Soil layers · dig force · slope stability
Simplified 2D model. Typical textbook values, not a real site.
Each layer has a unit weight γ, cohesion c and friction angle φ. Typical values: topsoil γ 16 kN/m³, c 2 kPa, φ 28°; clay γ 18.5, c 20, φ 18°; gravel γ 19, c 0, φ 36°; bedrock γ 24 (the claw cannot bite it).
σ' = σ − u and τf = c' + σ' tan φ'. Water in the pores carries pressure u = γw (z − zw) below the water table, so it lowers the strength.
σ' = σ − u, τf = c' + σ' · tan φ'A bite needs force F = τf · A · k. Here A is the cut area (bite width × 0.5 m) and k = 1 for the claw. Simplified: no bucket dynamics. If F is above your Force setting the claw stalls.
F = τf · A · k (example: z 1 m, γ 18, φ 30°, c 5 kPa, A 0.3 m² → 4.62 kN)Spoil piles into a cone with slope angle θ (gravel 34°, topsoil 28°, wet clay 22°). Example: 3 m³ at θ 34° gives h 1.09 m and r 1.62 m.
h = (3V tan²θ / π)^(1/3), r = h / tan θ, V = π r² h / 3Active coefficient Ka = tan²(45° − φ/2). An unsupported vertical cut stands up to H_c. The weakest layer in the wall sets H_c here. This is a simplification: no stress redistribution.
Ka = tan²(45° − φ/2) H_c = (4c/γ) · tan(45° + φ/2) z0 = H_c / 2
Example: c 10 kPa, φ 20°, γ 18 kN/m³ → H_c = 3.17 m. The inverse-tangent form gives 1.56 m, which is wrong.
Lateral force per metre of wall: P_a = ½ Ka γ H² (plus water force). A block must hold P_s ≥ 1.5 · P_a. Wood gives 25 kN/m, steel 60 kN/m. Example: H 2 m, γ 18, φ 30° → P_a = 12.0 kN/m.
P_a = ½ · Ka · γ · H²Unit weights scale with g, so H_c is proportional to 1/g and digging force to g. On the Moon (1.62 m/s²) clay stands about 36 m, which is why the Moon preset is wide.
Never enter an unsupported trench in real life. The depth at which shoring is required depends on local rules. Check your local standard (for example the UK HSE or US OSHA guidance on trench support). No people are simulated here.
One 2D cross-section (3D looks are visual only). No water CFD. No claw dynamics. The water table is static. H_c uses the weakest layer. Wet soil is modelled as a cut in strength, not as flow.
This simulation needs WebGL. Try a current browser with hardware acceleration on.