effective values.

Seismic analysis runs under a seismic design situation, and the design-standard picker narrows to those while the toggle is on. Five standards carry their own earthquake combination — Eurocode 7 (Earthquake, EN 1998-5:2004), Eurocode 7 2024 (Seismic), AASHTO Extreme Event I, AS 4678 (Earthquake, I20.1) and CSA S6 ULS 5 — and switching seismic on keeps you within the standard you were already using, moving from its static combination to its seismic one rather than to the generic set. Codes without an earthquake combination fall back to Seismic (pseudo-static). Choosing a seismic situation switches the toggle on for you, since solving one at k_h = 0 would simply reproduce the static answer.

Two different judgements are shown, and they answer different questions. The verdict is the design requirement: Pass when the factor of safety reaches the required value for the factor set in use, Fail when it does not. That threshold comes from the standard — 1.5 for a classic global factor of safety, 1.0 for partial-factor and LRFD sets where the margin already lives in the factors, 1.1 for the generic pseudo-static set.

Beside it is the reserve: how far above (or below) that requirement the result sits, as a percentage. A reserve under 10% is coloured amber as a prompt to look closer. That threshold is this application's convention and not a provision of any design standard — design codes are binary at the requirement and define no comfortable tier above it. A Eurocode result of 1.05 against a required 1.00 fully satisfies its code while sitting in the amber band.

A result that did not converge carries no verdict at all: the number is shown, marked as unconverged, and left unjudged.

After a solve, the critical surface is drawn with its FoS and verdict, and several overlays can be toggled: the spectral minimum-FoS heat map, the FoS-tinted cloud of all trial surfaces, a ranked list of the best surfaces, a materials legend and a live analysis summary. The slice inspector opens a SLOPE/W-style free body for any slice — every force annotated, plus the interslice ledger and base distribution charts. With reinforcement in the model, a per-element results table reports each element's available force, its tangential and normal components at the crossing, and the governing failure mode.

Beyond the single deterministic factor of safety, the probabilistic panel turns a solved model into a full distribution. Randomize each material's γ, c, φ (or c_u for undrained materials) and, when seismic loading is enabled, k_h — each variable gets a distribution (normal, lognormal or uniform) and a spread expressed as either a coefficient of variation or a standard deviation, with optional bounds. For Mohr–Coulomb materials you can also set a c–φ correlation coefficient. A seeded Monte-Carlo or Latin-Hypercube sweep then runs the requested number of realizations — up to 10,000 sampled against the frozen critical surface, or up to 1,000 with the search re-run per realization for the overall-slope target — and the same seed, settings and model reproduce the identical run every time. The results are the full FoS distribution: probability of failure P_f, reliability index β under both normal and lognormal conventions, a histogram of the sampled factors, and a tornado chart ranking which parameter's spread drives the risk most. Probabilistic runs sample around a fresh deterministic solve, so solve first; the run panel explains why the button is disabled otherwise.

A report can cover more than the single configuration you last solved: choose the trial-surface sources, methods and design factor sets to include, and the report re-runs the whole matrix when it is generated. Every combination gets its own numbered sub-section with its own critical surface and slice charts — a search re-runs its entire sweep for each design situation, so two factor sets can settle on genuinely different critical surfaces.

Seismic design situations are chosen beneath the seismic toggle, alongside the coefficients they use. Those coefficients are the same ones the Analyze workspace holds — editing them here changes them there, because the report re-runs your analysis rather than keeping a second copy of its settings.

A design standard in the matrix may need a project choice the Analyze standard does not — a consequence class, f_n category, structure class or geotechnical category. The report panel then shows Design choices used by these runs with a picker for each. Each is one project choice, shared with Analyze, so changing it here changes it there.

The governing case is marked in every results table and headlined on the cover — separately for static and seismic, because a seismic case is usually the more critical of the two and a single figure would quietly become that one.

It is the case with the least reserve over its own requirement, not the smallest factor of safety. Those are different questions once several factor sets are in play, because a factor of safety is only comparable against the target it was solved for: a result of 1.63 against a required 1.50 has 8.6% in hand, while 1.50 against a required 1.00 has 50%. The lower number is the more comfortable case, and reading the table by magnitude would nominate exactly the wrong row. Where a report uses a single factor set every run shares a requirement, and the two coincide.

Each row carries a Pass or Fail against its factor set's required value, and the reserve over that requirement as a percentage. The verdict is the standard speaking; the percentage is how comfortably. A reserve below 10% is coloured amber as a prompt to look closer — that threshold is this application's convention and not a provision of any design standard, and the tables say so in a footnote. A run that did not converge carries no verdict at all.

Under each factor set the report prints its partial factors, and each row has a Use column: "Used", "Not used: reason" or "Not given: reason". The reason comes from your model — no material takes the factor, no load of that kind, a static design situation, no element of that type, calculated bond on the interface route, or no material with that soil condition. The soil factors the interface route reads count as used, whatever the surrounding soil's strength model. A reinforcement mode the standard does not factor is printed as "Not given", with "—" for its value.

Each set the report ran, and each check set, also states its source, interpretation and validation status, and lists its limitations. A custom copy states its origin, how many factors it changed and whether the preset has since been revised; the preset's evidence and limitations are labelled as the base preset's.

An element factors table beneath the partial factors lists every element whose product or model factor the set reads, by its canvas tag, with the value used.

A probabilistic analysis section ends with the partial factors, status and element factors of the standard the batch solved under — the Analyze workspace's active set, even when that set is not a report variant.

In the results summary a U column gives each check's utilisation, and the governing check is marked "(governs)".

Under BS 8006-1 and AS 4678 the checks table also lists Direct sliding, with a paragraph naming the layer, its coefficients and the factor applied; manual-surface and probabilistic sections say that they do not include it.

For an EC7 DA2 run, the slice data and the slice-detail modal state that the forces are in the resistance-scaled frame (resistance ÷ γ_E,eff), not E_d.

The contents panel controls what the package includes: cover page with project information, table of contents, a list of figures and a list of tables, model and results figures, slice-by-slice tables for each method and variant, and — when the model has reinforcement — per-element reinforcement tables grouped by type. Figures are rendered true to scale with the same styling as the canvas.

Each index is independently switchable and prints only when it has something to list, so a report with its figures turned off never carries an empty page. The generated PDF also carries bookmarks: every numbered section appears in the reader's navigation pane, with each analysis sub-section nested beneath its section — which matters once a matrix of methods and factor sets runs to dozens of figures.

Press Generate to build the PDF, preview it in place, and download it. Page size can be A4 or Letter. A small package builds in seconds; a wide matrix of methods and factor sets takes longer, since every combination is solved and drawn. Regenerate at any time after model or settings changes — the report notices when it has gone out of date and says so.

Projects group related designs — typically one project per site or scheme. Each design is a complete slope model; opening it loads the editor, and edits autosave to your account shortly after you make them, with the save status visible in the header. Designs can also be exported and imported as JSON project files, which is the interchange path with local, offline work.

Material and reinforcement libraries let you define common soils and support elements once and reuse them across designs. Materials and reinforcement templates are stored in SI and converted to the design's unit system when inserted. The Materials tab also lists the built-in soil presets, which can be copied into your library.

Standards. The Standards tab holds your custom design factor sets. Save a custom set from a design's partial-factors window (Save to library), or start one here from a built-in standard (New set). Insert a library set into any design with the From library button beside the design factor set picker: the design keeps its own copy, so later library edits never change it — the partial-factors window says when the library copy has changed since. Edit a library set's factors on its page; Save asks before replacing changes made elsewhere. See also the built-in standards reference.

Recently opened designs are cached on your device, so a dropped connection does not interrupt editing; pending changes are written back to your account when the connection returns. Analysis always runs locally in the browser — the cloud stores your data, it never computes your factor of safety.