Recorded slope comparisonReal solver outputCoupled THM + material-point

Landslide analysis

Two degrees separate stability from 32 m of run-out

The same overconsolidated clay slope is run twice with the same engine and settings. Only the warming ramp changes.

+1 °C…

+3 °C…

Replay position

00.0s / 11.1s

Excess pore pressure on the native 0.5 m MPM background grid (view 0–120 m of the 240 m domain, no vertical exaggeration). All 47 published frames are retained in order; replay timing is a web presentation aid and the in-frame “sim t” clock is the recorded simulation time. No motion interpolation was added.

L-01

Recorded comparison

At +1 °C the slope holds. At +3 °C the slip band crosses 60 kPa at sim t = 12 s and the recorded deposit front runs out 32 m.

Source: 47 published frames
Grid: Native 0.5 m
Threshold: 60 kPa
Run-out: 32 m

Method

Excess pore pressure is cell-averaged on the solver's own 0.5 m background grid. The colour scale is anchored at the 60 kPa threshold; the 32 m value is the recorded final deposit front, defined by the furthest grid column holding at least 1 m of material above the terrace.

Every published frame remains in source order. The MP4 adds no motion interpolation; its playback clock is presentation time and the in-frame “sim t” clock is simulation time. Read the render receipt (new tab) and transcode receipt (new tab).

Open full evidence (new tab)

Recorded outcome

Stable scenario: +1 °C Holds
Failure scenario: +3 °C Fails
Threshold crossing: 60 kPa
Computed run-out: 32 m

Mechanism trace

From warming to computed run-out

The values describe this recorded comparison. They are not general landslide thresholds.

01Warming
The controlled ramp changes from +1 °C to +3 °C.
02Slip-band pressure
Thermal pressurisation crosses 60 kPa in the warmer run.
03Computed run-out
The failed mass travels 32 m in this scenario.

Recorded mechanism gallery

Six recorded slope mechanisms, one engine

Each board below is a recorded run of the same THM core and material-point engine, rendered in the same figure language as the comparison above: metre axes on the solver's native background grid, verdicts typeset only from recorded motion, and a render receipt beside every asset. Three concepts were re-run on 2026-08-26 from their archived decks; none is field validation.

Luxembourg demonstration

Final state of the Luxembourg cliff pair: the dry-winter panel holds while the saturated-joint panel shows the released slab and 23.8 m of debris, with pressure, front-advance and kinetic-energy traces

L-02

A Luxembourg sandstone cliff — the wet winter decides

A 32 m Grès-de-Luxembourg-type face with a bridged decompression joint: the 40 kPa dry-winter cleft dose holds; the 160 kPa saturated joint releases the slab, which crosses the bench and runs out 23.8 m.

Controlled variable: Cleft-water dose
Dry winter: 40 kPa · holds
Saturated joint: 160 kPa · 23.8 m
Status: Demonstration

Method

Demonstration pair inspired by the Mullerthal and Alzette cliff setting — model-scale strengths, NOT site-calibrated. The cleft-water dose is a prescribed pressure on the painted joint; the joint is bridged top and bottom, so both bridges must break before the slab releases. Render receipt (new tab)

Open full evidence (new tab)

Controlled comparisons

Final state of the feedback pair: both +5 °C slopes failed, with temperature, pore-pressure and front-advance traces separating the feedback run from its control

L-03

Shear heating is the amplifier

Two identical +5 °C runs, one switch: with shear-heating feedback the slip band reaches 11.87 °C and 712 kPa; the control without feedback stops at exactly 5.000 °C and 300 kPa.

Feedback on: 11.87 °C · 712 kPa
Control: 5.000 °C · 300 kPa
Run-out: 42.6 vs 40.1 m

Method

Recorded cases dT5 and dT5_nofb of the same HPC campaign as the headline comparison; asymmetric colour scale anchored at the 120 kPa analytic trigger. Render receipt (new tab)

Open full evidence (new tab)

Final state of the two-storm pair: the drained hillside intact, the undrained hillside with its surface layer flowed onto the terrace

L-04

Two storms, one hillside — drainage decides

The same 2 m surface layer under two storm doses: 10 kPa drained dissipates and the slope holds; 36 kPa undrained accumulates past the ~30 kPa layer threshold and the skin flows 13.0 m.

Drained dose: 10 kPa · holds
Undrained dose: 36 kPa · 13.0 m
Layer threshold: ~30 kPa

Method

Re-run 2026-08-26 from the archived concept deck; the threshold reading belongs to this layer and slope, not to storms in general. Render receipt (new tab)

Open full evidence (new tab)

Mechanism singles

Final state of the Case A slope with the emergent basal rupture band highlighted in the softening-strain field

L-05

Birth of a slip surface

A softening slope with no predefined failure plane: the basal rupture surface emerges as a localisation band, and the mass runs out 10.0 m before arresting.

Particles: 31,960
Run-out: 10.0 m
Field: Softening strain κ

Method

The bundled Case A record of the material-point engine — the same run family as the Lube/Bui collapse validation named in the evidence boundary. Render receipt (new tab)

Open full evidence (new tab)

Final state of the permafrost cliff with the detached block arrested on the terrace and the joint-pressure trace at its 225 kPa dose

L-06

Permafrost is the glue

A 32 m rock cliff with an ice-cemented joint: a +4.5 °C thaw dose of 225 kPa collapses the joint's cohesion, a coherent block detaches, topples and arrests 12.0 m out.

Thaw dose: 4.5 °C × 50 kPa/K
Joint pressure: 225 kPa
Run-out: 12.0 m

Method

Re-run 2026-08-26 from the archived concept deck; the thaw dose is a prescribed pressurisation of the painted joint. Render receipt (new tab)

Open full evidence (new tab)

Paraglacial sequence

Final state of the debuttressed wall with the melted wedge remnant at the toe and raveled debris across the terrace

L-07

The glacier was the buttress

A rock wall held by an ice wedge: a 50 kPa melt dose softens the wedge, it flows off, and the freshly exposed face ravels — debris reaching 24.0 m across the terrace.

Melt dose: 5 °C × 10 kPa/K
Wedge: Ice, ρ = 900
Debris front: 24.0 m

Method

Re-run 2026-08-26 from the archived concept deck; the end state reproduces the originally published animation, which verifies the recovery. Render receipt (new tab)

Open full evidence (new tab)

Evidence boundary

Recorded scenario. The two runs use the same overconsolidated clay slope, engine and settings. Only the warming ramp changes. The 60 kPa crossing and 32 m run-out belong to this scenario.

Figure and timing. Excess pore pressure is shown on the native 0.5 m grid under a threshold-anchored colour scale. The “fails” verdict appears from the frame that records the crossing at sim t = 12 s; “holds” appears in the terminal frames. The 47-frame encode preserves order and adds no motion interpolation.

Method checks. The analytic oedometric K₀ check and the Bui (2008) rod-collapse and Lube (2005) column-collapse comparisons remain separate method evidence.

Platform architecture. An implicit core handles transient coupled thermo-hydro-mechanical response; an explicit material-point method handles collapse and run-out. This recorded comparison is not field validation of a specific landslide site.

Bring a case

Have a slope problem to test?

Tell us which decision, failure mechanism or evidence threshold your current workflow needs to support.

Discuss a landslide case