1 The Use of Soil Nails to Upgrade Loose Fill Slopes Interim Summary Report Background and Objectives The application of soil nails in loose fill slopes has been a subject of considerable debate loc al ly . Bas ed on a pre limi nary st udy, de si gn gui de lin es were put for wa rd by a su b- committee appointed by the Geot echnical Division of the Hong Kong Institution of Engineers (H KIE) in 2003. Fur ther to the above, the GEO issue d the recommende d de sign methodology for upgrading sub-standard loose fill slopes using soil nails in 2003, which was endorsed by the Slope Safety Technical Review Board (SSTRB). Since then, much experience has accrued in respect of the design and construction of soil nails for upgrading loose fill slopes. Publications related to the subject have also been released locally and overseas, including studies on more fundamental issues, for example, how soil na ils interact with the surro unding soil s. It is the refore con sidered timely by the Standards and Testing (S&T) Division of the Geotechnical Engineering Office (GEO) to conduct a review of the current design methodology, with due regard to the experience gained and advances made in the technical u nderstanding of the subject in recent years. The GEO has commi ssioned AECOM to carry out a stu dy on the subjec t. The objec tives ofthis study are to review the current guidelines and refine the design methodology as appropriate. A task group compris ing representatives from the S&T Div ision of the GEO andthe Geotechnical Division of the HKIE has been formed to oversee the study. Scope of Study To fa cil it at e ma ppi ng out th e key dire cti on and focus of the st udy , th e GEO has sol ici tedviews/comments from practitioners/researchers in the local geotechnical profession (i.e. geotechnical consultants, universities and the relevant GEO Divisions) on the following items in relation to the use of soil nails to upgrade loose fill slopes: (a) gr ound in vest ig ation and labor at ory tes ti ng ; (b) de si gn met hodology, analy se s, de ta il in g; (c) construction issue s; (d ) geot echnic al control , a nd (e) fi el d perf orman ce of l oos e fil l sl opes upg ra de d by me ans of so il n ail s. Based on the collected information, the following key issues have been identified and are addressed in this study: No. Issue 1 Design conceptand methodol ogy 2 Th e de si gn pr essure a ct in g o n t he g ril lag e f aci ng structu re 3 Ef fect of nail orien tati on on the stabili si ng mechanisms 4 Be havi our of t he soi l-n ail ed s ys tem f or a ta per ed fi ll pro fi le 5 Re qui re ment of e mbeddin g t he g ril la ge bene ath the sl ope s urf ac e 6 As sessment of s teady state undr ai ne d sh ear stren gt h of lo ose f ill mater ia l 7 Pos si bl e use of so il n ai ls f or fi ll sl ope s with relative compaction less tha n 75% 8 Be havi our of l oose fi ll ma ter ial s at low confi ni ng stresses 9 Pa ra me te rs f or onse t of undr ai ne d co ll apse behavio ur 10 Soil nail design for loose fill derived from completely decomposed volcani cs (C DV) 11 Maximum size of openings for the grillage 12 Bond strength of soil nails in fill material
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
The Use of Soil Nails to Upgrade Loose Fill Slopes
Interim Summary Report
Background and Objectives
The application of soil nails in loose fill slopes has been a subject of considerable debatelocally. Based on a preliminary study, design guidelines were put forward by a sub-
committee appointed by the Geotechnical Division of the Hong Kong Institution of Engineers(HKIE) in 2003. Further to the above, the GEO issued the recommended design
methodology for upgrading sub-standard loose fill slopes using soil nails in 2003, which wasendorsed by the Slope Safety Technical Review Board (SSTRB).
Since then, much experience has accrued in respect of the design and construction of soil
nails for upgrading loose fill slopes. Publications related to the subject have also been
released locally and overseas, including studies on more fundamental issues, for example,
how soil nails interact with the surrounding soils. It is therefore considered timely by the
Standards and Testing (S&T) Division of the Geotechnical Engineering Office (GEO) to
conduct a review of the current design methodology, with due regard to the experience
gained and advances made in the technical understanding of the subject in recent years.
The GEO has commissioned AECOM to carry out a study on the subject. The objectives of this study are to review the current guidelines and refine the design methodology as
appropriate. A task group comprising representatives from the S&T Division of the GEO and the Geotechnical Division of the HKIE has been formed to oversee the study.
Scope of Study
To facilitate mapping out the key direction and focus of the study, the GEO has solicited views/comments from practitioners/researchers in the local geotechnical profession (i.e.
geotechnical consultants, universities and the relevant GEO Divisions) on the following itemsin relation to the use of soil nails to upgrade loose fill slopes:
(a) ground investigation and laboratory testing;
(b) design methodology, analyses, detailing;(c) construction issues;
(d) geotechnical control, and
(e) field performance of loose fill slopes upgraded by means of soil nails.
Based on the collected information, the following key issues have been identified and areaddressed in this study:
No. Issue
1 Design concept and methodology
2 The design pressure acting on the grillage facing structure
3 Effect of nail orientation on the stabilising mechanisms
4 Behaviour of the soil-nailed system for a tapered fill profile5 Requirement of embedding the grillage beneath the slope surface
6 Assessment of steady state undrained shear strength of loose fill material
7 Possible use of soil nails for fill slopes with relative compaction less than 75%
8 Behaviour of loose fill materials at low confining stresses
9 Parameters for onset of undrained collapse behaviour
10 Soil nail design for loose fill derived from completely decomposed volcanics (CDV)
To address the issues listed above, the following tasks are being carried out:
Task 1 – Review of design concept and methodology
In this task, the design concept and methodology put forward by the HKIE in 2003 have beenreviewed. Loose saturated fill is contractive and possesses a structure that could lead to
“strain softening”, with a significant post-peak drop in strength in a drained-undrained failuremode. For the purposes of analysing the ultimate limit state to ensure an adequate safety
margin, the current design approach assumes that the fill would reach the steady stateundrained shear strength by the time the nails are working as required. An alternative
approach would be to use the peak shear strength of the saturated loose fill materials coupled
with a large factor of safety. The large factor of safety is to ensure that the mobilised
deformation in the soil is small enough such that post-peak strain softening would not occur.
Local fill materials derived from CDG however exhibit very brittle behaviour. The suddenreduction in shear resistance may lead to a progressive failure mechanism. In addition, the
strain required to reach the peak strength varies tremendously depending on a number of
factors, including the grading of the fill material and the initial stress conditions. This would require an unrealistically large factor of safety to warrant a safe design. Considering theseuncertainties associated with the use of peak strength, the current approach of using the large
strain steady state undrained shear strength with a relatively low factor of safety represents a
robust and practical design method.
Task 2 – Study on facing pressure and nail orientation
This task aims to address Issues 2, 3, 4 and 5. A series of 2-D numerical analyses have beenconducted using FLAC. The analyses assume a 10m high, 35 o slope with 3 m of fully
saturated loose fill that have collapsed under undrained shearing and reached the steady stateundrained shear strength (Figure 1). The fill is therefore modelled by an elasto-plastic model
with a Mohr-Coulomb failure criterion. The input parameter c corresponds to the design
undrained shear strength of the fill (cu) and the friction angle is taken as zero. Following
HKIE (2003), different values of cu/p' ratio (cu/p'=0.2, cu/p'=0.3, and cu/p'=0.4) are used,
where p' is the mean effective stress. The modelling approach for the behaviour of the fill
material is essentially a total stress analysis. The soil beneath the 3m collapsible loose fill is
modelled as typical CDG (i.e. c' = 5 kPa and ' = 35°). Two sets of nail orientation are
examined: (i) perpendicular to slope surface (Figure 1a) and (ii) 20o to the horizontal (Figure
1b). The nail lengths are varied in order to achieve a factor of safety of 1.1. The mobilised
nail forces and facing pressure distributions are examined. The role of the vertical nail at the
slope toe and the behaviour of a tapered (Figure 2) fill layer are also investigated.
Two additional analyses have been conducted to approximate the behaviour of the nailed
loose fill slope under working conditions for the two selected nail orientations. In these twoanalyses, the loose fill is assumed to be saturated but undrained collapse is not triggered. The
model parameters are taken as c' = 0 kPa and ' = 26° or 30° (2 cases for parametric study).
The preliminary findings are:
(a) Soil nails installed at 20o to horizontal are more effective than those perpendicular to
the slope surface for both working conditions and ultimate conditions. The analyses
show that smaller nail forces are required and a lower pressure is acting on the facing
if the soil nails are installed at 20o to the horizontal for the same factor of safety of
1.1. This is due to the contribution of a larger horizontal nail force component, which
is more effective for the stabilisation of the critical failure mass. Therefore, it is more
appropriate to construct the nails at an angle closer to horizontal, say 10o - 20
o to the
horizontal, following usual practice in cut slopes. It is recommended to model the soil
nails using line forces in slope stability calculations instead of a pressure block actingon the slope surface to take due account of the contribution of the stabilising forces in
the direction of the nails.
(b) When the soil nails are installed at 20o to horizontal, the pressure distribution on the
facing is closer to a trapezoid than a triangle (Figure 3). It is therefore recommended
to design the soil nails such that the nail forces are more evenly distributed.
(c) For tapered fill profile, the failure mechanism does not extend to the slope toe, but
daylights at mid-height of the slope (Figure 4). This suggests that the soil nails at theupper part of the slope may play a more important role than those near the toe in some
geological profiles. Allowing the nail forces to be more evenly distributed could
tackle this type of failure mechanism.
(d) In the numerical analysis with soil nails installed at 20o, the omission of the vertical
nail does not affect the slope stability. The original purpose of constructing a row of
vertical nails is to take up the unbalanced force if the upper soil nails are
perpendicular to the slope surface. This unbalanced force could be due to construction
deviation, which leads to a small driving force in the potential sliding direction of thefailure mass. With the proposed new orientation of the soil nails (i.e. 10
o - 20
o to
horizontal), this component of unbalanced force is taken up by the inclined soil nailsdirectly. Therefore, the vertical soil nails are not necessary.
(e) The shear resistance along the soil-grillage interface has little influence on the
stability mechanism. Therefore, embedding the grillage deeply into the soil surfacemay not be necessary as long as a good contact can be ensured. It is recommended to
provide a minimum embedment of 200 mm into the slope surface.
(f) Under working conditions, the movement in the nailed loose fill slope is much
smaller when the nails are constructed at 20° to the horizontal, compared to nails thatare perpendicular to the slope surface. This suggests that larger ground movement is
induced before soil nail forces can be mobilised in the case of steeply inclined nails.
Task 3 – Review of triaxial test data
This task aims to address Issues 6, 7, 8, 9 and 10. The triaxial test data conducted at Public
Works Central laboratory (PWCL) from 2003 to 2008 have been collected. A review has
been carried out on this data set, together with the data reported in HKIE Report (2003). The
main objectives are to review the recommendations put forward by HKIE (2003) in terms of (i) design shear strength of fully saturated loose fill subjected to structural collapse under
undrained shearing, (ii) minimum requirement on relative compaction for the use of soil nails
in loose fill slopes, (iii) behaviour of loose fill material at very low confining stresses, (iv)
minimum stress ratio (or mobilised friction angle) for the onset of undrained collapse
behaviour, and (v) undrained collapse potential of CDV.
A total of 12 sets of test data have been collected from PWCL, among which 11 sites are of
CDG origin and the remaining one is CDV. The preliminary findings are:
(a) The css/p' peak ratios of most of the data points are above 0.2 with a few exceptions
(Figure 5). The outliers are tests conducted using samples taken at very shallow
depths (<0.5m). The dry densities of these test specimens are extraordinarily low (1.3-1.4 Mg/m3) due to the requirement to replicate the measured in-situ dry density.
Notwithstanding that these test specimens have very similar dry densities as that inthe field conditions, the low stress level (<10 kPa) could not be reproduced in the
triaxial machine. This gives rise to doubts as to whether the observed behaviour inthese tests is representative of the field conditions.
(b) In general, the css/p' peak ratio of 0.2 is applicable to local fill materials derived from
completely decomposed granite (CDG) only, provided that the tests are conducted
under the same conditions, in terms of dry density and stress level, as those in thefield. The ratio of 0.2 represents a conservative lower bound. With the gain in
experience in conducting triaxial tests on remoulded specimens, designers are
strongly encouraged to conduct site-specific laboratory tests, which would provide a better estimate of the undrained steady state strength of the fill.
(c) An attempt has been made to define the undrained shear strength based on its voids
ratio (or dry density) as an alternative method to determine the design shear strength.
According to critical state soil mechanics theory, undrained shear strength is a unique
function of the voids ratio of the soil for a given set of soil grading and mineralogy(Figure 6). As illustrated using the data from Stubbs Road (Figure 7), the steady state
line for local fill materials is not unique on the e-log p' plane for materials obtained from the same origin, despite that they show a unique relationship on the p'-q plane.
This is different from the critical state soil mechanics theory. It also suggests that thesteady state undrained shear strength of the local loose fill is not uniquely related to
voids ratio or dry density (Figure 8).
(d) Based on the observations from experienced laboratory personnel, soil specimens at a
relative compaction lower than 75% are very difficult to prepare due to the excessivevolume change of the soil specimen upon saturation. It is therefore unlikely that the
results from conventional routine testing on these extremely loose specimens would be reliable and representative. It is therefore recommended that for upgrading sub-
standard fill slope with soil nails, the minimum requirement of 75% relativecompaction of the loose fill should be maintained.
Due to insufficient information for CDV, test data from commercial laboratories are being
collected at the moment. The extended database will be used to address the remaining issuesfor this task.
Task 4 – Study of maximum size of grillage openings
This task aims to address Issue 11. A series of 3-D numerical analyses have been conducted
using FLAC3D. A single unit of the grillage is modelled in the analysis (Figure 9). The stress
conditions (i.e. the soil pressure exerted on the facing) obtained from 2-D FLAC analyses
(Task 2) are modelled as applied pressure boundaries. The analyses are conducted to find out
the maximum size of grillage openings beyond which the fill subjected to structural collapse
would “squeeze” out. The investigation covers different undrained shear strengths and
loading conditions. This task is still in progress; the results will be reported at a later stage of
the study.
Task 5 – Review of soil nail design in deep fill
This task aims to address Issue 12. The triaxial test data conducted at PWCL from 2003 to
2008 and those collected from commercial laboratories are being reviewed. The reviewfocuses on the behaviour of fill materials at high confining pressures, in particular whether
the collapse behaviour changes with consolidation pressure. In addition, the pull-out testsconducted by the University of Hong Kong will be reviewed to investigate if the loose soil
structure would lead to soil arching casing a lower-than-expected soil/nail resistance.