1693

Proceedings of the XVI ECSMGEGeotechnical Engineering for Infrastructure and DevelopmentISBN 978-0-7277-6067-8

© The authors and ICE Publishing: All rights reserved, 2015doi:10.1680/ecsmge.60678

showed that the FoS and ’ have a very limited influ-ence on h. The decisive parameter for the depth h is c´, with increasing effective cohesion the tension cut-off zone increases significantly. The influence of on h is again rather small.

Figure 8. Slip surfaces obtained with LEA and SRFEA (c´=40kPa, ’=15°)

CONCLUSION

In geotechnical engineering practice, limit equilibri-um analyses are widely used to calculate factors of safety for geostructures despite their obvious short-comings. An alternative is the so called strength re-duction technique performed employing standard displacement based finite element codes. The paper shows that in some cases the two procedures yield relatively high deviations in safety factors. It is shown that the reason for this is the occurrence of tensile stresses during strength reduction. It is further shown that the cohesion is the governing parameter for the size of the tension zone. This yields in combi-nation with steep slope inclination to significant dif-ferences in FoS. However, if a tension crack is intro-duced into the LEA the FoS compares well with the finite element solution. On the other hand, if tension is allowed to develop in the soil in the SRFEA the same FoS as for LEA (ignoring the tension crack) is obtained. Finally finite element limit analyses, which provide rigorous upper and lower bounds on the FoS, were performed to give reference solutions for com-parisons with the results obtained from SRFEA and

LEA. These calculations confirmed the importance of cutting off the possible development of tensile stress-es during the strength reduction procedure. Although not discussed in detail in this paper the studies per-formed revealed that for high friction angles and steep slopes with low FoS (' > 35°) the flow rule (associated or non-associated) may also have a sig-nificant influence on the calculated factor of safety which is in contradiction to the common perception that in slope stability analysis the flow rule does not matter too much. Furthermore it seems that in these cases LEA do not give conservative results because they are more close to SRFEA factors of safety em-ploying an associated flow rule which are of course higher than for non-associated flow.

REFERENCE

Bishop, A.W. 1955. The use of slip circles in the stability analysis of earth slopes. Geotechnique 5 (1), 7-17. Brinkgreve, R.B.J., Bakker, H.L. 1991. Non-linear finite element analysis of safety factors. Proc. Int. Conf. Comp. Meth. Adv.Geomech., 1117-1122. Balkema, Rotterdam. Brinkgreve, R.B.J., Swolfs, W.M., Engin E. 2011. PLAXIS 2D 2011-User Manual. Plaxis bv, Delft, The Netherlands. Chen W.F. 2008. Limit Analysis and Soil Plasticity. Fort Lauder-dale: J.Ross Publishing. Cheng, Y.M., Lansivaara, T. & Wei, W.B. (2007). Two dimen-sional slope stability analysis by limit equilibrium and strength re-duction methods. Computers and Geotechnics 34, 137–150. Drucker, D.C., Greenberg, W. & Prager, W. 1951. The safety fac-tor of an elastic plastic body in plane strain. Trans. ASME J. Appl. Mech. 73, 371–378. Drucker, D.C., Prager, W. & Greenberg, H. J. 1952. Extended lim-it design theorems for continuous media. Q. J. Appl. Math. 9 (4) 381–389. Janbu, N. 1954. Application of composite slip surface for stability analysis. Proceedings of the European conference on stability of earth slopes, Stockholm, Vol. 3, 43-49. Krabbenhoft, K., Lyamin, A.V., Hjiaj, M. & Sloan, S.W. 2005. A new discontinuous upper bound limit analysis formulation. Int. J. Numer. Methods Engng 63, (7), 1069–1088. Morgenstern, N.R. & Price, V.E. 1965. The Analysis of the Stabil-ity of General Slip Surfaces. Geotechnique 15, 79-93. Sloan, S.W. 1988. Lower bound limit analysis using finite ele-ments and linear programming. Int. J. Numer. Anal. Methods Ge-omech. 12 (1), 61–77. Sloan, S.W. 1989. Upper bound limit analysis using finite ele-ments and linear programming. Int. J. Numer. Anal. Methods Ge-omech. 13 (3), 263–282. Sloan S.W. 2013. Geotechnical stability analysis. Geotechnique 63 (7), 531–572. 51st Rankine Lecture. Sloan, S.W. & Randolph, M.F. 1982. Numerical prediction of col-lapse loads using finite element methods. International Journal for Numerical and Analytical Methods in Geomechanics, 6 (1), 47-76.

Effects of roots and mycorrhizal fungi on the stability of slopes

Effets des racines et des champignons mycorhiziens sur la stabilité des pentes

A. Yildiz*1, A. Askarinejad2, F. Graf1, C. Rickli3 and S.M. Springman4 1 WSL Institute for Snow and Avalanche Research SLF, Davos Dorf, Switzerland

2 Delft University of Technology, Delft, The Netherlands 3 Swiss Federal Institute for Forest, Snow and Landscape Research WSL, Birmensdorf, Switzerland

4 ETH Zurich, Zurich, Switzerland

* Corresponding Author

ABSTRACT Heavy rainfall periods initiate not only floods and debris flows, but may also trigger shallow landslides on both scree and vegetated slopes. This has had serious consequences in recent years in Switzerland, causing considerable damage to infrastructure, ecosys-tem, goods and services, even loss of lives. Although vegetation provides significant improvement in stabilizing steep slopes, conventional slope stability analyses generally neglect this entirely. Moreover, mycorrhizal fungi associated with plants contribute to stability, too, by promoting plant, and particularly root growth, and by supporting soil aggregate formation, which results in a further increase in apparent cohesion. It is intended to quantify the effect of biological interventions on the stability of soil and slopes in this study. Specimens consist-ing of combinations of a tree (Alnus incana), legume (Trifolium pratense), and grass (Poa pratensis) were prepared to investigate the root reinforcement effect. Direct shear tests were conducted in an inclinable large-scale direct shear apparatus on specimens with and without roots, following a six month growth period. The results of some preliminary direct shear tests are presented.

RÉSUMÉ Les périodes de fortes précipitations ne provoquent non seulement des crues et des laves torrentielles, mais peuvent aussi déclen-cher des glissements de terrain superficiels sur des éboulis ou des versants couverts de végétation. Ceci a récemment entraîné de lourds dé-gâts en Suisse, allant de dommages à l'infrastructure et aux biens et services écosystémiques jusqu'à la perte de vies. Bien que la végétation améliore la stabilité de pentes raides de façon significative, ceci est généralement négligé dans les analyses de stabilité conventionnelles. Les champignons mycorhiziens contribuent également à la stabilité en favorisant la croissance de plantes, plus particulièrement de racines, et la formation d'agrégats du sol, ce qui mène à une augmentation de la cohésion apparente. Le but de cette étude est de quantifier l'effet d'interventions biologiques sur la stabilité de sols et de pentes. Des spécimens de sol ont été plantés d'une combinaison d'arbres (Alnus in-cana), de légumineuses (Trifolium pratense) et d'herbacées (Poa pratensis) pour examiner l'effet de renforcement des racines. Des essais de cisaillement ont été conduits sur des spécimens de sol plantés et non plantés sur un appareil de cisaillement direct inclinable ä grande échelle, suivant une période de croissance de six mois. Les résultats de quelques essais de cisaillement direct préliminaires sont présentés.

1 INTRODUCTION

Effects of vegetation on the stability of slopes have been recognized and studied extensively over recent decades. Vegetation, in general, has both positive and adverse effects on the hydrological and mechanical aspects of slopes with respect to triggering mecha-nisms of shallow landslides. Increased interception and evapotranspiration are among the most promi-nent hydrological effects of vegetation, whereas root reinforcement has been considered as the main me-

chanical contribution of vegetation to the stability. Roots help to stabilize a slope by acting as fibres that are transferring the shear stresses developed in the soil matrix into tensile resistance via the interface friction along their surface (Gray & Barker 2004). Root reinforcement, as the main stabilizing agent, has been investigated based on different approaches. These include force-equilibrium models of root-soil interaction (Wu et al. 1979), laboratory or in-situ shear tests with root-permeated soil or soils with fi-bre inclusions (Askarinejad & Springman 2014; Gray

Geotechnical Engineering for Infrastructure and Development

1694

& Ohashi 1983), and further investigations based on the Fibre Bundle Model (Pollen et al. 2004).

Basic analytical models assume that roots act per-pendicular to the slip surface, and that the tensile strength of the roots is fully mobilised. However, it can easily be proven that the assumption of roots crossing the slip surface only in a perpendicular di-rection is unrealistic in nature. Furthermore, not only will roots pass through the shear surface but roots ex-tending laterally will also contribute to the strength (Schwarz et al. 2010). Pollen et al. (2004) demon-strated differences in the uptake of strain in roots and soil, showing that the roots are not necessarily stressed to their maximum tensile strength when the soil reaches peak shear stress. Therefore, it is essen-tial to quantify the contribution of roots to the shear strength of soil, when determining the shear strength of the soil matrix with root inclusions. For this rea-son, a robust Inclinable Large-scale Direct Shear Ap-paratus (ILDSA) was built for testing root-permeated soils. The design and mechanical aspects of the appa-ratus are presented herein, as well as results from some preliminary tests.

2 BACKGROUND Direct shear tests have been used in the practice and research of geotechnical engineering since the para-digm shift attributed to contributions of Hvorslev (1937). This is largely due to the simplicity of the test and ease in the procedure. There has been a wide va-riety of testing equipment in use throughout the life-span of the direct shear test. The most common and commercially available type of apparatus accommo-dates 60x60 mm square samples. Size limitations on particle size to key equipment dimensions, as de-scribed by ASTM D 3080, led to development of larger sized apparatuses for testing soil (Bareither et al. 2008; Springman et al. 2003), and also in tests on marginal materials, such as fly ash pellets (Baykal & Döven 2000) and municipal solid waste (Zekkos et al. 2010).

Either laboratory or in situ large-scale direct shear testing is widely employed to determine the shearing behaviour of root-permeated soil. Operstein & Fryd-man (2000) conducted direct shear tests on 200 mm diameter samples of chalky soil reinforced with Pis-tacia lentiscus and Myoporum parvifolium and con-cluded that the increase in shear strength due to roots

was 0.25Tr, where Tr represents the tensile strength of the roots. This is significantly less than the sugges-tions of Wu et al. (1979), in which the increase in shear strength is defined as 1.2Tr with the assumption of full mobilisation of tensile strength of the roots. Goldsmith (2006) performed direct shear tests on samples obtained from the field with a dimension of 250x250x250 mm. The experiments were conducted under saturated conditions and a nominal normal load was applied with a 2 kg metal plate. The shear stress values for specimens with and without roots at a pre-defined value of shear displacement were compared, and it was found out that specimens planted with Sa-lix nigra and Panicum virgatum yielded shear stress values of 36.0 kPa and 37.8 kPa at a shear displace-ment of 7 cm, respectively, whereas the shear stress of specimens without roots was 6.6 kPa at the same shear displacement. 3 TESTING PROGRAMME 3.1 Soil All specimens are prepared with a moraine obtained from a subalpine landslide location, Hexenruebi, in Central Switzerland, which is classified as SP-SM according to Unified Soil Classification System. The index properties of the soil used in the study are summarized in Table 1. Table 1. Classification of Hexenruebi moraine.

Gravel (%) 26 LL – PL (%) 13.3 – 16.8 Sand (%) 63 PI (%) 0 Fines (%) 11 Gs 2.69 Cc - Cu 4.1-70 emin - emax 0.19 – 0.42

The soil collected from the field is oven-dried at

105 ºC for 24 hours, and, subsequently, sieved so that particles greater than 20 mm can be removed. Oven-dried soils, with particles that are smaller than 20 mm, are filled into shear boxes (500x500x400 mm), and compacted in three layers up to a height of 300 mm, by applying 15 blows per layer using a 4.5 kg compaction rammer with a drop height of 460 mm.

3.2 Reinforcement by vegetation

Specimens with roots are prepared with a combi-nation of a tree (Alnus incana), legume (Trifolium pratense), and grass (Poa pratensis). Plant breeding

δ

Direction of Movement

Opposing ForcePlate

Planting Plots

Shear Box

Load Cell

*not to scale

includes the following steps: germination pots of 100-mm-diameter are filled with a peat-sand mixture with high water retention capacity and a mixture of two commercially available mycorrhizal fungi inocu-lum (INOQ Forst, INOQ Agri). Seeds, obtained from the seedbank of the Swiss Federal Institute for Forest, Snow and Landscape Research WSL, are distributed randomly on the surface and covered with an extra layer of 1-2 mm thick peat-sand mixture. Eight plots are marked on the surface of the soil in the shear box, as shown in Figure 1, and four individual plants from each species are transferred to each plot following a growth period of 6-8 weeks in the germination pots. Additional inoculation, Paxillus involutus and Mela-nogaster variegatus s. l. from WSL culture collec-tion, for Alnus incana is added after the transfer of individual plants into shear boxes. Specimens with roots are maintained in shear boxes inclined at 30º in order to simulate the natural growth of roots on a slope. Both the germination pots and shear boxes are kept under controlled conditions in a climate cham-ber, where the temperature and humidity are main-tained at 24ºC and 70%, respectively, between 05:00 and 20:00, and at 17ºC and 55% for the rest of the day.

3.3 Preparation of combined test

The specimens with roots are tested in the ILDSA after a total growth period of 6 months, including the time in germination pots. As shallow landslides are triggered generally after a heavy rainfall period due to saturation and loss of strength, the specimens are subjected to an artificial rainfall event according to Gehrmann et al. (2007) prior to shearing at a constant

intensity of 100 mm/h for 60 minutes, followed by a 30-minute break, and a final rainfall with the same intensity for 30 minutes. Matric potential is moni-tored using tensiometers (UMS T5x) during the rain-fall and brought as close as to zero around the shear plane. Direct shear tests are conducted both horizon-tally, and at an inclination of 30º from the horizontal axis, on specimens without roots at a constant rate of horizontal displacement of 1 mm/min and at different normal stress levels, ranging between 6 kPa, and 18 kPa. Specimens with roots are also tested at the same rate of horizontal displacement and same normal stresses, which are chosen to be comparable to typi-cal depths of shallow landslides (Rickli & Graf 2009; Springman et al. 2003). 4 DESIGN PRINCIPLES OF THE ILDSA It has been shown that the inherent mechanical as-pects of the apparatuses have direct effects on the measured shear strength parameters (Wu et al. 2008), and these effects may be exacerbated in a large appa-ratus. Shibuya et al. (1997) describes three main types of direct shear apparatuses in use, with respect to the freedom of the upper half of the shear box and the loading platen. Both the top platen and the upper half of the shear box are allowed to move freely in the vertical axis and to rotate in type A, whereas they are connected in type B so that they move vertically together and the rotation of the loading platen can be adjusted. The upper half of the box is prevented from movement and rotation in type C, and also a fixed loading platen, which can move vertically but cannot rotate, is used in this configuration. It is suggested that a type C configuration shear box with a non-rotating loading platen would be optimal. The normal load can be measured either conventionally with an external load cell, which will yield the value Wupper, or it can be measured at the bottom of the box, which will yield the value Wlower. It has also been shown that peak friction angle values calculated with the normal load value that has been measured conven-tionally are overestimated in a type C shear box. Therefore, the conventional load measurement will not be appropriate with the optimum shear box, espe-cially at the residual state and large shear displace-ments (Wu et al. 2008).

The ILDSA was designed and built taking the suggestions of Shibuya et al. (1997) into considera-

Figure 1. Top view of a specimen with roots prior to shearing.

1695

& Ohashi 1983), and further investigations based on the Fibre Bundle Model (Pollen et al. 2004).

Basic analytical models assume that roots act per-pendicular to the slip surface, and that the tensile strength of the roots is fully mobilised. However, it can easily be proven that the assumption of roots crossing the slip surface only in a perpendicular di-rection is unrealistic in nature. Furthermore, not only will roots pass through the shear surface but roots ex-tending laterally will also contribute to the strength (Schwarz et al. 2010). Pollen et al. (2004) demon-strated differences in the uptake of strain in roots and soil, showing that the roots are not necessarily stressed to their maximum tensile strength when the soil reaches peak shear stress. Therefore, it is essen-tial to quantify the contribution of roots to the shear strength of soil, when determining the shear strength of the soil matrix with root inclusions. For this rea-son, a robust Inclinable Large-scale Direct Shear Ap-paratus (ILDSA) was built for testing root-permeated soils. The design and mechanical aspects of the appa-ratus are presented herein, as well as results from some preliminary tests.

2 BACKGROUND Direct shear tests have been used in the practice and research of geotechnical engineering since the para-digm shift attributed to contributions of Hvorslev (1937). This is largely due to the simplicity of the test and ease in the procedure. There has been a wide va-riety of testing equipment in use throughout the life-span of the direct shear test. The most common and commercially available type of apparatus accommo-dates 60x60 mm square samples. Size limitations on particle size to key equipment dimensions, as de-scribed by ASTM D 3080, led to development of larger sized apparatuses for testing soil (Bareither et al. 2008; Springman et al. 2003), and also in tests on marginal materials, such as fly ash pellets (Baykal & Döven 2000) and municipal solid waste (Zekkos et al. 2010).

Either laboratory or in situ large-scale direct shear testing is widely employed to determine the shearing behaviour of root-permeated soil. Operstein & Fryd-man (2000) conducted direct shear tests on 200 mm diameter samples of chalky soil reinforced with Pis-tacia lentiscus and Myoporum parvifolium and con-cluded that the increase in shear strength due to roots

was 0.25Tr, where Tr represents the tensile strength of the roots. This is significantly less than the sugges-tions of Wu et al. (1979), in which the increase in shear strength is defined as 1.2Tr with the assumption of full mobilisation of tensile strength of the roots. Goldsmith (2006) performed direct shear tests on samples obtained from the field with a dimension of 250x250x250 mm. The experiments were conducted under saturated conditions and a nominal normal load was applied with a 2 kg metal plate. The shear stress values for specimens with and without roots at a pre-defined value of shear displacement were compared, and it was found out that specimens planted with Sa-lix nigra and Panicum virgatum yielded shear stress values of 36.0 kPa and 37.8 kPa at a shear displace-ment of 7 cm, respectively, whereas the shear stress of specimens without roots was 6.6 kPa at the same shear displacement. 3 TESTING PROGRAMME 3.1 Soil All specimens are prepared with a moraine obtained from a subalpine landslide location, Hexenruebi, in Central Switzerland, which is classified as SP-SM according to Unified Soil Classification System. The index properties of the soil used in the study are summarized in Table 1. Table 1. Classification of Hexenruebi moraine.

Gravel (%) 26 LL – PL (%) 13.3 – 16.8 Sand (%) 63 PI (%) 0 Fines (%) 11 Gs 2.69 Cc - Cu 4.1-70 emin - emax 0.19 – 0.42

The soil collected from the field is oven-dried at

105 ºC for 24 hours, and, subsequently, sieved so that particles greater than 20 mm can be removed. Oven-dried soils, with particles that are smaller than 20 mm, are filled into shear boxes (500x500x400 mm), and compacted in three layers up to a height of 300 mm, by applying 15 blows per layer using a 4.5 kg compaction rammer with a drop height of 460 mm.

3.2 Reinforcement by vegetation

Specimens with roots are prepared with a combi-nation of a tree (Alnus incana), legume (Trifolium pratense), and grass (Poa pratensis). Plant breeding

δ

Direction of Movement

Opposing ForcePlate

Planting Plots

Shear Box

Load Cell

*not to scale

includes the following steps: germination pots of 100-mm-diameter are filled with a peat-sand mixture with high water retention capacity and a mixture of two commercially available mycorrhizal fungi inocu-lum (INOQ Forst, INOQ Agri). Seeds, obtained from the seedbank of the Swiss Federal Institute for Forest, Snow and Landscape Research WSL, are distributed randomly on the surface and covered with an extra layer of 1-2 mm thick peat-sand mixture. Eight plots are marked on the surface of the soil in the shear box, as shown in Figure 1, and four individual plants from each species are transferred to each plot following a growth period of 6-8 weeks in the germination pots. Additional inoculation, Paxillus involutus and Mela-nogaster variegatus s. l. from WSL culture collec-tion, for Alnus incana is added after the transfer of individual plants into shear boxes. Specimens with roots are maintained in shear boxes inclined at 30º in order to simulate the natural growth of roots on a slope. Both the germination pots and shear boxes are kept under controlled conditions in a climate cham-ber, where the temperature and humidity are main-tained at 24ºC and 70%, respectively, between 05:00 and 20:00, and at 17ºC and 55% for the rest of the day.

3.3 Preparation of combined test

The specimens with roots are tested in the ILDSA after a total growth period of 6 months, including the time in germination pots. As shallow landslides are triggered generally after a heavy rainfall period due to saturation and loss of strength, the specimens are subjected to an artificial rainfall event according to Gehrmann et al. (2007) prior to shearing at a constant

intensity of 100 mm/h for 60 minutes, followed by a 30-minute break, and a final rainfall with the same intensity for 30 minutes. Matric potential is moni-tored using tensiometers (UMS T5x) during the rain-fall and brought as close as to zero around the shear plane. Direct shear tests are conducted both horizon-tally, and at an inclination of 30º from the horizontal axis, on specimens without roots at a constant rate of horizontal displacement of 1 mm/min and at different normal stress levels, ranging between 6 kPa, and 18 kPa. Specimens with roots are also tested at the same rate of horizontal displacement and same normal stresses, which are chosen to be comparable to typi-cal depths of shallow landslides (Rickli & Graf 2009; Springman et al. 2003). 4 DESIGN PRINCIPLES OF THE ILDSA It has been shown that the inherent mechanical as-pects of the apparatuses have direct effects on the measured shear strength parameters (Wu et al. 2008), and these effects may be exacerbated in a large appa-ratus. Shibuya et al. (1997) describes three main types of direct shear apparatuses in use, with respect to the freedom of the upper half of the shear box and the loading platen. Both the top platen and the upper half of the shear box are allowed to move freely in the vertical axis and to rotate in type A, whereas they are connected in type B so that they move vertically together and the rotation of the loading platen can be adjusted. The upper half of the box is prevented from movement and rotation in type C, and also a fixed loading platen, which can move vertically but cannot rotate, is used in this configuration. It is suggested that a type C configuration shear box with a non-rotating loading platen would be optimal. The normal load can be measured either conventionally with an external load cell, which will yield the value Wupper, or it can be measured at the bottom of the box, which will yield the value Wlower. It has also been shown that peak friction angle values calculated with the normal load value that has been measured conven-tionally are overestimated in a type C shear box. Therefore, the conventional load measurement will not be appropriate with the optimum shear box, espe-cially at the residual state and large shear displace-ments (Wu et al. 2008).

The ILDSA was designed and built taking the suggestions of Shibuya et al. (1997) into considera-

Figure 1. Top view of a specimen with roots prior to shearing.

Yildiz et al.

Geotechnical Engineering for Infrastructure and Development

1696

tion in order to test large-scale specimens with roots in direct shear. Sample size is a crucial consideration, not only in terms of soil mechanics, but also in terms of plant growth. A small container filled with an inert solid medium or soil implies a limited amount of wa-ter and nutrients and higher chance of pot-bound roots, leading to undesirable secondary consequenc-es. It is suggested to use a container that is large enough to produce a value of the total plant dry mass per unit rooting volume less than 2 g/l to prevent this (Poorter et al. 2012). 5 MECHANICAL SETUP OF THE ILDSA The ILDSA was built to test soil blocks having di-mensions of 500x500x400 mm, which are placed in split wooden boxes. The schematic drawing of the ILDSA, describing the main components of the appa-ratus, is shown in Figure 2. The normal load is ap-plied directly on a steel plate, which is fixed against rotation but allowed to move freely along a vertical axis, by a linear driving unit consisting of a linear ac-tuator (Thomson T 60), coupled with a synchronous servomotor (Beckhoff AM3032-0C40-0000). It is monitored during testing with a load cell (HBM C2). The maximum normal load that can be applied is 10 kN. The movement of the loading platen is measured with a laser displacement sensor (Baumer OADM 13U6475/S35A) and the inclination of the loading platen is monitored with a 2D inclination sensor (Mi-cronor MR401-3).

The shear load application system, with a capacity of 20 kN, is similar to that of the normal load. A lin-ear actuator (Thomson T 90) is coupled with a syn-chronous servomotor (Beckhoff AM3042-0E40-0000) to apply the shear load. A load cell with a ca-pacity of 20 kN is used to measure the shear force during testing. The horizontal displacement is meas-ured with a potentiometric linear transducer (Mega-tron MMS33). The shear box is placed on a base plate, equipped with two load cells at the front and two load cells at the back, and supported with an op-posing force plate. A load cell is mounted into the opposing force plate to monitor the force that has been applied to the plate by the shear box.

6 RESULTS & DISCUSSION

6.1 Mechanical friction of the apparatus The shear box is placed on a base plate, equipped with four load cells, and supported with an opposing force plate. A frictional force develops between the shear box and the base plate in order to eliminate the gap between the bottom half of the shear box and the opposing force plate, which is shown in Figure 1 as δ. Start of shearing is considered as the instant when the shear box contacts the opposing force plate. A correction force is defined in order to take the fric-tional force into account in calculating the shear force. The readings of the load cell embedded into the opposing force plate were evaluated and thresh-old values of the opposing force have been chosen as 100 N for horizontal tests, and 600 N for tests in-clined at 30º, to ensure that contact has occurred. The correction force to eliminate the gap can be formulat-ed, based on data from horizontal and inclined tests on soil, as follows:

F = 420.03δ + 302.81 (1) where F is the correction force in N and δ is the gap between the bottom half of the shear box and the op-posing force plate in mm. 6.2 Tests on specimens without roots A series of direct shear tests was conducted on spec-imens without roots that were prepared with the same compaction energy to have medium dense samples with relative densities ranging between 47% and 64% in order to study the effect of inclination of the box on the transfer of applied normal load to the shear

Shear Box (Lower)

Linear Driving UnitLinearDriving Unit

BasePlate 4 x Load Cells

Opposing ForcePlate

Loading Platen

*not to scale

Shear Box(Upper)

Figure 2. Schematic drawing of the ILDSA.

surface. The load values measured beneath the shear box are compared for horizontal and inclined tests, and illustrated for applied normal loads of 1500 N and 4000 N under dry conditions in Figures 3a and 3b, respectively. It can be seen that, although the val-ues measured with the load cells embedded into the base plate were different at the front and back of the shear box for different configurations, they showed similar trends during shearing.

6.3 Tests on specimens with roots Figure 4 shows the results of inclined direct shear tests that were conducted on specimens with roots and without roots. Specimens without roots in this se-ries of experiments were saturated with the same rainfall intensity and over the same duration as the

specimens with roots, unlike the previous tests on specimens without roots, which were conducted un-der dry conditions. The experiments were carried out at two different normal load levels, namely 1500 N and 4000 N, which results in applied normal stresses of 6 kPa and 16 kPa, respectively, excluding the weight of the soil above the shear surface. The results indicate that the peak shear force was 65% greater for the specimen with roots tested under saturated condi-tions at 1500 N applied load than the peak shear force of the specimen without roots tested, also, un-der saturated conditions at the same normal load. The increase in peak shear stress was 35% when both of the specimens were tested at 4000 N applied load un-der saturated conditions. The decrease in contribution of root reinforcement to the peak shear force with in-creasing normal stress under saturated conditions, thus a bilinear failure curve, is a trademark of all fi-bre reinforced soils regardless of the type of test or reinforcement (Maher & Gray 1990).

0 50 100 150 200

Shear Displacement (mm)

0

2000

4000

6000

8000

Shea

rFor

ce(N

)

N = 1500 N

N = 4000 N

with rootswithout roots

Figure 4. Shear force – displacement graphs for specimens with, and without roots

The peak shear strength parameters, drained cohe-

sion intercept, c', and angle of internal friction, ϕ', for specimens with and without roots are summarized in Table 2. Calculations were performed both with the applied normal load values and also the values meas-ured with the load cells at the bottom of the shear box. The failure envelope of root-permeated soil was forced to have the same inclination as that of soil. When the sample is dilating, soil grains will move relative to the side walls of the shear box, thus, due to the wall friction, Wlower will be higher than Wupper.

0 20 40 60 80 100Shear Displacement (mm)

0

500

1000

1500

2000

2500

Load

Cel

lRea

ding

satt

heB

otto

m(N

)

(a)

(b)

1000

2000

3000

4000

5000

Load

Cel

lRea

ding

satt

heB

otto

m(N

)

Horizontal - Back

Inclined - Back

Horizontal - Front

Inclined - Front

Horizontal - Back

Inclined - Back

Horizontal - Front

Inclined - Front

0 20 40 60 80 100Shear Displacement (mm)

0

Horizontal - Back

Inclined - Back

Horizontal - Front

Inclined - Front

Horizontal - Back

Inclined - Back

Horizontal - Front

Inclined - Front

Figure 3. Measured values of applied normal load from the load cells at the bottom of the shear box a) 1500 N b) 4000 N.

1697

tion in order to test large-scale specimens with roots in direct shear. Sample size is a crucial consideration, not only in terms of soil mechanics, but also in terms of plant growth. A small container filled with an inert solid medium or soil implies a limited amount of wa-ter and nutrients and higher chance of pot-bound roots, leading to undesirable secondary consequenc-es. It is suggested to use a container that is large enough to produce a value of the total plant dry mass per unit rooting volume less than 2 g/l to prevent this (Poorter et al. 2012). 5 MECHANICAL SETUP OF THE ILDSA The ILDSA was built to test soil blocks having di-mensions of 500x500x400 mm, which are placed in split wooden boxes. The schematic drawing of the ILDSA, describing the main components of the appa-ratus, is shown in Figure 2. The normal load is ap-plied directly on a steel plate, which is fixed against rotation but allowed to move freely along a vertical axis, by a linear driving unit consisting of a linear ac-tuator (Thomson T 60), coupled with a synchronous servomotor (Beckhoff AM3032-0C40-0000). It is monitored during testing with a load cell (HBM C2). The maximum normal load that can be applied is 10 kN. The movement of the loading platen is measured with a laser displacement sensor (Baumer OADM 13U6475/S35A) and the inclination of the loading platen is monitored with a 2D inclination sensor (Mi-cronor MR401-3).

The shear load application system, with a capacity of 20 kN, is similar to that of the normal load. A lin-ear actuator (Thomson T 90) is coupled with a syn-chronous servomotor (Beckhoff AM3042-0E40-0000) to apply the shear load. A load cell with a ca-pacity of 20 kN is used to measure the shear force during testing. The horizontal displacement is meas-ured with a potentiometric linear transducer (Mega-tron MMS33). The shear box is placed on a base plate, equipped with two load cells at the front and two load cells at the back, and supported with an op-posing force plate. A load cell is mounted into the opposing force plate to monitor the force that has been applied to the plate by the shear box.

6 RESULTS & DISCUSSION

6.1 Mechanical friction of the apparatus The shear box is placed on a base plate, equipped with four load cells, and supported with an opposing force plate. A frictional force develops between the shear box and the base plate in order to eliminate the gap between the bottom half of the shear box and the opposing force plate, which is shown in Figure 1 as δ. Start of shearing is considered as the instant when the shear box contacts the opposing force plate. A correction force is defined in order to take the fric-tional force into account in calculating the shear force. The readings of the load cell embedded into the opposing force plate were evaluated and thresh-old values of the opposing force have been chosen as 100 N for horizontal tests, and 600 N for tests in-clined at 30º, to ensure that contact has occurred. The correction force to eliminate the gap can be formulat-ed, based on data from horizontal and inclined tests on soil, as follows:

F = 420.03δ + 302.81 (1) where F is the correction force in N and δ is the gap between the bottom half of the shear box and the op-posing force plate in mm. 6.2 Tests on specimens without roots A series of direct shear tests was conducted on spec-imens without roots that were prepared with the same compaction energy to have medium dense samples with relative densities ranging between 47% and 64% in order to study the effect of inclination of the box on the transfer of applied normal load to the shear

Shear Box (Lower)

Linear Driving UnitLinearDriving Unit

BasePlate 4 x Load Cells

Opposing ForcePlate

Loading Platen

*not to scale

Shear Box(Upper)

Figure 2. Schematic drawing of the ILDSA.

surface. The load values measured beneath the shear box are compared for horizontal and inclined tests, and illustrated for applied normal loads of 1500 N and 4000 N under dry conditions in Figures 3a and 3b, respectively. It can be seen that, although the val-ues measured with the load cells embedded into the base plate were different at the front and back of the shear box for different configurations, they showed similar trends during shearing.

6.3 Tests on specimens with roots Figure 4 shows the results of inclined direct shear tests that were conducted on specimens with roots and without roots. Specimens without roots in this se-ries of experiments were saturated with the same rainfall intensity and over the same duration as the

specimens with roots, unlike the previous tests on specimens without roots, which were conducted un-der dry conditions. The experiments were carried out at two different normal load levels, namely 1500 N and 4000 N, which results in applied normal stresses of 6 kPa and 16 kPa, respectively, excluding the weight of the soil above the shear surface. The results indicate that the peak shear force was 65% greater for the specimen with roots tested under saturated condi-tions at 1500 N applied load than the peak shear force of the specimen without roots tested, also, un-der saturated conditions at the same normal load. The increase in peak shear stress was 35% when both of the specimens were tested at 4000 N applied load un-der saturated conditions. The decrease in contribution of root reinforcement to the peak shear force with in-creasing normal stress under saturated conditions, thus a bilinear failure curve, is a trademark of all fi-bre reinforced soils regardless of the type of test or reinforcement (Maher & Gray 1990).

0 50 100 150 200

Shear Displacement (mm)

0

2000

4000

6000

8000

Shea

rFor

ce(N

)

N = 1500 N

N = 4000 N

with rootswithout roots

Figure 4. Shear force – displacement graphs for specimens with, and without roots

The peak shear strength parameters, drained cohe-

sion intercept, c', and angle of internal friction, ϕ', for specimens with and without roots are summarized in Table 2. Calculations were performed both with the applied normal load values and also the values meas-ured with the load cells at the bottom of the shear box. The failure envelope of root-permeated soil was forced to have the same inclination as that of soil. When the sample is dilating, soil grains will move relative to the side walls of the shear box, thus, due to the wall friction, Wlower will be higher than Wupper.

0 20 40 60 80 100Shear Displacement (mm)

0

500

1000

1500

2000

2500

Load

Cel

lRea

ding

satt

heB

otto

m(N

)

(a)

(b)

1000

2000

3000

4000

5000

Load

Cel

lRea

ding

satt

heB

otto

m(N

)

Horizontal - Back

Inclined - Back

Horizontal - Front

Inclined - Front

Horizontal - Back

Inclined - Back

Horizontal - Front

Inclined - Front

0 20 40 60 80 100Shear Displacement (mm)

0

Horizontal - Back

Inclined - Back

Horizontal - Front

Inclined - Front

Horizontal - Back

Inclined - Back

Horizontal - Front

Inclined - Front

Figure 3. Measured values of applied normal load from the load cells at the bottom of the shear box a) 1500 N b) 4000 N.

Yildiz et al.

Geotechnical Engineering for Infrastructure and Development

1698

Both the conventional measurement of normal load and the measurements underneath the shear box yielded similar results in the case of the specimens without roots. The apparent cohesion due to root re-inforcement is slightly larger if it is calculated based on the conventional normal load measurements.

Table 2. Peak shear strength parameters of specimens with, and without roots

Without roots With roots

c' (kPa) ϕ' (º) c' (kPa) ϕ' (º)

Wlower 0 40.4 6.3 40.4

Wupper 0 39.8 7.5 39.8

7 CONCLUSIONS An inclinable large-scale direct shear apparatus was built in order to test the shear strength of root-permeated soil, and the design principles of the appa-ratus, as well as the initial results are presented in this study. The preliminary experiments showed that the normal load transferred to the bottom of the box is changing during shearing, with a similar trend both for the horizontal and inclined tests, although the ap-plied normal load is regulated at a constant value throughout the test. An apparent cohesion, calculated with the conventional measurement of normal load, is slightly larger compared to the value calculated with the measurements of the load cells at the bottom of the shear box. This study contains only the results of the preliminary experiments, and the further work, based on the preliminary results, will be performed and analysed with the same methodology. ACKNOWLEDGEMENT This study is funded by Swiss National Science Foundation within the framework of the National Re-search Programme "Sustainable Use of Soil as a Re-source" (NRP 68, Project no. 143122). REFERENCES Askarinejad, A. & Springman, S.M. 2014. Centrifuge modelling of the effects of vegetation on the responses of a silty sand slope sub-jected to rainfall. Computer Methods and Advances in Geome-chanics (Eds: Oka, F. Murakami, A. Uzuoka, R. & Kimoto, S.), 1339-1344. Taylor & Francis Group, London. ASTM Standard D 3080-04. 2004. Standard Test Method for Di-rect Shear Test of Soils Under Consolidated Drained Conditions, ASTM International, West Conshohocken, PA.

Bareither, C.A. Benson, C.H. & Edil, T.B. 2008. Comparison of shear strength of sand backfills measured in small-scale and large- scale direct shear tests, Canadian Geotechnical Journal 45 (9), 1224-1236. Baykal, G. & Döven, A.G. 2000. Utilization of fly ash by pelleti-zation process: theory, application areas and research results, Re-sources, Conservation and Recycling 30, 59-77. Germann, P. Helbling, A & Vadilonga, T. 2007. Rivulet approach to rates of preferential infiltration, Vadose Zone Journal 6, 207–220. Goldsmith, W. 2006. Soil strength reinforcement by plants. 37th Conference of the International Erosion Control Association, 3-18. International Erosion Control Association, Colorado. Gray, D.H. & Barker, D. 2004. Root-soil mechanics and interac-tions. Riparian Vegetation and Fluvial Geomorphology (Eds: Bennett, S.J. & Simon, A.), 113-123. American Geophysical Un-ion, Washington, D.C.. Gray, D.H. & Ohashi, H. 1983. Mechanics of fiber reinforcing in sand, Journal of Geotechnical Engineering 109 (3), 335–353. Hvorslev, M.J. 1937. Ueber die Festigkeitseigenschaften gestörter bindiger Böden. Ingeniorvidenskabelige Skrifter A, Nr. 45, Koben-havn. Maher, M.H. & Gray, D.H. 1990. Static response of sand rein-forced with fibres, Journal of Geotechnical Engineering 116 (11), 1661-1677. Operstein, V. & Frydman, S. 2000. The influence of vegetation on soil strength, Ground Improvement 4, 81–89. Pollen, N. Simon, A. & Collison, A. 2004. Advances in assessing the mechanical and hydrologic effects of riparian vegetation on streambank stability. Riparian Vegetation and Fluvial Geomor-phology (Eds: Bennett, S.J. & Simon, A.), 125-139. American Ge-ophysical Union, Washington, D.C.. Poorter, H. Fiorani, F. Stitt, M. Schurr, U. Finck, A. Gibon, Y. Usadel, B. Munns, R. Atkin, O.K. Tardieu, F. & Pons, T.L. 2012. The art of growing plants for experimental purposes: a practical guide for the plant biologist, Functional Plant Biology 39, 821–838. Rickli, C. & Graf, F. 2009. Effects of forests on shallow landslides - case studies in Switzerland, For. Snow Landsc. Res. 82 (1), 33-44. Schwarz, M. Lehmann P. & Or, D. 2010. Quantifying lateral root reinforcement in steep slopes – from a bundle of roots to tree stands, Earth Surface Processes and Landforms 35, 354–367. Shibuya, S. Mitachi, T. & Tamate, S. 1997. Interpretations of di-rect shear box testing of sands as quasi-simple shear, Géotech-nique 47 (4), 769-790. Springman, S.M. Jommi, C. & Teysseire, P. 2003. Instabilities on moraine slopes induced by loss of suction: a case history, Géotechnique 53 (1), 3–10. Wu, P.K. Matsushima, K. & Tatsuoka, F. 2008. Effects of speci-men size and some other factors on the strength and deformation of granular soil in direct shear tests, Geotechnical Testing Journal 31 (1), 45-64. Wu, T.H. McKinnell, W.P. & Swanston, D.N. 1979. Strength of tree roots and landslides on Prince of Wales Island, Alaska, Cana-dian Geotechnical Journal 16, 19-33. Zekkos, D. Athanasopoulos, G.A. Bray, J.D. Grizi, A. & Theodo-ratosand, A. 2010. Large-scale direct shear testing of municipal solid waste, Waste Management 30, 1544-1555.

Fiber optics applied for slope movements monitoring Utilisation des fibres optiques pour la surveillance des mouvements

de pente J. Zalesky*1, M. Zalesky2, L. Sasek3 and K. Capova1

1 Czech Technical University in Prague, Faculty of Civil Engineering, Prague, Czech Republic 2 ARCADIS CZ a.s., division Geotechnika, Prague, Czech Republic

3 SAFIBRA, s.r.o., Ricany, Czech Republic * Corresponding Author

ABSTRACT The paper presents fiber optics measurement and instruments of own construction developed within two past research pro-jects. After a period of development, laboratory testing and measurements in underground laboratory, the field measurements started. Dif-ferent means of geotechnical instrumentation for subsoil deformation monitoring were carried out in a former open-cast-mine in the Czech Republic in order to test our instrumentations and deliver more information about a side-slope of the former mine, which is under reclama-tion. Slow slope movements are still observed. The instrumentations were done both with fiber optics and common geotechnical defor-mation measurement tools, so that the results can be cross-checked. Two methods have been used: Fiber Bragg Gratings (FBG) point or line-wise deformation sensors and Brillouin Optical Time Domain Analysis (BOTDA) for distributed deformation measurements. These in-strumentations designed according to our proposals should have longer serviceability compared to common measurement casings used with i.e. inclinometer and sliding deformeter. Following applications are described: field measurement results as well as advantages and disad-vantages are discussed: FBG sub-surface extensometer for shallow slope movement monitoring, borehole instrumentation with FBG chains in Glass Fiber Reinforced Polymer (GFRP) protective sleeve for axial deformation measurements and BOTDA borehole instrumentation for subsoil deformation measurement. FBG modular borehole extensometer developed by the authors is presented.

RÉSUMÉ Le document présente l'utilisation des fibres optiques et les instruments conçus et mis au point par les auteurs dans le cadre de deux projets de recherche antérieurs. Après une période de développement, d'essais au laboratoire et de mesures en laboratoire souterrain, les mesures sur site ont commencé. Différents types d'instrumentation géotechnique pour la surveillance des déformations de pente ont été mis en place dans une ancienne mine à ciel ouvert en République tchèque afin de tester nos instrumentations et de fournir plus d'informations sur le talus de l'ancienne mine, qui est en cours de réhabilitation. Des mouvements de terrain de faible vitesse y sont observés encore à ce jour. Les instrumentations ont été réalisées en utilisant à la fois des fibres optiques et des outils de mesure de déformation habituels, de sorte que les résultats peuvent être recoupés. Pour les fibres optiques, deux méthodes ont été utilisées: Fibre Bragg Gratings (FBG) pour mesures de déformation en points particuliers ou en ligne, et Brillouin Optical Time Domain Analysis (BOTDA) pour mesures de déformation distribuées. Ces instrumentations de notre conception devraient avoir une durée de vie supérieure par rapport aux tubes de mesure utilisés habituellement, par ex. inclinomètre ou déformètre de forage. Les applications suivantes sont décrites et les résultats de mesures sur site ainsi que leurs avantages et inconvénients sont discutés: l' « extensomètre FBG sous-surface » pour la surveillance des pentes à faible profondeur ; l'instrumentation du trou de forage avec des chaînes FBG en manchon de protection Glass Fiber Reinforced Polymer (GFRP) pour mesures de déformation axiale ; et l'instrumentation BOTDA du trou de forage pour mesures de déformation du milieu géologique. L' « extensomètre FBG modulaire de forage » développé par les auteurs est présenté.

1 FIBER OPTICS FOR DEFORMATION MONITORING

Two research and development projects focused on geotechnical monitoring of deformations with the use of different fiber optics measurement methods were carried out from 2011 to 2013. The aim was to de-

velop, test and approve measurement instruments, in-strumentation procedures and data processing.

One R&D project was focused on point-wise or line-wise measurements with sensors based on Fiber Bragg Gratings (FBG). The objective of the other project was the Brillouin Optical Time Domain Analysis method (BOTDA) for distributed sensing.