Multiscale modeling of natural hazards

MPM and LBM-DEM


Krishna Kumar, krishnak@utexas.edu
Rei Hosseini, Qiuyu (Amber) Wang, Chihun Sung and Thiago Araujo
University of Texas at Austin



2nd INLGEO

Tongji University, 17th Sep 2021

Scales of modeling

Oso landslide (2014)

8 million cubic meters of glacial deposits and water-filled debris material transported to a distance of 1 km (Haugerud., 2014).

MPM Towards exascale simulations

What is Ray Tracing?

In-situ visualization

MPM Oso landslide rendering

In-situ visualization of Oso with MPM and Galaxy

Microscale modeling of submarine flows


Loose v dense: Runout distance

Loose
Dense

Loose v dense: Initiation phase

Loose
Dense

Pore-pressure distribution along the failure plane during initiation.

Loose v dense: Runout phase

Attack angle ('a' of 0.8) $t = 3 \tau_c $

Loose v dense: Runout phase

Loose
Dense
Water entrainment front (~15d length) at a slope of 5*

Effect of volume: 1k v. 10k particles


LBM Nodes = 50 Million : DEM grains = 10000 discs

Effect of volume on runout: Inhibiting factors

Negative pore pressures
Hydrodynamic forces

Effect of volume on runout: Enhancing factors

Effective stress
Volume v. Runout

Faiure of unsaturated slopes

Hirotoshi Mori - Public Works Research Institute (PWRI), Japan

Multiphase LBM: Effect of varying the contact angle

Fully hydrophilic surface
(θ = 0°)
Fully hydrophobic surface
(θ = 180°)
Neutral surface
(θ = 90°)

Multiphase LBM: Hysteresis in Hamburg sand

Multiphase LBM: Capillary structures

Neutral surface
(θ = 90°)

Multiphase LBM: Hysteresis

Multiphase LBM: Origin of Hysteresis








Krishna Kumar

krishnak@utexas.edu







View the Geoelements website for more information and software tools