Use of Soil Nails to Upgrade Loose Fill Slopes Professor John Endicott Dr Johnny Cheuk January 4, 2010
Acknowledgements GEO/S&T Ir Ken Ho Ir Anthony Lam Ir Patty Cheng Ir Thomas Hui GEO/LPM1 Ir W.K. Pun Ir Dr Alex Li Ir Carie Lam AECOM Ir Dr S.L. Chiu Ir Dr Axel Ng Mr David Mak Mr King Chan HKIE-GDC Ir Dr Eric Li Comments received from: GEO/LPM1, GEO/LPM2, GEO/LPM3, GEO/MW, GEO/ME, GEO/Island, GEO/GP, Arup, CM Wong & Asso., Fugro, Halcrow, Greg Wong & Asso., James Lau & Asso., Dr Robert Lo and Ir Dr Victor Li, Ir W.K. Pun, Ir Dr H.W. Sun Page 2
Content 1. Background Current design approach 2. Scope Comments from practitioners Identified tasks to address key issues 3. Methodology and Preliminary Findings Task 1: Review of overall design approach Task 2: Numerical study on stabilising mechanisms Task 3: Review of triaxial test results 4. Proposed Modifications to Design Approach 5. Concluding Remarks Page 3
Part 1 - Background
Background HKIE report GEO invited HKIE Geotechnical Division to make recommendations on the use of soil nails in loose fill slopes Final Report entitled Soil Nails in Loose Fill Slopes - a Preliminary Study was published (HKIE, 2003) Other reports and guidelines Design Division Technical Guideline (DDTG) No.10 Use of soil nails to stabilise loose fill slopes under the LPM programme (GEO, 2003) [superseded by Geoguide 7] Detailed design methodology and design charts were developed in an LPM Special Task: Design of soil nails and structural facing/grillage in stabilising loose fill slopes (Black & Veatch, 2003) Geoguide 7 (2008) Section 5.8 [based on HKIE (2003)] Page 5
Current Design Approach Step 1: Determine if upgrading works are required (i.e. c=0, φ'=26 ) Step 2: Determine the required nail forces as shown below, with the following assumptions (i) use steady state strength for fill (prescriptive: c=0.2p') (ii) use triangular surface pressure with a small basal shear resistance to achieve FoS=1.1 Other requirements: (i) Vertical nails at toe (ii) Grillage embedment of 0.5m (iii) Check squeezing out [previous guideline in DDTG No.10: Maximum grillage opening = 50%] Page 6
Part 2 - Scope
Major Comments Comments 1 2 3 4 5 6 Large deformation is required for the fill material to reach the steady state. Can the peak strength with a large safety factor be used for design? The assumed triangular pressure distribution leads to very long nails near the lower portion of the slope. The assumption of the surface pressure being normal leads to steeply inclined nails which are not effective. The fill profile may resemble an inverted triangle due to end-tipping. The behaviour may be different from that of a uniform fill. The 0.5m embedment of the grillage beams may damage the tree roots and cause significant construction difficulties. The use of c ss /p' peak =0.2 is not founded on a strong theoretical basis and are too conservative in many cases. Page 8
Major Comments (cont d) Comments 7 8 9 10 11 12 It is not clear why the use of soil nails is restricted to fill with relative compaction > 75%. Loose fill at low confining stress shows dilative behaviour, the liquefaction potential at this stress level should be explored. It is not clear if site-specific laboratory test results can be used to check possibility of undrained collapse, instead of using c=0, φ'=26. The design method is developed for fill derived from CDG. A review should be carried out to explore design parameters for other loose fill materials. The requirement of covering 50% of the slope surface resulted in very large grillage beams. For fill slopes with a significant fill depth, guidance on the need to assume a steady state for the whole fill stratum would be beneficial. Page 9
Identified Tasks to Address Key Issues Today s focus Task 1: Review of the design concept and methodology Task 2: Numerical study on stabilising mechanisms Task 3: Review of triaxial test data Task 4: Study of size of grillage opening Task 5: Design of soil nails in deep fill Page 10
Part 3 - Methodology and Preliminary Findings
Task 1 - Review of the design concept and methodology Issue: use of steady state strength Page 12
Review of Overall Design Methodology Current Approach Use steady state strength of the fill Factor of Safety = 1.1 Alternative Approach Use peak strength of the fill Factor of Safety >>> 1.1 HKIE (2003) Page 13
Task 2 - Numerical study on stabilising mechanisms (FLAC analyses) Issues: (a) nail orientation (b) facing pressure (c) role of vertical nails (d) behaviour of tapered fill (e) grillage embedment depth requirement Page 14
Numerical Study - Soil nails perpendicular to slope surface Page 15
Numerical Study - Soil nails at 20 to horizontal Grillage structure 3m saturated loose fill 10m 35 In-situ material Case 2: Soil nails at 20 to horizontal Page 16
Numerical Study - Finite Difference Model Saturated loose fill at steady state undrained strength (modelled by Mohr-Coulomb) [c u =0.2p'] 10m In-situ material (modelled by Mohr-Coulomb) [c=5kpa, φ=35 ] Procedure: 1. Establish initial stress conditions using c=5kpa and φ'=35 2. Install soil nails and facing, solve for equilibrium again 3. Reset displacements 4. Change the fill to a constant strength material (i.e. c u =0.2p') 5. Solve for equilibrium 6. Change the nail lengths until FOS=1.1 7. Examine the mechanisms, mobilised nail forces and facing pressure Page 17
Numerical Study - Model Parameters Input parameters: Parameter Loose fill In-situ soil (CDG) Dry density, ρ d (kg/m 3 ) 1800 1800 Bulk modulus, K (MPa) 1x10 7 9.6 Shear modulus, G (MPa) 1.7 21 c parameter (kpa) c u = 0.2p' c' = 5 Friction angle, φ' ( ) 0 35 Note: (1) Interface between nails and soil is assumed to have the same strength of the soil (2) Nail spacing is 1.5m (vertical) x 1.5m (horizontal) Page 18
Numerical Study - Summary of Analyses and Results Case Parameters for loose fill Nail orientation (special geometry) Nail length (m) FoS 1 c=0.2p', φ'=0 Perpendicular 9.5 1.07 2 c=0.2p', φ'=0 20 6.6 1.11 3 c=0.2p', φ'=0 20 (with a 6m vertical nail) 6.6 1.10 4 c=0.2p', φ'=0 20 (tapered fill) 6.6 1.11 5 6 c=0, φ'=26 (working conditions) c=0, φ'=26 (working conditions) Perpendicular 6.6-20 6.6 - Page 19
Numerical Study - Nail Forces and Facing Pressure Case 1 perpendicular nails c=0.2p' L=9.5m FOS=1.07 Total nail force=569kn Max. disp.=0.17m Case 2 20 o c=0.2p' L=6.6m FOS=1.11 Total nail force=279kn Max. disp.=0.23m Page 20
Numerical Study - Soil and Nail Movements Case 1 perpendicular nails c=0.2p' L=9.5m FOS=1.07 Total nail force=569kn Max. disp.=0.17m Case 2 20 o c=0.2p' L=6.6m FOS=1.11 Total nail force=279kn Max. disp.=0.23m Page 21
Numerical Study - Vertical Nails Page 22
Numerical Study - Vertical Nails Case 2 20 o c=0.2p' L=6.6m FOS=1.11 Total nail force=279kn Max. disp.=0.23m Case 3 20 (with vertical nails) c=0.2p' L=6.6m FOS=1.10 Total nail force=278kn Max. disp.=0.23m Page 23
Numerical Study - Tapered Fill Profile Case 4 20 (tapered fill profile) c=0.2p' L=6.6m FOS=1.10 Total nail force=102kn Max. disp.=0.04m Page 24
Numerical Study - Comparison of Facing Pressure * Stresses are compressive in all cases Page 25
Numerical Study - Working Conditions Case 5 perpendicular nails c=0, φ'=26 L=6.6m Total nail force=96kn Max. disp.=26.2mm Case 6 20 c=0, φ'=26 L=6.6m Total nail force=50kn Max. disp.=4.2mm Page 26
Numerical Study - Key Observations 1. Soil nails at 20 to the horizontal provide a more effective stabilising mechanism compared to nails that are perpendicular to the slope surface. This is reflected in the smaller total nail force required and a smaller facing pressure. 2. When nails are at 20, the relative movement between the loose fill and the nails mobilises tensile forces in the nails and directly increases the global stability of the fill slope. 3. When nails are at 20, the facing pressure distribution is closer to trapezoidal than triangular. Page 27
Numerical Study - Key Observations (cont d) 4. When the stabilising mechanism begins to operate, the stress along the grillage-soil interface is compressive. Deep embedment to ensure a good contact is not needed. 5. When nails are at 20, the vertical nails near the slope toe are not needed. 6. For a tapered fill profile, the failure mechanism may not extend to the slope toe. 7. When nails are at 20, the slope movement at working conditions is smaller due to effective mobilisation of nail resistance. Page 28
Task 3 - Review of triaxial test data Issues: (a) onset of undrained collapse (b) assessment of steady state strength (c) behaviour of fill materials at relative compaction < 75% (d) behaviour of fill derived from CDV (e) behaviour of fill materials at low confining stress Page 29
Triaxial Test Data - Overview of Data Sets Origin of material HKIE (2003) Data Set PWCL No. of sites 11 12 No. of tests 122 317 CDG fill 10 5 CDV fill 1 1 Cannot be confirmed 0 6 Remarks: (1) Very limited data on CDV (2) Minimum relative compaction at preparation is 75% (3) Minimum confining pressure is 20kPa Page 30
Triaxial Test Data - Onset of Undrained Collapse (HKIE, 2003) Page 31
Triaxial Test Data - Onset of Undrained Collapse (PWCL) Page 32
Triaxial Test Data - c ss /p' peak ratio (HKIE, 2003) 1.2 ss/p' peak c s 1 0.8 0.6 0.4 0.2 Diamond Hill (HKU) CLCY(HKU) Valley Estate (HKU) Diamond Hill CLCY Valley Estate Ho Man Tin Stubbs Road Lai Ping Road Shau Kei Wan (UST) Beacon Hill (UST) 0 1 10 100 1000 Consolidation pressure (kpa) Page 33
Triaxial Test Data - c ss /p' peak ratio (PWCL) Outliers are extremely shallow (<0.5m) samples Page 34
Triaxial Test Data - Critical State Soil Mechanics Framework e 1 2 3 Critical state line log p specimen 1 specimen 2 specimen 3 q Critical state line 2c 3 ss =q ss 2c ss =q ss 1, 2 p Page 35
Triaxial Test Data - c ss vs e (HKIE, 2003) 1000 100 s (kpa) c ss 10 Diamond Hill (HKU) CLCY(HKU) Valley Estate (HKU) Diamond Hill CLCY Valley Estate Ho Man Tin Stubbs Road Lai Ping Road Shau Kei Wan (UST) Beacon Hill (UST) 1 0 0.2 0.4 0.6 0.8 1 1.2 Voids ratio Page 36
Triaxial Test Data - c ss vs e (PWCL) Page 37
Triaxial Test Data - CSL on p' q plane (HKIE, 2003) s (kpa) qss 500 400 300 200 Diamond Hill (HKU) CLCY(HKU) Valley Estate (HKU) Diamond Hill CLCY Valley Estate Ho Man Tin Stubbs Road Lai Ping Road Shau Kei Wan (UST) Beacon Hill (UST) 100 0 0 50 100 150 200 250 300 p' ss (kpa) Page 38
Triaxial Test Data - CSL on p' q plane (PWCL) Page 39
Triaxial Test Data - CSL on e log p' plane (HKIE, 2003) 1.2 1 Vo oids ratio 0.8 0.6 0.4 0.2 0 Diamond Hill (HKU) CLCY(HKU) Valley Estate (HKU) Diamond Hill CLCY Valley Estate Ho Man Tin Stubbs Road Lai Ping Road Shau Kei Wan (UST) Beacon Hill (UST) 1 10 100 1000 p' ss (kpa) Page 40
Triaxial Test Data - CSL on e log p' plane (PWCL) Page 41
Vo oids ratio Triaxial Test Data - Data from Stubbs Road 1.3 1.2 1.1 1 0.9 0.8 0.7 0.6 0.5 0.4 Preparation (Stubbs Rd) After consolidation Final (Steady State) Different voids ratios, similar strength? 1 10 100 p' (kpa) Page 42
Triaxial Test Data - Data from Anderson Road Vo oids ratio 0.75 0.70 0.65 0.60 0.55 Public Works Central Laboratory Data - Anderson Road e = 0.515 e = 0.530 e = 0.560 e = 0.580 e = 0.600 e = 0.670 e = 0.720 0.50 0.45 0.40 0 50 100 150 200 250 300 P'consolidation (kpa) Page 43
Triaxial Test Data - Key Observations 1. The lower bound mobilised friction angle at the onset of undrained collapse can be taken as 26, which is the same as that suggested in HKIE(2003). 2. The c ss /p' peak ratio of 0.2 is a suitable lower bound design value. However, the data is very scattered. Site specific testing may result in a much higher value. 3. The critical state soil mechanics framework does not apply in local loose fill materials in that the undrained shear resistance is not uniquely related to voids ratio. Page 44
Triaxial Test Data - Key Observations (Cont d) 4. Soil specimens at a relative compaction lower than 75% are very difficult to prepare due to the excessive volume change of the soil specimen upon saturation. 5. Very limited data is available for fill derived from CDV. 6. The behaviour of fill materials at low confining stress cannot be evaluated by conventional triaxial tests. Page 45
Task 4 - Study of size of grillage opening (In progress) Issue: Maximum size of grillage opening to avoid squeezing out of the fill Vertical stress distribution Shear strain 3D finite difference grid Page 46
Task 5 - Design of soil nails in deep fill (In progress) Issues: (1) Design parameters for fill at depths (2) Pull-out resistance of nails in fill materials Non-collapsible fill c=0, φ'=26 Review of pull-out tests done at HKU Page 47
Part 4 - Proposed Modifications to Design Approach
Proposed Design Approach Page 49
SLOPE/W Results Using the Old Approach Page 50
SLOPE/W Results Using the New Approach Page 51
Part 5 - Concluding Remarks
Concluding Remarks 1. The approach of using the steady state undrained shear strength coupled with a low factor of safety is considered suitably robust given the uncertainties of the undrained collapse behaviour of local loose fill materials. 2. Soil nails at an angle closer to horizontal are more effective in increasing the global stability of a loose fill slope compared to nails that are perpendicular to slope surface. 3. When the soil nails are at an angle closer to horizontal, the vertical nails at the toe are not needed. The corresponding pressure distribution is closer to trapezoidal than triangular. Page 53
Concluding Remarks 4. The grillage embedment can be reduced. A nominal value of 200-300mm is recommended. 5. The lower bound mobilised friction angle at the onset of undrained collapse is approximately 26 6. The design strength of the fill may be conservatively taken as 0.2p' in the absence of site specific data. However, site specific testing is strongly encouraged. 7. The nail forces should be modelled as line forces in slope stability calculations in order to take account of the contribution of the stabilising effects. 8. The design nail force distribution needs not be triangular. Page 54
Thank You January 4, 2010
Numerical Study - Tapered Fill Profile Page 56