Liquefaction Mitigation using GeoComposite Vertical Drains Kyle Rollins and Joshua K.S. Anderson Brigham Young University Civil & Environmental Engineering Dept Provo, Utah, USA
Acknowledgements NCHRP-IDEAS Program Nilex, Inc. ConeTec, Inc.
Liquefaction Mitigation by Densification Versus… Vibro-Compaction Stone Columns Dynamic Compaction Compaction Grouting
Liquefaction Mitigation by Drainage Liquefiable Sand Drains
EQ Drain with and without Filter Fabric
Drain Installation EQ Drain Hollow Steel Mandrel Anchor Plate
Potential Problems with Conventional Densification Expensive and time consuming effort Cost increases and success decreases as fines content increases. Cost of improvement increases as initial blow count increases. Improved density may be overestimated by conventional penetration correlations.
Potential Advantages of Earthquake Drains Reduced cost of installation Shorter installation time Greater flow capacity than stone columns Densification during drain installation May provide mitigation for silty sands that are difficult to densify
PROJECT OBJECTIVES Evaluate ability of drains to dissipate excess pore pressures. Evaluate ability of drains to reduce liquefaction-induced settlement. Provide case histories to validate/calibrate computer models.
Test Site Locations Vancouver, B.C. Treasure Island, CA
Treasure Island Test Site Downtown San Francisco Test Site
Installation Induced Settlement
Blast-Induced Settlement in Untreated Area
Blast-Induced Settlement in Treated Area Cluster 1 Cluster 2 (Wick Drains) Cluster 4 Cluster 5 Cluster 6 Cluster 7 Cluster 3 Cluster 8 Settlement Stakes Blast Holes
Blast-Induced Settlement Cluster 1 Cluster 2 (Wick Drains) Cluster 4 Cluster 5 Cluster 6 Cluster 7 Cluster 3 Cluster 8 Settlement Stakes Blast Holes 20 40 60 80 100
Pore Pressure Response
Vancouver BC Test Site Vancouver CANLEX Test Site EQ Drain Test Site MasseyTunnel
Typical CPT Profile
Layout for EQ Drain Test Areas 1.22 m 4 Blast Holes at 5 m radius
Pipe Mandrel with Minimum Densification
Finned-Mandrel for Maximum Densification
Avg. Installation-Induced Settlement
Video of EQ Drain Test Click on picture to see second earthquake drain test in vancouver, BC
Pore Pressure Response -High Vibration
Blast Induced Settlement for EQ Drains Relative to Untreated Test Area
Comparison of qc with Time Low Vibration High Vibration
ANALYSIS OF TEST RESULTS Calibrate model with measured response from blast event. Compute expected response from earthquake event.
Input Parameters for FEQDrain Analysis Soil Layering Hydraulic Conductivity, K Modulus of Compressibility, Mv Drain Properties Nq/NL, Stress Cycle Ratio Td, Earthquake Duration
Variation of Compressibility (Mv) with Ru Keep Ru in this Range (Seed et al,1976)
Reality Check on Input Parameters
Measured and Computed Pore Pressure (Treasure Island)
Measured and Computed Pore Pressure (Vancouver)
Measured and Computed Settlement (Vancouver)
Drain Performance for Various Earthquake Events and Drain Spacings Magnitude Duration (sec) Nq/Nl Drain Spacing (m) Maximum Ru Settlement (mm) Blast (Vancouver) 8 4.0 1.22 1.0 310 6.0 2.0 0.91 0.40 31 6.75 17 0.47 35 3.0 0.61 48 7.5 0.65 53
Conclusions Relative to Drains Significant densification provided. Rate of dissipation increased. Settlement can be reduced for low Ru. Drain layout must be designed for anticipated earthquake.