Beruflich Dokumente
Kultur Dokumente
net/publication/327384560
CITATIONS READS
0 68
4 authors, including:
Andy Fourie
University of Western Australia
185 PUBLICATIONS 1,985 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Shah Neyamat Ullah on 02 September 2018.
Abstract 3D finite element simulations are conducted to better understand the soil-structure interaction
of rectangular foundations resting on a sand embankment over soft mine tailings. To account for large
strains, an updated Lagrangian formulation is used in which the stiffness matrix is updated with the
non-linear change of the soil geometry. The sand is modelled using a modified Mohr-Coulomb (MMC)
soil model where the friction and dilation angles are varied with deviatoric plastic strains. It is shown
that failure over a moderate slope angle of 22 and normalised sand embankment height (H/B) of 0.26
(H is the embankment height and B is the footing width) involves a deep seated circular slip plane,
involving the sand and clay layers, where the foundation was close to the crest of the embankment. The
failure pattern turns gradually to typical punch-through with an inclined sand shear plane as the
distance of the footing edge exceeds 1.13B from the crest. The results of this study agree well with those
from model centrifuge experiments, and can be utilised in developing better analytical models and
equipment capable of live monitoring of slope deformation.
Keywords: Foundations; soil-structure interaction; slope stability, finite element; mine tailings; punch-
through, sand over clay
1 INTRODUCTION
Mining activity generates a large volume of wastes known as mine tailings, which is a
combination of fine-grained solid material (remaining after the recoverable metals and minerals
have been extracted from the crushed and ground mine ore), processing chemical and process
water remaining (MRG, 2009). As tailings can be relatively soft, making it difficult for
earthmoving equipment to operate, a sand cover is usually placed progressively over the
backfilled tailings. This placed sand layer forms a slope at the verge and earthmoving
equipment is required to travel close to the crest of the slope for placing the next layer of sand
(Hossain et al. 2015). This causes instability of the structure due to large deformation of the
soft tailings and possible punch-through and rotational failure of the structure (see Reinke et al.
2010; Fitton et al. 2010; Hossain et al. 2015).
While a great number of research exist for performance of foundations near a slope (Selvadurai
and Gnanendram 1989, Narita and Yamaguchi 1990; Huang et al. 1994, Castelli and Lentini
2012, Chakraborty and Kumar 2013), most deal with single layer sand or clay slopes with strip
footings. Hossain and Fourie (2013) report strip footing failure mechanisms using the
centrifuge-PIV (particle image velocimetry) technique on a sand slope resting on mine tailings.
The mechanism indicated significant rotation of the footing placed near the crest with inclined
shear planes within the sand layer.
ULLAH et al.
Hossain et al. (2015) report centrifuge experimental results for a skirted and non-skirted
rectangular foundation (idealised earthmoving equipment) on a sand embankment over soft
mine tailings. Vertical load-penetration responses for a range of tailing strength, sand height
and setback distance (distance of the footing edge from the crest) was explored. No
corresponding numerical studies have emerged to improve understanding on the failure
mechanics, predict the load-penetration response and slope displacements. Accurate peak load
dictates the type of equipment that can be used in a tailing storage facility (TSF) (such as light
or heavy weight equipment), on the other hand, magnitude of slope movements and field slope
monitoring may contribute to developing an early warning system.
Over the years a number of experimental, numerical and analytical studies have emerged for
foundation on layered soil (see works of Kenny and Andrawes, 1997; Teh et al. 2010, Lee et
al. 2013, Hu et al. 2015, Valore et al. 2017, Ullah and Hu, 2017; Ullah et al. 2017).
However, most of them use either strip, circular or conical foundation with no sand slope (i.e.
β = 0) and therefore the solutions obtained cannot be applied directly to the current problem.
Furthermore, the numerical simulations reported in the literature mostly do not account for
rotation as the foundation penetrates from the top sand layer into the bottom clay.
2
ULLAH et al.
2.75 m) are chosen to represent typical earthmoving equipment and is identical to Hossain et
al. (2015) for direct comparison between centrifuge and numerical results.
Both structured and unstructured mesh are considered to comprehensively assess the resistance
and soil failure mechanism (Figure 2). Following recommendations of Ullah et al. (2016), a
minimum boundary distance of 2B is maintained as shown in Figure 1 along with a distance of
2.36B from the toe in order to avoid boundary effects. After establishing the initial geostatic
stresses, a vertical displacement in the Z direction is applied at the foundation reference point
initiating the penetration process. The foundation is modelled as smooth as is the model for
centrifuge testing (Hossain et al. 2015). However, the effect of roughness is also briefly
discussed to highlight the influence of the potential roughness of the earthmoving equipment.
(a) (b)
Figure 3 a) Schematic of the friction model in MMC and b) Effect of the modified Mohr-Coulomb
model on dense sand over clay (b = 0.19B, H = 0.7 m, B = 2.65 m, peak φ = 43.5o, ψ = 15.6o)
3
ULLAH et al.
modulus ratio of E/su = 350. Poisson’s ratio for clay is 0.48. The sand and clay effective unit
weights were taken as 10 and 6 kN/m3 respectively. The effect of sand model is highlighted in
Figure 3. The penetration resistance profiles using the MMC (the friction and dilation angles
varies according to the calculated deviatoric strain) and MC (the friction and dilation angles are
constant) models, but with constant other parameters (b = 0.19B, H = 0.7 m, B = 2.65 m, peak
φ = 43.5o, ψ = 15.6o), are shown in Figure 3b. The MMC model measures a lower resistance
with the difference being ~ 13%. For this low normalised sand height of H/B = 0.264, the clay
strength dictates the mobilised resistance. The difference in resistance between the MMC and
MC model is expected to be greater for higher sand height where the sand layer shares the
greater portion of the load.
The load penetration response obtained from the numerical simulations is compared with the
centrifuge test results of Hossain et al. (2015) in Figure 4. The sand height is fixed at 0.7 m
(H/B = 0.264). The setback distance b is varied as 0.19B, 0.57B and 1.13B. The peak friction
angle for the sand layer was estimated according to Bolton’s (1986) strength-dilatancy
relationships by assuming the dilatancy contributing parameter m = 3 and replacing the mean
stress with the experimental peak vertical stress (Ullah et al. 2017). These relationships are
derived from tri-axial compression tests on sand and are not strictly applicable for the current
complex loading conditions, where deformation is asymmetric, and due to rotation of the
foundation some elements may experience a greater lateral stress than vertical stress. To
account for different unit weights of sand and clay and OCR (over consolidation ratio) in the
experiments, the clay undrained strength was estimated using Ladd et al.’s (1977) expression
by back fitting the T-bar strength profile of Hossain et al. (2015). The back fitted a and n
parameters were found to be a = 0.14 and n = 0.7 (see Figure 1 for the equation, σ'vo = initial
vertical effective stress). The parameter a is lower than that suggested by Gourvenec et al.
(2009) (a = 0.185) but is close to the lower bound suggested for UWA kaolin clay (Ullah, 2017).
The undrained shear strength su at the sand-clay interface was 3.06 kPa, nonlinearly increasing
to 7.53 kPa at the bottom of the clay layer.
4
ULLAH et al.
distance b is increased to 0.57B, the peak capacity is exactly predicted by the unstructured mesh
and only slightly (~12%) overpredicted by the structured mesh. For setback distance 1.13B,
both structured and unstructured mesh predict the resistance well for shallow penetration (d/B
< 0.07), but underpredict for deeper penetrations, with the structured mesh being close to the
measured resistance. Out of interest, an additional analysis is carried out preventing the
foundation rotation, leading to only a slight increase of resistance (Figure 4c). Effect of
foundation roughness is discussed later.
Figure 5 Soil flow mechanism for b = 0.19 B: a) vectorial displacements, b) plastic strain magnitudes
5
ULLAH et al.
(a) (b)
Figure 7 Response of rough and smooth foundations: a) soil reaction moment and b) foundation
rotation.
Figure 7a shows that a major portion of reaction moment is generated at only a small rotation
of 2-3 degrees for both rough (roughness coefficient of 0.4) and smooth (roughness coefficient
of 0) foundation. Any further rotation occurs under a near perfect plastic state of soil without
encountering additional rotational resistance. Figure 7b shows that the rate of rotation is high
during the early stage of foundation penetration. A relatively rough foundation requires a
greater penetration depth to achieve the same rotation angle of the smooth foundation. The
corresponding vertical load penetration response for b = 0.19B is included in Figure 4a.
Remote field monitoring can be a useful way of preventing equipment failure in tailings storage
facilities. Radar based ground slope monitoring is well established in the mining industry (see
Harries and Roberts, 2007). Slope toe and crest movement is maximised when the setback
distance is minimum. For b = 0.19B, the slope toe and crest movement is plotted against
foundation penetration depth in Figure 8. It is seen that immediate movement of the toe is
observed in the negative X direction i.e. movement away from foundation (square markers in
6
ULLAH et al.
Figure 8), whereas a penetration of 2% of the diameter B is required to record any vertical
movement of the toe. The crest movement is seen to be downward and away from the
foundation throughout the penetration depth studied.
Figure 9 Mobilised sand friction angle: a) initial touchdown and b) at a penetration depth d/B =
0.14
The comparison of vertical penetration resistance between a rough and smooth foundation
suggest that foundation roughness increases the penetration resistance by about 7% (Figure 4a).
This contrasts the results for a circular foundation in sand over clay deposit, suggesting
negligible influence of foundation roughness (Shiau et al. 2003; Ullah and Hu, 2017). Note,
due to convergence problems a fully rough rectangular foundation (infinite friction coefficient)
could not be modelled in this study.
A range of friction angle is mobilised within the soil according to the plastic deviatoric strains.
Figure 9a shows the friction angles at a low penetration depth after touchdown for b = 0.19B
(see Figure 4 for load-penetration response). The peak friction angle can be calculated as 39.48o
and the critical state value is 31o. At the last recorded penetration depth the sand reaches critical
state along the full length of the shear plane with greater deviatoric strains recorded at the sand-
clay interface with the sand friction angle reaching the critical state value of 31o (Figure 9b).
7 CONCLUSION
This paper presents three-dimensional finite element simulation for a rectangular foundation
on a sand embankment over soft clay tailings. The results show that choice of a structured or
unstructured mesh affects the peak resistance as well as failure mechanism in comparison to
centrifuge test results. The structured mesh computes a greater resistance and foundation rotates
towards the slope as in classical slope stability methods. The opposite is true for an unstructured
mesh. The bearing resistance decreases as the foundation is placed closer to the crest of the
slope. The failure mechanism involves inclined shear planes within the sand layer and a deep
circular slip surface within the soft tailings, extending near the toe of the slope. A modified
Mohr-Coulomb sand model is implemented and shown to improve the modelling the effects of
sand density. A comparison with the standard Mohr-Coulomb model suggests a 13% difference
in resistance for a dense sand (peak friction angle of 43.5o) of normalised height of 0.264B.
Field monitoring of toe or crest slope can be used as an indicator to failure of earth moving
equipment in tailings storage facilities (TSF).
7
ULLAH et al.
ACKNOWLEDGEMENTS
The research presented here was undertaken with support from CQUniversity's eResearch
facilities. The work was performed using the High Performance Computing facility
(www.cqu.edu.au/hpc). This support is gratefully acknowledged.
REFERENCES
ABAQUS (2017). Abaqus analysis user's manual. Version 2017: Dassault Systemes Simulia Corp: 2017.
Bolton, M. D. (1986). The strength and dilitancy of sands. Géotechnique, 36(1), 65-78.
Bishop, A.W., (1955), “The Use of the Slip Circle in the Stability Analysis of Slopes”, Geotechnique, Vol. 5, pp
7 - 17.
Castelli, F., & Lentini, V. (2012). Evaluation of the bearing capacity of footings on slopes. International Journal
of Physical Modelling in Geotechnics, 12(3), 112-118.
Chakraborty, D., & Kumar, J. (2013). Bearing capacity of foundations on slopes. Geomechanics and
Geoengineering, 8(4), 274-285.
Fitton, T. G., Seddon, K. D., & Alexander, M. G. (2010). Geosynthetic capping of a large tailings storage facility.
Paper presented at the Proceedings of the First International Seminar on the Reduction of Risk in the
Management of Tailings and Mine Waste, Perth.
Harries, N. J., & Roberts, H. (2007, May). The use of slope stability radar (SSR) in managing slope instability
hazards. In Rock Mechanics: Meeting Society’s Challenges and Demands, Proceedings of the 1st
Canada–US Rock Mechanics Symposium, Vancouver. Taylor & Francis, London(pp. 53-60).
Hossain, M. S., & Fourie, A. (2013). Stability of a strip foundation on a sand embankment over mine tailings.
Géotechnique, 63(8), 641-650.
Hossain, M. S., Fourie, A., & Yi-Wen, B. P. (2015). Rectangular Foundations on a Sand Embankment over Mine
Tailings. International Journal of Geomechanics, 15(3), 04014054.
Hu, P., Wang, D., Stanier, S. A., & Cassidy, M. J. (2015). Assessing the punch-through hazard of a spudcan on
sand overlying clay. Géotechnique, 65(11), 883-896. doi:10.1680/jgeot.14.P.097
Huang, C.-C., Tatsuoka, F., & Sato, Y. (1994). Failure mechanisms of reinforced sand slopes loaded with a footing.
Soils and foundations, 34(2), 27-40.
Kenny, M. J., & Andrawes, K. Z. (1997). The bearing capacity of footings on a sand layer overlying soft clay.
Géotechnique, 47(2), 339-345.
Ladd, C. C., Foot, R., Ishihara, K., Poulos, H. G., & Schlosser, F. (1977). Stress-deformation and strength
characteristics. Paper presented at the Proc. 9th International Conference on Soil Mechanics and
Foundation Engineering (ICSMFE).
Lee, K. K., Cassidy, M. J., & Randolph, M. F. (2013). Bearing capacity on sand overlying clay soils: experimental
and finite-element investigation of potential punch-through failure. Géotechnique, 63(15), 1271-1284.
MRG. (2009). MG5 Guidelines for miners: tailings and tailings storage facilities in South Australia, version 1.4,
September, Minerals regulatory guidelines, Dept. of Primary Industries and Resources, South Australia,
Adelaide , SA.
Narita, K., & Yamaguchi, H. (1990). Bearing capacity analysis of foudations on slopes by use of log-spiral sliding
surfaces. Soils and foundations, 30(3), 144-152.
Reinke, G., Soliman, A., & Purkis, A. (2010). Rockfill embankment construction on tailings—Realistic, cost-
effective and reliable application at Xstrata, Mount Isa, Australia. Paper presented at the Proc., 1st Int.
Seminar on the Reduction of Risk in the Management of Tailings and Mine Waste, A. Fourie and R.
Jewell, eds., Australian Centre for Geomechanics, Crawley, WA, Australia.
Selvadurai, A. P. S., & Gnanendran, C. T. (1989). An experimental study of a footing located on a sloped fill:
influence of a soil reinforcement layer. Canadian Geotechnical Journal, 26(3), 467-473.
Shiau, J. S., Lyamin, A. V., & Sloan, S. W. (2003). Bearing capacity of a sand layer on clay by finite element limit
analysis. Canadian Geotechnical Journal, 40(5), 900-915.
Teh, K. L., Leung, C. F., Chow, Y. K., & Cassidy, M. J. (2010). Centrifuge model study of spudcan penetration
in sand overlying clay. Géotechnique, 60(11), 825-842. doi:10.1680/geot.8.P.077
Ullah, S. N. (2017). Jackup Foundation Punch Through in Clay with Interbedded Sand. PhD thesis, University of
Western Australia,
Ullah, S. N., & Hu, Y. (2017). Peak punch-through capacity of spudcan in sand with interbedded clay: numerical
and analytical modelling. Canadian Geotechnical Journal, 54(8), 1071-1088.
Ullah, S. N., Hu, Y., Stanier, S., & White, D. (2016). Lateral boundary effects in centrifuge foundation tests.
International Journal of Physical Modelling in Geotechnics, 17(3), 144-160.
Ullah, S. N., Stanier, S., Hu, Y., & White, D. (2017). Foundation punch-through in clay with sand: analytical
modelling. Géotechnique, 67(8), 672-690.