Find where the prediction changes.
Explore the combined effects of groundwater depth and normalized CPT resistance bounds. Identify the governing layer, the PGA threshold, and where more precise resistance information could reduce the scenario envelope.
Layer model
| Layer | Bottom, m | γ moist, kN/m³ | γ sat, kN/m³ | qc1Ncs | Candidate? |
|---|
Layers start at 0 m and are contiguous. Candidate = evaluate using a sand-like triggering framework. An excluded layer is not certified non-hazardous.
What is different about this prototype?
The workflow combines an inverse PGA threshold, governing-layer switches, a bounded decision map, and layer-by-layer resistance-information sensitivity. The candidate research question is whether those diagnostics improve investigation decisions over a single nominal FS profile. Research novelty and field performance remain unestablished. The calculations do not constitute a new triggering law.
Investigation ranking is the average contraction of the critical-PGA envelope when one layer’s resistance bound is ideally resolved at its nominal value. It is averaged uniformly over the displayed groundwater grid, not weighted by probability. It excludes test cost, measurement error, spatial variability, and groundwater uncertainty reduction. It is a sensitivity diagnostic, not a formal value-of-information calculation or a drilling recommendation.
Method and applicability
The deterministic normalized-CPT branch uses Boulanger and Idriss (2014/2016): CRR7.5,1atm(qc1Ncs), magnitude scaling, Kσ, and the magnitude-dependent rd relationship. FS = CRR × MSF × Kσ / CSR. Critical PGA = CRR × MSF × Kσ / [0.65 × (σv/σ′v) × rd × target FS]. Atmospheric pressure is 101.3 kPa and water unit weight 9.81 kN/m³.
Hydrostatic pore pressures and layerwise moist/saturated unit weights update with groundwater depth. Layers are subdivided and split at the water table. Only saturated candidate cells are evaluated. The limiting PGA concerns first subcell threshold crossing, not ground manifestation or failure. Refine the mesh before interpreting thin layers.
qc1Ncs is already normalized and clean-sand-equivalent. It is held fixed as a state descriptor, with explicit bounds. The app does not re-normalize raw CPT data or predict changes in fabric, density, aging, consolidation, suction, capillary saturation, or site response. Surface PGA is a supplied independent demand. Static shear correction is omitted for the level-ground scope.
Prototype bounds: depth ≤20 m, Mw 5.5–8.5, qc1Ncs and its bounds 21–211, and no slope or foundation loading. These are implementation limits, not a statement that validation is uniform within them. Shallow low-stress extrapolation is flagged. Excluded soils may require cyclic-softening or other assessments. No sediment ejection or settlement is predicted.
References and research alignment
Boulanger & Idriss, CPT-based liquefaction triggering procedure (2016). Equations and method assumptions provide the calculation basis. The companion 2014 report is UCD/CGM-14/01.
Cubrinovski & Dhakal (2025), Simplified procedure for estimating liquefaction-induced settlement, and Dhakal & Cubrinovski (2025), Key considerations for a simplified evaluation of liquefaction-induced settlement, motivate a later consequence-focused extension. Their settlement equations are not implemented here.
This independent prototype is not endorsed by Dr. Dhakal or UBC. See the published research and validation plan for the proposed study, current checks, and outstanding work.