Prediction of a Hypothetical Debris Flow using LS-DYNA Second JTC1 - PowerPoint PPT Presentation

Prediction of a Hypothetical Debris Flow using LS-DYNA Second JTC1 Workshop Arthur Cheung Geologist Hong Kong 4 December 2018

Brief Self-introduction ➢ Work as a geologist in Arup ➢ Have been working in natural terrain hazard study since 2011 ➢ Interested in the development of LS-DYNA in debris mobility modelling and debris-barrier interaction

Acknowledgement Yuli Huang Jack Yiu Richard Sturt AT&R UK and San Francisco HK Geotechnics

Outline ➢ Brief Introduction of LS-DYNA ➢ Benchmarking Exercise – Case D1 ➢ Model Setup ➢ Model Parameters ➢ Results ➢ Summary

The Numerical Model – LS-DYNA

What is LS-DYNA? ➢ LS-DYNA is a multi- purpose finite element program for linear and non-linear mechanics

What is LS-DYNA? ➢ LS-DYNA is a multi- purpose finite element program for linear and non-linear mechanics

What is LS-DYNA? ➢ Track records on impact analyses ➢ Able to model large deformation and movement of debris mass with the use of advanced finite element meshing method ➢ Able to simulate soil-structure interaction

Development since 2012 2012 - 2014 2014 - 2016 2016 - current The Energy Balance Engineering Interactions of Flexible Application between Barrier System Perspective Landslide Debris and Flexible Barriers Adopt in Research Validation Real Work

Debris Mobility Rock Fall 1 2 Modelling Interaction between 3 LS-DYNA the flexible barrier and the landslide Capabilities debris Energy Balance Optimisation in 4 5 Design of Flexible and Rigid Barriers PRIVATE AND CONFIDENTIAL

LS-DYNA Results Validated by Well-documented Cases Volkwein rock fall tests in Switzerland

LS-DYNA Results Validated by Well-documented Cases 1 3 4 2 1 Yu Tung Road Landslide 2 Fei Tsui Road landslide 3 Sham Tseng San Tsuen landslide 4 Kwun Yam Shan landslide

LS-DYNA Results Validated by Well-documented Cases Illgraben flexible barrier Full-scale field test flexible field test barrier in Veltheim

LS-DYNA Debris Mobility Modelling in Different Regions MAINLAND CHINA HONG KONG SEOUL Area: Mount. Umyeon, Seoul Area: Conghua Training Centre, China Area: Fo Tan, Hong Kong HONG KONG SIERRA LEONE HONG KONG Area: Victoria Public Mortuary and Cavern Area: Ocean Park, Hong Kong Area: Freetown, Sierra Leone

Hypothetical Landslide – Case D1

Case D1 – A historical landslide catchment in Hong Kong

Model Setup – The Topography ➢ Modelled by rigid shell elements ➢ Resolution = 5m x 5m ➢ No. of elements = 11,550 Landslide Source Boulders Dam Topography

Model Setup – Debris Modelling ➢ Volume = 10,000 m³ ➢ Debris modelled as ALE solid Landslide ➢ Unit Weight = 1900 kg/m³ Source ALE Container Topography

Model Setup – Debris Mass and Basal Friction ➢ Debris mass movement captured at regular 0.5s interval Rheology Material Property Adopted Remarks input parameters Internal friction 15° Arup (2014) angle, ϕ i Basal friction angle, Voellmy 8° GEO TGN 29 ϕ b (Adverse Site Setting Turbulence Parameters) 500 m/s² coefficient

Model Setup – Gravity Initialisation ➢ Debris mass initialised at source location by applying gravity load within first 2 seconds ➢ An artificial rigid tube and lid, acting as the gate, to confine and restrict the debris mass Gate movement during initialization Topography ➢ The gate has very low contact friction with Lid the debris mass to minimize disturbance Debris

Model Setup – Gravity Initialisation Gate ➢ Debris mass initialised at source location by applying gravity load within first 2 seconds ➢ An artificial rigid tube and lid, acting as the Lid gate, to confine and restrict the debris mass movement during initialization Topography ➢ The gate has very low contact friction with the debris mass to minimize disturbance Debris

Model Animation

Model Results

Model Results – Velocity Hydrograph A B C Point A Point B Point C Maximum 9.7 8.2 7.3 Velocity (m/s) Time 40.5 47.5 50.0 Recorded (s)* *Debris are allowed to slide downslope at t=2s

Model Results – Debris Thickness Hydrograph Point A Point B Point C Maximum Thickness 3.0 3.0 3.0 (m) Time Recorded 40.0 72.5 52.0 (s)* *Debris are allowed to slide downslope at t=2s

Model Animation – Debris Thickness at Point A

Model Results – Time History of Debris Frontal Velocity Average Velocity vs. Chainage 18 16 Boulders Dam 14 12 Velocity (m/s) 10 8 6 4 2 0 0 100 200 300 400 500 600 700 800 900 1000 Approximate Chainage (m)

Possible Mitigation Measures ➢ Rigid Barrier ➢ Flexible Barriers ➢ Baffles Debris-flexible barrier interaction using LS-DYNA Example of multiple barriers option design using LS-DYNA

Possible Mitigation Measures Rigid Barrier

Summary ➢ LS-DYNA are found to produce generally reasonable results for the benchmarking case ➢ LS-DYNA using ALE formulation to provide a continuum-based numerical solution can provide realistic motions of landslide debris using Voellmy rheology with conventional parameters ➢ LS-DYNA has the capability to carry out soil-structure interaction to visualise and optimise design mitigation measures

Thank you A.K.C. Cheung, J. Yiu, H.W.K. Lam & E.H.Y. Sze (2018) “ Advanced Numerical Analysis of Landslide Debris Mobility and Barrier Interaction ” HKIE Transaction Theme Issue on Landslides and Debris Flow – Theory and Design, Mitigation, Stabilisation and Monitoring, 2018, Hong Kong Institution of Engineers.