All technical data shown has been generalized/anonymized for public representation.
Work · Arup / GEO-CEDD · 2024

Natural Terrain Mitigation — Back-analysis of Pinehill Village and Lei Pui Street Landslides

3D numerical back-analysis of two rainstorm-induced landslides using LS-DYNA, quantifying mobility and failure scenarios to inform risk-reduction measures for government stakeholders.

3DLS-DYNA numerical back-analysis
2Landslide sites analysed
OHF → CDFFailure-transition modelling
PythonModel set-up & figure pipeline
01

Challenge

Aerial photo of the Hong Chi Pinehill Village landslide
General view of the landslide (Hong Chi Pinehill Village). (a) Aerial photo.
Top-down view of the topography modelled in LS-DYNA
(b) Top-down view of the topography modelled in LS-DYNA (t = 0 s).

Two rainstorm-induced landslides on natural terrain: how far and how fast would the failed material have travelled and deposited? Working for GEO / CEDD, Bonnie back-analysed the mobility of the Pinehill Village and Lei Pui Street events using the 3D numerical tool LS-DYNA. The work quantified failure scenarios and informed risk-reduction measures for government stakeholders, including an additional run assessing the Pinehill Village event without the rigid debris-resisting barrier in place.

02

Method & Standards

⛰️ Numerical modelling (LS-DYNA)

The analysis reproduced the runout and mobility of the failed material across complex natural terrain, using an additional barrier-free run to isolate the effect of the reinforced concrete structure on the observed outcome.

🧮 Python tooling & method development

Bonnie wrote Python tooling for efficient processing of topographical data and model set-up, and developed an approach for modelling the transition between open hillside failure (OHF) and channelised debris flow (CDF) — a key behaviour where failure initiates on open slopes before concentrating into drainage lines.

03

Visual / Demo

Photograph of the rigid debris-resisting barrier
Figure 6. Debris deposition in the rigid barrier (Hong Chi Pinehill Village). (a) Photography, looking at the rigid barrier.
Top-down view of debris deposited in LS-DYNA
(b) Top-down view of the debris deposited in LS-DYNA (t = 50 s).
Recommended apparent friction angle for open hillside failure
Recommended apparent friction angle (φa) for open hillside failure (OHF). The back-analysis used the actual landslide event volume to check whether the current government technical guidance note value range is appropriate. Source: CEDD Technical Guidance Note 34.

She extracted and post-processed results from d3plot output and produced the analysis figures in Python, turning the simulation output into the visual evidence used in reporting and review.

04

Impact

By matching the back-analysed runout and deposition against the observed events — including the effect of the debris-resisting barrier — the work gave government stakeholders an evidence base for risk-reduction measures, delivered through a repeatable Python-driven modelling and figure pipeline.

📄 This work contributed to the published study "Successful interception of natural terrain landslides by reinforced concrete debris-resisting barriers at two sites in Hong Kong." The team helped write the paper but was not credited in the author list — full attribution goes to the published authors. Read it on ResearchGate →