ISSN 1854-0171 L Z fi / // S / « ^ ViT ^ o. # with R2=0.7. It was stated that such correlations, although developed for particular types of soil, could be applicable worldwide, but were actually found to be site-specific due to a number of factors pertaining to local soils (particle distribution, plasticity, soil texture, etc.) and thus should be used with caution. Given Vg - Ndph correlations should not be thought of as a substitute for seismic (geophysical) measurements, they can be useful for verifying the measured values of the shear-wave velocity Vg or in supplementing seismic (geophysical) testing due to some impossibilities (space constraints, structure of terrain, environmental noise, etc.). In this paper we present statistical empirical correlations between the shear-wave velocity Vg (geophysical seismic surface MASW method at low strains yg~ 10-6, Fig. 1) and the soil penetration resistance Ndph (Dynamic Probing Heavy test DPH at high strains yg~ 10-1, Fig. 1), in shallow quaternary alluvial sediments for a local geological site-specific case study at four locations in Varazdin using the power-law relation: Vg = A-NDpHB (1) From measured DPH data, the shear-wave velocity (Vg), the shear modulus (Gmax or Go) and the Young's modulus (Eo and Er) were estimated. The main aim of this paper is not to give an empirical correlation between Vg - Ndph as a substitute for seismic measurements, but to show a reliable way of comparing two different experimental approaches at low strain vs. high strain for the determination of site-specific dynamic soil properties. The suggested correlation between Vg and Ndph can be used to make rough estimates of Vg from Ndph (they are site-specific, and so not applicable worldwide), particularly for preliminary studies and/or non-critical projects that are under consideration. 2 DYNAMIC SOIL PROPERTIES A medium is said to be elastic if it possesses a natural state (strains and stresses are at zero) to which it will revert to when applied forces are removed. Under the influence of applied loads, the stress and strain will change together, and the relation between them, known as the constitutive relation, is an important characteristic of the medium [13]. A determination of the dynamic soil properties is a critical task in any solution of geotechnical earthquake-engineering problems. The selection of testing techniques for the measurement of dynamic soil properties requires a careful consideration and understanding of the specific O "D O Geophysical Tests _ . ^ 9 ■ Range for P Deformation Analyses Region for Bearing _J Capacity and Stability Calculations Unload-Reload PMT O qFlat DMT Initial Loading PMT'CL; J. J_ J. Penetration Tests 10-6 10-5 1Q-4 10-3 10-2 Shear Strain, Ys 10-1 10° Figure 1. The starting point of the stress-strain curve (at low strain ys~10-6) can be accurately established from seismic tests and its end point (at high strain ys~10-1) by dynamic tests [12]. 10. Acta Geotechnica Slovenica, 2016/1 S. Strelec et al.: Empirical correlation between the shear-wave velocity and the dynamic probing heavy test: Case study, Varazdin, Croatia problem at hand. Low strain tests generally operate at strain levels that are not large enough to induce significant nonlinear stress-strain behaviour in the soil (below ys<10-3) and are based on the theory of wave propagation in the material. High-strain tests are used to measure the high-strain characteristics of soils (elastoplastic behavior), such as soil strength and soil deformation during irrecoverable deformations in the soil [1]. The dynamic modulus is the ratio of stress to strain under vibratory conditions (calculated from data obtained from either free- or forced-vibration tests, in shear, compression or elongation), the so-called low-strain modulus. The shear modulus describes the material's response to shear stress. Most seismic geophysical tests induce low shear strains and the measured shear-wave velocity Vg can be used to compute the maximum dynamic shear modulus or stiffness of the soil Gmax or Go with a knowledge of the soil density p [1, 13, 14, 15]: Go = p-YS2 (2) The soil density p can be evaluated from the measured shear-wave velocity Ys and the depth h [12]: p = 0.85-log(VS ) - 0.16 ■ log(h) (3) A small strain shear modulus is the key benchmark and establishes the highest, achievable soil stiffness that other moduli can be compared to on a relative basis. The Young's modulus Eo describes the material's response to linear stress. Its relation to the shear modulus Go is defined as: Eo = 2-Go(1+u) (4) where v is the Poisson's coefficient. With Go and Eo being operable at non-destructive small strains and a shear strength Tmax at failure strains, it is possible to relate small - high strain modulus (two opposite ends of the strain spectrum). For monotonic static loading a modified degradation hyperbola (Fig. 1) relates the normalized shear modulus G/Go and the normalized Young's modulus E/Eo to the mobilized shear stress t/tmax in the form [12]: GGo='-f or for the Young's modulus: EEo=1-f ' MAX ' MAX (5) (6) where f and q are empirical parameters controlling the rate of modulus decay (Fig. 1). For practical problems, the mobilized shear stress can be considered as the reciprocal of FS (factor of safety). Burns and Mayne [16] found that f = 1 and q = 0.3 provide reasonable first-guess values for the degradation parameters in non-structured and non-cemented soils. The Young's modulus E in Eq.6 is used in deformation analyses as a reference modulus of elasticity (E = Er) because it is directly dependent on the implemented stress and failure stress. The shear-wave velocity Vg is one of the most important parameters for the determination of the dynamic soil properties and ground-response analyses. Seismic geophysical methods are effective in preliminary field tests since they can cover large volumes of soil at low strains. To induce soil deformation similar to that induced by an earthquake, dynamic probing tests are suited to studying a discrete point of interest in a large field area based on preliminary seismic tests. In this way very useful information about dynamic soil properties can be extracted for further ground-response analyses and the study of local site effects. The measured dynamic soil properties (Vs, Go, Eo, Er) can be used for the seismic design of structures based on Eurocode 8 - EC8 [17, 18, 19], site response studies [2] and settlement analyses. In a liquefaction assessment, Vg is also widely used [1]. 3 FIELD MEASUREMENTS_ Field measurements were carried out at four test locations in and nearby Varazdin (Fig. 2, marked with a black circle). The geophysical seismic surface method MASW (Multichannel Analysis of Surface Waves) was used to obtain the shear-wave velocity (Vg) and in-situ dynamic probing tests DPH (Dynamic Probe Heavy) were used to obtain the penetration resistance of the soil (NDpH). 3.1 Geological setting Varazdin is a baroque city located in the NW part of Croatia in shallow quaternary alluvial sediments of the Drava River basin. In this seismic zone there have been a couple of strong earthquakes in the past. The archives mention a strong earthquake in 1459, with its epicentre near Varazdin (I0 = IX0 MCS). The largest earthquake in modern times occurred in 1982 beneath the Ivancica Mt. (Ml = 4.7, I0 = VIIo MCS), and there were also a number of smaller-intensity events in the past according to [20]. The geological characteristics of the Varazdin area are quaternary alluvial soft sediments of the Drava river basin (Fig. 2). It was formed during the Pleistocene and 10. Acta Geotechnica Slovenica, 2016/1 S. Strelec et al.: Empirical correlation between the shear-wave velocity and the dynamic probing heavy test: Case study, Varazdin, Croatia Figure 2. Geology (Quaternary: Pleistocene and Holocene) of Varazdin region at four test locations (1. Landfill - Kneginec, 2. Water well - Varazdin, 3. Zagreb street - Varazdin, 4. Biogas facility Moticnjak - Trnovec)[21]. Holocene as a result of the accumulation processes of the Drava river, as mentioned by Prelogovic [21]. It is composed of gravel and sand with variable portions of silt and clay [22, 23, 24]. The thickness of the sand-gravel deposits in the NW area are less than 5 metres, gradually increasing in the downstream direction (WE) to reach more than 60 metres thick in the eastern part. The water table is relatively high at 2-6 m. The general groundwater flow direction is NW-SE and is parallel to the Drava river. The basin is cut by several faults in an SW direction towards the Ivancica Mt. and a WE direction parallel to the Drava River. 3.2 Multichannel analysis of surface waves Spectral Analysis of Surface Waves (SASW) was introduced in the early 1980s by Nazarian and Stokoe [25, 26, 27] to make use of the spectral analysis of surface waves generated from an impact source like a hammer on a surface. It is used for shallow, shear-wave velocity characterization in which only two receivers (geophone) are used and the spacing between the receiver and the source is changed many times to cover the desired range of the investigation depth. Since this repetition takes several hours, the Multichannel Analysis of Surface Waves by Park et al. [28] and Xia et al. [29] was introduced. The Multichannel Analysis of Surface Waves (MASW) is a non-destructive seismic method used to evaluate the material layer thickness, the shear-wave velocity Vg (1D or 2D profile with depth), the Poisson's ratio and the soil density [28, 29, 30, 31, 32, 33, 34]. This method is fast and simple because of the strong nature of the surface wave (Rayleigh waves) energy that can be generated by a simple impact source (e.g., a sledgehammer) and the entire range of the investigation depth can be covered with multiple receivers (geophones) without the need to change the receiver's configuration [28, 29, 30, 31, 32, 33, 34]. A schematic diagram of an active MASW field survey is shown in Fig. 3. The MASW profile consist of 24 geophones with a natural frequency of 4.5 Hz in a linear array (69 m spread length, 3 m spacing and 6 m offset), a sledgehammer of 10 kg, a hammer metal plate, and a 24-channel Geode seismograph with supporting software for the data acquisition. The maximum investigation depth (zmax) depends on the length of the geophone spread (D) and the source impact energy, while the resolution depends on the minimum geophone spacing (dx). The data analysis of the captured surface waves (Rayleigh waves) created by the sledgehammer impact on the metal plate was analysed using Seislmager (OYO Corp.) and SurfSeis (Kansas Geological Survey) software packages to obtain a 1D and 2D shear-wave velocity profile with depth. The data analysis consisted of three 10. Acta Geotechnica Slovenica, 2016/1 S. Strelec et al.: Empirical correlation between the shear-wave velocity and the dynamic probing heavy test: Case study, Varazdin, Croatia Figure 3. Schematic diagram of active MASW field survey data acquisition [34]. steps: 1) preparation of a MASW record (field file or shot gather), 2) dispersion analysis of Rayleigh waves -fundamental mode and 3) inversion for Vg profiles [28, 29, 30, 31, 32, 33, 34]. The 1D soil profiles for all the test locations obtained from the MASW are shown in Fig. 4., and the 2D profiles are shown in Fig. 5 for the test locations 2.Water well - Varazdin and 4. Moticnjak - Trnovec. The top first layer in all four locations consists of clay having a Vg range of 110-170 m/s from 0 up to 4 m in depth. The second layer consists of sand having a Vg range of 190-260 m/s from 0 up to 9 m depth. The third layer consists of gravel having a Vg range of 270-400 m/s from 5 up to 20 m and above depth. The maximum investigation depth was cut to 20 m for the Vg - Ndph correlation analysis. Based on the EC8 classification of subsoil classes according to Table 3.1 [17, 18, 19] all the investigated area can be classified as ground type C (Vg,30 = 180-360 m/s) (Fig. 4). 1. Vs [m/s] 100 150 200 250 300 350 400 I ' £ 10 a. a> a 12 14 15 18 20 h- ■l 'l - S wave velocity model (inverted) average Vs30=312.6 m/s 2. Vs [m/s] 100 150 200 250 300 350 400 ■ Depth [m] oooCT^Nooooj-ttroc \ L 1 / S wave velocity model (Inverted) average Vs30=288.3 m/s 3- Vs [m/s] 100 150 200 250 300 350 400 B \ - S wave velocity model (inverted) average Vs30=268.8 m/s Vs [m/s] 100 150 200 250 300 350 400 S 10 a. o> a 12 14 16 18 20 L 1 h i i i \ - S wave velocity model (Inverted) average Vs30=286.1m/s Figure 4. 1D Vg profile vs. depth. Location: 1) Landfill - Kneginec, 2) Water well - Varazdin, 3) Zagreb street - Varazdin, 4) Moticnjak - Trnovec. 10. Acta Geotechnica Slovenica, 2016/1 S. Strelec et al.: Empirical correlation between the shear-wave velocity and the dynamic probing heavy test: Case study, Varazdin, Croatia [2 I S wave velocity cross-section: Water we I 12 17 22 27 32 37 42 Distance [m] 72 77 f 5 wave velocity a a:- o seo Mi 3(1 at 130 4 S wa ve ve I ocity cross-secti on: Moti c n j a k oi slan c? [m] 5 10 15 20 25 3D 35 40 45 5D 55 GO 65 70 75 [nVi] Suave velocity I tin XI 3fifi 334 »2 SO aa isa 1EJ 132 1C0 [mfc] Figure 5. 2D Vs profile vs. depth. Location: 2) Water well - Varazdin, 4) Moticnjak - Trnovec. 3.3 Dynamic probing heavy (DPH) test The purpose of an in-situ dynamic probing test is to determine the resistance of soils and semi-rocks in-situ via the dynamic penetration of a cone in order to characterize the soil layers by their depth, position, nature, density or thickness [35, 36]. It can be used to determine the strength and deformation properties of soils or to distinguish non-cohesive soils from cohesive soils [37]. In our study we used one type of dynamic penetrometer, Dynamic Probe Heavy (DPH - Fig. 6), which is more economically efficient and faster than SPT (borehole is needed), in line with EN ISO 22476 - 2 [38] and the European German DIN 4094 - 2 standards [39]. DPH is performed so that the cone (base area of cone point 20 cm2) is driven into the soil (no borehole needed) by free-fall hammer strokes with a specified mass (50 kg) and height of the drop (50 cm). The numbers of strokes required for a 100-mm penetration of the cone Ndph are recorded. The penetration resistance is defined as the number of strokes NDpH per 100 mm of penetration. A continuous record is collected in accordance with the length of sinking (depth). The test does not allow for soil sampling. The results from the DPH test are presented as the number of strokes NDpH against the depth from the direct field record and are in the range of standardized values (usually 3 up to 50) (see Fig. 7). The "harsh" spikes that appear occur when encountering obstructions such as a boulder or very hard, dense soil. When the test is carried out in granular soils below the water table, the soil may become loosened, and the number of strokes NDpH can decrease. Groundwater influence (loosening) on the number of strokes NDpH is significant, and corrections have to be made according to EN ISO22476 - 2 and DIN 4094 - 3 [38, 39]. A groundwater-level correction was made for sand: NdphCORR = 1.3-Ndphorig + 2 and for gravel: NDPHCORR = 1.2-NdphOrig + 4.5 [40, 41, 42], but was not necessary for clay, which was above the groundwater level. Remov- 10. Acta Geotechnica Slovenica, 2016/1 S. Strelec et al.: Empirical correlation between the shear-wave velocity and the dynamic probing heavy test: Case study, Varazdin, Croatia Figure 6. Geotool dynamic probing rig (Dynamic Sampling Ltd, UK). The rig is capable of dynamic probing (DPL-light, DPM-medium, DPH-heavy and DPSH-super heavy) in accordance with DIN 4094. Specification: Width 78 cm, Length 78.5 cm, Height 23.4 cm, Drop Height 50 cm, Rod/Tool Length 120 cm (max) and Basic Weight 125 kg. Reference penetration every 100 mm. Base area of cone 20 cm2. Cone diameter 43.7 mm. Rods diameter 32 mm. Hammer mass 50 kg. Drop height 50 cm. Angle of the cone point 90°. Torque wrench for impact of rod skin friction. Figure 7. Ndph profile vs. depth.Location: 1) Landfill - Kneginec, 2) Water well - Varazdin, 3) Zagreb street - Varazdin, 4) Moticnjak - Trnovec (depth to 15 m). WL - water level mark. ing the effect of water increases the number of recorded strokes (Fig. 7). Fig. 7 shows how the numbers of strokes Ndph are changing with depth. The top first layer is clay with a Ndph range from 1 to 5 (without groundwater correction), the second layer sand with Ndph range from 4 to 25 (corrected 7-38), while the third layer is classified as gravel with Ndph from 12 to 27 (corrected 20-40). 4. EMPIRICAL CORRELATION BETWEEN THE SHEAR-WAVE VELOCITY AND THE DYNAMIC PROBING HEAVY TEST In this section we present an empirical correlation between the shear-wave velocity Vg and the dynamic probing test DPH in shallow quaternary alluvial sediments for a local geological site-specific case study at four locations in Varazdin using the power law given by Eq. 1. From the Vg and Ndph profiles we extracted good data for the statistical correlations with depth. The following example (location 3 - Zagreb street) of the measured data shown in Fig. 4 and 7 depicts the development of the Vg - Ndph empirical correlations. Example: for a depth of 10 m, the measured shear-wave velocity is Vg = 320 m/s and the number of blows Ndph = 17 are taken as a mean value between depths of 9-10 m. Since at this depth the soil is classified as gravel below the water table, we performed a groundwater-level correction: 10. Acta Geotechnica Slovenica, 2016/1 S. Strelec et al.: Empirical correlation between the shear-wave velocity and the dynamic probing heavy test: Case study, Varazdin, Croatia NDphCORR = 1-2-Ndphorig + 4.5 = 25. In this way we extracted more than 200 points from all the measurements at all four locations with two different field methods (MASW and DPH). In Fig. 8 empirical Vg - Ndph correlations shown for a particular test location are presented. Vg - Ndph correlations for the original (black) and corrected (red) Ndph data are shown on all four graphs. It is evident that the data correction due to the groundwater effect on Ndph causes the curve to shift down on the Vg axis (red vs. black) and that in all four cases R2 (determination coefficient) is higher than in the original Ndph data. The Vg - Ndph correlation is statistically better for a groundwater-level correction on Ndph. It can also be seen that R2 is above 0.85 at two locations, i.e., 1. Landfill and 3. Zagreb street, irrespective of the groundwater correction (WL = 4 m and 2 m of depth). For the other two locations, i.e., 2. Water well and 4. Moticnjak, R2 is comparatively low and shows a significant change from the groundwater correction. Since all the field measurements were performed in shallow quaternary alluvial sediments, the main aim of this study is to propose a site-specific Vg - Ndph correlation, rather than more general correlations. Fig. 9 presents empirical Vg - Ndph correlations for the site-specific region of Varazdin compared to a few previously published empirical Vg - NgpT correlations for all soils [5]: Original Ndph. Vg = 97.839-NDPH 0395; R2 = 0.723 Corrected NDPH: Vg = 92.998-NDPH 0363; R2 = 0.815. 5 DISCUSSION From the empirical Vg - Ndph correlation (Fig. 9) it can be seen how corrections for the groundwater effect on DPH measurements can "enhance" the correlation (R2 is higher). But, to see real meaning in this correction, we compared the MASW measured Vg,meas with the estimated Vg,orig and Vg,corr from the original Ndphorig and the corrected Ndphcorr data (Fig. 10). There is a small difference between the MASW measured Vg,meas and the estimated Vg,orig and Vg,corr from the Vg - Ndph correlation. The residuals are from 0 to 50 m/s, and this difference falls Vs = 78.029 N0-4849 . r2 = 0.868 Landfill Vs = 84.426 N0-4003 R2 = 0.893 2. Vs = 117.09n0-3073 r2=o.492 Waterwell Vs = 102.36 n0-3143 r2 = 0.665 3 Vs = 108.55 n0-3633 vs = 102.65 n0-3317 r2 = o.923 Zagreb st. r2 = 0.945 4. Vs = 85.015n0-4608 r2 = 0.709 Moticnjak Vs = 82.968 n0-4139 r2 = 0.810 15 20 25 n - dph Figure 8. Vs - Ndph correlations by test location: 1) Landfill - Kneginec, 2) Water well - Varazdin, 3) Zagreb Street - Varazdin, 4) Moticnjak - Trnovec. Black - original Ndph data. Red - corrected Ndph data. 10. Acta Geotechnica Slovenica, 2016/1 S. Strelec et al.: Empirical correlation between the shear-wave velocity and the dynamic probing heavy test: Case study, Varazdin, Croatia Vs = 97.839 N0395 R2 = 0.723 Vs - Ndph Vs = 92.998 N0 363 R2 = 0.815 Athanasopoulos (1995) VS = 107.6 NSPT036, Hasancebi and Ulusay (2007) VS = 90 NSPT-Imai (1977) VS = 91 N^/337, Jafari et al. (1997) VS = 22 NSPT085 Athansopoulos (1995) Hasancebi & Ulusay (2006) Imai (1977) Jafari & al. (1997) Maheshwari (2010) 10 15 20 N - DPH 25 30 35 40 Figure 9. Vg - Ndph correlations for site-specific region of Varazdin (all locations, all soils; black line: original Ndph, red line: corrected Ndph). Also presented are a few previously published Vg - Ngpf correlations for all soils [5]: blue dotted and dashed lines). 550 500 450 400 „ 350 1 300 Vs 250 200 150 100 50 0 5 within statistical error bounds and so the given results can be used as a rough estimation of Vg from Ndph. A groundwater-level correction is a good way to "statistically enhance" an empirical correlation, but as seen in Fig. 10 there is a very small difference between the Vg,orig estimated from Ndph°rig and the Vg,corr estimated from Ndphcorr. Both empirical correlations are sufficient for a rough estimation of Vg from Ndph and the groundwa-ter-level correction can be applied to lower the statistical error, but the correlation with raw (original) data can also be effective in estimating Vg from Ndph. From the field measurements and empirical correlations, the shear-wave velocity Vg was estimated and dynamic elastic moduli (Go, Eo and Er) based on that velocity were calculated, Fig. 11, 12, 13. In the calculation of the shear modulus Go (Eq. 2), the Young's modulus Eo (Eq. 4) and the reference Young's modulus Er (Eq. 6), the Figure 10. Comparison between measured Vs (blue: VSmeas), Vs estimated from Ndph°RIG (black: VSorig) andVg estimated from NDPHCORR (red: Vgcorr). 10. Acta Geotechnica Slovenica, 2016/1 S. Strelec et al.: Empirical correlation between the shear-wave velocity and the dynamic probing heavy test: Case study, Varazdin, Croatia soil density was evaluated using Eq. 3, while Poisson's ratio with values of v = 0.20 and v = 0.45 for dry and soil under the water table was used. From Fig. 11, 12 and 13 the shear modulus Go varies from 20 MPa (MN/m2) to 292 MPa, the Young's modulus Eo varies from 49 MPa to 846 MPa, and reference Young's modulus Er varies from 9 MPa to 158 MPa for the Varazdin region. The estimated elastic moduli Go, Eo and Er were crosschecked at low (MASW) and high (DPH) strains and they made a good match (especially for test location 3. Zagreb street). High deviated results (test locations 1. Landfill, 2. Water well and 4. Moticnjak) can be explained by a number of factors pertaining to local soils, such as the particle distribution, plasticity, soil texture, loose soil, sticking cone into the boulder, etc., and thus the Vg - Ndph correlation should be used with caution. Two different approaches (low vs. high strain) were compared, and the results found to be in good agree- 1 Go [MPa] 100 200 300 400 Go [MPa] 100 200 300 400 2 ' tt 3 ■ 4 ■ —T 9 5 ■ 7 ■ \\ ¥ 8 ■ J= 9 ' S- 101 a 11 . 12 ■ —¿ft— 13 ■ 14 ■ Tr 15 ■ il— 16 ■ rf^I— 17 ■ I 18 ■ flf 19 ' il 20 ■ G - Vs,meas ■ G - Vs.orig • G - Vs.corr Figure 11. Shear modulus Go (low strain) calculated from measured Vs (blue: VS meas), Vs estimated from NDpH (black: VSorig) and VS estimated from NDPHCORR (red: VS corr). Figure 12. Young's modulus Eo (low strain) calculated from measured Vs (blue: VSmeas), Vs estimated from NDpHORIG (black: VS orig) and VS estimated from NDPHCORR (red: VS corr). 10. Acta Geotechnica Slovenica, 2016/1 S. Strelec et al.: Empirical correlation between the shear-wave velocity and the dynamic probing heavy test: Case study, Varazdin, Croatia ment when the relative difference between the velocities is small and smooth, Fig 14. The relative difference of velocities (RD(o/m) or RD(c/m)) is defined as the ratio of the absolute difference between the measured and estimated velocity and the measured velocity for some depth: RD — abs(l — VS,estimated) (7) V S,measured One can find that the local extremes of the RD curves, Fig. 14, match the local extremes of the N - DPH curves, Fig. 7. Moreover, if the relative difference of the velocities is small and the curve is smooth, Fig. 14 location 3, then the local RD maximum indicates a local N - DPH minimum, and vice versa. When there is a significant spike in the N - DPH data we can expect a significant dispersion of the Vg - Ndph data, Fig. 14 location 4. Regardless of the field measurements, at low (MASW) or high (DPH) strains, it is necessary to evaluate the Figure 13. Young's modulus Er (high strain) calculated from measured Vs (blue), Vs estimated from NDph°rig (black) and Vs estimated from Ndphc°rr (red). Figure 14. Relative difference between the estimated and measured velocities. RD(o/m or c/m) is a relative difference of Vs estimated from NDph°rig or Vs estimated from NDphC°rr and measured MASW velocity Vs (black and red line respectively). 10. Acta Geotechnica Slovenica, 2016/1 S. Strelec et al.: Empirical correlation between the shear-wave velocity and the dynamic probing heavy test: Case study, Varazdin, Croatia appropriate dynamic properties of the materials in the soil deposit. A precise determination of the dynamic soil properties is a somewhat difficult task when solving geotechnical engineering problems. It is clear that the choice of technique depends on the problem to be solved. Dynamic properties play a vital role in the design of earthquake-resistant structures or structures subjected to dynamic loads. 6 CONCLUSION In this paper we present empirical correlations between the shear-wave velocity Vg and the dynamic probing test (DPH) for s case study in a site-specific region of Varazdin. The geological characteristics of Varazdin are quaternary alluvial soft sediments of the Drava river basin (Fig. 2). Field measurements (low strain - MASW, high strain - DPH) were carried out at four different test locations inside and outside the city of Varazdin. Since Varazdin sits on alluvial soft sediments, the penetration resistance of the Ndph groundwater-level correction was taken into consideration. An empirical correlation was developed between the measured shear-wave velocity Vg and the raw (original) Ndphorig and groundwater-level-corrected Ndphcorr data. The results of the developed empirical correlation (Fig. 9) for NDPHORIG and for NDPHCORR were: VS = 97.839-Ndph0395; R2 = 0.723, VS = 92.998-Ndph0363; R2 = 0.815. The groundwater-level correction was found to be effective in "statistically enhancing" the empirical correlations, which can be seen in Fig. 10. There is a very small difference between Vg,orig estimated from Ndphorig and Vg,corr estimated from Ndphcorr. These developed empirical correlations proved to be sufficient for rough estimations of Vs from Ndph and the groundwater-level correction can be applied in order to lower the statistical error, but a correlation with raw (original) data can also be effective in a good way for estimating Vs from Ndph. The shear-wave velocity Vg is one of the most important parameters for a determination of the dynamic soil properties and the ground response analysis. From the field measurements and empirical correlations, the shear-wave velocity Vg, the dynamic elastic moduli of the shear modulus (Go) (Eq. 2) and the Young's modulus (Eo and Er) (Eq. 4 and 6) were estimated (Fig. 11, 12, 13). Two different approaches (low vs. high strain) were compared, and the results found to be in good agreement when the relative difference between the velocities is small and smooth, Fig 14. The measured dynamic soil properties (Vg, Go, Eo, Er) can be used for the seismic design of structures based on Eurocode 8 [17, 18, 19], site response studies [2], settlement analysis and Vg is also widely used in liquefaction assessments [1]. Preliminary seismic tests proved that dynamic probing tests are effective when looking at a discrete point of interest in a large field area. The suggested correlation Vg - Ndph can be used to make rough estimations of Vg from Ndph (they are site-specific, and so not applicable worldwide). It should not be used as a substitute for seismic measurements, but used as a comparison and verification of the measured shear-wave velocity and dynamic probing test for site-specific dynamic soil properties. In this way very useful information about the dynamic soil properties can be extracted for ground-response analyses and the study of local site effects, particularly for preliminary studies and/or noncritical projects that are under consideration. REFERENCES [1] Kramer, S.L. 1996. Geotechnical Earthquake Engineering. Prentice Hall, New Jersey. [2] Reiter, L. 1990. Earthquake Hazard Analysis. Columbia University Press, New York. [3] Barton, N. 2007. Rock Quality, seismic velocity, attenuation and anisotropy. Taylor & Francis Group, London. [4] Hanumantharao, C., Ramana, G.V. 2008. Dynamic soil properties for microzonation of Delhi, India. J. Earth Syst. Sci. 117 S2, 719-730. [5] Marto, A., Soon, T.C., Kasim, F. 2013. A correlation of shear wave velocity and standard penetration resistance. EJGE, 18, 463-471. [6] Jafari, M.K., Shafiee, A., Razmkhah, A. 2002. Dynamic properties of fine grained soils in south of Tehran. Soil Dyn. and Earthquake Eng. 4, 25-35. [7] Uma Maheswari, R., Boominathan, A., Doda-goudar, G.R. 2010. Use of Surface Waves in Statistical Correlations of Shear Wave Velocity and Penetration Resistance of Chennai soils. Geotechnical and Geology Engineering 119-137. [8] Kuo, C.H., Wen, K.L., Hsieh, H.H., Chang, T.M., Lin, C.M., Chen, C.T. 2011. Evaluating Empirical Regression Equations for Vg and Estimating VS30 in Northeastern Taiwan. Soil Dyn. and Earthquake Eng. 31, 431-439. [9] Akin, M.K., Kramer, K.L., Topal, T. 2011. Empirical Correlations of Shear Wave Velocity (VS) and Penetration Resistance (SPT-N) for Different Soils in an Earthquake prone Area (Erbaa-Turkey). Engineering Geology 119, 1-17. [10] Anbazhagan, P., Kumar, A., Sitharam, T.G. 2013. 10. Acta Geotechnica Slovenica, 2016/1 S. Strelec et al.: Empirical correlation between the shear-wave velocity and the dynamic probing heavy test: Case study, Varazdin, Croatia Seismic Site Classification and Correlation between Standard Penetration Test N Value and Shear Wave Velocity for Lucknow City. Indo-Gangetic Basin. Pure and Applied Geophysics 170, 299-318. [11] Iyisan, R. 1996. Correlations between Shear Wave Velocity and In-Situ Penetration Test Results. Digest. 96, 371-374. [12] Mayne, P.W., Schneider, J.A. 1999. Small and large strain soil properties from seismic flat dilatometer tests. Proc. Pre-failure Deformation Characteristics of Geomaterials, Jamiolkowski et al. editors, Torino, 419-426. [13] Aki, K., Richards, P.G. 2009. Quantitative Seismology. University Science Book, California. [14] Mavko, G., Mukerji, T., Dvorkin, J. 2003. The Rock Physics Handbook. Cambridge University Press, Cambridge. [15] Ohsaki, Y., Iwasaki, R. 1973. Dynamic shear moduli and Poisson's ratio of soil deposits; Soils and Foundn. 13, 61-73. [16] Burns, S.E., Mayne, P.W. 1996. Small and high strain soil properties using the seismic piezocone. Transport. Research Record 1548, National Acad. Press, Wash. DC, 81-88. [17] CEN 2004. Eurocode 8 - design of structures for earthquakes resistance. Part 1 - general rules, seismic actions, design and rules for building. European standard EN 1998 -1, December 2004, European Committee for Standardization, Brussels. [18] Hrvatski zavod za norme, 2011a. HRN EN 19981:2011 hr, Eurokod 8: Projektiranje potresne otpornosti konstrukcija -1. dio: Opca pravila, potresna djelovanja i pravila za zgrade (EN 1998-1:2004+AC:2009). HZN - Oglasnik za normativne dokumente 6/2011. [19] Hrvatski zavod za norme, 2011b. Nacionalni doda-tak za niz normi HRN EN 1998 - 1:2011/NA:2011, Eurokod 8: Projektiranje potresne otpornosti konstrukcija - 1.dio: Opca pravila, potresna djelovanja i pravila za zgrade. [20] Markušic, S., Herak, M. 1999. Seismic zoning of Croatia. Natural Hazards 18, 269 - 285. [21] Prelogovic, E., Velic, J. 1988. Quaternary tectonic activity in western part of the Drava depression. Geološki vjesnik 41, 237 - 253. [22] Babic, Ž., Čakarun, I., Sokač, A., Mraz, V. 1978. On geological features of quaternary sediments of Drava basin on Croatian territory. Geološki vjesnik 30(1), 43 - 61. [23] Urumovic, K. 1971. On quaternary aquifer complex in Varaždin area. Geološki vjesnik 24, 183 - 191. [24] Urumovic, K., Hlevnjak, B., Prelogovic, E., Mayer, D. 1990. Hydrogeological conditions of Varaždin aquifer. Geološki vjesnik 43, 149 - 158. [25] Nazarian, S., Stokoe II, K.H. 1984. In situ shear [26 [27 [28 [29 [30 [31 [32 [33 [34 [35 [36 [37 [38 [39 [40 [41 [42 wave velocity from spectral analysis of surface waves, 8th World Conference on Earthquake Engineering Proceedings, 31-38. Stokoe II, K.H., Nazarian, S. 1983. Effectiveness of ground improvement from spectral analysis of surface waves, 8 th Euro. Conf. on Soil Mech. and Found. Engin. Proceedings, 91-94. Stokoe II, K.H., Wright, G.W., Bay, J.A., Roesset, J.M. 1994. Characterization of geotechnical sites by SASW method, in Geophysical characterization of sites, ISSMFE Technical Committee #10, edited by R.D. Woods, Oxford Publishers, New Delhi. Park, C.B, Xia, J., Miller, R.D. 1998b. Imaging dispersion curves of surface waves on multichannel record, 68th Ann. Internat. Mtg., Soc. Expl. Geophys., Expanded Abstracts, 1377-1380. Xia, J., Miller, R.D., Park, C.B. 1999. Estimation of near-surface shear-wave velocity by inversion of Rayleigh wave: Geophysics 64, 691 - 700. Park, C.B, Miller, R.D, Xia, J. 1999. Multichannel analysis of surface waves. Geophysics 64, 3, 800-808. Park, C.B., Miller, R.D., Xia, J., Hunter, J.A., Harris, J.B. 1999. Higher mode observation by the MASW method: Technical Program with Biographies, SEG, 69th Annual Meeting, Houston, TX, 524-527. Xia, J., Miller, R.D., Park, C.B., Hunter, J.A., Harris, J.B. 1999. Evaluation of the MASW technique in unconsolidated sediments: Technical Program with Biographies, SEG, 69th Annual Meeting, Houston, TX, 437-440. Miller, R.D, Xia, J., Park, C.B., Ivanov, J.M. 1999. Multichannel analysis of surface waves to map bedrock. Leading Edge 18, 12, 1392-1396. http://www.masw.com/DataAcquisition.html/ <1.12.2014.> Waschkowski, E. 1982a. Dynamic Probing and Site Investigation. Proc. ESOPT 2, Amsterdam. Waschkowski, E. 1982b. Dynamic Probing and Practice. Proc. ESOPT 2, Amsterdam. Hack, R., Azzam, R., Charlier, R. 2004. Engineering geology for infrastructure planning in Europe: a European perspective. Springer, Berlin. EN ISO 22476 - 2: 2005: Geotechnical Engineering - Field testing - Part 2: Dynamic probing. DIN 4094: 2002: Subsoil - Field Investigations. Part 3: Dynamic probing. BS EN 1997 - 2: 2007. Eurocode 7 - Geotechnical design - Part 2: Ground investigation and testing, London. Stenzel, G., Melzer, K. J. 1978. Soil Investigations by Penetration Testing According to DIN 4094. Tief bau 20, 155-160, 240-244. Czado, B., Pietras, J.P. 2012. Comparison of the cone penetration resistance obtained in static and dynamic field test. AGH Journal of Mining and Geoengineering 36, 1, 97-105. 10. Acta Geotechnica Slovenica, 2016/1 VREDNOTENJE PLASTIČNE KRITIČNE GLOBINE ZA PROBLEME SEIZMIČNIH AKTIVNIH BOČNIH ZEMELJSKIH PRITISKOV Z METODO NAPETOSTNIH KARAKTERISTIK Izvleček Plastična kritična globina ali običajno imenovana globina natezne razpoke ima precejšen vpliv na aktivne bočne zemeljske pritiske v koherentnih tleh. F prispevku je globina za zemljino določena za primer potresa z uporabo napetostnih karakteristik ali metodo porušnic. Plastična kritična globina je bila izračunana na podlagi teorije metode napetostnih-karakteristik in ob upoštevanju horizontalnih in vertikalnih psevdo-statičnih potresnih koeficientov. Predlagana rešitev obravnava črto diskon-tinuitete v mreži napetostnih karakteristik. F analizi so bili prav tako upoštevani naklon zaledja in zidu, kohezija in notranji kot trenja ter tudi adhezija in kot trenja na površini med zemljino in zidom. Rezultati kažejo, da so plastične kritične globine te študije manjše od drugih metod in so bližje modificirani metodi Mononobe-Okabe. Amin Keshavarz Persian Gulf University, School of Engineering Shahid Mahini Street, Bushehr, Iran E-posta: keshavarz@pgu.ac.ir Ključne besede plastična kritična globina, značilnosti napetosti, aktivni bočni zemeljski pritisk, potresno 16. Acta Geotechnica Slovenica, 2016/1 EVALUATION OF THE PLASTIC CRITICAL DEPTH IN SEISMIC ACTIVE LATERAL EARTH PRESSURE PROBLEMS USING THE STRESS-CHARACTERISTICS METHOD Abstract The plastic critical depth or the conventional tension crack depth has a considerable effect on the active lateral earth pressure in cohesive soils. In this paper the depth for soils has been evaluated in the seismic case using the stress-characteristics or slip-line method. The plastic critical depth was calculated on the basis of the theory of the stress-characteristics method and by considering the horizontal and vertical pseudo-static earthquake coefficients. The proposed solution considers the line of discontinuity in the stress-characteristics network. The earth slope, wall slope, cohesion and friction angle of the soil and the adhesion and the friction angle of the soil-wall interface were considered in the analysis as well. The results show that the plastic critical depths of this study are smaller than those of the other methods and are closer to the modified Mononobe-Okabe method. The effects of the wall and the backfill geometry, the mechanical properties of the soil and the pseudo-static coefficients were studied. Amin Keshavarz Persian Gulf University, School of Engineering Shahid Mahini Street, Bushehr, Iran E-mail: keshavarz@pgu.ac.ir Keywords plastic critical depth, stress characteristics, active lateral earth pressure, seismic 1 INTRODUCTION In cohesive soils, the computed active lateral earth pressure can be negative from the ground surface to some depth. The plastic critical depth is the depth where the computed soil pressure is negative from the ground surface to that depth. To calculate the active lateral earth pressure many engineers assume that the lateral earth pressure is zero from the ground surface to the plastic critical depth. Therefore, the calculation of the plastic critical depth is important in any evaluation of the active lateral earth pressure. Numerical examples showed that if the plastic critical depth is considered, the static lateral earth pressure can be more than 20 to 40 percent of the lateral earth pressure, without taking into account the plastic critical depth [1]. Peng [2] assumed that the failure surface is planar and evaluated the static active lateral earth pressure using the limit-equilibrium method. He considered the plastic critical depth and the surcharge in his study. Peng [3] modified the Mononobe-Okabe method to calculate the seismic active lateral earth pressure. He presented some equations to calculate the lateral earth pressure and the plastic critical depth in soils. Nian and Han [4] calculated the seismic active lateral earth pressure against rigid retaining walls. They used the Rankin theory and proposed an equation to estimate the plastic critical depth. In their study the retaining wall is vertical, but the ground surface can be inclined. Also, Iskander et al. [5] expanded the Rankin solution for seismic cases and proposed an equation for the plastic criti- 16. Acta Geotechnica Slovenica, 2016/1 A. Keshavarz: Evaluation of the plastic critical depth In seismic active lateral earth pressure problems using the stress-characteristics method cal depth. Ma et al. [6] evaluated the lateral earth pressure and the plastic critical depth using the pseudo-dynamic method. Recently, Lin et al. [7] used the slice-analysis method to compute the nonlinear distribution of the seismic active earth pressure of cohesive-frictional soil. The stress-characteristics or slip-line method is one of the famous methods for analysing geotechnical problems. This method was presented by Sokolovski [8, 9]. Reece and Hettiaratchi [10] and Kumar and Chitikela [11] used the stress-characteristics method to compute the passive lateral earth pressure. Cheng [12] proposed a rotation of the axes in the stress-characteristics method to determine the seismic lateral earth pressure. Peng and Chen [13] applied this method to compute the active lateral earth pressure in the static case. The stress-characteristics method has also been used to evaluate the stability of reinforced soil structures [14, 15]. Table 1 summarizes the different methods for calculating the plastic critical depth. The parameters used in the table will be defined in the next sections. It is clear that most of the methods proposed closed-form solutions. Some of them need trial and error, and three of them consider all the parameters in the solution. In this paper the plastic critical depth is studied using the stress-characteristics method. Although Peng and Chen [13] used this method to compute the active lateral earth pressure, they did not consider the seismic effects Table 1. Summary of different methods to calculate the plastic critical depth. No._Proposed by Theory Parameters considered Comments E E 1 Rankin Simple Rankin e E ß E - The simplest equation (Eq. (25)) kh E kv E q 0 cw 0 SW 0 2 Peng, 2012 [3] Modified e 0 ß 0 - Closed form without trial and error Mononobe-Okabe kh 0 kv 0 q 0 cw 0 Sw 0 3 Ma et al., 2012 [6] Pseudo-dynamic e 0 ß E -No closed form kh 0 kv 0 - Needs optimization q 0 cw 0 Sw 0 -Closed form 4 Peng and Chen, 2013 Slip line e 0 ß 0 -Needs trial and error [13] kh E kv E -When ¿3=0 solution can be found without q 0 trial and error cw E Sw E 5 Iskandet et al., 2013 [5] Expanded Rankin e 0 ß 0 - Closed form without trial and error kh 0 kv 0 q E cw E Sw E 6 Nian and Han, 2013 [4] Modified Rankin e E ß 0 - Closed form without trial and error kh 0 kv 0 q E cw 0 Sw 0 7 Lin et al., 2015 [7] Slice analysis method e 0 ß 0 -Closed form kh 0 kv 0 -Needs trial and error q 0 cw 0 Sw 0 -Closed form 8 This study Slip line e 0 ß 0 -Needs trial and error kh 0 kv 0 -When 3=kh=kv=0, solution can be found q 0 without trial and error c w 18. Acta Geotechnica Slovenica, 2016/1 A. Keshavarz: Evaluation of the plastic critical depth In seismic active lateral earth pressure problems using the stress-characteristics method and the discontinuity line in the stress field. In this study, the estimation of the plastic critical depth is explained clearly, and several analyses are made in different cases. 2 THEORY_ The backfill is a c-0 soil, where c and 0 are the cohesion and the friction angle of the soil, respectively. The soil obeys the Mohr-Coulomb failure criterion. The retaining wall is rigid, and the soil-wall interface has the adhesion cw and the friction angle Sw. Figure 1 shows a soil element in the plane-strain case. There are two families of failure orientations, PA and PB, known as the negative and positive characteristics, and make a stress field. As shown, the stress-characteristics lines make the angle m = p /4 - f / 2 with the orientation of the principal stress ffj [9]. Each point in the soil media has four features, x, z, p and f, where x and z are the coordinates of the point and p is the average stress in the Mohr's circle and f is the angle between ffj and the horizontal axis (Figure 1). If the body forces are zero, the equilibrium equations along the stress characteristics can be written as [9]: 2(p tanf + c)dy + dp = 0 (1) 2(ptan f + c)dy - dp = 0 (2) 2.1 Boundary conditions Figure 2 shows the geometry of the problem. DE is the ground-surface boundary and has the surcharge q. ft is the ground-surface angle with the horizontal axis and 6 is the angle between the wall and the vertical axis. The positive signs of these angles are shown in the figure. kh and kv are the horizontal and vertical pseudo-static earthquake coefficients, respectively. The positive directions of the seismic accelerations are shown in the figure. To solve the problem, the boundary condition of the wall and ground surface must be calculated. Figure 2. The geometry of the problem. The Mohr's circles of stress along the ground surface and wall boundaries are shown in Figure 3. The normal and shear stresses on the ground surface can be written as So = q cosb[(1- kv) cosb - kh sinb] = Afl (3) r0 = q cosb[(1- kv )sinb + kh cosb] = A2q (4) where q is the equivalent surcharge (Eq. (18)) and (1-kv) cos b cos (d + b) A A cos 6 (5) (1- kv) cos b sin(d + b) cos 6 and Figure 1. The orientation of the positive (a+) and negative (a-) characteristics and the Mohr's circle of stress [16]. tand = 1- kv (6) 18. Acta Geotechnica Slovenica, 2016/1 A. Keshavarz: Evaluation of the plastic critical depth In seismic active lateral earth pressure problems using the stress-characteristics method a) b) Figure 3. Mohr's circle of stress of the a) wall and b) ground surface. The radius of the Mohr's circle on the ground surface can be written as (Figure 3b) R0 = p0sinf + ccosf = ^/(0 -p0) 2 + V (7) From Eq. (7), the average stress on the ground surface, Po, is Po s0 + ccosf sinf — tJ(&0 sinf + ccosf )2 — (r0 cosf )2 cos2 f (8) and using the Mohr's circle, we can write ho = p/2 + b - p s0 = P0 +(p0sinf + c cos f)cos2h0 t0 = (p0 sin f + c cos f )sin2h0 The angle f0 can be obtained from Eq. (9) as p0 sin ((5 + b) (9) y0 =-+0.5 ß — d — sin-1 p0sinf + c cos(; (10) Referring to Figure 3a, on the retaining wall boundary hf=yf—p/2-q s f = Pf — (pf sinf + c cos f)cos2hf (11) tf = cw + sf tan= (pf sin f + ccos f )sin2hf The angle f on the wall, f can be found from Eq. (11) as pf sin 6W + cw cos yf=2+9+0.5 —d„ + sin pf sinf + ccosf (12) 2.2 Calculating the plastic critical depth Since the stresses on the left- and right-hand side of point O are different, this point is a singularity point. To obtain the depth of the plastic critical depth without computing the whole network, the singularity point must be solved. If ff>f0, the stresses are continuous everywhere, but when fff0, from Eq. (1), pf can be found as If f=0 : pf = p0 — 2c(yf — A0) else : pf =— c cotf + (p0 +c cot f)exp j—2tanf (pf — y0) (13) But if ff 0 (24) In the static case, when ^=0, the plastic critical depth can be computed without trial and error. When ff >f0, and for the static case, the results of this paper are the same as those of Peng and Chen [13]. To calculate the plastic critical depth in active lateral earth pressure problems, an equivalent surcharge can be used [13]. This uniform equivalent surcharge on the OA boundary (Figure 2) can be written as q = q+gzc (18) where zc is the plastic critical depth and y is the unit weight of the soil. The equivalent surcharge q must be computed such that the normal stress on the wall is zero. When Of = 0, from Eq. (11) and (12) [13] f Vc pf = c tan f + - ic2 - c,.2 cosf (19) If ff>f0 , the value of p0 can be obtained from Eq. (13) as If 0=0 : Po = pf + 2c(yf - y) else : p0 =— c cotf + (pf +c cot f)exp |2tanf(yf —yo ) (20) and if ff pf + ccotf Po =-c cot f + sin2(y f-Wo ) . , -^ ^ sin2(yo - wo where from Eq. (16) Wo = 0.5 yf + y - cos Isinf coslyf - y f - yo (21) (22) 3 RESULTS Based on the algorithm described in the previous section, a computer code was prepared. In this part of the paper different parametric analyses were made for the plastic critical depth, and the results were compared to the results of the other studies. In the static case (kh=kv=0) and for 8w=cw=Q=fi=0, the solution leads to the following equation, which is the simple Rankin formula for the plastic critical depth 2c g zc = — tan p f - + - 4 2 - ^ (25) g Nian and Han [4] developed the Rankin theory to calculate the seismic active lateral earth pressure. They neglected the friction angle and the adhesion of the soil-wall interface and assumed that the wall is vertical (i.e., Sw=cw=Q=0). Their equation for zc is 2c(sinf(1- tan¡3tanS) + yj(1 — tan¡3tanS)2 + 4tan2 S I g (1- K )cos f (1- tan ß tan S)2 + 4tan2S cos2 f (26) Also, Peng [3] developed the Mononobe-Okabe theory for the seismic case by taking into account the effect of the plastic critical depth and proposed the following equation to calculate q zc = 18. Acta Geotechnica Slovenica, 2016/1 A. Keshavarz: Evaluation of the plastic critical depth In seismic active lateral earth pressure problems using the stress-characteristics method _ c cos(0 — ß )cos f + cw sin(aa + ß )sin(aa + 0 — f) q = I — (27) ^kh2 +(1 — kv)2 cos(aa + 0)sin(aa — f — S)cosß where tana sin(f + S — 0) + m0 sin0 cos 0 + ^ 1 — m0 cos (0 + f + S) cos(0 — f — S) + m0sin 0 Cw cos(ß + S) (28) (c + cw )cos(0 — ß)cos f In addition, aa can also be calculated using the following equations: if ccos(q -b)cosf + cw sin(f + b + fi)sin($ + fi) > 0, then sin(f + fi — q) + no sin(f + fi)cos(f + fi) + yjl + no cos(q + f + fi) cos(0 — f — S) + nocos (f + S) cw cos(ß + S) (29) ccos(0 — ß)cosf + cw sinff + ß + S)sin(0 + S) if ccos(0 — b)cosf + cw sin(f + b + fi)sin(0 + fi) = 0, then aa = f + fi A comparison of the results of this paper with the results of the other methods is shown in Table 2. The percentage relative errors between the results of this paper and the other methods are shown in the table as well. In the last row of the table the average percentage errors for all the cases are written. It is clear that the least and the most errors belong to Peng's limit-equilibrium method [3] (average error 3%) and Eq. (25) (average error 38%), respectively. In addition, the depths computed from the stress-characteristics method are lower than those of the other methods. In the static case and for q=9=0, a comparison was made between the results of this study for zc and those of others, as shown in Figure 5. In this case the results of Eq. (25), Nian and Han [4] and Iskander et al. [5] are the same. It is clear that the results of this study are very close to Peng [3] (average difference is about 1.4%). Eq. (25) predicts smaller values of zc (about 15% smaller) and also the results of Lin et al. [7] are about 29% smaller than the results in this paper. To evaluate the effects of the parameters on the plastic critical depth, several parametric analyses were made. Figure 6 shows the effects of the pseudo-static horizontal (kh) and vertical (kv) coefficients on the plastic critical depth zc. The horizontal axis is kh and varies from -0.5 to 0.5. It is clear that zc increases with an increase in kh When kh is positive, its variation has more influence on increasing zc. For example, if kv=0, when kh changes from 0 to -0.5, zc decreases by about 58%, but when kh changes from 0 to 0.5, the increase in zc is about 93%. These differences in the negative values of kv are less than those of the positive values. --this study ..............................Peng [3] --------------Ref. [4-5] ------------------Lin et al. [7] H=8 m c=20 kPa cw=c/2, ^=»/3 0=0 ß= -15° q=0 Y=18 kN/m3 kh=0 K=0 20 0 (deg) 30 Figure 5. A comparison between the results of this study and those of others. m 0 a n Table 2. Comparison of the results of this study with other methods. (0=-10o, ¿3=0, Sw=0/2, cw=0.5c, kh=-0.2, kv=0.1, y=18 kN/m3). 0 (deg) c (kPa) q (kPa) Zc (m) Absolute relative error (%) Ma et al. 2012, [6] Eq. (25) Nian & Han, 2013, [4] Peng, 2012,[3] This study Ma et al. 2012, [6] Eq. (25) Nian & Han, 2013, [4] Peng, 2012, [3] 20 0 10 0 0 0 0 0 0 0 0 0 20 10 0 1.424 1.592 1.315 1.224 1.206 15 24 8 1 20 10 10 0.904 1.032 0.759 0.668 0.651 28 37 14 3 40 10 0 1.68 2.384 1.837 1.442 1.382 18 42 25 4 40 10 10 1.168 1.832 1.281 0.886 0.826 29 55 36 7 20 20 0 2.6 3.168 2.629 2.447 2.412 7 24 8 1 20 20 10 2.072 2.624 2.074 1.892 1.857 10 29 10 2 40 20 0 3.048 4.768 3.674 2.884 2.764 9 42 25 4 40 20 10 2.52 4.208 3.118 2.328 2.208 12 48 29 5 Average error: 16 38 19 3 18. Acta Geotechnica Slovenica, 2016/1 A. Keshavarz: Evaluation of the plastic critical depth In seismic active lateral earth pressure problems using the stress-characteristics method Also, increasing kv leads to an increase in zc. Similar to kh, when kv is positive, the variations are more rapid. The average changes in zc for kv= -0.5 relative to kv=0 are about 22% and 40% in the negative and positive values of kh, respectively. These changes for kv=0.5 relative to kv=0 are 39% and 144% in the negative and positive values of kh, respectively. The results of Peng [3] are also shown in Figure 6. It is clear that the values of results of this study are lower than those of Peng. For the selected parameters shown in the figure, the average relative errors between these two methods are 1% for the positive values of kh and kv and 30% for the negative values of kh and kv. The effects of the soil friction angle (0) and ground-surface slope angle (fi) values on zc are shown in Figure 7. It is clear that zc increases with an increase in 0. The results of the analyses indicate that the soil-wall interface friction angle (Sw) has very little influence on zc and its effect can be ignored. Increasing fi leads to an increase in the plastic critical depth. When fi changes from -30° to 30°, zc increases by about 17%. Figure 8 shows the influence of the soil cohesion and soil-wall interface adhesion on zc. The horizontal axis is the soil cohesion (c) and several graphs have been plotted for different values of cw/c. It is obvious that by increasing c, the plastic critical depth increases. Increasing cw/c also leads to an increase in zc, but its influence is less than that of c. For the assumed parameters indicated in the figure, the difference between the results of cw/c=0 and cw/c=1 for all values of c is about 30%. The change of zc with c is linear. The slope of this -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 kh Figure 6. The effects of the kh and kv values on the plastic critical depth. q=0 0=0 Y=18 kN/m' c=20 kPa cw=0.75c 8=0.75<|> 20 . .30 0 (deg) Figure 7. The effects of the 0 and fi values on the plastic critical depth. 20 30 c (kPa) Figure 8. The effects of the c and cw values on the plastic critical depth. 8 - 4 - q=0 *=«■/7y*/ > 9=0 / / / X : K=K=o / / / 20^ y=18 kN/m3 c=20 kPa c„=0.75c > fi„=0.75<|> ^— <|>=0' ■ 0 6 (deg) Figure 9. The effects of the 6 values on the plastic critical depth. 18. Acta Geotechnica Slovenica, 2016/1 A. Keshavarz: Evaluation of the plastic critical depth In seismic active lateral earth pressure problems using the stress-characteristics method line changes from 0.19 to 0.25 m/kPa when cw/c changes from 0 to 1. Figure 9 shows the influence of the wall angle (9) on zc. The horizontal axis shows the wall angle and changes from -30° to 30°. Different graphs have been plotted for different values of the soil friction angle. We can see that by increasing 9, the value of zc increases. zc varies more rapidly with 9 for the larger values of 0. For example, when 0=45°, by increasing 9 from -30° to 0, zc increases by about 202%. This increase for 0=0, when changing 9 from -30° to 0 it is about 55% and when changing 9 from 0 to 30° it is about 34%. 4 CONCLUSIONS_ The stress-characteristics method has been used to evaluate the plastic critical depth. The most important conclusions from this study are: - The values of the results of the stress-characteristics method described in this paper for the plastic critical depth, zc, are smaller than those of the pseudo-dynamic [6], modified Rankin [4] and modified Mononobe-Okabe [3] methods and are closer to the modified Mononobe-Okabe method. The method of Lin et al. [7] under predicts the plastic critical depth. - Increasing the horizontal and vertical pseudo-static coefficients (based on their assumed directions) can increase zc. The percentage changes relative to the static case can be more than 100%. - The values of zc increase with an increase in the soil friction angle, but the soil-wall interface friction angle has very little influence on zc and can be ignored. - For the assumed direction of the ground-surface angle ft, by increasing it, zc increases. For the sample parameters selected in this paper, by changing ft from -30° to 30°, the average increase in zc is about 17%. - The cohesion of the soil (c) has a considerable effect on zc. The variation of zc with c is linear. The slope of this line increases with an increase in cw/c. - Increasing the wall angle (for the assumed direction of this study) leads to an increase in zc. This effect is greater for larger values of 0. The influence of 9 on zc is more than that of ft. Acknowledgements The author would like to thank the reviewer for very valuable comments and suggestions that were helpful in developing the paper. REFERENCES [1] Shukla, S.K., Gupta, S.K., Sivakugan, N. 2009. Active earth pressure on retaining wall for c-^> soil backfill under seismic loading condition, Journal of Geotechnical and Geoenvironmental Engineering 135, 5, 690-696. [2] Peng, M.X. 2010. Analytical solution of plastic critical depth for active earth pressure on retaining wall (in Chinese), Rock and Soil Mechanics 13, 10, 3179-3183. [3] Peng, M.X. 2012. Coulomb's solution to seismic active earth pressure on retaining walls, Chinese Journal of Rock Mechanics and Engineering 31, 3, 640-648. [4] Nian, T., Han, J. 2013. Analytical solution for Rankine's seismic active earth pressure in c-^> soil with infinite slope, Journal of Geotechnical and Geoenvironmental Engineering 139, 9, 1611-1616. [5] Iskander, M., Chen, Z., Omidvar, M., Guzman, I., Elsherif, O. 2013. Active static and seismic earth pressure for c-^> soils, Soils and Foundations 53, 5, 639-652. [6] Ma, S.J., Wang, K.H., Wu, W.B. 2012. Pseudo-dynamic active earth pressure behind retaining wall for cohesive soil backfill, Journal of Central South University 19, 11, 3298-3304. [7] Lin, Y.L., Leng, W.M., Yang, G.L., Zhao, L.H., Li, L., Yang, J.S. 2015. Seismic active earth pressure of cohesive-frictional soil on retaining wall based on a slice analysis method, Soil Dynamics and Earthquake Engineering 70, 133-147. [8] Sokolovskii, V.V. 1960. Statics of soil media, London, UK: Butterworth. [9] Sokolovskii, V.V. 1965. Statics of granular media, Oxford, UK: Pergamon Press. [10] Reece, A., Hettiaratchi, D. 1989. A slip-line method for estimating passive earth pressure, Journal of Agricultural Engineering Research 42, 1, 27-41. [11] Kumar, J., Chitikela, S. 2002. Seismic passive earth pressure coefficients using the method of characteristics, Canadian Geotechnical Journal 39, 2, 463-471. [12] Cheng, Y. 2003. Seismic lateral earth pressure coefficients for c-^> soils by slip line method, Computers and Geotechnics 30, 8, 661-670. [13] Peng, M.X., Chen, J. 2013. Slip-line solution to active earth pressure on retaining walls, Geotech-nique 63, 12, 1008-1019. [14] Jahanandish, M., Keshavarz, A. 2005. Seismic bearing capacity of foundations on reinforced soil slopes, Geotextiles and Geomembranes 23, 1, 1-25. [15] Keshavarz, A., Jahanandish, M., Ghahramani, A. 2011. Seismic bearing capacity analysis of rein- 18. Acta Geotechnica Slovenica, 2016/1 A. Keshavarz: Evaluation of the plastic critical depth In seismic active lateral earth pressure problems using the stress-characteristics method forced soils by the method of stress characteristics, Iranian Journal of Science and Technology, Transaction of Civil Engineering 35, C2, 185-197. [16] Anvar, S.A., Ghahramani, A. 1997. Equilibrium equations on zero extension lines and their application to soil engineering, Iranian Journal of Science and Technology, Transaction B-Engineering 21, 1, 11-34. [17] Lee, I.K., Herington, J.R. 1972. A theoretical study of the pressures acting on a rigid wall by a sloping earth or rock fill, Geotechnique 22, 1, 1-26. [18] Liu, F.Q., Wang, J. H. 2008. A generalized slip line solution to the active earth pressure on circular retaining walls, Computers and Geotechnics 35, 2, 155-164. 18. Acta Geotechnica Slovenica, 2016/1 MIKRO ODZIVI ZRNATIH MATERIALOV PRI RAZLIČNIH VERTIKALNIH TLAKIH Z UPORABO METODE DISKRETNIH ELEMENTOV Ključne besede mikrostrukture, vertikalni pritisk, tekstura, mikro topologija, usklajevalno število, makro-mikro odnos Izvleček Z uporabo metode diskretnih elementov (DEM) so na sestavljenih ovalih potekali dvoosni tlačni preizkusi, s katerimi so se preučevali odzivi zrnatih materialov na mikro ravni pri različnih vertikalnih tlakih. Numerični vzorci so bili sestavljeni iz 8450 ovalov v pravokotnem okvirju brez prekrivanja. Pripravljeni so bili štirje vzorci različne gostote iz začetno ustvarjenega razstresenega vzorca pri različnih vertikalnih tlakih. Rezultati simulacij podajajo napetostno-deformacijsko-dilatacijsko obnašanje, podobno kot pri opazovanih peskih pri različnih vertikalnih tlakih. Ugotovljen je bil razvoj različnih mikro-parametrov in njihovih medsebojnih odnosov. Ko je vertikalni tlak razmeroma visok, je razlika med koordinacijskim številom in učinkovitim koordinacijskim številom zelo majhna; vendar pa je razlika očitna za nizki vertikalni tlak. Opisana je tudi mikro topologija zrnatega sestava za več pomembnih stanj prestriga. Omeniti je potrebno, da je topološka razporeditev zrnatih materialov odvisna od vertikalnih tlakov. Normalizirano število praznih por je minimalno pri najnižjem vertikalnem tlaku, medtem ko je enako število največje pri najvišjem vertikalnem tlaku. Linearno razmerje je opaziti med normaliziranim številom praznih por in učinkovitim koordinacijskim številom, ne glede na vertikalne tlake. Merjen je razvoj deviatorične teksture za različne vertikalne tlake in predstavljen makro-mikro odnos. Md. Mahmud Sazzad Rajshahi University of Engineering & Technology, Department of Civil Engineering Rajshahi-6204, Bangladeš E-pošta: mmsruet@gmail.com 16. Acta Geotechnica Slovenica, 2016/1 MICRO-SCALE RESPONSES OF GRANULAR MATERIALS UNDER DIFFERENT CONFINING PRESSURES USING THE DISCRETE ELEMENT METHOD Keywords microstructures, confining pressure, fabric, microtopol-ogy, coordination number, macro-micro relationship Abstract Biaxial compression tests were carried out on assemblies of ovals to study the micro-scale responses of granular materials under different confining pressures using the discrete element method (DEM). A total of 8450 ovals were generated in a rectangular frame without any overlap. Four dense samples were prepared from the initially generated sparse sample under different confining pressures. The simulated results yield a stress-strain-dilatancy behaviour similar to that observed in sands under different confining pressures. The evolution of the different microparameters and their inter-relationships are established. When the confining pressure is relatively high, the difference between the coordination number and the effective coordination number is very small; however, the difference is apparent for a low confining pressure. The microtopology of the granular assembly at several important states of shear is also reported. It is noted that the topological distribution of the granular materials is confin-ing-pressure dependent. The normalized void-cell number is a minimum under the lowest confining pressure, whereas the same number is a maximum under the highest confining pressure. A linear relationship is observed between the normalized void-cell number and the effective coordination number, regardless of the confining pressures. The evolution of the deviatoric fabric for different confining pressures is measured and the macro-micro relationship is presented. Md. Mahmud Sazzad Rajshahi University of Engineering & Technology, Department of Civil Engineering Rajshahi-6204, Bangladesh E-mail: mmsruet@gmail.com 1 INTRODUCTION The responses of granular materials are greatly influenced by the confining pressure. This is evident from numerous experimental studies [1-8]. These studies, in general, indicated that the deviatoric stress increases or the angle of the internal friction, f = sin-1 [[ — s2) /(s1 + s2) ,decreases with an increase of the confining pressure, where o1 and a2 are the stresses in the x1- and x2-directions, respectively. For example, Fukushima and Tatsuoka [2] considered extremely low to high confining pressures (5-400 kPa) in triaxial compression tests and indicated that $ does not change a great deal when the confining pressure is low and the change is apparent when the confining pressures are relatively higher. In a similar study, Tatsuoka et al. [3] conducted a series of drained plane-strain compression tests under confining pressures of 4.9 to 392 kPa considering the variation of bedding angles and reported that the dependency of $ is very small when the confining pressure is lower than 50 kPa. Nevertheless, these characterizations have been made from experimental studies based on macroscopic measurements where the micro-scale information is not known. To explore the microscopic responses for different confining pressures, the discrete element method (DEM) [9] can be used. Only a few studies considering the confining pressures using DEM have been reported in the literature. Mirghasemi et al. [10] studied the 16. Acta Geotechnica Slovenica, 2016/1 M. M. Sazzad: Micro-scale responses of granular materials under different confining pressures using the discrete element method effects of confining pressures on the shear strength and dilatancy of simulated rockfill using DEM and demonstrated that the simulated stress-dilatancy behaviour is similar to data obtained from experiments on rockfills. Sitharam [11] studied the effect of confining pressures on the micro-scale responses using DEM and indicated that the increase in the average coordination number and the accompanying decrease in the fabric anisotropy reduce the shear strength at higher confining pressures. It is important to understand the micro-scale information, even from a simpler simulation using DEM, before using this knowledge to develop physically sound continuum models. The objectives of the present paper are: (i) to simulate the macro-mechanical responses of assemblies of oval particles under different confining pressures (0-100 kPa) using DEM and (ii) to carry out a comprehensive study in order to explore the evolution of different micro-quantities under different confining pressures using DEM. 2 NUMERICAL EXPERIMENTS_ In this section, the numerical method used in the simulation, the sample-preparation procedures using ovals and the simulation of the biaxial compression tests are briefly discussed. 2.1 Brief overview of DEM The numerical simulations were carried out using DEM. The basic idea in DEM is simple. Each particle in DEM can make and break contacts with its neighbours. The accelerations of a particle are computed using Newton's second law of motion, as follows: fS > f" tanf (4) ii = £ f. i = 1,2 , ië= £M , (2) (1) where m is the mass, Xj are the translational acceleration components, f are the force components, I is the moment of inertia, 0 is the rotational acceleration and M is the moment of a particle. The velocities and displacements are obtained by integrating the accelerations over time. The increments of the normal and shear displacements are computed by comparing the two successive time increments and used in the force-displacement law to obtain the increments of the normal and shear forces as follows: Af" = k"DU", Afs = ksAus (3) where kn and ks are the normal and shear contact stiffnesses, respectively. Slipping between particles occurs as soon as the following condition is satisfied: where ^ is the interparticle friction angle, fn is the normal force and fs is the shear force. 2.2 Sample preparation Numerical samples consisting of 8450 ovals in eleven different sizes (i.e., widths) ranging from 1 to 2 mm were prepared. A schematic diagram of an oval with reference axes is shown in Fig. 1. In the present simulation the computer program OVAL [12] is used. In OVAL, a simple contact model consisting of two linear springs in the normal and tangential directions is incorporated. The initial sample was created by randomly placing the ovals (height-to-width ratio of 0.60) on grids of a rectangular frame with no contact. The initial sample generated at this stage was very loose. This sample was then compacted isotropically with 15, 25, 50 and 100 kPa in different stages using the periodic boundaries, a boundary condition in which the periodic cells are surrounded by identical cells. A particle that sits astride a periodic boundary has a numerical image on the opposite boundary. A deformation rate tensor is used to apply the global deformation uniformly. The interparticle friction coefficient, defined as m = tan f , was intentionally turned off (set to zero) during the isotropic compaction to densify the assembly. When the target confining stress was reached, setting y to zero in different stages during the isotropic compression, the sample was allowed to adjust with y equal to 0.5 for a few steps before the start of the simulation. Then, y was set to 0.50 [25] during the simulation of biaxial compression tests for all the confining pressures. An isotropically compressed dense sample compacted with 100 kPa with reference axes is shown in Fig. 2. The void ratio of the sample after the isotropic compression for different confining pressures was the same (0.126). •V] x2 height Figure 1. Schematic diagram of an oval with an inclination angle a. 28. Acta Geotechnica Slovenica, 2016/1 M. M. Sazzad: Micro-scale responses of granular materials under different confining pressures using the discrete element method Xi Table 1. DEM Parameters used in the simulations - Figure 2. Isotropically compressed dense sample compressed to 100 kPa with reference axes. 2.3 Simulation of a biaxial compression test Simulations of drained biaxial compression tests were performed using the strain-control condition. During shear, the sample height decreased vertically with a very small strain increment of 0.00002% in each step by keeping the stress in the lateral direction constant (i.e., 15 or 25 or 50 or 100 kPa, whichever applicable). The parameters used in the simulations are presented in Table 1. Note that the coefficients of global-type viscous damping used in the present study (Table 1) are maintained sufficiently small to keep the effect of the numerical damping to a minimum and to provide more stable solutions. The quasi-static condition during the simulation was examined by monitoring a non-dimensional index, Iuf, defined as follows [13]: Iuf — Np ^ (unbalanced forces)2 / Np Nc -x100(%), (5) contact forces)2 / Nc where Np and Nc are the numbers of particles and contacts, respectively. The unbalanced force in Eq. (5) is the resulting force acting on a particle. Lower values of Iuf are desirable because they are associated with a higher simulation accuracy [13] and indicative of the lowest effect of numerical damping. In the present study, Iuf remains reasonably small (the average value is less than 0.4%) during shear, regardless of the confining pressures. DEM parameters Value Normal contact stiffness, kn (N/m) 1 x 108 Shear contact stiffness, ks (N/m) 1 x 108 Mass density (kg/m3) 2650 Increment of time step (s) 1 x 10-6 Interparticle friction coefficient, p 0.50 Damping coefficients 0.05 3 MACRO-MECHANICAL RESPONSES Fig. 3(a) shows the relationship between the deviatoric stress, q = (s1 — s2) / 2, and the axial strain, e1, while Fig. 3(b) shows the relationship between q/p and e1 for different confining pressures, ranging from 15 to 100 kPa. Here, the mean stress, p = (s1 + s2) / 2. Note that q increases with an increase in the confining pressures. The classic behaviour of granular materials such as sand 200 150 ■ C3 % 100 ■ 50 f\ i V i i --15 kPa .......25 kPa ---50 kPa ----lOOkPa V (a) 4 6 £i(%) 10 (b) 0.7 0.6 0.5 0 4 0 3 0.2 0 1 0 0 15 fcPa lOOkPa 25 kPa 50 kPa ■ 0 s 2 4 6 £1(%) Figure 3. (a) Relationship between q and £1; (b) q/p and £1 for different confining pressures. 10 28. Acta Geotechnica Slovenica, 2016/1 M. M. Sazzad: Micro-scale responses of granular materials under different confining pressures using the discrete element method on 4 -7 25 50 75 100 Confining pressure (kPii) 125 Figure 4. Relationship between $ at the peak stress state and the confining pressure. for a dense sample under plane-strain conditions (i.e., hardening is followed by softening) is observed. The relationship between $ at the peak stress and the confining pressure is shown in Fig. 4. Note that the change in $ at the peak stress is negligible when the confining pressure is lower than 50 kPa. Above a confining pressure of 50 kPa, very little decrease in $ is observed. The simulated behaviour reported in Fig. 4 is qualitatively the same as that reported by Tatsuoka et al. [3] for planestrain compression and Fukushima and Tatsuoka [2] for a triaxial compression test using Toyoura sand, in which it was found that the change $ with the confining pressure is very small when the confining pressure is lower than approximately 50 kPa. This confirms the ability of DEM to capture the real behaviour of sands qualitatively using ovals, even under very small confining pressures. The relationship between the volumetric strain, ev, and £j is shown in Fig. 5. ev is defined as £v=dV/V, where dV is the change in volume and V is the initial volume of the sample prior to the shear. A positive value of ev denotes compression, while a negative value denotes dilation. Note that the dilation is suppressed as the confining pressure increases. The sample exhibits only little compression at small strains for a confining pressure of 100 kPa (maximum one), which is followed by a huge dilation. In contrast, the sample exhibits dilation from the beginning of the shear for a confining pressure of 15 kPa (minimum one) due to the lower lateral confinement. Note also that the evolution of ev is almost the same for relatively lower values of the confining pressures (i.e., 15 and 25 kPa), which is the same as reported in Tatsuoka et al. [3]. The relationship between the dilatancy index, DI=-dev/de1, and e1 is shown in Fig. 6 for different confining pressures, where dev is the change in ev and de1 is the change in e1. Note that the relationship is dependent on the confining pressures even for a small strain. -15 kPa .......25 kPa ---50 kPa ----100 kPa -1 1 0 10 2 4 6 8 £l(%) Figure 5. Relationship between ev and £1 for different confining pressures. Figure 6. Evolution of dilatancy index -dev/de1 with £1 for different confining pressures. 4 MICRO-MECHANICAL RESPONSES_ In this section the evolution of different micro-parameters such as the coordination number, the effective coordination number and the slip coordination number and their inter-relationship are presented. 4.1 Coordination number The change in the coordination number, Z, as a function of e1 is shown in Fig. 7. Z is defined as follows [16]: Z = 2Nc N„ (6) A considerable reduction in Z is observed for the initial stage of the simulation due to the reshuffling of the initial fabric regardless of the confining pressures. The 28. Acta Geotechnica Slovenica, 2016/1 M. M. Sazzad: Micro-scale responses of granular materials under different confining pressures using the discrete element method N 0 -15 kPa -------25kPa - - - 50 kPii ----100 kPa S 10 2 4 6 £i(%) Figure 7. Evolution of Z as a function of £j for different confining pressures. 5 - IN Figure 8. Evolution of Z as a function of £j for different confining pressures. behaviour is similar to that observed in other DEM studies [11, 14]. The reduction in Z is the highest for the lowest confining pressure. This is directly linked to the higher dilation under lower confining pressures, as is clear in Fig. 5. The lateral boundaries for lower confining pressures (i.e., 15 and 25 kPa) have less restriction to move outward from the sample, causing relatively more disintegration of the contacts in the lateral direction. 4.2 Effective coordination number The effective coordination number is defined by considering the particles that effectively participate in the load-bearing framework, as reported in Kuhn [15]. The non-participating particles are neglected in computing the effective coordination number. The effective coordination number is defined in a similar way to the coordination number, as follows [15-16]: z=2N N (7) where Nr and Np are the total number of contacts and particles, respectively, that share in the effective load-bearing framework. The change in Z as a function of £j is shown in Fig. 8. Note that the behaviour of Z is similar to that of Z. A comparison between Z and Z for confining pressures of 15 kPa and 100 kPa is shown in Fig. 9 to show their relative importance and comparable evolution during shear. Note that the evolution of Z is higher than Z for both the confining pressures, even though their initial values are same. This indicates that the number of non-participating particles increases with the increase of e1 during shear. Note also that the difference between Z and Z is higher for a confining pressure of 15 kPa than for 100 kPa. IN} K Z (LOO kPa) \ 7,(100 kVa) ~/r - Z ( 15 kPa) 7(15 kPuj 4 6 £l(%) 10 Figure 9. A comparison between Z and Z for confining pressures of 15 and 100 kPa. 4.3 Slip coordination number The slip coordination number can be defined in a similar way to the coordination number, as follows [16]: 2N„, S = - N (8) where Nsj is the total number of sliding contacts. The change in S as a function of e1 is shown in Fig. 10. Note that S reduces rapidly after an initial increase up to the peak, regardless of the confining pressures. The reduction is greater for a confining pressure of 100 kPa than for 15 kPa. Note also that the evolution of S has no similarity with Z and Z up to the peak stress. 4.4 Microtopology The topological distribution of a granular system can be represented as a planar graph by connecting the branch 28. Acta Geotechnica Slovenica, 2016/1 M. M. Sazzad: Micro-scale responses of granular materials under different confining pressures using the discrete element method -] 5 kPa .......25 kPa ---50 kPa ----100 kPa 0 1 I 2 4 10 £i(%) Figure 10. Evolution of S as a function of Ej for different confining pressures. vectors of those particles that effectively participate in the load-bearing framework. Such a topological distribution of a granular assembly, considering only the effective contacts and particles, was proposed by Kuhn [15]. All sorts of non-participating particles are neglected in the planar graph in which each polygonal micro-domain is referred to as a void cell. Following the same approach as reported in Kuhn [15], the topological distributions of granular materials in the initial state prior to shear (e1=0), at the peak stress state and at large strain (e1=10%) for confining pressures of 15 and100 kPa are shown in Fig.11. The number of void cells is a maximum at zero strain, indicating a compact contact network for the initial state of the simulation prior to shear. The sizes of the void cells are also smaller at this stage. When the deformation starts, the number Peak (15 kPa) Peak (100 kPa) £1=10% (15 kPa) £1=10% (100 kPa) Figure 11. Snapshots of the topological distributions of granular materials at ^=0, at the peak stress state and £j=10% for confining pressures of 15 and 100 kPa. 28. Acta Geotechnica Slovenica, 2016/1 M. M. Sazzad: Micro-scale responses of granular materials under different confining pressures using the discrete element method of void cells starts decreasing and the size of the void cells starts getting larger. At the peak stress state, the distributions of the void cells reshuffle and the change is apparent. Many void cells become elongated, almost parallel to the ^-direction due to the disintegration of contacts in the x2-direction. In addition, large voids are observed in several regions. The number of void cells continuously decreases and the size of the void cells further increases as the deformation continues. Large voids are observed at e1=10% for a confining pressure of 100 kPa. The formation of large void cells in several regions is a consequence of excessive contact disintegration due to the rotation of the particles and the collapse of the force chains. Although the void cells appear to evolve in a similar way for all the confining pressures, the difference is evident in Fig. 12, where the number of void cells (Nv) is normalized by the initial number of void cells (Nvi) prior to shear (e1=0) and plotted against e1. Note that the normalized void-cell number is same at zero strain for all the confining pressures; however, it reduces as the deformation progresses. The minimum value Nv /Nvi is observed for a confining pressure of 15 kPa, whereas the maximum value is observed for a confining pressure of 100 kPa. The minimum number of Nv/Nvi for a confining pressure of 15 kPa is associated with the maximum dilation during shear. The relationship between Z and Nv/Nvi for different confining pressures is shown in Fig. 13. The relationship between Nv/Nvi and Z regardless of the confining pressures can be mathematically expressed as follows: 'N.," Z = 3.45 N + 2.1 (9) ei(%) Figure 12. Relationship between Nv / Nvi and ej for different confining pressures. 5 in 4 ■ O 15 kPa 6 O 25 kPa n □ a 50 kPa « a 100 kPa / o p* 0.? 0,6 0.9 Nv/Nvi 1,2 Figure 13. Relationship between Z and Nv / Nvi for different confining pressures. 4.5 Deviatoric fabric The fabric in a granular system is highly disordered on the grain scale. The fabric is usually characterized by contact normals [17-19]. The fabric, Hj, can be quantified in term of contact normals as follows [20]: 1 Nc H ■■ = —Vnana ij Nc c a=1 i, j = 1- 2 (10) where nf is the i-th component of the unit contact normal at the a-th contact. Further characterization of the contact normals based on the strong and weak contacts would be interesting [21-22]. In the present study, a contact is defined as strong if it carries a contact force (f greater than the average contact force fa) and defined as weak if it carries a contact force smaller than, or equal to fa. The average contact force is given by fa = Nc Vfk k=1 N (11) Two additional fabric tensors for strong and weak contacts can be defined similar to Eq. (10) as follows [22]: N, i, j = 1 - 2 , (12) 1 H, = —Vnsns j N ' ) ivc s=1 Hw = — V j Nch N ni n, 1 j i, j = 1-2 , (13) where ns and nW are the i-th component of the unit contact normals for the ,-th strong and w-th weak contact, respectively, N, is the number of strong contacts, Nw is the number of weak contacts and Nc : N, + Nw 2 28. Acta Geotechnica Slovenica, 2016/1 M. M. Sazzad: Micro-scale responses of granular materials under different confining pressures using the discrete element method The evolution of the deviatoric fabric, Hn - H22 , for different confining pressures is shown in Fig. 14 as a function of e1. H11 - H22 is zero at e1=0 and increases with e1 up to a peak, regardless of the confining pressures and reaches the same value at large strain. The evolution of the fabric anisotropy is dominant for the lowest confining pressures. The highest fabric anisotropy with the lowest confining pressures is associated with the highest dilatancy, as observed in Fig. 5. The dilatancy is suppressed by the increasing confining pressures resulting in the suppression of the fabric anisotropy. Similar to H11 - H22 , the evolution of Hs11 — Hs22 and HWi — HW as a function of e1 is also shown in Figs. 15(a) 0.3 0.2 3= ■-f 0.1 0.0 15 kPa 25 kPa 50 kPa / 100 kPa 6 9 £i(%) 12 15 Figure 14. Evolution of Hu - H22 as a function of £1 for different confining pressures. and 15(b), respectively. H1s1 — H222 peaks at a small strain. At a large strain, H1s1 — H222 reaches almost the same value for different confining pressures. The evolution of HW1 — H22 is rather different from that of H1s1 — Hs22 . HW1 — H22 is opposite to H1s1 — H22 near the peak, indicating a reverse fabric evolution of weak contacts, compared to strong contacts. However, H^ — H2W2 approaches to positive values with the increase of ex. The phase-change point of HJ1 — H22 from negative to positive is confining-pressure dependent. At large strain, H^ — H2W2 also reaches the same value for different confining pressures. 4.6 Macro-micro relationship Several approaches in the literature related to the macro-quantity (stress ratio) and the micro-quantity (fabric) computed using the single micro-parameter (contact normal) depending on the varieties of simulation conditions, such as particle shape, stress paths, sample LA^-J 0.3 0.2 - 0 1 - (a) 0.0 100 kPa X\S 50kPa 15 LPa 25 kfa 6 9 £i(%) 12 15 (b) -0.1 - -0.2 -0.3 6 9 £i(%) 12 15 Figure 15. Relationship between a) H^ — H^2 versus and b) HJ1 — H22 versus ^ for different confining pressures. density, etc. [14, 22, 23-24] were reported. These studies indicated that the micro-quantity relates strongly to the macro-quantity when the contact normal vectors of strong contacts are considered in quantifying the fabric tensors. These studies also observed a unique macromicro relationship regardless of the conditions used in their studies. Following a similar approach, the relationship between q/p and (Hn - H22)/(Hn + H22) at different confining pressures is shown in Fig. 16 (a), while the relationship between q/p and (H1s1 — Hs22) / (H1s1 + Hs22) is shown in Fig. 16 (b). Note that the relationship between q/p and (Hn - H22)/ (Hn + H22) is pressure dependent when all the contacts are considered. In contrast, q/p is almost equal to 1 — H 22) / (H11 + H 2 2) for relatively lower confining pressures (i.e., 15-50 kPa) when strong contacts are 28. Acta Geotechnica Slovenica, 2016/1 M. M. Sazzad: Micro-scale responses of granular materials under different confining pressures using the discrete element method 0.7 0.6 t 0.5 0.4 (a) 0.3 0.7 0.6 a, 0.5 0.4 (b) 0.3 0.0 0 1 0.2 0.3 (H11 - H22)/(H11 + H22) jrty a J)Hf &> Y o 15 kPa C- 25 kPa o ¿50 kPa a 100 kPa 0.3 0.4 0.5 0 6 (H11 - H22)/(H11 + H22) 0.7 Figure 16. Relationship between a) q/p and (H11 - H22)/(H11 + H22) versus £1 and b) q/p and (H11 — H22) / (H11 + H22) for different confining pressures. parameters and their inter-relationship for different confining pressures, even from simpler simulations, before using this knowledge in developing physically sound and micro-mechanic-based continuum models. This micro-information may also be helpful for the comprehensive understanding of the complex behaviour of a granular system. Several interesting and important findings of this study are summarized below. (i) The difference between the coordination number and the effective coordination number is very small when the confining pressure is relatively high; however, the difference is apparent when the confining pressure is very low. (ii) The slip coordination number reduces sharply after an initial increase up to the peak stress, regardless of the confining pressures. At large strain, the slip coordination number under low confining pressures takes precedence over the larger confining pressures. (iii) The topological distribution of granular materials is confining-pressure dependent. The minimum value for the normalized void-cell number is observed for the lowest confining pressure, whereas the maximum value is observed for the highest confining pressure. (iv) A linear relationship between the normalized void-cell number and the effective coordination number is observed, regardless of the confining pressures. (v) A single micro-parameter related to the contact normals of all the contacts is not sufficient to establish a unique macro-micro relationship between the stress ratio and the fabric ratio, regardless of the confining pressures. REFERENCES considered. Note also that a unique relationship, regardless of the confining pressures considered in the present study, is not available, even when strong contacts are considered in quantifying the fabric tensors. 5 CONCLUSIONS_ A numerical simulation was carried out to investigate the evolution of microstructures under different confining pressures using the DEM. The simulated stress-strain-dilatancy behaviour is qualitatively similar to that observed in experimental studies using sands for different confining pressures. Different micro-parameters are measured and their inter-relationship is discussed. It is important to understand the evolution of these micro- [1] Ponce, V.M., Bell, J.M. 1971. Shear strength of sand at extremely low pressures. J. SME Div, Proc of ASCE 97(SM4), pp. 625-638. [2] Fukushima, S., Tatsuoka, F. 1984. Strength and deformation characteristics of saturated sand at extremely low pressures. Soils and Foundations 24, 4, 30-48. [3] Tatsuoka, F., Sakamoto, M., Kawamura, F. 1986. Strength and deformation characteristics of sand in plane strain compression. Soils and Foundations 26, 1, 65-84. [4] Yamamuro, J.A., Lade, P.V. 1996. Drained sand behavior in axisymmmetric tests at high pressures. Journal of Geotechnical Engineering 122, 2, 109-119. [5] Yamamuro, J.A., Lade, P.V. 1997. Instability of granular materials at high pressures. Soils and Foundations 37, 1, 41-52. 28. Acta Geotechnica Slovenica, 2016/1 M. M. Sazzad: Micro-scale responses of granular materials under different confining pressures using the discrete element method [6] Alshibli, K.A., Batiste, S.N., Sture, S. 2003. Strain localization in sand: plane strain versus triaxial compression. Journal of Geotechnical and Geoen-vironmental Engineering 129, 6, 483-494. [7] Koseki, J., Yoshida, T., Sato, T. 2005. Liquefaction properties of toyoura sand in cyclic tortional shear tests under low confining stress. Soils and Foundations 45, 5, 103-113. [8] Kumar, J., Madhusudhan, B.N. 2010. Effect of relative density and confining pressure on poisson ratio from bender and extender elements tests. Geotechnique 60, 7, 561-567. [9] Cundall, P.A., Strack, O.D.L. 1979. A discrete numerical model for granular assemblies. Geotech-nique 29, 1, 47-65. [10] Mirghasemi, A.A., Rothenburg, L., Matyas, E.L. 1997. Numerical simulations of assemblies of two-dimensional polygon-shaped particles and effects of confining pressure on shear strength. Soils and Foundations 37, 3, 43-52. [11] Sitharam, T.G. 1999. Micromechanical modeling of granular materials: effect of confining pressure on mechanical behavior. Mechanics of Materials 31, 10, 653-665. [12] Kuhn, M.R. 2003. Smooth convex three-dimensional particle for the discrete element method. Journal of Engineering Mechanics 129, 5, 539-547. [13] Ng, T.T. 2006. Input parameters of discrete element methods. Journal of Engineering Mechanics 132, 7, 723-729. [14] Sazzad, M.M., Suzuki, K. 2012. A comparison between conventional triaxial and plane-strain compression on a particulate system using 3D DEM. Acta Geotechnica Slovenica 12, 2, 17-23. [15] Kuhn, M.R. 1999. Structured deformation in granular materials. Mechanics of Materials 31, 6, 407-429. [16] Sazzad, M.M. 2014. Micro-scale behavior of granular materials during cyclic loading. Particuology. doi:10.1016/j.partic.2013.12.005. [17] Oda, M., Nemat-Nasser, S., Konishi, J. 1985. Stress-induced anisotropy in granular masses. Soils and Foundations 25, 3, 85-97. [18] Nouguier-Lehon, C., Cambou, B., Vincens, E. 2003. Influence of particle shape and angularity on the behavior of granular materials: a numerical analysis. International Journal of Numerical Analytical Methods in Geomechanics 27, 14, 1207-1226. [19] Sazzad, M.M., Suzuki, K. 2010. Micromechanical behavior of granular materials with inherent anisotropy under cyclic loading using 2D DEM. Granular Matter 12, 6, 597-605. [20] Satake, M. 1982. Fabric tensor in granular materials. In Vermeer PA and Luger HJ (eds), Proc IUTAM Symposium on Deform Fail of Granular Materials, Delft, Balkema, pp. 63-68. [21] Radjai, F., Wolf, D.E., Jean, M., Moreau, J.J. 1998. Bimodal character of stress transmission in granular packings. Physical Review Letter 80, 1, 61-64. [22] Sazzad, M.M., Suzuki, K. 2013. Density dependent macro-micro behavior of granular materials in general triaxial loading for varying intermediate principal stress using DEM. Granular Matter 15, 5, 583-593. [23] Antony, S.J. 2000. Evolution of force distribution in three-dimensional granular media. Physical Review E 63, 1, 1-13. [24] Antony, S.J., Momoh, R.O., Kuhn, M.R. 2004. Micromechanical modelling of oval particulates subjected to bi-axial compression. Computation Material Science 29, 4, 494-498. [25] Kuhn, M.R. 2003. Heterogeneity and patterning in the quasi-static behavior of granular materials. Granular Matter 4, 155-166. 28. Acta Geotechnica Slovenica, 2016/1 16. Acta Geotechnica Slovenica, 2016/1 TEMELJNA ŠTUDIJA O UČINKOVITOSTI NA MESTU IZDELANEGA SAMOSTOJNEGA BETONSKEGA PILOTA X PREREZA Izvleček Na mestu izdelani betonski piloti Xprereza (XCC) se uporabljajo, ker imajo višjo nosilnost kot na mestu izdelani betonski piloti krožnega prereza (CCC) z enako površino prereza. Čeprav je bila preučevana nosilnost XCC pilotov, uspešnost samostojnih XCC pilotov, predvsem glede mehanizma prenosa napetosti v odvisnosti od geometrijskih učinkov, še vedno ni povsem jasna. V prispevku je podana primerjava terenskih statičnih obremenilnih preizkusov za XCC in CCC samostojna pilota z enako površino prečnega prereza. Za boljše razumevanje so bile dodatno izvedene tridimenzionalne numerične povratne analize. Izmerjeni in izračunani rezultati kažejo, da ima XCC samostojni pilot približno 25% višjo mejno nosilnost kot CCC samostojni pilot. To je zato, ker ima XCC samostojni pilot približno 20% višji skupni stranski odpor, ki je zaradi 60% večjega obsega prereza pilota in nekoliko manjše stranske odpornosti na enoto površine. V zemljini so se razvili bočni ločni učinki, ki povzročajo neenakomerno razporeditev efektivne normalne in strižne napetosti po obsegu XCC samostojnega pilota. Ugotovimo lahko, da imajo XCC samostojni piloti višjo učinkovitost z vidika privarčevanega materiala v primerjavi s CCC samostojnimi piloti. Yu Wang PLA University of Science and Technology, State Key Laboratory of Disaster Prevention & Mitigation of Explosion & Impact Nanjing 210007, Kitajska E-posta: lgdxwangyu@163.com Yaru Lv PLA University of Science and Technology, State Key Laboratory of Disaster Prevention & Mitigation of Explosion & Impact Nanjing 210007, Kitajska E-posta: yaru419828@163.com Dongdong Zhang (vodilni avtor) PLA University of Science and Technology, College of Field Engineering Nanjing 210007, Kitajska E-posta: zhangdodo_163@163.com Jieying Zhou PLA University of Science and Technology, College of Defense Engineering Nanjing 210007, Kitajska E-posta: zhoujieying666@126.com Ključne besede na mestu izdelan betonski pilot X prereza, terenski statični obremenilni preizkus, tridimenzionalni numerična povratna analiza, mehanizem prenosa napetosti, stranska odpornost, ločni efekt zemljine 16. Acta Geotechnica Slovenica, 2016/1 A FUNDAMENTAL STUDY OF THE PERFORMANCE OF X-SECTION CAST-IN-PLACE CONCRETE SINGLE PILES Keywords X-section cast-in-place concrete pile, field static load test, three-dimensional numerical back-analysis, stresstransfer mechanism, side resistance, soil arching Abstract X-section cast-in-place concrete (XCC) piles are used because they have a higher bearing capacity than circular-section cast-in-place concrete (CCC) piles of the same cross-sectional area. Although the bearing capacity of XCC piles has been studied, the performance of XCC single piles, especially for the stress-transfer mechanism dependent on the geometrical effects, is still not fully understood. This paper reports two comparative field static load tests on an XCC and a CCC single pile of the same cross-sectional area. In addition, corresponding three-dimensional numerical back-analyses are performed to provide a fundamental understanding. The measured and computed results reveal that the XCC single pile has an approximately 25% higher ultimate bearing capacity than the CCC single pile. This is because the XCC single pile has an approximately 20% larger total side resistance, which is caused by a 60% larger pile perimeter and a slightly smaller unit side resistance. Lateral soil arching effects are developed, causing a non-uniform effective normal stress and a shear stress across the circumference of the XCC single pile. It is suggested that XCC single piles have a higher efficiency in terms of material saving compared with CCC single piles. 1 INTRODUCTION In order to improve the bearing capacity of a single pile, it is necessary to either increase its side resistance, end resistance or both [1]. By manipulating the cross-sectional geometry of a pile, its cross-sectional perimeter can be increased without increasing the cross-sectional area (material waste), which, in effect, alters the side resistance. Many studies have been conducted on the behaviour of special cross-sectional piles, such as the barrette [2], the belled pile [3], the pipe pile [4-5], the large diameter cast-in-place pipe pile [6], the tapered pile [7], and the H pile [8-9]. These piles alter the cross-sectional shapes that are designed to achieve the same objective, which is improving the pile performance, particularly the bearing capacity. In recent years a new cross-sectional-shaped pile, the X-section cast-in-place concrete (XCC) pile, has been implemented in China (see Fig. 1). Since 2008, these new XCC piles, constructed as a piled raft system (including pile, gravel and/or geogrid), have been used to reinforce the soft ground for motorways in Nanjing, for example, two linkage sections between the expressway and the Fourth Bridge of the Yangtze River, two sections of the Jiangshan Street, and a section of the 312 National Motorway. The utilization of XCC piles in those projects has resulted in cost savings of more than 20%. Yu Wang PLA University of Science and Technology, State Key Laboratory of Disaster Prevention & Mitigation of Explosion & Impact Nanjing 210007, China E-mail: lgdxwangyu@163.com Yaru Lv PLA University of Science and Technology, State Key Laboratory of Disaster Prevention & Mitigation of Explosion & Impact Nanjing 210007, China E-mail: yaru419828@163.com Dongdong Zhang (corresponding author) PLA University of Science and Technology, College of Field Engineering Nanjing 210007, China E-mail: zhangdodo_163@163.com Jieying Zhou PLA University of Science and Technology, College of Defense Engineering Nanjing 210007, China E-mail: zhoujieying666@126.com 16. Acta Geotechnica Slovenica, 2016/1 Y. Wang et al.: A fundamental study of the Performance of X-sectlon cast-ln-place concrete Single piles To advance these techniques, mainly for the design and construction of XCC piles, comparative field tests and three-dimensional (3D) numerical back-analyses between an XCC piled raft and a circular-section cast-in-place concrete (CCC) piled raft of the same cross-sectional area and configuration have been conducted by Lv et al. [10-12]. The results indicate that, under the same load conditions, approximately 66% and 46% of the applied load is carried by the XCC and CCC pile, respectively, and simultaneously, approximately 45% and 24% of the applied load is taken by the side resistance of the XCC and CCC pile, respectively. In other words, the larger bearing capacity of the XCC piled raft contributes to the cross-sectional geometry, leading to a larger pile perimeter and arching effects. In addition, a modified analytical solution, dependent on the cross-sectional geometry of XCC piles, was derived from the solution of Mindlin and Geddes by Lv et al. [13]. Corresponding parametric studies were conducted to explore the geometrical effects. In addition, the behaviour of the negative skin friction on the surface of XCC piles has been studied and reported by Kong et al. [14]. However, the stress-transfer mechanism of a single pile is quite different from that of a piled raft [15-17]. This is because the applied load is entirely carried by the pile for single piles, but it is carried by the piles and their surrounding soil for piled rafts. Nevertheless, the performance of XCC single piles, especially the stresstransfer mechanism between XCC single piles and their surrounding soil, is still not fully understood. This paper aims at revealing the fundamental stress-transfer mechanism for XCC single piles by field tests and 3D numerical back-analyses. For comparison, a CCC single pile of the same cross-sectional area is involved. The pile-soil interactions were considered by means of setting slip elements at the pile-soil interfaces. 2 BRIEF REVIEW OF XCC PILES_ The cross-section and corresponding parameters of the XCC piles are shown in Fig. 2. The circle drown by the dashed line is defined as the surrounding circle and the arcs cutting into the surrounding circle at four symmetrical positions are defined as the cutting arc. The geometrical parameters of an X-shaped cross-section are 2a, 2R, 9, where 2a is the distance between two adjacent arcs; 2R is the diameter of the surrounding circle; and 9 is the degree of the central angle of the cutting arcs. Thus, the radius of the cutting arcs r can be obtained by: d R2 - a2 - a) r = (i) sin 2 For given geometrical parameters, the cross-sectional area of an XCC pile (Ax) is calculated as: Ax =kR 2 - 2 [ r - sin(0 )r2 ]-: 7T a \ 2 i a \ 2 -2 | — - 2arcsin(—) R - cos 2arcsin(—) R 2 R ) \ R ! Flat face (2) Figure 1. Constructed: (a) XCC single pile, and (b) XCC pile group. Figure 2. Cross-section and parameters of an XCC pile. 40. Acta Geotechnica Slovenica, 2016/1 0 Y. Wang et al.: A fundamental study of the Performance of X-sectlon cast-ln-place concrete Single piles The cross-sectional perimeter of an XCC pile (Cx) is calculated as Cx = 8 arcsin(—)R + 40 r (3) The X-shaped cross-sections are formed to enlarge the pile perimeter by changing the directions of the arcs. Hence, ideally, the side resistance can be increased without decreasing the end resistance. 3 METHODOLOGIES The performance of XCC single piles was studied using field tests and numerical simulations. Two field static load tests were conducted on an XCC single pile and a CCC single pile of the same cross-sectional area to explore the bearing capacity of XCC single piles. Meanwhile, corresponding 3D numerical back-analyses were carried out, not only to analyze the field tests but also to provide a deeper understanding. 3.1 Brief review of field static load tests Two field static load tests were conducted in Nanjing, China. The XCC testing pile was cast in place using typical C20 concrete with dimensions of 2R = 0.53 m, 2a = 0.11 m, 6 = 130° and a pile length (L) = 10.5 m, giving a cross-sectional area of 0.112 m2 (Eq. 2). As a reference, a CCC single pile with the same cross-sectional area (pile diameter D = 0.38 m) was produced. Thus, the perimeters of the XCC cross-section and the CCC cross-section are 1.93 m and 1.19 m, respectively (Eq. 3). The perimeter of the XCC single pile is approximately 60% larger than that of the CCC single pile. The site of the field tests is a relatively flat terrain. Original landscape features are retained and belong to the Yangtze River floodplain topography, according to the geomorphologic unit. The geotechnical properties of the soil profile are shown in Fig. 3. The ground-water table is located at a depth of approximately 2.2 m below the ground surface. Cone penetration tests (CPTs) were conducted to obtain the stiffness of the soil layers. The soil layer from the ground surface to a depth of approximately 3.5 m is soft; this is a fill layer. From a depth of 3.5 m to 8.0 m, the penetration resistance is more than 10 MPa, indicating a stiff fine sand layer. From a depth of 10.5 m to 20.0 m, the penetration resistance is approximately 7 MPa, suggesting a sand layer. Between the stiff fine sand and the sand layer, there is a soft inter-layer with a thickness of 2.5 m, which is silty clay. Some soil samples were retrieved for laboratory testing. The properties of the soil layers, including the water content, the unit weight, the effective friction angle, and the void ratio, are summarized in Table 1. The test instrumentation consisted of the load apparatus and the detecting instruments. The load apparatus included a hydraulic jack and dial indicators. Different pressures, controlled by the load apparatus, were applied to the pile heads using a slow maintained load testing method. The pressure was read from the hydraulic table Moisture content (%) Penetration resistance (MPa) 0 20 40 0 7 14 Shear strength (kPa) 0 40 80 ju 12 g 13 S18 24 ci o D ■ Plate test O Triaxial 0 -6 1-12 JB § £ -18§ 24 Figure 3. Soil layers and properties. 0 6 Acta Geotechnica Slovenica, 2016/1 41 Y. Wang et al.: A fundamental study of the Performance of X-sectlon cast-ln-place concrete Single piles Table 1. Properties of the soil layers. From /m To /m Type of soil Mean water content /% Void ratio Unit weight /(kN/m3) Young's modulus /MPa Cohesion */kPa Effective friction angle /° 0.0 3.5 Fill 29.4 0.96 17.8 6 0 20 3.5 8.0 Fine sand 26.3 0.76 18.9 24 0 20 8.0 10.5 Silty clay 38.6 1.11 17.6 6 0 31 10.5 20.0 Sand 26.1 0.75 18.9 24 0 31 * denotes 1 Pa in ABAQUS and the settlement was recorded using dial indicators with an accuracy of 0.01 mm. Each instrument was calibrated before the tests. More details about the field tests can be found in the literature [11]. 3.2 Introduction of 3D numerical back-analyses In this paper, 3D numerical back-analyses were carried out using the finite-element program ABAQUS [18]. For consistency, a full solid was simulated without taking advantages of the symmetry. In other words, the dimensions of the numerical CCC single pile and the numerical XCC single pile coincide well with those of the corresponding testing pile. Fig. 4a shows the meshes of the 3D pile foundations. Reduced integrated 8-node brick elements were created. Duplicated nodes were used to form a zero-thickness interface between the piles and their surrounding soil, as shown in Fig. 4b, adjacent to which the fine elements were generated. The interface friction angle (S) was estimated using the equation proposed by Randolph and Wroth [19]: 8 = tan[sin+ sin2 ')] (4) where is the effective friction angle. The lateral earth-pressure coefficients (K0) were estimated using the equation: K = 1 - sin t (5) The external boundaries were positioned 5 m (approximately 13D) laterally from the center of the single piles and 20 m (approximately 1.9L) below the ground surface. Since the piles located at a distance of 6D or greater can be considered as a single acting pile [20], the distance chosen (13D) is deemed as sufficient to ensure no boundary effects. The boundary conditions are as follows: the radial curved plane was allowed to move freely in the tangential and z directions, but was fixed in the radial direction; for the bottom plane (i.e., z = 20 m), all of the movements were restrained. A drained analysis was used in the numerical simulations. The piles (typical concrete) were simulated to be 20 m (a) Pile length L= 10.5 m Elato-plastic slip 20 m (b) Pile length L = 10.5 m 10 m Figure 4. Finite-element mesh of: (a) 3D pile, and (b) elasto-plastic slip. linear-elastic, with the Young's modulus being Ep (= 20 GPa) and the Poisson's ratio being v (= 0.2). The soil profile was simulated as an elasto-plastic material using the composite Mohr-Coulomb criterion. Fig. 5 shows the yield surface of the Mohr-Coulomb criterion [18]. The yield function is given by 40. Acta Geotechnica Slovenica, 2016/1 Y. Wang et al.: A fundamental study of the Performance of X-sectlon cast-ln-place concrete Single piles F = Rmcq - p tan^- c = 0 (6) where p is the normal stress; q is the deviatoric stress; and Rmc(©,$) is a measure of the shape of the yield surface in the deviatoric plane and is defined as Rmc = sin(© + x/3)j (V3 cos ) + cos(© + ^/3)tan^/3 (7) where $ is the slope of the Mohr-Coulomb yield surface in the Rmcq-p stress plane, which is commonly referred to as the friction angle of the soils and ranges from 0 to 90°; c is the cohesion of the material; and © is the deviatoric polar angle, which is defined as cos(3©) = r V q3 (8) where r denotes the third stress invariant. To eliminate the influence of apparent cohesion, the values of the cohesion for all of the soil layers were set to zero (1 kPa in ABAQUS). The effective friction angles were given according to the laboratory tests. The Young's moduli of the soil layers were estimated based on the CPT tests. The Poisson's ratios were taken to be 0.35 for all of the soil layers. The groundwater table was set at 2.2 m below the ground surface as the monitored results in field. The soil properties are summarized in Table 1. Rmcq A (a) © = 0 Figure 5. Yield surface of the Mohr-Coulomb criterion in: (a) the meridional plane, (b) the deviatoric plane [18]. To simulate the static load tests, twelve steps including eight loading steps from p = 100 kN to 800 kN and four unloading steps from p = 800 kN to 0 kN were applied to the CCC single pile. A corresponding fifteen steps, including ten loading steps from p = 100 kN to 1000 kN and five unloading steps from P = 1000 kN to 0 kN, were applied to the XCC single pile. The load of each step was applied to the pile heads using linear amplitude. 4 RESULTS AND ANALYSES 4.1 Measured and computed load-settlement relationships The measured and computed pile settlements versus the applied loads are shown in Fig. 6. Since the pile tips were located at the bottom sand layer, the measured settlement curves do not show a well-defined ultimate load. The computed settlement curves illustrate obvious elastic and plastic stages. The plastic stages of the CCC single pile and the XCC single pile start at approximately P = 220 and 310 kN, respectively. In addition, the computed settlement of the CCC single pile is underestimated during the loading stages. Meanwhile, the computed settlement of the XCC single pile is underestimated for the former loading stages, but overestimated for the latter loading stages. The differences between the measured and the computed results could be induced by overestimating the soil stiffness and idealizing the constitutive model. However, good agreements in the trend indicate the validation of the numerical back-analyses. The ultimate bearing capacity that was adopted in this study was obtained using the settlement criterion s/D = 5% [21], where s denotes the pile settlement and D denotes the equivalent pile diameter. Hence, the Applied load, P (kN) Figure 6. Measured and computed load-settlement relationships. 40. Acta Geotechnica Slovenica, 2016/1 Y. Wang et al.: A fundamental study of the Performance of X-sectlon cast-ln-place concrete Single piles measured ultimate bearing capacities of the CCC single pile and the XCC single pile are P = 350 and 500 kN, respectively. Subsequently, using the same criterion, the computed ultimate bearing capacities of the CCC single pile and the XCC single pile are approximately P = 400 and 500 kN, respectively. It is obvious that the computed XCC single pile has an approximately 25% higher ultimate bearing capacity than the CCC single pile of the same cross-sectional area. In other words, XCC single piles have a higher efficiency in the cross-sectional area than the CCC piles. The contributions of the higher bearing capacity of the XCC piles will be explored step by step. 4.2 Load sharing between the side resistance and the end resistance Fig. 7a and 7b show the computed axial forces of the CCC single pile and the XCC single pile, respectively. Both the CCC single pile and the XCC single pile have a similar trend, i.e., the axial force increases with an increasing applied load. Since the maximum loads of 800 kN and 1000 kN were applied to the CCC single pile and the XCC single pile in field tests, respectively, four loading steps (i.e., 200 kN, 400 kN, 600 kN, and 800 kN) are involved in the computed CCC single pile, and five loading steps (i.e., 200 kN, 400 kN, 600 kN, 800 kN, and 1000 kN) are included for the computed XCC single pile in this figure. Additionally, both the CCC single pile and the XCC single pile have another similar trend, i.e., the axial force decreases with the increasing depth. The axial force at the pile heads equals the applied load (P), while the axial force at the pile tips indicates the end resistance (Qn). Therefore, the slope of the axial force is the unit side resistance and the difference between the applied load and the end resistance (P-Qend) is the total side resistance. It was observed that the slope of the axial force of the XCC single pile is smaller than that of the CCC single pile, demonstrating that the side resistance of the XCC single pile should be larger than that of the CCC single pile. In other words, the end resistance of the XCC single pile is smaller than that of the CCC single pile. This is because the XCC single pile has 60% more pile perimeter than the CCC single pile, although it has an identical cross-sectional area. Based on the above description, the load sharing of the end resistance (aenJ) and the side resistance (asije) is obtained using Eq. 9 and Eq. 10, respectively. Fig. 8 shows the proportion of the load sharing of the side resistance and the end resistance with respect to the applied load. No matter what the cross-sectional shapes, aend decreases with the applied load for the initial loading stages, but the trend becomes negative during the later loading stages, resulting in a minimum peak value at approximately P = 300 kN. Meanwhile, asije increases with the applied load for the initial loading stages, but the trend becomes negative for the later loading stages, inducing a maximum peak value at the same load. This phenomenon is governed by the mobilization of the side resistance and the end resistance. It is found that during the initial loading stages, more than 80% of the applied load is carried by the end resistance. This indicates that the applied load is mainly resisted by the pile tips because the trend of the relative motion between the pile and soil has not been developed. Then, asije exceeds 50% at approximately P = 200 kN. After the full mobilization of the side resistance (relative motion is approximately 5 to 10 mm) [22] , aside decreases, but aenj increases again. This is because the end resistance is fully mobilized at approximately s/D = 5%, which is always later than the full mobilization of the side resistance [22]. Axial force, ß(kN) 400 800 Axial farce, Q (kN) 1200 0 P=200kN P=400kN P=600kN P=800kN P=1000kN Figure 7. Computed axial force of: (a) CCC single pile, and (b) XCC single pile. 40. Acta Geotechnica Slovenica, 2016/1 Y Wang et al.: A fundamental study of the performance of X-section cast-in-place concrete single piles 100- 20- CCCpile^) CCCpile^ -XCC pile (o^) - XCCpile (c^.jds) 200 400 600 800 Applied load, P(kN) 1000 Figure 8. Computed load sharing of the side resistance and the end resistance. Qer = ±erL (9) P a.. = 1 - a = side end P O'er. P (10) In addition, it was found that aside is exceeded by aend at P = 400 and 700 kN for the CCC single pile and the XCC single pile, respectively. This illustrates that the side resistance of the XCC single pile plays a more important role in resisting the applied load than the CCC single pile in serviceability conditions [22]. For example, aside of the XCC single pile is always larger than that of the CCC single pile. The average side resistance of the XCC single pile is about 120% that of the CCC single pile. 4.3 Unit side resistance and total side resistance In order to investigate the larger aside of the XCC piles, the unit side resistance and the total side resistance at P = 500 kN are derived in Fig. 9a and 9b, respectively. In this study, the unit side resistance is back-calculated using the axial force. The total side resistance of a cross-section is obtained by the unit side resistance multiplied by the pile perimeter and the effective length of the segment. From the figure it is clear that the unit side resistance of the XCC single pile is smaller than that of the CCC single pile, especially for the lower part of the pile shaft. Below z/L = 0.7, the discrepancy between the two piles increases as the depth is increased. The unit side resistance of the CCC single pile is more than three times that of the XCC single pile at z/L = 0.85. Since the pile perimeter is normalized, the difference is caused by geometrical effects. After multiplying the unit side resistance by the pile perimeter, the total side resistance of the XCC single pile is about 120% that of the CCC single pile. This means that the pile perimeter plays an important role in improving the side resistance of XCC piles. The 60% higher pile perimeter induces an about 20% total side resistance, indicating that XCC single piles have a lower efficiency in the pile perimeter than the CCC piles. To reveal the lower efficiency of the pile perimeter for XCC piles, the shear stress at different locations, i.e., X1, X2, X3, and X4, is derived and shown in Fig. 9c. As a reference, the shear stress estimated by the ft-method [23] is also plotted in the same figure, including ft = 0.1, 0.2, and 0.3. It was found that the shear stress developed on the concave surfaces of the XCC single pile (that can be described by ft = 0.28) is larger than that developed a Unit side resistance (kPa) Total side resistance (kN) Shear stress, Ts(kPa) 200 0 30 Computed: —®— CCC pile -■-XCC pile (XI) —XCC pile (X2) —□— XCC pile (X3) —x— XCC pile (X4) I Calculated: ß = 0.1 ß = 0.2 - ß = 0.3 0.0 -0.2 -0.4 ■Ja & T3 -0.6: T3 u N o £ 0.8 1.0 Figure 9. Computed: (a) unit side resistance, (b) total side resistance, and (c) non-uniform shear stress under P = 500 kN. Acta Geotechnica Slovenica, 2016/1 45. Y. Wang et al.: A fundamental study of the Performance of X-sectlon cast-ln-place concrete Single piles on the cylindrical surfaces of the CCC single pile (which can be described by ft = 0.20), which is in turn larger than that developed on the outward flat surfaces of the XCC single pile (which can be described by ft = 0.18). This observation provides two findings: firstly, for XCC piles, the ft-method always overestimates the shear stress of the flat surfaces but underestimates the shear stress of the concave surfaces; secondly, the non-uniform shear stress induces a lower efficiency in the pile perimeter of the XCC single piles. In addition, at the intersection of the fill and fine sand (z/L = 0.33) and the fine sand and the silty clay (z/L = 0.76), the shear stress is unusual due to the special characteristics of the pile-soil interfaces. 4.4 Effective normal stress Since the magnitude of the shear stress is closely related to the effective normal stress, the effective normal stress in the surrounding soil elements is derived on the basis of polar coordinates. Fig. 10 shows the effective normal stress distribution for the section z/L = 0.5 at P = 500 kN. It was found that the effective normal stress acting on the concave surfaces of the XCC pile is the largest of all sections, resulting in the largest shear stress at position X1 (see Fig. 9c). Meanwhile, the effective normal stress acting on the flat surfaces of the XCC pile is smaller than that on the cylindrical surface of the CCC pile and the concave surfaces of the XCC pile, resulting in a small shear stress of X4. This is because specific soil arching effects are developed across the circumference of the XCC piles, inducing stress redistribution. This is the fundamental reason for the larger ultimate bearing capacity of XCC single piles. In order to illustrate the distribution of the effective normal stress along the depth, the effective normal stress acting on the pile shafts along the depth under P = 500 kN is shown in Fig. 11. Different locations, i.e., X1, X2, X3, and X4, are also involved. It was found that no matter whether it is the CCC single pile or the XCC single pile, the effective normal stress increases along the upper depth and then decreases along the lower depth. There are peak values at the intersections between two adjacent soil layers (i.e., z/L = 0.33 and 0.76). This trend agrees well with the distribution of the non-uniform shear stress (see Fig. 9c). Governed by the geometrical effects, the effective normal stress acting on X1 is the largest, but that acting on X3 is the smallest. In other words, the nonuniform shear stress on XCC pile shafts is induced by the non-uniform effective normal stress, which in turn results from the stress concentration and the principal stress rotation. There is an interesting phenomenon that the effective normal stress of X4 is obviously larger than that of X3, but the shear stresses of X3 and X4 are almost identical. This result demonstrates that the flat surfaces are easier to fully mobilize than the concave surfaces. Effective normal stress (kPa) 0 20 40 60 80 1.0 Figure 11. Computed effective normal stress along the depth under P = 500 kN. 0' 35 * o W tl o 11 tf *90 0 —Circular pile 180 —XCC pile Figure 10. Computed effective normal stress at z/L = 0.5 under P = 500 kN. 40. Acta Geotechnica Slovenica, 2016/1 5 CONCLUSIONS Two field static load tests and the corresponding 3D numerical back-analyses were performed on an XCC single pile and a CCC single pile of the same cross-sectional area to investigate the performance of XCC single piles with respect to the geometrical effects. Based on the measured and computed load-settlement relationships, it is observed that the XCC single pile has an approximately 25% higher ultimate bearing capacity than the CCC single pile. This is mainly attributed to the geometrical effects of XCC piles, including a larger pile perimeter and a special stress-transfer mechanism, resulting in a large total side resistance. Although the unit side resistance of the XCC single pile is slightly smaller than that of the CCC single pile under Y. Wang et al.: A fundamental study of the Performance of X-sectlon cast-ln-place concrete Single piles the same applied load, the total side resistance of the XCC single pile is approximately 120% that of the CCC single pile. This is caused by two opposite factors: firstly, the perimeter of the XCC single pile is approximately 160% that of the CCC single pile; secondly, the unit side resistance acting on the XCC single pile is smaller than that on the CCC single pile due to geometrical effects. A non-uniform effective normal stress was observed across the circumference of the XCC singe pile, leading to a non-uniform shear stress. Thus, the shear stress acting on the concave surfaces of the XCC single pile is larger than that on the cylindrical surface of the CCC single pile, which is larger than that on the flat surfaces of the XCC single pile. This is the special stress-transfer mechanism of XCC single piles, which is induced by soil arching effects around the X-shaped cross-section. XCC single piles have a higher efficiency in the cross-sectional area (material saving) than CCC single piles of the same cross-sectional area since they have a higher bearing capacity. Meanwhile, XCC single piles have a lower efficiency in the cross-sectional perimeter than CCC single piles of the same cross-sectional area because the increment of the bearing capacity (25%) is smaller than the increment of the pile perimeter (60%). Acknowledgement The authors acknowledge the support of the National Nature Science Foundation of China (51408607), the Central Universities fund operating expenses (No: 2013B01214). REFERENCES [1] Poulos, H.G., Davis, E.H. 1980. Pile foundation analysis and design. New York Wiley. [2] Gifford, A.B., Green, G.E., Buechel, G.J., Feldman, A.I. 1986. In situ tests and design of a cylinder pile wall. Use of in situ tests in geotechnical engineering (GSP 6) (ASCE), New York, pp. 569-587. [3] Ng, C.WW., Yau, T.L., Li, J.H., Tang, WH. 2001. Side resistance of large diameter bored piles socketed into decomposed rocks. Journal of Geotechnical and Geoenvironmental Engineering (ASCE) 127, 8, 642-657. [4] Lee, J.H., Salgado, R., Paik, K. 2003. Estimation of load capacity of pipe piles in sand based on cone penetration test results. Journal of geotechnical and geoenvironmental engineering (ASCE) 129, 5, 391-403. [5] White, D.J. 2002. An investigation into the behaviour of pressed-in Piles. PhD dissertation, University of Cambridge. [6] Ding, X.M., Zheng, C.J., Liu, H.L. 2014. A theoretical analysis of vertical dynamic response of large-diameter pipe piles in layered soil. Journal of Central South University 21, 3327-3337. [7] Ghazavi, M. 2007. Analysis of kinematic seismic response of tapered piles. Geotechnical and Geological Engineering 25, 1, 37-44. [8] William, G.D., Thomas, S., Sarah, A., John, D., Christopher, L. 2010. Field-measured response of an integral abutment bridge with short steel H-piles. Journal of Bridge Engineering 15, 1, 32-43. [9] Yang, J., Tham, L.G., Lee, P.K.K., Yu, F. 2006. Observed performance of long steel H piles jacked into sandy soils. Journal of Geotechnical and Geoenvironmental Engineering (ASCE) 132, 1, 24-35. [10] Lv, Y., Ding, X., Liu, H. 2011. In-situ tests on cast-in-place concrete X-section pile for bearing capacity of single-pile composite foundation. Geotechnical Special Publication No. 220 ASCE, Changsha, 39-47. [11] Lv, Y., Liu, H., Ding, X., Kong, G. 2012. Field tests on bearing characteristics of X-section pile composite foundation. Journal of Performance of Constructed Facilities (ASCE) 26, 2, 180-189. [12] Lv, Y., Liu, H., Ng, C.W.W., Ding, X., Gunawan, A. 2014. Three-dimensional numerical analysis of the stress transfer mechanism of XCC piled raft foundations. Computer and Geotechnics 55, 365-377. [13] Lv, Y., Liu, H., Ng, C.W.W., Gunawan, A., Ding, X. 2014. A modified analytical solution of soil stress distribution for XCC pile foundations. Acta Geotechnica 9, 3, 529-546. [14] Kong, G., Zhou, H., Liu, H., Ding, X., Liang, R. 2014. A simplified approach for negative skin friction calculation of special-shaped pile considering pile-soil interaction under surcharge. Journal of Central South University 21, 3648-3655. [15] Clancy, P., Randolph, M. F. 2001. Analysis of piled raft foundations using a variational approach. International Journal for Numerical and Analytical Methods in Geomechanics 17, 12, 849-869. [16] Poulos, H.G. 2001. Piled raft foundations-design and applications. Géotechnique 50, 2, 95-113. [17] Kalpakci, V., Ozkan, M.Y. 2012. A simplified approach to the settlement estimation of piled rafts. Acta Geotechnica Slovenica 1, 77-85. [18] ABAQUS standard user's manual, Ver. 6.6. 2006. Hibbitt, Karlsson & Sorensen. Inc., Pawtucket, R.I. [19] Randolph, M.F., Worth, C.P. 1980. Application of the failure state in undrained simple shear to the shaft capacity of driven piles. Géotechnique 31, 1, 143-157. [20] Lee, C.J., Chen, C.Z. 2002. Negative skin friction on ground piles. Proc. Int. Conf. Physical Modeling in Geotech., St. John's Newfoundland, Canada, 679-684. [21] O'Neill, M.W., Reese, R.C. 1999. Drilled shaft: construction procedures and design methods. Federal Highway Administration, Washington, D.C. [22] O'Neill, M.W. 2001. Side resistance in piles and drilled shafts. Journal of Geotechnical and Geoenvi-ronmental Engineering (ASCE) 127, 1, 3-16. [23] Meyerhof, G.G. 1976. Bearing capacity and settlement of pile foundations. Journal of Geotechnical and Geoenvironmental Engineering 102, 3, 197-227. 40. Acta Geotechnica Slovenica, 2016/1 OBNAŠANJE ZEMLJINE OJAČANE S PASIVNIMI SIDRI IN BRIZGANIM BETONOM TER SIDRANIH ZIDOV NA PODLAGI PODATKOV TERENSKEGA MONITORINGA V RAZLIČNIH POGOJIH TAL V ISTANBULU Izvlecek y študiji so, iz več kot 90 terenskih opazovanj, zbrani in analizirani podatki o bočnih premikih sten 28 globokih izkopov v Istanbulu. Najbolj pogosto uporabljeni podporni zidovi za izvedbo globokih izkopov v Istanbulu so zidovi iz pasivnih sider in brizganega betona (SNPWs) in kontinuirane pilotne stene (CPWs). Druge vrste podpornih zidov vključujejo sidrane stene iz mikro pilotov (MPWs), diafragme (DWs), betonske stene (CWs) in pilotne stene s prekrivanjem pilotov (SPWs). Analize so bile izvedene za podatke za zidove iz pasivnih sider in brizganega betona ter podatke za sidrane zidove. Vrednosti maksimalnih bočnih pomikov zidu Sh,max za projekte iz pasivnih sider in brizganega betona se splošno gibljejo od 0.05%H do 0.35%H, s povprečno vrednostjo 0.20%H. Medtem, ko se vrednosti maksimalnih bočnih pomikov zidu Sh,max za projekte s sidranimi pilotnimi stenami splošno gibljejo od 0.043%H do 0.32%H, s povprečno vrednostjo 0.155%H. Nekateri dejavniki, ki vplivajo na deformacije globokih izkopov, kot so gostota pasivnih sider, kvocient vpetja stene in togost stene, so bili proučevani na osnovi podatkov terenskega opazovanja. Premiki zidu so bili primerjani tudi z rezultati opazovanj bočnih premikov nekaterih primerov varovanja izkopov iz celega sveta. Dogan Qetin Yildiz Technical University, Civil Engineering Department Davutpasa Campus, Esenler, Istanbul, Turcija E-posta: dogancetin56@hotmail.com Ključne besede globoki izkop, premik zidu, sestava tal, zaščita iz pasivnih sider in brizganega betona, sidrani zid 16. Acta Geotechnica Slovenica, 2016/1 PERFORMANCE OF SOIL-NAILED AND ANCHORED WALLS BASED ON FIELD-MONITORING DATA IN DIFFERENT SOIL CONDITIONS IN ISTANBUL Dogan Qetin Yildiz Technical University, Civil Engineering Department Davutpasa Campus, Esenler, Istanbul, Turkey E-mail: dogancetin56@hotmail.com Keywords deep excavation, wall displacement, soil formation, soil-nailed wall, anchored wall Abstract In this study, data on the lateral wall displacements of 28 deep excavations in Istanbul soil from over 90 field monitors were collected and analysed. The most commonly used retaining walls for deep excavations in Istanbul are soil-nailed shotcrete walls (SNPWs) and contiguous pile walls (CPWs). Other types of retaining walls include anchored micropile walls (MPWs), diaphragm walls (DWs), concrete walls (CWs) and secantpile walls (SPWs). The analyses were performed under two main headings: analysis of soil-nailed walls' data and analysis of anchored walls' data. The data of S^max generally range from 0.05%H to 0.35%H with an average value of 0.20%H in the soil-nailed projects. The data of Sh,max generally range from 0.043%H to 0.32%H with an average value of 0.155%H in anchored pile wall projects. Some factors affecting the deformation of deep excavations, such as nail density, embedded wall ratio and wall stiffness, were studied based on field-monitoring data. Wall movements were also compared with observations in some case histories from around the world. 1 INTRODUCTION Due to the city's rapid economic growth, a number of deep excavation projects for high-rise buildings, shopping malls and metro systems have been carried out in Istanbul during the past two decades [1]. Estimates of the wall movements and ground deformations induced by deep excavations are critically important in urban areas in order to be able to provide safety for adjacent structures and buildings [2, 3]. Understanding the behaviour of excavation bracing is a complicated issue because of the number of factors affecting the performance of wall systems, such as soil type, base stability, compression and rebound, consolidation, wall-system stiffness, construction procedures, and workmanship [3, 4]. Any of these factors can contribute to the overall movement of a supported excavation. Therefore, great concern has been shown with respect to deep excavations since the report initiated by Peck [5]. Numerous researches have been performed to understand and evaluate the behaviour of retaining walls. Because of the difficulties in making a direct analysis of the performance of wall systems and the behaviour of ground deformation [3,6,7], case studies have been a better way to obtain information about deep excavations and retaining walls. Case studies performed by Ma et al. [3], Hwang and Moh [6], Finno and Bryson [8], Ou et al. [9], Hou et al. [10], Wong and Chou [11] can be given as examples. 16. Acta Geotechnica Slovenica, 2016/1 D. Qetin: Performance of soil-nailed and anchored walls based on field-monitoring data in different soil conditions in Istanbul Databases prepared by Goldberg et al. [12], Mana and Clough [13], Clough and O'Rourke [14], Ou et al. [15], Wong and Patron [16], Carder [17], Fernie and Suckling [18], Wong et al. [19], Hsieh and Ou [20], Long [21], Yoo [22], Moormann [23] and Leung and Ng [24] present comprehensive information about the performance of in-situ walls constructed in urban areas in terms of different soil conditions, adjacent structures and utilities. In this study, data from field monitors on the lateral displacement of the walls from some 28 deep excavations in Istanbul soil were collected. The notation for the symbols and parameters used in this study is shown in Figure 1. The magnitudes of the lateral displacement for different types of retaining systems were compared with each other and also compared with relevant case histories worldwide. Furthermore, some factors affecting the deformation of deep excavations were studied. Figure 1. Definition of the wall and support parameters in a deep excavation. H is the excavation depth, Hemb is the embedded wall length, sv is the vertical distance between the anchors or soil-nailing levels, Hw is the depth of the ground-water level from the surface, Sh,max is the maximum wall displacement and ft is the wall inclination from the vertical in Fig.1. 2 GENERAL SOIL CONDITIONS IN PROJECT FIELDS A very comprehensive and detailed geotechnical survey carried out by the Metropolitan Municipality of Istanbul between 2005 and 2009 for very large areas of both the Asian and European sides of the city [25, 26] was used to explain the soil profile and general soil properties in the project fields. As a result of that intensive work, 24 different local soil formations were identified on both the Asian and European sides. If the project fields are placed on the map of the microzonation work, it can be seen that 7 of the 24 local formations were encountered in deep-excavation projects. These local formations can be called Fill, Alluvium, Gungoren, Gurpinar, Bakirkoy, Cukurcesme and Trakya. The alluvium in the investigated areas mainly consists of clay-type materials of soft-to-medium stiffness. Sand layers are also present at different depths. The Gungoren formation is composed of clay and sand layers that are very stiff and dense. The over-consolidated clay layers in this formation show claystone properties. The Gurpinar formation generally consists of consecutive clay-claystone and/or sand-sandstone layers. The Bakirkoy formation contains limestone layers, but this formation also has clay-sand layers within the limestone layer. The Cukurcesme formation in the investigated areas is mostly composed of gravel and/or sand of different fractions and forms. The Trakya formation is the general name of the claystone-sandstone-mudstone in Istanbul. The upper level of this formation can be seen mostly as highly fractured and/or weathered. Therefore, the upper levels in the Trakya formation can be defined as clay or sand, depending on the main fraction of the material. Figure 2 shows the borders of the investigation areas. The investigation work can be classified as in-situ tests and laboratory tests. According to the soil-investigation works carried out by the related department of the Metropolitan Municipality of Istanbul, the ground-water level is generally between 1 m and 20 m from the ground surface in the investigated area depending on the altitude of the point at which the measurement of the ground-water level was made. 2.1 In-situ Tests The geological investigation and geophysical measurements were carried out as in-situ tests. The geological investigation consists of normal borehole, deep borehole, liquefaction borehole, landslide borehole and CPT. Table 1 shows the content of the geological survey. Table 2 shows the content of the geophysical investigation. 50. Acta Geotechnica Slovenica, 2016/1 D. Qetin: Performance of soil-nailed and anchored walls based on field-monitoring data in different soil conditions in Istanbul Table 1. Content of the geological survey. Type of investigation work Number of borehole Number of borehole points for Asian Side points for European Side Total quantity of boreholes for Asian Side (m) Total quantity of boreholes for European Side (m) Normal Borehole 4426 2830 98681 86840 Deep Borehole 122 27 11594 4201 Liquefaction Borehole 433 764 6683 12344 Landslide Borehole 10 608 227 18144 CPT Test ? 663 4076 8769 Table 2. Content of the geophysical investigation. Number points measured for Number points measured for Measurement method Asian Side European Side Total quantity (m) Seismic refraction 4132 2762 435093 Seismic reflection - - 34105 PS-Logging 504 201 10997 Array microtremor 30 30 - Electric resistance 80 2625 - Figure 2. Location of the projects and the border of the geotechnical investigation area in Istanbul. 2.2 Laboratory Tests Laboratory tests included whole basic soil tests such as Atterberg limits, sieve analysis, unaxial and triaxial compressive strength tests, natural water content tests, hydrometer tests and consolidation tests. According to the reports of the microzonation works [25, 26], for natural water content tests ASTM D2216, for sieve analysis ASTM D422, for hydrometer tests ASTM D4221, for Atterberg Limits tests ASTM D4318, for unconfined compression tests ASTM D2166 and for Triaxial Compression tests ASTM D2850 standards were followed. Table 3 shows the content of the laboratory tests. 50. Acta Geotechnica Slovenica, 2016/1 D. Qetin: Performance of soil-nailed and anchored walls based on field-monitoring data in different soil conditions in Istanbul Table 3. Content of the laboratory tests. Side Natural water content tests Determination of the grain size distribution Sieve analysis Hydrometer Tests Atterberg Limits Unaxial Compressive Tests Triaxial Tests Consolidation Asian 29072 29071 6867 21608 572 214 1339 European 53938 53938 124 46432 1120 462 2315 Table 4. Average soil properties of the local soil formation. Particle size distribution Local formation name Definition of the formation Wn (%) WL (%) Wp (%) Ip (%) Clay and Silt (%) Sand (%) Gravel (%) Cu (kPa) Fill New and old fill material 23.8 46.5 13.2 33.3 48.4 28.3 23.3 85 Alluvium Sandy soft to stiff clay 32.2 50.3 13 37.3 68.1 26.4 5.5 69 Trakya Mudstone, sandstone, claystone (from fractured to extensively fractured) 14.6 35.5 15.3 20.2 35.2 37.7 27.1 93 Gurpinar Alternating sandy stiff clay-claystone 24.6 56.1 15.7 40.4 73.6 23.4 3 127 Gungoren Soft-to-medium sandy clay 28.9 60.6 16 44.5 79.1 17.7 3.2 98 Bakirkoy Limestone with clay strip 24.1 47.1 15.2 31.9 62.8 18 19.2 78 Cukurcesme Clayey gravely sand 17.8 41.3 13.5 27.8 36.6 54.2 9.2 164 As a result of a comprehensive geotechnical investigation, 24 different local soil formations were identified on the Asian and European sides of Istanbul. As can be seen from Figure 2, most of the projects were carried out bordering the geotechnical investigation area. Therefore, the soil formations encountered in the project files were identified using the results of the geotechnical investigation. Selected projects were conducted in 7 different local soil formations. The properties of the soil formation and a brief explanation of the local soil formation are given in Table 4. 3. SUMMARY OF INFORMATION FROM COLLECTED DATA For the database, data from 28 individual projects in Istanbul were collected. Almost all of these projects were carried out in the past two decades. The data were mainly extracted from conference proceedings, PhD theses and a company database that contains a lot of valuable unpublished data. In each case, the measured wall displacement was analysed in terms of the wall type, the support system and its geometric configuration, the embedded wall ratio, the soil-nailing density, the wall stiffness and the soil type. The most common method for measuring wall deflection is the use of vertical inclinometers, instruments for measuring the relative horizontal displacements through a borehole casing placed behind the wall. The bottom must be embedded to the zero displacement zone. An inclinometer is read using a probe. The probe measures the tilt, which can be converted to a horizontal movement. The inclinometer probe is connected to a power source and readout unit to enable the measurements. The electrical cable linking the probe to the readout device is usually marked in 0.5-m increments so the shape of the casing can be measured at consistent depths. The measurements are taken starting at the bottom of the inclinometer. Subsequent readings are made of the casing as the probe is raised incrementally, usually at 0.5-m intervals, to the top of the casing. This process is conducted shortly after the casing is installed to determine the initial shape of the casing, i.e., to obtain the zero reading. The difference between the zero and the subsequent readings is used to determine the change in the shape and position of the initially vertical casing [27]. Two sets of casing grooves allow the inclinometer probe to be oriented in either of two planes set at 90° to each other. Therefore, the horizontal components of movement, both transverse and parallel to any assumed direction of movement, can be computed from the inclinometer measurements [28]. The measurement accuracy is generally to hundredths of a mm. The most commonly used retaining walls for deep excavations in Istanbul are soil-nailed shotcrete walls (SNPWs) and contiguous pile walls (CPWs). Other types of retaining walls include anchored micropile walls (MPWs), diaphragm walls (DWs), concrete walls (CWs) and secant pile walls (SPWs). Summarized information about the projects is given in Table 5. 50. Acta Geotechnica Slovenica, 2016/1 D. Qetin: Performance of soil-nailed and anchored walls based on field-monitoring data in different soil conditions in Istanbul ►B-axls vA-axis Suction ji-x (a) Figure 3. (a) Principles of an inclinometer. Configuration of inclinometer parts. (b) Illustration of inclinometer process.[28] (b) A L"Li-LLLI almr.mL'nl of guide casing (exaggerated) - Table 5. Summary information of the projects. No. Project District Name Support .. Soil , H Hw „ Wall type formation System profile name Hemb Sv Sh n (m/ (m) (m) (m) (m) (m) m2) EI (kNm2 n /m) h max Reference (mm) Baumax , ., 1 Avcilar AVM Multi anchored 80 cm Clayey diameter sand-bored pile Hard clay 7.9 Gung-oren 15.8 6.7 2.4 312500 4 25.3 18.9 36.3 Unpublished data Karabulut 2 Textile Kagith-Factory ane Project Multi an-chored+ strut 65 cm diameter Silty soft secant clay pile Alluvium 8 -2 2 1.2 172300 2 18.4 Unpublished 30 data 21.6 3.7 49.7 Busi- Bahce-lievler Multi anchored 65 cm Clayey sand-clay Gung-oren 21.6 3.7 68.5 Unpub- 3 nessPort diameter 21.6 - 3.7 2.5 1.8 - 115000 5 43.2 lished AVM bored pile 19 4.1 34.2 data 18 4.5 19.8 4 Taksim Sisli Multi 25 cm diameter Mud-stone-sandstone Trakya 19.65 - 1070 2 19.65 Cetin et Residence anchored micro pile 18.70 16.8 al. [29] Multi anchored 25 cm Fractured sandstone 26 2.5 43.3 5 Bomonti Sisli diameter micro pile Trakya 22.8 2.1 2.5 1.5 - 2350 2 42.5 Unpublished Soil nailing Shotcrete Fractured wall sandstone Trakya 22 1.5 1.5 5.5 1 14.60 data 25 45 Tekfen Soil nail- 6 _ Levent lowers ing Highly Shotcrete fractured 24 wall gray-wacke Trakya 20 1.5 2 2.85 14.4 10 19.50 24 9.75 19.2 " Yilmaz et - al. [30] 7 Otoport Esenyurt Multi anchored 80 cm Clayey diameter sand-san-bored pile dy clay Gung-oren 25 7.2 21 -6 2.3 2 312500 2 51.7 Unpublished data Setat 8 2002 LeVent Soil nailing Shotcrete wall Dense fill Fill 9.50 10 10 13 7.6 1.5 2 2.4 10 10 _ Yilmaz et al. [30] 22.1 1 6 5 6 4 50. Acta Geotechnica Slovenica, 2016/1 D. Qetin: Performance of soil-nailed and anchored walls based on field-monitoring data in different soil conditions in Istanbul Table 5. Summary information of the projects (Continued). No. j District Support Name System Wall type Soil profile Soil formation H Hw Hemb Sv Sh n (m/ (m) (m) (m) (m) (m) m2) EI (kNm2 n /m) h, max Reference (mm) T. , „ . 100 cm .y Cuku- Ideal Gazios- Multi .. sand . . . . diameter . . rcesme Hospital manpasa anchored overlying bored pile Trakya sandstone 29.2 4.4 2.92.5-2 1.2 699850 1 Unpub-18.9 lished data Metal Co. 10 Factory Catalca Project Multi anchored 65 cm diameter secant pile Sandy gravely ? clay 9.25 5.25 6.75 2.5 1.8 3.75 1.7 123240 2 16.2 Unpublished 4.1 data 11 B;K FUya Besiktas Complex Soil nail- tag Exten-Shotcrete sively wall fractured sandstone Trakya 18.50 25 32.5 3.4 1.5 1.8 5.3 ? 3 4.9 14.8 22.5 Keskin ~ [1] Saglik 12 HotUep Bey0glu Project Single prop 65 cm diameter secant pile 7.4 Alluvium Alluvium 6.1 5.9 172300 2 7.9 Unpublished data Erdem „ .. Multi 13 . . Gunesli , , Hospital anchored 65 cm diameter bored .l Medium- Gung-hard clay oren reinforced concrete wall 16.3 16.3 0.91.8 122570 2 45.5 13.7 Unpublished data Filiz Con-14 struction íkitelli Co. Multi anchored 80 cm diameter bored pile Medium- Gung-hard clay oren 18.2 312500 4 Unpub-7.95 lished data Besler 15 Co. Ware- Pendik house Soil nailing Shotcrete wall Extensively fractured Trakya grey-wacke 14.7 16.2 18.4 2.3 2.4 2.7 10.3 17.8 Keskin 12.9 [1] Ozbas Cons. Co. 16 Kadikoy Residence Project Multi anchored 25 cm diameter micro-pile+ 40 cm concrete wall Highly fractured sandstone overlying Highly fractured limestone Trakya Bakirkoy 22.9 25.9 25.9 22.7 9.1 2.1 1.8 233940 4 30.7 Unpub- - lished 40.3 data 35.7 10 1.7 27 10 2.8 22 17 Isti- Istinye-nyepark Sariyer Soil nailing Shotcrete wall Extensively fractured Trakya gray-wacke 12 12 14 1.5 2 16 18 20 22 2.9 2.7 3.1 3.2 3.4 ? 9 37.2 25.2 19.6 44.8 55.8 80 96.8 Keskin [1] 14 3.1 28 15.7 18.8 3.1 18 Kanyon Levent Soil nailing Shotcrete wall 21.3 Extensively fractured Trakya 25.3 gray-wacke 1.5 2 26.3 28.3 28.3 28.3 4.3 4.7 4.9 5.0 5.0 5.1 ? 9 45.5 31.9 68.2 58.2 86.8 53.7 67.9 96.2 Keskin [1] 9 2 3 1 3 3 2 2 3 4 50. Acta Geotechnica Slovenica, 2016/1 D. Qetin: Performance of soil-nailed and anchored walls based on field-monitoring data in different soil conditions in Istanbul Table 5. Summary information of the projects (Continued). No. project District Name support Wall s°ii ^maL h Hw Hemb sv sh n (m/ ^ System profile (m) (m) (m) (m) (m) m2) , . '___name _ /m) Sh, max Reference (mm) Zeytin-burnu Municipality 19 Under- Z.eytin-burnu ground Parking Lot Project Lime.. 45 cm Multi stone „ , . , ,, „ . . diameter . . . Bakirkoy 11.3 anchored . ... with clay bored pile . layers B.7 2.6 2.1 29240 6 28.5 34 6.9 Unpub-8 8 lished - data 26.6 20.1 Alper Co. 20 Company Umra-Residence niye Project Soil nail- Shotcrete ing wall Fractured Sand- Trakya stone 15.2 15.2 1.6 1.4 24.9 Unpub- - lished 34.5 data Mashat-21 tan Resi- Maslak dans Soil nail- Shotcrete ing wall Extensively fractured Trakya 18.3 gray-wacke 1.5 1.6 2.8 1 58.6 Keskin [1] 22 Terrace- Esenyurt Mix 80 cm soft- 11.5 12 Multi . diameter medium Gurpinar- anchored , . .. . r 13.4 bored pile clay 12 9.6 2.2 1.25 312500 4 13.4 9.6 10.4 15.3 26.9 17.2 Unpublished data Exten- 7 2.4 5.6 Tepe Soil nail- Shotcrete sively 9 5.3 16.2 Keskin 23 Shopping Maltepe Mall ing fractured Trakya wall 9 1.5 1.8 3.1 4 15.3 [1] gray-wacke 10 5.3 24 24 Cultural Sultan- Center gazi Multi anchored 65 cm diameter Fill -Sand- Fill 17.6 bored pile stone Trakya 17.6 2 1.7 146000 2 4.5 Unpub- - lished 6.7 data 25 Hotel Kucuk- Soil nail- Shotcrete Radisson cekmece ing wall clay-sand-limestone- Gung-oren 7.5 over consoli- Bakirkoy 7.5 dated clay 2 2 15.4 38.4 Durgu-noglu and Olgun [31] Axis 26 Shopping Mall Kaig-thane single . . Dia- „ r . Gung-anchored . Soft clay phragm oren 17 +multi „ overlying _ . wall Trakya propped 1280000 1 Durgu-100 noglu et al. [28] 27 Faco Medicine Levent-Factory Besiktas Project 65 cm diameter Multi bored pile Gray-anchored 25 cm wacke diameter micropile Trakya 22 18 1.5 3.2 1.95 115000 1 1.5 2.7 2.1 2350 1 45 36 Adatepe et al. [33] Siya-vuspasa Under-28 ground Parking Lot Project Bahce-lievler . . 80 cm Fill „.„ „ Multi .. , . Fill Gur- diameter overlying 11 anchored pinar bored pile Hard clay 3 2.25 1.7 - 312500 4 Unpub-10.1 lished data n 2 3 3 2 50. Acta Geotechnica Slovenica, 2016/1 D. Qetin: Performance of soil-nailed and anchored walls based on field-monitoring data in different soil conditions in Istanbul H is the excavation depth, Hw is the depth of the underground water level from the surface, Hemb is the embedded wall length, Sv is the vertical distance between the anchor or soil-nailing levels, Sh is the horizontal distance between the anchor or the soil nailing, q is the nail density (m/m2), EI is the wall stiffness, " is number of the inclinometers in a project, Sh,max is the maximum lateral wall displacement in Table 5. The anchor properties in the anchored wall projects are given in Table 6. 4 DATA ANALYSIS_ 4.1 Analysis of soil-nailed wall data Soil nailing is a useful technique for constructing excavations by reinforcing with fully bonded steel bars. The reinforced surface is stabilized with sprayed concrete, which creates a zone similar to a conventional gravity retaining wall. However, this sprayed concrete face does not play a vital role in the structural stability of the soil-nailed wall system. The inclination to the vertical of the wall, the soil condition, the excavation depth and the nail density per unit length of the wall (m2/m) can be considered important factors affecting the performance of the wall system. Figure 4 shows the relationship between the maximum lateral displacement of the wall (Sh,max) and the excavation depth (H) for soil-nailed shotcrete wall systems. The data shown in Figure 4 are from projects 5, 6, 8, 11, 15, 17, 18, 20, 21, 23 and 25. There is a clear relationship between Sh,max the maximum lateral displacement and the excavation depth. The data of Sh,max generally range from 0.05%H to 0.35%H Table 6. Anchor Properties for anchored walls. Excavation Depth Number Anchor Length Tendon EA (kN/m) Allowable Project No Project Name (m) of Anchor (m) Diameter Tensile Ca- Min. Max. Rows Min. Max. ('') pacity (KN) 1 Baumax AVM 15.80 6 20 25 0.6 112000 728 2 Karabulut Textile Factory 8 3 20 14 0.6 84000 545 3 BusinessPort 18 21.60 5-8 12 22 0.6 84000 545 4 Taksim Residence 18.7 19.65 ? ? ? 0.6 84000 545 Bomonti Hotel 5 and Congress Center 22.8 26 11 12 28 0.6 84000 545 7 Otoport 21 25 10 35 0.6 84000 545 9 Ideal Hospital 29.2 11 18 32 0.6 140000 910 10 Metal Co. Factory 5.25 9.25 3-4 13 19 0.6 84000 545 13 Erdem Hospital 16.3 7 19 30 0.6 84000 545 14 Filiz Co. Factory 18.2 6 14 25 0.6 84000 545 16 Ozbas Residence 22.7 25.9 12 13 35 0.6 84000 545 19 Zeytinburnu Parking lot 11.3 4 14 20 0.6 84000 545 22 Terramix 11.5 13.4 4-5 14 26 0.5 59200 385 24 Cultural Center 8-17.6 3-8 12 19 0.6 84000 545 27 Faco Medicine Factory 18 22 7-9 13 22 0.6 84000 545 28 Siyavuspasa Parking lot 11 4 14 21 0.5 59200 385 50. Acta Geotechnica Slovenica, 2016/1 D. Qetin: Performance of soil-nailed and anchored walls based on field-monitoring data in different soil conditions in Istanbul Figure 4. Relationship between maximum wall displacement and excavation depth for soil-nailing data. with an average value of 0.20%H. Except for projects 8 and 23, all the data shown in Figure 4 were obtained from projects carried out in the Trakya formation that is mainly composed of claystone-sandstone-mudstone, from extensively fractured to highly fractured. With a few exceptions, it can be said that the distribution of the data is within very reasonable limits. The data outside the limit are from Projects 17 and 25. Project 17 was conducted in the Gurgoren formation, which, composed of soft clay layers, is a weak soil formation in comparison to the Trakya formation. The excavation depth and the maximum wall displacement as a percentage of excavation depth were 7.5 m and 0.51%H, respectively, for project 17. It can be said that the wall movement beyond the maximum limit was due to the weak soil condition. In project 25 there were 9 inclinometers. The data from 7 inclinometers are within the limits, while the data from only 2 inclinometers are outside the limits. Sometimes, unexplained, excessive wall movement can be seen in projects that have very long facades. Generally, the inclination of the wall varies between 5° and 11° to the vertical in soil-nailing projects. This inclination has some positive effects on the performance of the wall, such as a lower earth pressure, a lower soil density and a better safety factor in comparison with the condition of a zero wall inclination to the vertical. Mittal [34] conducted a study on the effects of the variation of some design parameters, such as the nail length, the nail inclination and the wall inclination, on the safety of the wall. It was found that the safety factor of the wall system increases uniformly with an increasing wall inclination up to 20° to the vertical. Beyond this angle, the dimension of the increase of the safety factor of the wall system is very high. Figure 5. Relationship between nail density and excavation depth. 4.1.1 Influence of nail density The nail density can be calculated as follows: q = L/sv sh [1, 35] (1) where q is the nail density, L is the average nail length, sv is the vertical distance between the nail levels and Sh is the horizontal distance between the nails. Figure 5 shows the relationship between the nail density and the excavation depth. The nail density increases with the height of the soil-nailed wall. With the exception of a few points, the data are evenly distributed on both sides of the slope line. Figure 6 (next page) shows the relationship between the nail density and maximum wall displacement. The data are scattered over a wide range. Increasing the nail density is a natural consequence of increasing the excavation depth. Therefore, the nail density should be considered in conjunction with the excavation depth. It can be said that increasing the nail density is not a decisive factor solely on the performance of the wall. 4.2 Analysis of anchored wall data The data for five different wall types were collected and analysed. Figure 7 shows the relationship between the maximum wall displacement and the excavation depth depending on the type of wall. Similar to the soil-nailing data, there is a clear relationship between the maximum lateral displacement and the excavation depth. The data of Sh,max generally range 50. Acta Geotechnica Slovenica, 2016/1 D. Qetin: Performance of soil-nailed and anchored walls based on field-monitoring data in different soil conditions in Istanbul 100 90 80 70 60 50 E «f 40 30 20 10 0 • • •a* v Nail density (m/m ) 0.45 0.4 0.35 0.3 1 0.25 : E 0.2 .c o 0.15 0.1 0.05 Nail density (m/m ) (b) Figure 6. Relationship between nail density and maximum wall displacement. from 0.043%H to 0.32%H with an average value of 0.155%H. With a few exceptions, it can be said that the distribution of the data is within very reasonable limits. The data from project 26 were obtained from soft clay overlying sandstone. The excavation depth was 17 m, which can be considered as a large depth. Although the wall type was a diaphragm wall, which has a high wall stiffness, the weak soil conditions combined with the large excavation depth may have caused high wall movements in practice. The secant pile wall and the diaphragm wall systems are used to prevent water flow into the excavation site if the underground water level is above the final excavation depth. Although the data for the secant pile wall (SCW) and the diaphragm wall (DW) are • Contiguous pile Micropile □ Concrete wall+micropile ♦ Diaphragm wall ▼ Secant pile ^h.m^min 10 15 20 Excavation Depth (m) Figure 7. Relationship between maximum wall displacement and excavation depth depending on the type of wall. 50. Acta Geotechnica Slovenica, 2016/1 D. Qetin: Performance of soil-nailed and anchored walls based on field-monitoring data in different soil conditions in Istanbul very limited, it can be seen in Fig. 7 that the Sh,max values from the SCW and DW projects are close to or over the max line. This limited data indicate that the underground water level should be taken into account very carefully in the design stage of the wall. MPW and MPW+CW data show that this type of wall has a very good performance in the soil conditions mentioned. As for contiguous pile walls (CPW), the data are generally between the maximum line and minimum line up to an excavation depth of H=20 m. However, for excavation depths greater than H=20 m, the data are above the average line and below the maximum line. An excavation depth of 20 m can be considered as a threshold for a CPW. Table 7 shows the results obtained from projects 1, 2, 7 and 13 and the results given by some researchers for soft clay soil conditions regardless of the construction method or wall type. Table 7. Comparison of this study's results with other studies for soft-to-stiff clay condition. Study $h,max / H Soil Con- Note Min. Avg. Max. dition This study (project 2, 7, 13, 22) 0.09 0.18 0.37 Soft-to-stiff clays Four different project results Goldenberg et al. [20] 0.90 Soft-to-stiff clays Case histories collection Bentler [37] 0.452 Soft-to-stiff clays Case histories collection Wang et al. [36] 0.452 Soft-to-stiff clays Case histories collection Kung et al. [38] 0.39 Taipei clay Case histories collection Ou et al. [15] 0.40 Taipei soft soil Case histories collection As can be seen in Table 7, the average limit of lateral movement for soft-to-stiff clays is 0.18%H. This is smaller than the values given by Goldenberg et al. [20], Bentler [37], Wang et al. [36], Kung et al. [38] and Ou et al. [15]. Long [21] presented a database consisting of about 300 case histories collected from different soil conditions with different wall and support types. As a result of an intensive analysis, it was found that the normalized maximum lateral movement values, S^max, are mostly between 0.05%H and 0.25%H, where H is the excavation depth. In this paper, for soil-nailing data, the maximum lateral movement values range from 0.05%H to 0.35%H with an average value of 0.20%H, and for anchored wall data, the maximum lateral movement values range from 0.043%H to 0.32%H with an average value of 0.155%H. The maximum lateral values presented by Long [21] and the maximum lateral values presented in this study differ only slightly from one another. Tomlinson [40] reviewed 34 case histories of deep excavations conducted with different methods. The results were grouped into three major categories of ground condition. For soft-to-firm NC clays, S^max ranges from 0.08%H to 0.58%H with an average value of 0.30%H, for stiff-to-hard OC clays, Sh,max ranges from 0.06%H to 0.30%H with an average value of 0.16%H and for sands and gravels, S^max ranges from 0.04%H to 0.46%H with an average value of 0.19%H. Projects 1, 10, and 14 were conducted in hard clay. For these projects, Sh,max ranges from 0.04%H to 0.23%H with an average value of 0.12%H, which are close to the values reported by Tomlinson [40] for hard clays. A total of 536 case histories were collected and analysed by Moorman [23]. These case histories were classified into five categories in terms of the ground condition. For stiff clays in Moorman's database, the upper limit of the maximum lateral movement is 1%H and most of the data are under the 0.5%H limit. These values are higher than the values obtained from projects 1, 10 and 14 in this study. For very limited data from a deep excavation conducted in rock, the upper limit of the maximum lateral movement is 0.5%H and most of the data are under the 0.25%H limit. In this study, most of the data for the maximum lateral movement obtained from projects conducted in a rock formation are under 0.20%H, which is very close to the value for a deep excavation conducted in a rock formation presented by Moorman [23]. 4.2.1 Influence of embedded depth ratio The embedded depth ratio can be expressed as follows, Remb = Hemb / H (2) where Hemb is the embedded depth of the wall and H is the excavation depth (m). This ratio is an important factor that reflects the economy of the wall [36]. Figure 8 shows the relationship between the maximum wall displacement and the embedded depth ratio of the wall. 50. Acta Geotechnica Slovenica, 2016/1 D. Qetin: Performance of soil-nailed and anchored walls based on field-monitoring data in different soil conditions in Istanbul Figure 8. Relationship between maximum wall displacement and the embedded depth ratio of the wall. The dimensions of the maximum wall displacement are shown in mm units in Fig. 8(a), while they are shown as a percentage of S^max IH in Fig. 8(b). Although the dimensions of the maximum wall displacement were expressed in two different ways, it can be seen that there is a clear trend that the embedded ratio increases while the lateral displacement, or S^max IH ratio, decreases in accordance with the inclination line (Fig.8 (a) and (b)). Projects 2 and 22 have a higher embedded ratio. Project 2 was conducted in very soft clay soil conditions. There was a risk that the anchors would not work because of the soft soil conditions. Therefore, the embedded ratio selected was very high. Similarly, project 22 was conducted in clay soil that had a landslide risk. Thus, the embedded ratio selected was more than 1. According to the displacement value of the wall, it can be said that a 110i- 100 90 80 70 . I 60 Is E 50- • " « 40 30 20 10 I1' 0.7 0.6 0.5 0.4 c 0.3 0.2 0.1 10 15 • •• J.! si 0 5 10 15 Wall stiffness, EI(kNm2/m) x -|05 Wall stiffness, EI(kNrrr/m) x 10 Figure 9. Relationship between maximum wall displacement and wall stiffness. 50. Acta Geotechnica Slovenica, 2016/1 D. Qetin: Performance of soil-nailed and anchored walls based on field-monitoring data in different soil conditions in Istanbul high embedded ratio is a strong parameter to prevent excessive displacements provided that selection of high wall stiffness in soft soil conditions. 4.2.2 Influence of wall stiffness Wall stiffness can be expressed as EI, where E is the elasticity modulus of the concrete and I is the moment of inertia of the wall. Figure 9 shows the relationship between the maximum wall displacement and the wall stiffness. There is no definite information about the thickness of the shotcrete walls. The wall stiffness of the anchored walls was considered in this analysis. Figure 9 shows the data obtained from projects with large-diameter pile walls. Except for projects 9 and 26, the wall stiffness of the anchored pile walls is less than 500,000 kNm2/m and the data from anchored pile walls vary from 0.05%H to 0.45%H. There is a large scatter in the data. It can be said that there is no direct or no linear relationship between the wall stiffness and the movements. Hence, it can be interpreted that the wall movements are mainly independent of the wall stiffness. This result is consistent with those reported by Moormann [23], Moormann [39] and Long [21]. This study is the first attempt to prepare a deep-excavation database for Istanbul. Although there are limited data for some analyses, this database provides a good idea about the dimensions of wall movement. This database can be improved by collecting additional data in the future. 4.2.3 Influence of soil type All the anchored wall data were plotted in Figure 10 in terms of the main soil type in which the projects were conducted. The data for the Gungoren formation are around the average line up to an excavation depth of 20 m, after which the data are between the maximum and average lines. For the alluvium data, although the excavation depths are low, the data are between the maximum and average lines. The data for the Gurpinar formation are around the average line. The data from the Trakya formation are far from the maximum line even though the excavation depths are greater than 20 m. Because the data for the fill overlying Trakya formation and the Trakya-Bakirkoy formation are very limited, no interpretation can be made. However, it can be said that the data for the Trakya formation are the best results, regardless of soil type, wall type and support type. 5. CONCLUSIONS On the basis of the results from this study, the following conclusions can be drawn: - The results obtained show that successful wall performance can be obtained under a wide variety of soil conditions provided that suitable wall and/or support systems are chosen in the design phase. 120 r 100 80 • Gungoren Formation ▼ Alluvium Gurpinar Formation □ Trakya Formation ♦ Bakirkoy Formation * Trakya+Bakirkoy Formations -p- Fiil +Trakya Formations (Ô. )max=0.32%H 10 15 20 Excavation Depth(m) Figure 10. Relationship between maximum wall displacement and excavation depth depending on soil type. 50. Acta Geotechnica Slovenica, 2016/1 D. Qetin: Performance of soil-nailed and anchored walls based on field-monitoring data in different soil conditions in Istanbul The maximum lateral displacement of the wall increases with an increasing excavation depth for all types of walls. The data for the SCWs and DWs are close to or over the max line. These limited data show that the underground water level should be taken into account very carefully in the design stage. The data for the MPWs and MPWs + CWs show that these types of walls perform well. The data for CPWs are generally between the max line and min line up to a H=20 m excavation depth. However, for an excavation depth over H=20 m, the data are above the average line and between the average and maximum lines. An excavation depth of H=20 m can be considered as a critical excavation depth for CPWs. The data for S^max generally range from 0.05%H to 0.35%H with an average value of 0.20%H in soil-nailed projects. The data for Sh,max generally range from 0.043%H to 0.32%H with an average value of 0.155%H in anchored pile wall projects The nail density increases with the height of soil-nailed walls. It should average 2.2 m/m2 for H=10 m, 3.5 m/m2 for H=20 m and 4.8 m/m2 for H=30 m, according to the data obtained from projects conducted in Istanbul. With regard to soil-nailing shotcrete wall systems, it can be said that a soil-nailed shotcrete wall system is a very suitable system in rock conditions if there is no adjacent structure or infrastructure. Wall inclination to the vertical is an important factor to be able to obtain good performance for soil-nailing projects. In light of the analysis, there is a clear trend that the maximum wall displacement decreases with an increasing embedded ratio. Especially in soft soil conditions, a high ratio should be chosen to be able to prevent excessive soil and wall movements. However, this ratio is also related to the economics of the project. Therefore, this ratio should provide both safety and economy at the same time. There is no linear relationship between the wall stiffness (EI) and the wall movements (Sh>max). A number of different Sh,max values can be seen for the same EI values. In this case, the excavation depth is a more important parameter than the embedded ratio. Soil type has an important influence on the dimensions of wall movement. The project conducted in the Trakya formation had the best results in terms of wall movement. Down to an excavation depth of 20 m, the wall movements for the projects conducted in the Gungoren formation are around the average line. However, beyond an excavation depth of 20 m designers should be careful in the design process for projects in the Gungoren formation. Acknowledgements The author would like to thank C.E. Mahmut Candan and Geoform Engineering and Consultancy Company for their help with the data collection. REFERENCES [1] Keskin, B. 2008. Analysis of the performance of retaining systems in deep excavations in graywackes. Ph.D. Thesis, Bogazici University, Istanbul, Turkey. [2] Hashash, Y.M.A., Song, H., Osouli, A. 2011. Three-dimensional inverse analyses of a deep excavation in Chicago clays. Journal for Numerical and Analytical Methods In Geomechanics 35, 1059-1075. [3] Ma, J., Berggren, B., Bengtsson, P.E., Stille, H., Hintze, S. 2008. Behavior of anchored walls in soils overlying rock in Stockholm. International Journal of Geoengineering Case Histories, 2(1), 1-23. [4] Yoo, C., Lee, D. 2008. Deep excavation-induced ground surface movement characteristics - A Numerical Investigation. Computers and Geotech-nics 35, 231-252. [5] Peck, R.B. 1969. Deep excavations and tunneling in soft ground state-of-the-art report. Proceedings of 7th International Conference on Soil Mechanics and Foundation Engineering, Mexico. [6] Hwang, R.N., Moh, Z.C. 2007. Deflection paths and reference envelopes for diaphragm walls in Taipei Basin. Journal of GeoEngineering 2(1), 1-12. [7] Cheng, M.Y., Wu, Y.W. 2009. Prediction of diaphragm wall deflection in deep excavations using evolutionary support vector machine inference model (ESIM). 26th International Symposium on Automation and Robotics in Construction, Austin, Texas, 176-182. [8] Finno, R.J., Bryson, S. 2002. Response of building adjacent to stiff excavation support system in soft clay. Journal of Performance of Constructed Facilities 16(1), 10-20. [9] Ou, C.Y., Teng, F.C., Wang, I.W. 2008. Analysis and design of partial ground improvement in deep excavations. Computers and Geotechnics 35, 576-584. [10] Hou, Y.M., Wang, J.H., Zhang, L.L. 2009. Finite-element modeling of a complex deep excavation in Shanghai. Acta Geotechnica 4, 7-16. [11] Wong, I., Chou, T.S.H. 1998. Ground movements due to pile driving in an excavation in soft soil. Canadian Geotechnical Journal 36(1), 152-160. [12] Goldberg, D.T., Jaworski, W.E., Gordon, M.D. 50. Acta Geotechnica Slovenica, 2016/1 D. Qetin: Performance of soil-nailed and anchored walls based on field-monitoring data in different soil conditions in Istanbul 1976. Lateral support systems and underpinning. Rep. FHWA-RD-75-128, Federal Highway Administration, Washington D.C. [13] Mana, A.I., Clough, G.W. 1981. Prediction of movements for braced cuts in clay. Journal of Geotechical Engineering 107(6), 759-777. [14] Clough, G.W., O'Rourke, T.D. 1990. Construction induced movements of in situ walls. Proc., ASCE Conf. on Des. and Perf. Of Earth Retaining Struct Geotech. Spec. Publ., ASCE, 25, 439-470. [15] Ou, C.Y., Hsieh, P.G., Chiou, D.C. 1993. Characteristics of ground surface settlement during excavation. Canadian Geotechnical Journal 30, 758-767. [16] Wong, L.W., Patron, B.C. 1993. Settlements induced by deep excavation in Taipei. Proc., 11th South Asian Geotechnical Conference, Singapore, 787-791. [17] Carder, D.R. 1995. Ground movements caused by different embedded retaining wall construction techniques. Transport Res. Lab. Rep. 172, Berkshire, U.K. [18] Fernie, R., Suckling, T. 1996. Simplified approach for estimating lateral movement of embedded walls in U.K. ground. Proc., Int. Symp. on Geo Aspects of Underground Construction in Soft Ground, London, England. [19] Wong, I.H., Poh, T.Y., Chuah, H.L. 1997. Performance of excavations for depressed expressway in Singapore. Journal of Geotechnical and Geoenvin-romental Engineering 123(7), 617-625. [20] Hsieh, P.G., Ou, C.Y. 1998. Shape of ground surface settlement profiles caused by excavation. Canadian Geotechnical Journal 35(6), 1004-1017. [21] Long, M. 2001. Database for retaining wall and ground movements due to deep excavations. Journal of Geotechnical and Geoenvinromental Engineering 127(3), 203-224. [22] Yoo, C. 2001. Behavior of braced and anchored walls in soils overlying rock. Journal of Geotechnical and Geoenvironmental Engineering 127(3), 225-233. [23] Moormann, C. 2004. Analysis of wall and ground movement due to deep excavation in soft soil based on a new worldwide database. Soils and Foundations 44(1), 87-98. [24] Leung, E.H.Y., Charles, W.W.Ng. 2007. Wall and ground movements associated with deep excavations supported by cast in situ wall in mixed ground conditions. Journal of Geotechnical and Geoenvironmental Engineering 133(2), 129-143. [25] Istanbul Metropolitan Municipality (IBB) 2007. Microzonation work for the southern European side of Istanbul. [26] Istanbul Metropolitan Municipality (IBB) 2009. The second stage of microzonation work for the Asian side of Istanbul. [27] Terzaghi, K., Peck, R.B. 1967. Soil mechanics in engineering practice, 2nd Edition, Wiley, NewYork. [28] Stark, T.D., Choi, H. 2008. Slope inclinometers for landslides. Lanslides 5, 339-350. [29] Cetin, E., Bayrakli, Y., Güler, E. 2004. An example of a deep excavation pit's project. Journal of Turkey Engineering News (Türkiye Mühendislik Haberleri Dergisi) 431(3), 27-32. [30] Yilmaz, S., Kulac, H.F., Durgunoglu, H.T. 2000. Behavior of the soil nailed wall in different soil condition. 8th National Congress of Soil Mechanics and Foundation Engineering, Istanbul, Turkey. [31] Olgun, C.G., Durgunoglu, H.T. 1996. Soil nail practice in Istanbul. 10th European Young Geotechnical Engineers Conference, Izmir, Turkey, 274-283. [32] Durgunoglu, H.T., Kulac, F., Ikiz, S., Gunay, Akcakal, O. 2012. Selecting of displacement controlled deep excavation retaining system in soft clay. 14th National Congress of Soil Mechanics and Foundation Engineering, Isparta, Turkey. [33] Adatepe, S., Gokalp, A., Duzceer, R. 2012. Performance of the different retaining systems in a deep excavation project carried out in Istanbul graywacke. 14th National Congress of SoilMechan-ics and Foundation Engineering, Isparta, Turkey. [34] Mittal, S. 2006. Soil nailing application in erosion control-an experimental study. Geotechnical and Geological Engineering 24, 675-688. [35] Phear, A., Dew, C., Ozsoy, B., Wharmby, N.J., Judge, K., Barley, A.D. 2005. Soil Nailing - Best Practice Guidance, CIRIA C637, London. [36] Wang, J.H., Xu, Z.H., Wang, W.D. 2010. Wall and ground movements due to deep excavations in Shanghai soft soils. Journal of Geotechnical and Geoenvinromental Engineering 136(73), 985-994. [37] Bentler, D.J. 1998. Finite element analysis of deep excavations. Faculty of Viriginia Polytechnic Institute and State University Blacksburg, Virginia. [38] Kung, G.T.C., Juang, C.H., Hsiao, E.C., Hashash, Y.M.A. 2010. Simplified model for wall deflection and ground-surface settlement caused by braced excavation in clays. Journal of Geotechnical and Geoenvinromental Engineering 133(6), 731-747. [39] Moormann, C. 2002. Trag-und verformungsverhalten tiefer Baugruben in bingigen Böden unter besonder Berücksichtigung der BaugrundTragwerk und der Baugrund-Grundwasser-Interaktion(in German), Ph.D. Thesis, Technical University Darmstadt, Germany. [40] Tomlinson, M.J. 1995. Foundation design and construction. 6th Edition, Longman, Harlow. 50. Acta Geotechnica Slovenica, 2016/1 METODA NAPOVEDOVANJA DEFORMACIJ NABREKLJIVIH GLINENIH ZEMLJIN IN OBLIKOVANJE PLITVIH TEMELJEV, KI SO PODVRŽENI DVIGANJU Izvleček Nabrekljive zemljine je mogoče najti v mnogih delih sveta. Stanje prakse na tem področju se spreminja v zadnjih desetletjih. Projektiranje temeljev v nabrekljivih tleh predstavlja pomemben izziv za inženirje. Pretirana škoda nastane deloma zaradi pomanjkanja ustreznega načrtovanja, zaradi česar je potreba po boljših orodij za projektante, da bi se ocenil vpliv nabrekljivih tal v tipičnih projektnih situacijah. Za natančno napovedovanje dvigovanja tal je potrebno pravilno merjenje tlaka nabrekanja. Predlagan je teoretični model, s katerim je opisan potencial nabrekljivih glinenih tal, karakteristike glin so pridobljene iz edometrskih preizkusov. Prispevek opisuje analizo obnašanja nabrekljive zemljine pri povečani vlažnosti pod zgradbami in objekti. Metode in načela, ki se trenutno uporabljajo za oblikovanje konstrukcij temeljev na nabrekljivih zemljinah, so zaradi neenakomerne deformacije teh zemljin pri obremenitvah, problematična. Pričujoča študija je bila izvedena za izračun dviganja plitvih temeljev na nabrekljivih tleh, z upoštevanjem spremembe vlažnosti zemljine, kot tudi porazdelitve kontaktnih tlakov pod temelji. Baheddi Mohamed University of Batna, Department of Civil Engineering, laboratory L.R.N.A.T. 05000, Alžirija E-pošta: baheddi_md@yahoo.fr Djafarov Mekhti Azerbaijan Civil Engineering University, Laboratory of Soil Mechanics, A. Sultanova st 5, Baku 370073, Azerbajdžan E-pošta: md13djafarov@gmail.com Charif Abdelhamid King Saud University, College of Engineering P. O. Box 800, Riyadh 11421, Savdska Arabija E-pošta: abdelhamid_charif@caramail.com Ključne besede nabrekljiva zemljina, velikost nabrekanja, pritisk nabrekanja, prosti edometrski preizkus nabrekanja, obremenitveni edometrski preizkus nabrekanja, plitvi temelji 64. Acta Geotechnica Slovenica, 2016/1 A METHOD FOR PREDICTING THE DEFORMATION OF SWELLING CLAY SOILS AND DESIGNING SHALLOW FOUNDATIONS THAT ARE SUBJECTED TO UPLIFTING Keywords swelling soil, swelling magnitude, swelling pressure, free-swell oedometer test, loaded-swell oedometer test, shallow foundations Abstract gwelling soils can be found in many parts of the world. The state of practice in this area has been changing over recent decades. The design of foundations for expansive soils is an impo rtant challenge facing engineers. The excessive damage is, in part, due to the lack of proper design, resulting in the need for better tools for practitioners in order to assess the impact of swelling soils in typical design applications. A correct measurement of the swelling pressure is required for an accurate prediction of the heave. A theoretical model is proposed to describe the swelling potential of clay soils on the basis of their characteristics obtained from oedometer tests. This paper describes analysis of the behavior of swelling soils when moistened under buildings and structures. The methods and principles currently used for the design of structure foundations on swelling soils involve important problems due to the non-uniform deformations of these soils when subjected to structural loads. The current study was conducted to compute the uplifting of shallow foundations on swelling soils considering the water-content change as well as the contact-pressure distribution under the footing. 1 INTRODUCTION In any geotechnical study relative to a construction project, the swelling of a soil is as important as the settlement. The dimensional variations that, result from this phenomenon, constitute a permanent challenge for design and geotechnical engineers. The durability of a structure constructed on swelling soils depends on an appropriate appreciation of the phenomenon. The swelling of clay soils, containing smectites or illites in different quantities, is the origin of numerous problems in buildings and large structures. These disturbances are frequent in regions with a dry climate like some parts of the Caucasus parts, in Kazakhstan, Algeria, Morocco, etc. The swelling of soils can provoke important material damages, or even partial to total rupture of the structure, when it is not considered in the design process. It is therefore important to foresee correctly the possible distortions of swelling soils, in terms of amplitude and the speed of evolution, and to analyze its influence on the serviceability and stability of the structure. Swelling soils have been a major concern for designers over many years. Some construction procedures have been developed to limit the effects of inflation on the constructions and can be found in the classic works, of., Lancelot L et al., Mouroux et al. (in French) [19] [22], Baheddi Mohamed University of Batna, Department of Civil Engineering, laboratory L.R.N.A.T. 05000, Algeria E-mail: baheddi_md@yahoo.fr Djafarov Mekhti Azerbaijan Civil Engineering University, Laboratory of Soil Mechanics, A. Sultanova st 5, Baku 370073, Azerbaijan E-mail: md13djafarov@gmail.com Charif Abdelhamid King Saud University, College of Engineering P. O. Box 800, Riyadh 11421, Saudi Arabia E-mail: abdelhamid_charif@caramail.com 64. Acta Geotechnica Slovenica, 2016/1 M. Baheddi et al.: A method for predicting the deformation of swelling clay soils and designing shallow foundations that are subjected to uplifting Sorochan E.A., Mustafaev A.A. (in Russian) [27] [23], and Chen F.H. (in English) [7]. Currently, an abundance of documentation explains the mechanisms for the swelling of clays, either at the microscopic level as in test tubes tested in the laboratory or in-situ soil. Nevertheless, a survey of the behavior of the structures in contact with swelling soils constitutes a complex task and the existing methods contain some insufficiencies. Most of the research carried out, is limited to the amplitudes of the inflation of the clay soils in nature, under loads of spread foundations. Little attention has been given the propagation of the inflation phenomena in the mass of the swelling soils as a function of time. 2 IDENTIFICATION OF THE MATERIAL. The studied swelling clay comes from the Urban District of Baku (Azerbaijan) where it has provoked many disturbances in the structures of a concrete channel. The tests were carried out on samples of undisturbed clay soil, of the Pliocenes era, collected from the Shamour channel "Apcheron", (Baku) in the soil mechanics laboratory of the Civil Engineering Institute of Baku (Figures 1 to 3). 2.1 Mineralogical analysis A diffractometry analysis using X-rays was performed to determine the mineralogy of the studied soils. The physical and mechanical properties are given in Table 1. A chemical analysis of the samples of inflating clays gave the following compositions: Table 1. Physical and mechanical characteristics of the studied clay soils. Figure 1. Example of disorders in the construction of the " Samour -Apcheron " concrete channel. Soil characteristics Symbols Unit Values Natural water content W % 10-16 Degree of saturation Sr % 83-91 Wet Unit Weight Yh kN/m3 21.6-22.5 Dry Unit Weight Yd kN/m3 17.7-18.3 Specific Unit Weight Ys kN/m3 27.3-27.4 Voids Ratio e % 49.6-54.2 Porosity n % 33-35 Liquid Limit Wl % 46-51 Plastic Limit WP % 24-32 Plasticity Index Ip % 19-22 Liquidity Index Il % -41.0 to -74.0 Coefficient of compressibility av 1/MPa 0.08 Modulus of distortion between 0.1- 0.2MPa E MPa - At natural water content 7.0-7.8 - After saturation 6.0-7.2 Cohesion C MPa - At natural water content 0.2-0.58 - After saturation 0.08-0.14 Internal friction angle f Degree - At natural water content 25-31 - After saturation 17-23 Grain size distribution - 0.5 - 0.25 mm % - - 0.25 - 0.1 mm % - - 0.1 - 0.05 mm % 18.26 - 0.05 - 0.01 mm % 23.58 - 0.01 - 0.005 mm % 11.79 - 0.005 - 0.001 mm % 46.37 66. Acta Geotechnica Slovenica, 2016/1 M. Baheddi et al.: A method for predicting the deformation of swelling clay soils and designing shallow foundations that are subjected to uplifting Table 2. Chemical composition of the studied clays. N° of the Units Denomination sample Na+K Ca++ Mg++ Cl SO4 hco3 CO3 PH 1 % 13.88 1.49 0.33 7.3 6.59 1.19 0.63 7.5 2 % 12.18 1.69 0.33 7.30 5.42 0.89 0.59 7.8 3 % 13.88 1.29 0.25 7.30 6.84 0.89 0.39 7.8 SiO2: 52.28 % Al2O3: 15.27 % Na2O: 2.73 % K2O: 2.59 % MgO: 2.45 % CaO: 6.70 % TiO2: 0.79 % MnO2: 0.10 % Fe2O3: 6.77 % The chemical analysis by using the method of dosage of the elements composing the swelling clay produced the results in Table 2. Figure 2. Test of inflation in a free cell oedometer (without piston). Figure 3. Swelling test in an oedometer under load. the concentration of salts of the interstitial water and a "matrix" component, governed by the initial negative interstitial pressure of the sample. In most cases "matrix" component plays, a major role. Numerous methods have been proposed in the literature to evaluate the inflation potential of a soil from measures of the plasticity parameters and grain-size distribution, for example, Delage P. et al [10]., Bigot G., Zerhouni M.I., [6], Zerhouni, M.I., et al. [31], Cokca E., [8], Djedid, A., et al., [12], Cuisinier O. [9], Marcial, et al. [20], Bekkouch et al., [4] [5], Erzin Y.[14], Kariuki PC. [17], Yilmaz I. [30], Nelson, J.D. et al. [24], Tomas PJ., et al. [29]. For these authors, a very high inflation potential corresponds to a free inflation (expressed in percentage) greater than 25%, a high potential to an inflation between 5 and 20%, a medium potential to an inflation between 1.5 and 5%, and a low potential to an inflation lower than 1.5%. Johnson L.D., [16], Komine H. Ogota N [18], Miao L. et al., [21], the corresponding pressures of inflation are, respectively, superior to 300 kPa (very high potential), varying between 200 and 300 kPa (high), between 100 and 200 kPa (medium) and lower than 100 kPa (low). Several methods exist to measure the pressure of inflation in oedometers, among which we note: - Method of Huder and Amberg (1970). - Method of inflation with constant volume, according to ASTM norm D 2487-06.[1]. - Method of inflation or settlement under constant load, which requires several identical samples. - Free-swelling method followed by reloading. 3 EXPERIMENTAL ANALYSIS AND INTERPRETATION OF THE RESULTS The swelling rate corresponds to the relative variation in volume (%) of a sample either unloaded or subjected to a small load (generally the weight of the piston in the oedometer) when it is put into contact with water without pressure. The pressure of swelling is consists of an "osmotic" component due to the difference in The experimental studies show that the percentage of soil expansion increases proportionally with its density, its liquid limit, its clay contents, its plasticity and shrinkage indexes, as well as its pressure of pre-consolidation Rao B.H. et al., [25], Thakur VKS, et al., [28], Derriche, Z., et al., [11], Ejjaouani, H., et al. [14], and Fleurau J.M., et al., [15]. A method using the slope of the line connecting the percentage swell for various values of inundation stress was described in Nelson et al. [24]. These same studies report that the pressure of the expanding soil is 66. Acta Geotechnica Slovenica, 2016/1 M. Baheddi et al.: A method for predicting the deformation of swelling clay soils and designing shallow foundations that are subjected to uplifting inversely proportional to its natural water content. The analysis of the experimental results [3], allowed us to draw the curves giving the variation of the swelling potential in function of time for different values the compression stresses (Fig 4). It also allowed us to establish various other relations: Variation of the water content after expansion as a function of the different values of the compression stresses, (Fig 5), Variation of the expansion potential as a function of the different compression stresses, (Fig.6), Variation of the inflation potential with the water content in oedometer tests (Fig 7). The following section is relative to the mathematical model describing the expansion mechanism, established from the oedometric tests. The dependence of the water content of an expansive soil, on the compression stresses (after swelling has taken place) can be expressed using an exponential function: W = W0 e(-gP' (1) — ' 1 ln (W g = sw P W \ sw ) with: WW : Water content without loading (P = 0) and - 1 (W0 (2) For the interval of stresses, generally between 50 and 400 kPa in civil and industrial constructions, the dependence between the water content of the expansive soil and the compression stresses can be approximated using the relation. Ww = W» - cP (3) with: £sw(%) 5 4,5 4 | 3,5 •a 3 i I 2'5 I 2 m 1,5 1 0,5 0 I- ► i--i V (.-} / -* *—> ä-€ ) 6 7 t [days] 8 9 10 11 12 13 —■—P = 0.05MPa --*-P=0.3MPa ■ -P = 0.1 MPa -P=0.4MPa -P=0.2MPa Figure 4. Variation of the swelling potential with time. Wsw : Initial value of the water content given by the straight line in Figure 5. X : Slope of the curve in Figure 5. The obtained values are: Wsw = 0.24 X = 0.2MPa- In this part, five models for the prediction of the clay expansion according to the geotechnical properties are used. Their formulations are given in Table 3. Some of these models (Seed et al. and the first model of Vijayvergiya and Ghazzaly) ignore the natural water content. These models are valid in the range of the expansion amplitude varying from the lower limit of 5 % (the point at which a soil is considered to be expansive) to the upper limit of 60% (the maximum observed expansion). This allows us to determine, from these models, minimum 0,1 0,05 o 0,05 0,1 0,15 0,2 0,25 0,3 0,35 0,4 0,45 Figure 5. Variation of the water content with different load levels. 1P, MPa 66. Acta Geotechnica Slovenica, 2016/1 M. Baheddi et al.: A method for predicting the deformation of swelling clay soils and designing shallow foundations that are subjected to uplifting £ 5 4,5 4 3,5 3 2,5 2 1,5 1 0,5 0 sw 1%) P, MPa 0 0,1 0,2 0,3 0,4 0,5 0,6 Figure 6. Variation of the inflation potential function of the different loads. £sw(%) 5- < 4 ¡3 o p. M a> 1 0,05 Wsw 0,1 0,15 Water content 0,25 -A— P=0.05 MPa -•—P=0.3 MPa -P= 0.1 MPa —♦— P= 0.2 MPa -P= 0.4 MPa Figure 7. Variation of the swelling potential as a function of the water content. Table 3. Tested models for predicting the inflation of the clays. Model Reference Mathematical expression Seed and al. Seed et al. (1962) e = 2.16.10-5 IP44 sw P Nayak and Christensen Nayak and Christensen (1971) C e = 2.29.10-21P'45^ + 6.38 sw P W 0 Vijayvergiya and Ghazaly-1 Vijayvergiya and Ghazzaly (1973) , , . 62.427, + 0,65Wl -130.5 lg(e ) = u L 8 sw 19.5 Vijayvergiya and Ghazaly-2 Vijayvergiya and Ghazzaly (1973) 0.4Wl -W0 + 5.5 lg(e ) =-L-0- sv sW !2 Johnson Johnson (1978) Pour IP<40 ew = -9.18 + 1.5546IP + 0.08424 Z + + 0.1 W0 - 0.0432W0 IP - 0.01215ZIP 66. Acta Geotechnica Slovenica, 2016/1 M. Baheddi et al.: A method for predicting the deformation of swelling clay soils and designing shallow foundations that are subjected to uplifting water content comparable to the shrinkage limit, and maximum water content comparable to the liquid limit. Mathematically, the prediction models should constitute functional relations between a dependent variable, the percentage of the expansion, and some independent explicative variables. 4 PROPOSITION OF A MODEL FOR THE EXPANSION POTENTIAL OF THE STUDIED CLAYS_ The analytical results obtained show that, the variation of the expansion potential is dependent on the value of the compression stresses and the variations of the water content during the swelling process, and is governed by: 1-- Wsw -W0 )-xP (4) This analytical expression (4) gives, with good precision, the value of the swelling potential of clay soils on the basis of their properties obtained from oedometer tests. This formula contains the specific features of the expansive soils: ^sw e° sw : Expansion potential. : Expansion potential without loading. Psw : Expansion pressure. Wo : Initial water content. Wsw : Initial value of water content given by the graph Wsw=f(P) (Fig 5). X : Slope of the curve in Figure 5. This expression for the expansion potential is different from that found in the literature, by including a good approximation of its nonlinear relation with loading. The power of the load P in equation (4) does not exceed 2. find the position of the boundaries of the swelling zones (II) (Fig. 6). The main method for determining the boundaries of the areas of the swelling soil is as follows: In general the depth is determined, as the sum of the stresses due to the action of the weight of the soil, the live load and the additional vertical stress due to the non-wetted portion of the soil mass balancing the swelling pressure. The probable swelling zone of the soil can be between the obtained depths. Taking this into account, the boundaries of the areas of the swelling soil are determined from the following equations: s = s z,tot z,p z g ■s P z,ad sw (5) Under the conditions of the in-situ pressure, the deformation due to swelling is taken through dimensional compression conditions within the swelling soil layer, where the condition az < P is satisfied. From the top of the soil layer, we divide the wet soil layer of thickness dz into elementary slices. The relative deformation of the swelling soils takes the following form: Ad, e = - - = e 1-- (Wsw - wn ) - x p (6) Based on the results of experimental research [3], the relation (6) can be simplified as follows: where: esw = K - hP + b2P2 b = e° (7) h = (Wsw - wn ) + xPw 5 DESIGN SHALLOW FOUNDATIONS ON EXPANSIVE SOILS_ The stress state of the soil foundation is determined by computations calculations based on rules that consider the effect of the pressure of the weight of soil az,g , the weight of the structure oz>p and the supplementary vertical pressure due to the influence of the weight of the non-wetted soil at the boundaries of the wet surfaces. In determining the value of the uplift of the foundations after swelling the soil depending on the swelling pressure and the stress state of the sub grade, it is necessary to b2 = — X Psw Note that the relation (7), which incorporates the coefficients b0, b1 and b2, includes important properties of the swelling soils. Under the conditions of natural tension, changes to the compressive stresses can be determined using the classic formulas of soil mechanics sz, = gww x z . Then, from (4) we will have: e = b - kg z + b2g2 z2 (8) sw o 1 I sw 2 ' sw where: ysw : is the: unit weight of the swelling soils. 66. Acta Geotechnica Slovenica, 2016/1 M. Baheddi et al.: A method for predicting the deformation of swelling clay soils and designing shallow foundations that are subjected to uplifting The final value of the absolute swelling strain of the soil under conditions of natural tension is determined using the following expression: Ssw, g = C (p)dz (9) ZB where: ZB et ZH : Represent the lower and upper limits of the swelling soil, respectively. Substituting (8) into (9) and integrating, we obtain: = K (zh - Zb )-1 bgw (Z2h - zi)+3 b2 g w (Z3h - zi ) (10) In the particular case, when the stress due to the self weight of the soil, does not exceed the value of the swelling pressure across the entire thickness of the swelling soil mass and with, (ffz,g < psw) along with (0 < Z < Hsw): For: Zb = 0; Zh = Hsw , expression (10) determines the deformation due to swelling in conditions of natural tension becomes: Ssw.g = KHsw -2bgHw + 3b2llHl (11) where : Hsw : Thickness of the layer of swelling soil. 5.1 Computations of the uplift of the foundation of a building or structure, founded on swelling clay soils. The general case must take into account the action of the three normal stress components. But to simplify the problem, as supported by the standards of literature, swelling deformation is determined in relation to the conditions of uniform compression. As noted is above, the deformation due to swelling in the foundations of buildings and structures occurs in the area of swelling, where it meets the following condition: s ,, < p z,tot -X sw Hence, the equations for determining the formulas, for the uplift of the soil mass, are the transformation of expression (4) into a polynomial. The compressive stress to the depth of the foundation is taken by Sadjin V.S. [26] as. s „ = s e z, p O - ^ It + ^ where a0 = om0y - ysw Df (12) Zone I Zone II Zone III Figure 8. Stress distribution in a foundation. 66. Acta Geotechnica Slovenica, 2016/1 M. Baheddi et al.: A method for predicting the deformation of swelling clay soils and designing shallow foundations that are subjected to uplifting umoy Df b, a Mean stress at the bottom of the foundation. Depth at the bottom of the foundation. Dimensions of the bottom of the foundation. In the particular case, when all the soil layer is a swelling clay, s tot < psw and in accordance with Zb=0 and Zh=Hsw , formula (15) takes the form: Thus the sum of the pressure within the studied depth of the foundation takes the following form: szo = P = Isw (Z + Df ) + se ~zf1 + 1 2 lb a (13) Taking into account (13), the expression (4) takes the following form: e = b - b,(Z + Df ) - bs SW o f ' 1c 1 z ( b + 1 ' 2 b a gw (Z + Df ) + s e - 2 z < 1 + 1 ' 2 b a (14) The final value of the deformation due to the uplift of the foundations of buildings and structures is determined using formula (9). Substituting (14) into (9), we obtain: Zh I S », p = /K - bl7sw(Z + Df) - bso e - 1 z ( b + 1 ) 2 b a Ssw,p =(b0 - b1gwDf + b2glwDf )« - ^ gsw (b1 - 2b2gswDf ) + 3 b2g2SWH 2s0 (b1 - 2 b2 gswDf) 1 + 1 b a , b2S0 1 + 1 b a 3 1 œ 1 1) + 4 b 2 gswS0 1 + 1 b a 1 - e « 1 + 1 b a 1 + 1 b a (16) Case of a square foundation + b- - 1 z ( 1 + 1 ) gsw(Z + Df) + s e 2 b a After necessary transformations, we obtain the relation for the uplift of rectangular foundation bearing on the swelling soil: Ssw,p = (( - bgSwDf + b2g2wDf )(ZH - ZB ) -1 gsw ( - 2 b2 gswDf )(( - ZB)+1 glb2 (zh - Z3 2So ( - 2 b2gswDf) 1 +1 b a b2s2 h!T+T b a 2 < 1 + 1 ) Z« 2 b a 1 ( 1 + 1 ) * ( 1 + 1 ) z b a - ( 1 + 1 ) ZH b a + 4 b g s 1 2 I sw c 1 f1 + 11 z, 2 lb J ' 2 fb + 11 z« 2 lb (15) For a square foundation a = b, and in accordance with the formula (15) we have: Ssw,p = (( - bxgmDf + b2glD)) - Zb ) - 2gsw ( - 2 b2gswDf)( - ZB) +3 b2 gW (z« - z, ) + bsc ( - 2 b2gswDf e " - e -bb.s2 2 2 2 2 z, b 2 zh b + 2 bb. g s 2 sw b) e > (17) - (zu b) In the particular case, when the entire thickness of the soil layer is a swelling clay, az,tot < psw ; and Zb=0, Zh=Hsw ; and formula (17) takes the form: S p =(b - Kg Df + b,g2 D2)H sw,p \ o 1 i sw j 2 i sw j I sw - 1 g (b, - 2 b2g Df)H2 + - b2g2 H3 2 ' sw y 1 2 ' sw j j sw ^ 2 ' sw sw + S (b1 - 2 b2 g wDf) - 1 +--bb2s 2 2 2 2 2 « 1 - e b + 2 bb g s 2 sw b - («sw + b) e (18) Hsw Z 72. Acta Geotechnica Slovenlca, 2016/1 M. Baheddi et al.: A method for predicting the deformation of swelling clay soils and designing shallow foundations that are subjected to uplifting Case of continuous footing For a continuous footing (a ^ ¥ formula (15) takes the form: Ssw.p = (( - bllmDf + b2glD2f )(Zh - ZB) - 2 7Sw (Z - 2 b2gwDf )( - ZB) + 3 b2 gW ( _ ZB + 2 bo (b. _ 2 b.g Df 1 o \ 1 2 I sw f Zk 2 b Za 2 b bbo2 2 _ Za _ e b _ e + 4 bb2gsw Oo 2 b) e (19) _ ( 2b) Zh 2 b In the particular case, when the entire thickness of the soil layer is a swelling clay. az,tot < psw ; Formula (19) takes the form: S p =(b - kg Df + b,g1 D2)h sw,p V o 1 i sw j 2 i sw f I sw - - g (b - 2 b.g Df )H + - b.g2 H3 i sw \ 1 2 i sw j I s^ ^ 2 i sw sw + 2 bo (b, _ 2 b2g Df 1 o \ 1 2 ' sw f + 4 bbg o 2 sw _ 1 + bb2o 2 1 _ e 2 b _ (Hsw + 2b) e Hsw Yb (20) Case of a circular foundation For circular foundations with diameter d, (b = a = 0. formula (15) takes the form: 886 d), SSw,P = (( _ b,gwDf + b2glD2f )(Zh _ ZB) _ 2 g sw ( _ 2b2 gwDf )(ZH _ ZB )+3 b2 g2w (ZH _ Z + 0.886 dOo (b1 _ 2 b2 gswDf) + 0.443 do 3 1.129 ZH d 1.129 ZB d + 1.772 b2 g o 2 sw 2.257 Zb 2.257 Zh d _ e d 0.886 d ) _ (ZH + 0.886 d) e (21) In the particular case, when the entire thickness of the soil layer is a swelling clay. ffz,tot < psw ; formula (21) takes the form: S p =(b - bg Df + b,g2 D2)H sw,p \ o 1 i sw j 2 i sw j I sw - 1 g (b - 2bg Df )H + 1 b.g2 H3 i sw \ 1 2 i sw j I s^ ~ 2 < sw sw + 0.886 do (b _ 2 b,g D, 1 o \ 1 2 I sw f 3 1.129 Hw _ 1 0.443 d bo2 2 - 2.257 Hsw 1 _ e ~ + 1.772 d b2gswoo 0.886 d _ (Hw + 0.886 d) e~ (22) 6 APPLICATION m For a continuous footing Data: gw = 2.125 t/m? ; Df = 1 m ; b = 1 m ; Hw = 12 o = 10 t/m2 ; w = 14.1% ; e° = 22% o / ' n 1 sw Computations of the coefficients: b0 , b1 and b2 : b = e wsw _ w. b = = 0.22 (0.212 _ 0.141) = 0.01562 (wsw _ wn ) + Xpw 0.22 2.18 [(0.212 _ 0.141) + 0.007x2.18] = 8.7.10-3 cm2/daN = 8.7.10- 4 m2/1 0 212_0 198 With : y = --:-= 0.007cm2/daN 2 _ 0 ' b, = y = — x 0.007 = 7.06.10- 4 C / daN2 Pw 2.18 ' = 7.06.10_ 6 m4/12 By substituting the values of the coefficients b0 , b1 and b2 into formula (17), we will obtain the uplift deformation of the foundation: 1.129 Z 1.129 Z 66. Acta Geotechnica Slovenica, 2016/1 M. Baheddi et al.: A method for predicting the deformation of swelling clay soils and designing shallow foundations that are subjected to uplifting Swp = (0.01562 - 8.7 • 10-4 x 2.125 x 1 + 7.06• 10-6 x 2.1252 x 12)12 -1 x 2.125 x (8.7-10-4 - 2 x 7.06-10-6 x 2.125 x1)122 +1 x 7.06• 102 3 x2.1252 x 123 + 2 x 1 x 10(8.7 • 10-4 -2 x 7.06• 10-6 x 2.125 x 1) 12 12 (e 2x1 -1) + 1x 7.06-10-6 x102(1 -e~T) + 4x1x7.06-10-6 7 CONCLUSIONS x2.125 x10 2 x1 - (12 + 2 x1)e 2x1 = 0.0404m => Sswp = 0.0404 m = 4.04 cm If we take b = 2m, the value of the uplifting becomes: => Ssw,p = 0.028 m = 2.80 cm Case of a square foundation Data: g = 2.125 tlm3 ; Df = 1 m ; b = 1 m ; H = 12 < SW i f ' SW s = 10 tlm2 ; w = 14.1% ; e° = 22% o / ' n 1 sw Computing the coefficients: b0 , b1 and b2: m b, = = 0.22 (0.212 - 0.141) = 0.01562 (wsw - wn ) + x p; 0.22 2.18 With : x = [(0.212 - 0.141) + 0.007x2.18] T3 cm2/daN = 8.7 = 0.007 cm2/daN = 8.7.10-3 cm2/daN = 8.7.10- 4 m2/t 0.212 - 0.198 2 - 0 0.22 b2 =x = — x 0.007 = 7.06.10- 4 cmV daN2 2 p 2.18 ' X sw =7.06.10- 6 m4/12 By substituting the values of the coefficients b0 , bj and b2 into formula (5) we will evaluate the uplift deformation of the foundation. Sswp = (0.01562 - 8.7 • 10~4 x 2.125 x 1 + 7.06• 10^6 x 2.1252 x 12)12 -1 x 2.125 x (8.7 •10-4 -2 x 7.06-10-6 x 2.125 x1)122 +1 x 7.06-102 3 x2.1252 x 123 +1 x 10(8.7 • 10-4 -2 x 7.06• 10-6 x 2.125 x 1) -12 1 2x12 (e 1 -1) + ^x 1 x 7.06-10-6 x102(1 - / 1 ) + 2 x1x 7.06^10-6 x2.125 x10 1 - (12 + 1)e 1 = 0.0477m Based on an analysis of the experimental results presented here, the following conclusions are can drawn. A theoretical model was formulated to describe the swelling potential of clay soils on the basis of their features obtained from oedometer tests. For expansive soils, designers are only interested in the measurable quantities which are generally the pressure and the amplitude of the expansion. A model has been proposed for the expansion potential of the studied clays. On the basis of the non linearity of the physical mechanism of distortion, the method developed in the present work, relates the phenomenon of expansion with the distribution of the contact pressure, the soil properties, as well as the type and the value of the external loading and allows the design of lifting shallow foundations on expansive soils, On the basis of the nonlinearity of the deformation mechanism of swelling clays, the method developed in the present work permits a calculation of the lifting surface of different types of foundations on expansive soils. The rise of the foundations, based on expansive soils, depends not only on the expansiveness of the clay soils but also on the magnitude of the external loads transmitted to the ground by the buildings. The obtained results open up new perspectives for methods to design foundations on expansive clay soils. REFERENCES => Ssw,p = 0.0477 m = 4.77 cm [1] ASTM. 2006. Standard practice for classification of soils for engineering purposes (Unified Soil Classification System). ASTM standard D2487-06. Annual Book of ASTM Standards, Vol. 04.18. American Society for Testing and Materials, West Conshohocken, pp. 395-408. [2] Baheddi, M., Djafarof, M.D.,Charif, A. 2007. Design method of flexible continuous footings on swelling clay soils. Building Research Journal, Institute of construction and Architecture Slovak Academy of Sciences, 55, 3, 169-184. [3] Baheddi, M., Mustafaev, A.A., Djafarof, M.D. 1990. Mechanism of clays swelling deformation under the effect of centrifugal forces and the design shallow foundations. Thesis Ph D. Moscow (in Russian). [4] Bekkouch, A., Aissa Mamoune, S.M., Houmadi, Y. 2003. Variabilité des paramètres de gonflement 66. Acta Geotechnica Slovenica, 2016/1 M. Baheddi et al.: A method for predicting the deformation of swelling clay soils and designing shallow foundations that are subjected to uplifting sous une fondation superficielle; Symposium international sur les fondations superficielles FONDSUP; (05 - 07 Novembre) Paris, France, pp. 113-122. [5] Bekkouch, A., Djedid, A., Aissa Mamoune, S.M. 2002. identification et prévision du gonflement des marnes de Tlemcen (Algérie). PARAM, Laboratoire Centrale des Ponts et Chassées de l'école Nationale des Ponts et Chaussées, pp. 125-132. [6] Bigot, G., Zerhouni, M.I. 2000. Retrait, gonflement et tassement des sols fins Bulletin des laboratoires des Ponts et Chaussées, 229, 105-114. [7] Chen, F.H. 1988. Foundations on expansive soils. Developments in Geotechnical Engineering, Elsevier Publishing Co., 54, 465 p. [8] Cokca, E. 2000. Comparison of suction and oedometer methods for the measurement of swelling pressure. Engineering Geology 33, 2, 141-147. [9] Cuisinier, O., Masrouri, F. 2005. Hydromechanical-behaviour of a compacted swelling soil over a wide suction range. Engineering Geology 81, 3, 204-212. [10] Delage, P., Howat, M.D., Cui, Y.J. 1998. The relationship between suction and swelling properties in a heavily compacted unsaturated clay. Engineering Geology 50, 1-2, 31-48. [11] Derriche, Z., Iguechtal, L., Tas, M. 1999. Comportement des ouvrages dans liest argiles expansives d'In-Aménas. Revue Française de Géotechnique, 89, 55-65. [12] Djedid, A., Bekkouche, A., Aissa Mamoune, S.M. 2001. Identification et prévision du gonflement de quelques sols de la région de Tlemcen (Algérie). Bulletin de liaison des Laboratoires des Ponts et Chaussées 233, 67-75. [13] Ejjaouani, H., Magnan, J.P., Shakhirev, V. 2000. Calcul des fondations sur sols gonflants, Revue Marocaine de génie civil, 89, 19P. [14] Erzin, Y., Erol, O. 2007. Swell pressure prediction by suction methods. Engineering Geology 92, 3-4, 133-145. [15] Fleurau, J.M., Verbrugge, J.C., Huergo, P.J., Correia, A.G., Kheirbek-Saoud, S. 2002. Aspects of the behavior of compacted clayey soils on drying and wetting paths. Canadian Geotechnical Journal 39, 1341-1357. [16] Johnson, L.D. 1978. Predicting potential heave and heave with time in swelling foundation soils. US Army Engineer Waterways Experiment Station, CE, Vicksburg, Miss. Technical report S-78-7. [17] Kariuki, P.C., van der Meer, F. 2004. A unified swelling potential index for expansive soils. Engineering Geology 72, 1-2, 1-8. [18] Komine, H., Ogata, N. 2004. Predicting swelling characteristics of bentonites. Journal of Geotechni- cal and Geoenvironmental Engineering 130, 8, 818-829. [19] Lancelot, L.,Shahrour, I., Khaddaj, S. 2002. Etude en laboratoire du gonflement de l'argile des Flandres, PARAM. Laboratoire Centrale des Ponts et Chassées de l'école Nationale des Ponts et Chaussées, 145-151. [20] Marcial, D., Delage, P., Cul, Y.J. 2001. Compressibility of two swelling clays under high pressure. In Adachi &Fukue (eds), Clay Science for engineering. Rotterdam: Balkerna. ISBN:90-5809-175-9, pp. 571-576. [21] Miao, L., Houston, S.L., Cui, Y., Yuan, J. 2007. Relationship between soil structure mechanical behavior for an expansive unsaturated clay. Canadian Geotechnical Journal 44, 2, 126-137. [22] Mouroux, P., Margon, P., Pinte, J.C. 1988. La construction économique sur les sols gonflants, BRGM, Manuels et méthodes, 14, 125 p. [23] Mustafaev, A.A. 1989. Foundations on collapsible and swelling soils (In Russian) Vysshayashkola, Moscow, 592 p. [24] Nelson, J.D., Chao, K.C., Overton, D.D. 2007. Definition of Expansion Potential for Expansive Soil. Proceedings of the 3rd Asian Conference on Unsaturated Soils, Nanjing, China. April. [25] Rao, B.H.,Venkataramana, K., Singh, D.N. 2011. Studies on Determination of Swelling properties of soils from suction measurements, CanadianGeo-technical Journal 48, 3, 375-387. [26] Sadjin, V.S. 1971. About the model of swilling soils. Foundations and underground structures on swelling soils 61. Strojizdat, Moscow (in Russian). [27] Sorochan, E.A. 1989. Construction of works on the inflating soils. Strojizdat, Moscow, 312 p. (in Russian). [28] Thakur, V.K.S., Singh, D.N. 2005. Rapid determination of swelling pressure of clay minerals. ASTM Journal of Testing and Evaluation 33, 4, 239-245. [29] Thomas, P.J., Baker, J.C., Zelazny, L.W. 2000. An expansive soil index for predicting shrink-swell potential. Soil Science Society of America Journal 64, 1, 268-274. [30] Yilmaz, I. 2006. Indirect estimation of the swelling percent and a new classification of soils depending on liquid limit and caution exchange capacity. Engineering Geology 85, 34, 295-301. [31] Zerhouni, M.I., Dhouib, A., Hubert, B. 2002. Paramètres de gonflement et retrait des argiles -essais normalisées en France PARAM, Laboratoire Centrale des Ponts et Chassées de l'école Nationale des Ponts et Chaussées, pp. 125-132. 66. Acta Geotechnica Slovenica, 2016/1 UPORABA PERTHOVEGA PENETROMETRA ZA PESKE (PSP) ZA DOLOČITEV INŽENIRSKIH PARAMETROV PESKOV Ključne besede dinamični točkovni indeks odpora qd , inženirski parametri, ponovljivost, statistične metode Izvleček Določanje in-situ inženirskih parametrov peščenih zemljin je bil vedno izziv za geotehnične inženirje, zaradi česar je bilo do sedaj razvitih več metod. Preizkus s Perthovim penetrometrom za peske (PSP) je ena od najbolj vsestranskih metod. Je bistveno hitrejše in cenejše orodje kot vrtalna oprema, še posebej, če je globina raziskovanja zmerna. V prikazani raziskavi je obravnavana metodologija za uporabo naprave PSP za ovrednotenje inženirskih parametrov peščenih zemljin v laboratorijskih pogojih in preučevana ponovljivosti rezultatov preizkusa. Najprej so bili izvedeni preizkusi na tipičnih teheranskih mladih naplavinah (slabo granulirana peščena zemljina, SP), na vzorcih petih različnih gostot, navlaženih s pršenjem ali zalivanjem. Dalje so bile proučene običajne in logNormal porazdelitve podatkov o preizkusih z uporabo Kolmogorov-Smirnov testa normalnosti porazdelitve. Nato pa so bile, na podlagi dobljenih rezultatov, določene vrste razmerij med dinamičnim točkovnim indeksom odpora qd in drugimi parametri, kot so relativna gostota (Dr), modul elastičnosti (E), strižni modul (G) in trenjski kot zemlljine. Rezultati kažejo, da so dobljeni odnosi pol-logaritmični in logaritmični, večina dobljenih eksperimentalnih enačb ima visok koeficient določanja (>90%). Za oceno natančnosti rezultatov je bil uporabljen interval zaupanja meritev pri 95% zanesljivosti. Rezultati kažejo, da so bili vsi dobljeni eksperimentalni odnosi ustrezni. Nazadnje je bila ponovljivost rezultatov testov ocenjena z izračunom koeficienta variacije, ki je bil za vse teste manjši od 30%. S.D. Mohammadi Bu-Ali Sina University, Faculty of Sciences, Department of Geology Mahdieh Ave., 65175-38695 Hamedan, Iran E-posta: d.mohammadi@basu.ac.ir 64. Acta Geotechnica Slovenica, 2016/1 USE OF A PERTH SAND PENETROMETER (PSP) DEVICE TO DETERMINE THE ENGINEERING PARAMETERS OF SANDS Keywords dynamic-point resistance index (qd), engineering parameters, repeatability, statistical methods Abstract Determining the in-situ engineering parameters of sandy soils has always been a challenge for geotechnical engineers, resulting in several methods having been developed so far. The Perth Band Penetrometer (PSP) test is one of the most versatile of these methods. It is a considerably faster and cheaper tool than boring equipment, especially when the depth of the exploration is moderate. In the present research, a methodology for the use of a PSP device to evaluate the engineering parameters of sandy soils in laboratory conditions is discussed and the repeatability of the test results is studied. First of all, the tests were performed on typical Tehran young alluvial deposits (poorly graded sandy soil, SP) consistently prepared to 5 densities using the sand raining orpluviation technique. Next, the normal and logNormal distributions of the test data using the Kolmogorov-Smirnov normality test were examined. After that, based on the obtained results, the relationship types between the dynamic point resistance index (qd) and other parameters, such as the relative density (Dr), the modulus of elasticity (E), the shear modulus (G) and the friction angle of the soil, were determined. The results show that the obtained relationships were semi-logarithmic and logarithmic, and most of the obtained experimental formulas had a high coefficient of determination (>90%). To evaluate the accuracy of the results, 95% confidence and prediction bands were also used and the results show that all the obtained experimental relationships were appropriate. Finally, the repeatability of the test results was evaluated by calculating the coefficient of variations, which was less than 30% for all the tests. S.D. Mohammadi Bu-Ali Sina University, Faculty of Sciences, Department of Geology Mahdieh Ave., 65175-38695 Hamedan, Iran E-mail: d.mohammadi@basu.ac.ir 1 INTRODUCTION The engineering parameters of sandy soil at an earthwork, such as a foundation design on sandy soils, are commonly assessed using penetrometer testing. The most commonly employed dynamic method is the Standard Penetrometer Test (SPT), which involves the driving of a split-spoon sampler using a 63.5 kg drop weight [1]. While such devices are useful tools for larger-scale site investigations, their applicability for large-scale projects is limited by the scale of the apparatus and its relative insensitivity for shallow depths. Dynamic probing is a continuous soil investigation technique and is assumed to be one of the simplest soil penetration tests. It basically consists of repeatedly driving a metaltipped probe into the ground using a drop weight with a fixed mass and travel. Testing is carried out continuously from ground level to the final penetration depth. The continuous sounding profiles enable the easy recognition of 64. Acta Geotechnica Slovenica, 2016/1 S.D. Mohammadi: Use of a Perth sand penetrometer (PSP) device to determine the engineering parameters of sands Table 1. Some of the dynamic probing devices and their specifications. Penetrometer specifications Mackintosh Dynamic Cone Penetrometer (DCP) Perth Sand Penetrom-eter (PSP) Dynamic Probing Light (DPL) Dynamic Probing Medium (DPM) Dynamic Probing High (DPH) Dynamic Probing Super High (DPSH) Hammer weight (kg) 4.5 8 9 10 30 50 63.5 Fall height (m) 0.3 0.575 0.6 0.5 0.5 0.5 0.75 Cone diameter (mm) 27.94 20 - 35.7 35.7 43.7 50.5 Area at the base of cone (cm2) 6.13 3.15 - 10 10 15 20 Diameter of rod (mm) 12.7 16 16 22 32 32 32 Tip angle (deg) 30 60 - 60 60 60 60 Standard range for number of blows 3-50 (for 10cm of penetration) <20 (for 5cm of penetration) <20 (for 5cm of penetration) 3-50 (for 10cm of penetration) 3-50 (for 10cm of penetration) 3-50 (for 10cm of penetration) 5-100 (for 20cm of penetration) References [5] [7] [2] [2] [11] [11] [11] dissimilar layers and even thin strata using the observed variation in the penetration resistance. There are many dynamic probe devices in the world, such as Dynamic Cone Penetrometer (DCP), Mackintosh probe, Dynamic Probing Light (DPL), Dynamic Probing Medium (DPM), Dynamic Probing High (DPH), Dynamic Probing Super High (DPSH), Perth Sand Penetrometer (PSP), etc. Table 1 shows some of the dynamic probing devices and their specifications. The Perth Sand Penetrometer (PSP) is a lightweight dynamic penetrometer, a tool that is considerably faster and cheaper than boring, particularly when the depth of exploration is low and the soils being investigated are not coarse gravel [2, 3]. Most of the larger cities in parts of the world such as in Southern Australia, Northern Iran and Central UAE are situated on the coast or desert and many have extensive coastal deposits of clean, aeolian and alluvial sands. In earthwork projects like the construction of roads, dwellings and lightweight commercial and industrial structures, a determination of the engineering parameters of sandy deposits is important. The Perth Sand Penetrometer is described in detail in Australian Standard AS1289.F3.2 [4]. Since then, it has been used for the site characterization of pavement layers and foundations in other countries such as Iran [2]. There are some correlations between the dynamic probing devices (except PSP) and the engineering parameters of different soils [e.g. 5, 6, 7, 8], thus the main objectives of this paper are to describe the capability of the PSP to study the engineering parameters of sandy soils. 2. THE PERTH SAND PENETROMETER (PSP) The Perth Sand Penetrometer is described in detail in Australian Standard AS1289.F3.3 [4]. The physical arrangement of the PSP is summarized in this part, and illustrated in Fig. 1. It consists of a 9kg sliding weight that delivers a measured quantity of energy by falling through a height of 600mm onto an anvil block. This energy is used to push a 16mm blunt ended steel rod into the ground. The steel rod is usually scribed at increments of 50mm and the results are expressed as the number of blows required to drive the rod through a distance of 150mm. The total mass of the device is less than 20 kg, making it relatively portable. Mohammadi et al. [9] explained that the results of dynamic penetrome-ter tests with a flat tip (e.g., the PSP) are more repeatable and accurate compared to other penetrometers such as the DCP. Thus, it is better to use the PSP device instead of other dynamic penetrometers. Raising and releasing of the weight is achieved by hand, with a certain amount of care required to ensure that - The weight is lifted through the full 600mm height, - There is negligible impact on the upper stop at the top of the lift, and - The weight is released cleanly and allowed to fall without interference. The sounding rods are configured so that once driven, the hammer can be removed and additional rods added to enable the testing to continue to depths of several meters. In the author's experience, the maximum practical depths are of the order of 5 or 6 m in exceptionally loose conditions. Beyond these depths, difficulties in the 78. Acta Geotechnica Slovenica, 2016/1 S.D. Mohammadi: Use of a Perth sand penetrometer (PSP) device to determine the engineering parameters of sands Stop i-j-j-i Slide hammer 9.0 kg Slide rod Anvil model to assess the data distribution. The function of the Normal distribution is [10]: Steel sounding rod Blunt tip hardened steel 50mm long Figure 1. Perth Sand Penetrometer [4]. retrieval of the rods and the risk of lost rods through damage become too great. The main advantages of the PSP include: (1) Speed of operation; (2) Use in difficult terrain where access is poor; (3) The requirement for minimal equipment and personnel; (4) Low cost of the equipment; (5) Simplicity of operation and data recording/analysis; (6) Use in the interpolation of soil strata and properties between trial pits and boreholes; (7) Reduction in the number of boreholes required. fx (x 1 I 1/ X - m, s) = —-j= expi-—(-q s-J2p [ 2 s (1) in which p is the mean and a2 the variance. The pdf is symmetrical about a mode at p, and falls off quickly as x deviates from the mean. The Normal pdf is usually tabulated or calculated numerically using its Standard Normal form, with a mean of zero and a variance of one, N (0,1). Sometimes probability distributions are chosen simply because they appear to fit the observed data. This rationale has nothing to do with physical law. This is often the case in fitting probability distributions to, say, soil data. One observes that when the frequencies of undrained strength measurements are plotted against the strength value they might exhibit a bell-shaped form characteristic of the Normal pdf, and thus a Normal distribution model is used to approximate them [10]. Various statistical methods can be used to test whether the match between the empirical data frequencies and the theoretical pdf model is close enough to assume that the differences between the observed and the modelled could simply be due to sampling variability [10]. The most common of these tests are the Chi-square goodness of fit and the Kolmogorov-Smirnov maximum deviation test. In many cases, however, the easiest way to assess the quality of the fit of a theoretical distribution to an observed distribution is to plot the observed distribution against the theoretical distribution. A probability grid is a specialized graph paper with horizontal and vertical scales designed such that the cumulative frequencies of a particular form of pdf (e.g., a Normal distribution) plot as straight lines. Thus, empirical data can be plotted on a probability grid and a visual determination made as to whether they are well approximated. The effect of the Normal probability grid can be achieved by using cumulative frequencies as arguments to an inverse standard Normal cumulative distribution function. 3 PROBABILISTIC AND STATISTICAL METHODS The application of probabilistic and statistical methods in geotechnical engineering has increased remarkably in recent years. In this paper, some probabilistic and statistical methods were used to evaluate the laboratory test results obtained from laboratory tests on studied river sandy soil. One of the methods used is the assessment of the distribution properties of the data. The Normal probability density function (pdf) is an appropriate Another graphical tool for comparing data with analytical distributions is the quantile-quantile (Q-Q) graph. A Q-Q graph shows the observed quantile of the data plotted against the theoretical quantile of the best fitting distribution. If the frequency distribution of the sample data is reasonably approximated, the empirical and analytical quantile should be approximately the same, and the data pairs should lie on a 45-degree line [11]. An assessment of the regression line confidence can be checked using confidence and prediction bands. To 78. Acta Geotechnica Slovenica, 2016/1 S.D. Mohammadi: Use of a Perth sand penetrometer (PSP) device to determine the engineering parameters of sands achieve that, 99% and 95% confidence bands are useful in geotechnical engineering. Finally, the correlation coefficient R2 between two types of data, such as soil parameters, is determined by [10]: E (x - x)(y - y) Table 2. The index properties of used soil. R = (-*- VIE (x - x )2 E (y - y )2 Where: x: the value of the obtained data y: the value of the measured data 4 MATERIALS AND METHODS ,)2 (2) In order to achieve the main objectives of this paper it was necessary to select a suitable sample. First of all, the appropriate sampling area was selected and the sampling was done. Then, the selected sample was prepared in the laboratory. 4.1 Sampling area The materials used in this research included typical Tehran young alluvial deposits (a poorly graded sandy soil) taken from the west of Tehran. Geologically, the Tehran young deposits comprise subrounded sand grains with 5% gravel. The X-ray analysis results show that it is comprised of quartz, feldspar, pyroxene and calcite. 4.2 Sample preparation The data used in this paper was obtained from laboratory tests undertaken by the author at the Geotechnical Engineering Laboratory of Bu-Ali Sina University, Hamedan, Iran. To prepare the soil sample for testing, alluvial deposits were oven dried and passed through sieve No. 4. Fig. 2 shows the gradation curves of the original soil and the Grain Size Distribution Curve (ASTM D 422-63) * i 50 0 0.01 '" \ alluvial deposits " j \ \ Parameter value emax (-) 0.97 emin (-) 0.46 Gs (-) 2.66 yd (max) (KN/m3) 17.85 yd (min) (KN/m3) 13.24 Cu (-) 1.16 Cc (-) 1 value of clay (%) 0 value of silt (%) 2 USCS soil classification SP sample after passing sieve No. 4, which is classified as poorly graded sand (SP) according to the Unified Soil Classification System. The index properties of the soil are shown in Table 2. To achieve a uniform sample, the sample in the testing mould was remoulded using a sand pluviation set that was designed and constructed at Bu-Ali Sina University. It has a height and diameter, both of 1m. The experimental arrangement used to generate a suitable sample for this study is schematically shown in Fig. 3. The sand was rained from the hopper, through holes ranging in size from 6mm to 15mm diameter, at a spacing (S) ranging from 20 to 100mm, with the holes arranged in a triangular pattern. Diameter (mm) Figure 2. Gradation curves of alluvial deposit and used soil. Figure 3. Experimental arrangement used in this study. 78. Acta Geotechnica Slovenica, 2016/1 S.D. Mohammadi: Use of a Perth sand penetrometer (PSP) device to determine the engineering parameters of sands Table 3. The index properties of used soil. Table 4. Testing program for laboratory investigations. Dr(%)* Mean of Water content (%) Dry unit weight (gr/cm3) 25 0.04 1.44 35 0.04 1.48 50 0.04 1.55 60 0.04 1.60 75 0.04 1.67 * Relative density In all cases, a single diffuser screen with an aperture size of 2.36mm was employed (except where the pluviation rate was so great that sand accumulated on the screen, at which point, the diffuser was removed). The fall heights (free fall distance travelled by the sand from the diffuser to the top of mould) considered were 61mm, 236mm, 481mm, and 641mm. A 1 liter mould (107mm in diameter and 114mm in height) was used to catch the sand for the density control. Details of the tests on samples having different densities are indicated in Table 3. To prepare the soil sample for the direct shear test, a circular shear box with 60 mm diameter and 25 mm height was used. To achieve a uniform compaction in the circular shear mould of the direct shear machine, tamping by a small circular steel plate with a 60 mm diameter was used. All the direct shear tests were also carried out in dry conditions. 4.3 Testing procedures Several tests including the Perth Sand Pentrometer (PSP) test, the Plate Load (PLT) and direct shear tests were undertaken on the compacted materials as described in the following sections. 4.3.1 The Perth Sand Pentrometer tests and its repeatability As previously mentioned, the PSP tests were used to determine the engineering parameters of sandy soil in laboratory conditions. To achieve some relation between the PSP results and the engineering parameters of sandy soils, several tests according to the ASTM standard were undertaken. Table 4 shows the program of laboratory tests. The result of the PSP tests is shown by the qd index, where qd is the dynamic point resistance (Pa), which is determined using [12]: W qd =^-T777-rd (3) W + W ' Where: qd is the unit point resistance value (Pa) W is the mass of the hammer (kg) h is the height (m) Dr (%) of sand samples PSP (number of tests) PLT* (number of tests) Direct shear (number of tests) 25 3 3 7 35 3 3 7 50 3 3 7 60 3 3 7 75 3 3 7 Total - 15 15 35 * Plate Load Test g is the acceleration due to gravity (m/sec2) A is the area at the base of the cone (m2) e is the average penetration per blow W is the total mass of the extension rods, the anvil and the guiding rods (kg) Fig. 4 shows the qd index values obtained from the laboratory PSP tests that were undertaken on sandy soil at the testing mould. qd (kPa) 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 100 200 . 300 ' 400 500 600 700 Figure 4. Average of qd index versus depth for studied soil at the testing mould. The repeatability of the PSP test results is an important consideration. To evaluate the repeatability, several tests were carried out. Each testing series included three PSP tests on studied soil with different relative densities (Loose, Medium and Dense). In order to study the repeatability of the results, it was important to choose a suitable parameter that represented the repeatability. The use of the standard deviation value, s, was not appropriate for this purpose because it is large for large values of NPSP [5]. However, the coefficient of variation (COV) can be considered as an indicative parameter because it represents a Normalized standard deviation. The COV parameter is calculated using Eq. (4): COV = s / X (4) 78. Acta Geotechnica Slovenica, 2016/ 1 S.D. Mohammadi: Use of a Perth sand penetrometer (PSP) device to determine the engineering parameters of sands Where: X : the average of NPSP at each depth s : the standard deviation of NPSP at each depth Table 5 shows the coefficient of variation (COV) for the results obtained from some of the dynamic probe devices reported by various researchers. Table 5. Coefficient of variation for results obtained from some dynamic probe devices. Dynamic probe devices Range of COV (%) Mean of COV (%) References Standard penetration test (SPT) 27-85 30 [12] Dynamic Cone Pen-etrometer (DCP) 3-58 17 [7] Mackintosh probe 0-39 11 [5] Dynamic Probing Light (DPL) 0-23 8 [11] Dynamic Probing Medium (DPM) 0-17 5 [11] Dynamic Probing High (DPH) 0-55 16 [11] Dynamic Probing Super High (DPSH) 0-44 13 [11] The sources of variability in soil properties differ, and accordingly the coefficients of variation differ for different dynamic devices [5]. It can be seen that the variation of the COV for the results of the Standard Penetration Test (N), which is basically a super heavy dynamic probe test, is higher than the other dynamic devices that are reported to be between 27% and 85% with a mean value of 30% [13]. The repeatability of the SPT test results could be used as a measure of the repeatability of the PSP results by comparing the COV values of the two methods. In the present research, the values of the COV have been determined for each depth in each series of tests. The average value of the COV was about 5.60% and its standard deviation was 9.51. In more than 68.7% of the tests, the value of the COV was 0%, and in 12.5% of the tests, this value was 20.28%. In the tests undertaken, the values of the COV varied between 0 and 28.3%, and for all cases it was less than 30%. Therefore, the results of the PSP tests for the three relative densities (Loose, Medium and Dense) can be considered as repeatable when compared with the values presented in Table 5. 4.3.2 Plate Load Test (PLT) The Plate Load Test (PLT) is a useful site investigation tool and has been used for the proof testing of pavement layers in a lot of European countries for many years. Currently, it is used for the evaluation of both rigid and flexible pavements [14]. The PLT, at full or reduced scale, is sometimes considered as the best means of determining the deformation characteristics of soils, but it is only used in exceptional cases because of the costs involved [15]. In the present research, a round plate of 230mm diameter was used. The PLT was used as a reference test to obtain the strength parameters of the soil under investigation. A loading frame was designed to fit the mould and its support. To perform the test, the bearing plate and hydraulic jack were carefully placed at the centre of the samples under the loading frame (Fig. 5). The hydraulic jack and the supporting frame were able to apply 60 tons load. For the measurement of deformations, dial gauges that are capable of recording a maximum deformation of 25.4 mm (1 in.) with an accuracy of 0.001 in. were employed. The ASTM-D1195 [16] standard method was followed to perform the test. The elastic modulus is always considered as a more important deformeability parameter for a geomaterial. As in the case for other stress-strain tests, different elastic moduli can be obtained from the PLT. The soil elastic moduli can be defined as: (1) the initial tangent modulus; (2) the tangent modulus at a given stress level; (3) reloading and unloading modulus and; (4) the secant modulus at a given stress level [14]. In this study, since the stress-strain curves had a clear peak point, the initial tangent modulus was determined for all the plate load tests. To determine the initial modulus (Eplt©), a line was drawn at a tangent to the initial segment of the stress-strain curve, then an arbitrary point was chosen on the line and the stress and deflection corresponding to this point Compacted Sample Load Plate-Cross Head Z^MJ 0.15 3.2 4.2 3.7 log qd(kPa) Figure 7. LogNormal distribution of the q, index shown by (a) histogram of data and (b) the Kolmogorov-Smirnov Normal test. 5.2 qdversus Dr(%) The relative density (Dr) is a useful parameter to describe the consistency of sands [19]. To obtain the correlation between the average qd and Dr, first of all, a statistical distribution of Dr values was investigated. Figs. 8a and 8b show the Normal histogram curve and the Normal probability plot of a Kolmo-gorov-Smirnov test for relative density, respectively. As shown, the distribution of the Dr values of the studied soil is Normal. After that, the correlation between the average qd and Dr was investigated. Because the distribution of Dr is Normal and the distribution of qd is logNormal, thus the equation for Dr-qd will be Semi-Logarithmic. Eq. (6) and fig. 9 suggest a good correlation between these two parameters. The determination coefficient (R2) of Eq. (6) is 0.98. 78. Acta Geotechnica Slovenica, 2016/1 S.D. Mohammadi: Use of a Perth sand penetrometer (PSP) device to determine the engineering parameters of sands C 03 £ IE Dr (%) Kolmogorov-Smirnov Normality Test .999 . Average: 48.1034 .99 StDev: 17.6480 N: 29 .95 - £ .80 . !5 m ■O .50 . s CL .20 . .05 . .01 . .001 - Approximate P-Value > 0.15 Dr (%) Figure 8. Normal distribution of the Dr shown by (a) histogram of data and (b) the Kolmogorov-Smirnov Normal test. Also, fig. 9 shows the 95% confidence and prediction bands of the line regression. As can be seen, all the data is limited to the prediction band area. Dr(%) = -181.18 + 406.38(Logqd) (R2=0.96) (6) 80 70 60 '50 40 30 20 10 Dr (%) = -181.178 + 406.385 log(Logqd(kPa)) S = 3.76417 R-Sq = 96.1 ' Regression 95% CI ----95% PI I-1-1— 3.2 3.7 4.2 Logqd(kPa) Figure 9. Correlation between qd index and Dr(%). 5.3 qdversus modulus of elasticity (E) As mentioned previously, the modulus of elasticity (E) is a useful parameter in earthworks. To evaluate the Normality of the modulus of elasticity, all the data obtained from the PLTs such as EpLT(i) and Eplt(r2) were tested using the Normal histogram curve and the Normal probability plot of the Kolmogorov-Smirnov test (figs. 10 and12). The results show that EpLT(i) and 54321 - (a) 3.8 3.9 4.0 4.1 4.2 4.3 LogEPLT(l)(kPa) Kolmogorov-Smirnov Normality Test .95 . & .80 . !5 m .50 . n P .20 . n .05 . .01 . .001 . Average: 4.09942 StDev: 0.172753 N: 20 (b) o ^o__^ 0 Approximate P-Value: 0.098 3.8 3.9 4.0 4.1 4.2 4.3 4.4 LogEPLT(i)(kPa) Figure 10. logNormal distribution of the EpLT(i) shown by (a) histogram of data and (b) the Kolmogorov-Smirnov Normal test. EPLT(R2) are according to the logNormal and Normal functions, respectively. For the data obtained in this study, the best correlations between the average qd, EpLT(i) and Eplt(r2) are presented in fig. 11 and 13 (Eqs. 7 and 8). However, there is a better correlation (Eq.8) between the average qd index and the PLT reloading modulus (Eplt(r2)) compared to the correlation with EPLT(i). Also, all the data and the regression lines are limited at the 95% confidence and prediction limits, respectively. LogEPLT (kPa) = 2.50 + 0.43Logqd (R2=0.90) (7) JPLT (R2) 4.45 — 4.35 — if 4.25- (kPa) = -33870.5 + 11407.1Logqd (R2=0.96) (8) LogEPLT(i)(kPa)= 2.49664 + 0.432990 Logqd(kPa) S = 0.0535711 R-Sq = 90.9 ° 4.15- 5] 4.05- 3.95 -3.85 -3.75 - 95% CI ---95% PI 3.2 3.7 Logqd(kPa) 4.2 Figure 11. Correlation between qd and EpLx(i). 84. Acta Geotechnica Slovenica, 2016/1 S.D. Mohammadi: Use of a Perth sand penetrometer (PSP) device to determine the engineering parameters of sands .999 .99 .95 ^ 80 1 -50 .a 2 .20 Q. .05 .01 .001 2000 4Ö00 ßäoo 8Ö00 10000 12000 14000 16000 EPLT(2)(kPa) Kolmogorov-Smirnov Normality Test Average: 8354.86 (b) StDev: 4420.61 N: 20 O n 0 --^ o o Approximate P-Value >0.15 5000 10000 EPLT(2)(kPa) 15000 Fig. 14 and Eq. (9) show the correlation between EpLT(i) and Eplt(r2), which has a linear trend. Eplt )(kPa) = -93444.2 + 24832.5LogEPLT (i)(kPa) (R2=0.94) (9) 5.4 qd versus shear modulus (G) Several methods are available to evaluate the shear modulus of coarse-grained and fine grained soils, such as geophysical methods, the Plate Load Test (PLT), etc., which are all costly. In this research, the PLT shear modulus (Gplt) of the studied sandy soil was determined. Like the sandy soils' modulus of elasticity (Eplt©) obtained in the present research, the Gplt values were according to the logNormal function (figs 15a and 15b). Several correlations between the average qd versus Gplt were investigated. The best correlation between the average qd and (Gplt) is presented in fig. 16 and Eq (10). All the data and regression lines are located at the 95% confidence and prediction limits, respectively. The results show that the shear modulus increases with an increase in the values of qd. This correlation is linear with a determination coefficient of 0.92. Figure 12. Normal distribution of the Eplt(r2) shown by (a) histogram of data and (b) the Kolmogorov-Smirnov Normal test. LogGPLT (kPa) = 0.39 + 0.897Logqd (R2=0.92) (10) 15000- EPLT(2)(kPa) = -33870.5 + 11407.1 Logqd(kPa) S = 869.077 R-Sq = 96.3 % O. LU Logqd(kPa) Figure 13. Correlation between qd index and eplt(r2). EPLT(2)(kPa) = -93444.2 + 24832.5 LogEPLT(i)(kPa) S = 1096.28 R-Sq = 94.2 % S 10000— CM Q. HI LogEPLT(i)(kPa) Figure 14. Correlation between EpLT(i) and eplt(r2)- a 3 0 3 cr £ 2- W 3.6 3.8 4.0 Log(GPLT)(kPa) Kolmogorov-Smirnov Normality Test .999 . .99 . .95 . & .80 . 15 ra .50 . -Q .20 . n .05 . .01 . .001 . Average: 3.70302 (b) StDev: 0.353770 N: 20 o ^_____0 ---- - o ^ ------- o Approximate P-Value > 0.15 3.2 3.7 Log(GPLT)(kPa) 4.2 Figure 15. logNormal distribution of the Gpu shown by (a) histogram of data and (b) the Kolmogorov-Smirnov Normal test. 78. Acta Geotechnica Slovenica, 2016/1 S.D. Mohammadi: Use of a Perth sand penetrometer (PSP) device to determine the engineering parameters of sands 4.5 - CO a, 4.0 - Log(GPLT)(kPa) = 0.396857 + 0.893155 Logqd(kPa) , S = 0.101384 R-Sq = 92.2 % Q. CD o' 3.5 - 3.0- Regression .......95% CI ---95% PI 3.2 -1- 3.7 Logqd(kPa) 4.2 Figure 16. Correlation between qd index and gplt. o> œ le 35 —| CL Phi (deg) = -10.5654 + 83.3116 log(Logqd(kPa)) S = 1.49383 R-Sq = 86.9 % o _ — ___ .X" ____ - —* _— - Regression ...... 95% CI ---95% PI 3.2 3.7 4.2 Logqd(kPa) Figure 18. Correlation between qd index and friction angle (0). 5.5 qd versus friction angle The friction angle (0) of sandy soils is an important engineering parameter for the design of the earth structure in earthworks. To assess the Normality of the effective friction angles (0) obtained in the present research, the Normal histogram curve and the Normal probability plot of a Kolmogorov-Smirnov test were used (figs. 17a and17b). The results show that the effective friction angle (0) is according to the Normal function. Jo S 5 S S Ï0 Ï2 Ï4 Phi (deg) Kolmogorov-Smirnov Normality Test .999 .99 Average: 36.6055 StDev: 4.01684 N: 20 (b) .95 _ O 80 _ 1 50 --Q 2 .20 CL .05 .01 _ „ oji---- -—-""""o O Approximate P-Value > 0.15 .001 _ 30 35 40 Phi (deg) Figure 17. Normal distribution of the friction angle (0) shown by (a) histogram of data and (b) the Kolmogorov-Smirnov Normal test. The correlation between the average qd and effective friction angle (0) is presented in Fig. 18 and Eq. (11): Phi(Deg) = -10.56 + 83.31Log(Logqd) (R2=0.87) (11) 6 SUMMARY AND CONCLUSIONS The PSP is a lightweight device that can be conveniently used for soil investigations in shallow depths. The results of PSP testing can be used to assess rapidly the variability of soil conditions, allowing different layers to be identified. Based on the results of the present research, correlations can be established between the qd index and the engineering parameters of sandy soils. A statistical approach was applied to find the best correlations of the results with a high coefficient of determination (R2). For the results obtained, the determination coefficients (R2) between qd and the engineering parameters were mostly greater than 0.90. The statistical methods indicated that the distribution of the data obtained from the PSP was according to logNormal function and the obtained experimental formulas were located in semi-logarithmic and logarithmic classes. Also, all the data and regression lines were located at the 95% confidence and prediction bands, respectively. Table 6 shows the summary of the equations obtained in this study. To control the repeatability of the results of the PSP tests, the values of the coefficient of variation (COV) were calculated. This coefficient varied between 0 and 28.3%. Therefore, it can be concluded that the results of the PSP tests for five relative densities (Dr) can be considered as repeatable when compared with the values presented by Lee [13]. REFERENCES [1] Palmer, D.J., Stewart, J.G. 1957. Some observations on the Standard Penetration Test and a correlation of the Test with a new Penetrometer. Proc. 4th Int. 78. Acta Geotechnica Slovenica, 2016/1 S.D. Mohammadi: Use of a Perth sand penetrometer (PSP) device to determine the engineering parameters of sands Table 6. Summary of developed equations in this paper. Parameters Correlation Type Equations Determination Coefficient (R2) Dr - qd Semi-Logarithmic Dr(%) = -181.18 + 406.38(Logqd ) 0.96 EPLT(i) - qd Logarithmic LogEPLT (i) (kPa) = 2.50 + 0.43Logqd 0.90 EPLT(R2) - qd Semi-Logarithmic Eplt (R2) (kPa) = -338705 + 114071Logqd 0.96 EPLT(i)-EPLT(R2) Semi-Logarithmic Eplt )(kPa) = -934442 + 248325LogEPLT (, (kPa) 0.94 GPLT - qd Logarithmic LogGPLT (kPa) = 0.39 + 0.897Logqd 0.92 V - qd Semi-Logarithmic Phi(Deg ) = -10.56 + 83.31Log (Logqd ) 0.87 Conf. on Soil Mechanics and Foundation Engineering, London, 1, 231-236. [2] Mohammadi, S.D., Fityus, S.G., Bates, L. 2012. Calibration of the Perth Sand Penetrometer (PSP) for Silica Sands. Iranian journal of Sciences and Technology, Transaction Civil Engineering, 36,13-23. [3] Sawangsuriya, A., Edil, T.B. 2005. Evaluating stiffness and strength of pavement materials. Geotech-nical Engineering, 157, 217-230. [4] AS1289.F3.3. 1984. Determination of the penetration resistance of a soil with a Perth Sand Penetrometer. Standards Association of Australia. [5] Fakher, A., Khodaparast, M., Jones, CJFP. 2006. The use of the Mackintosh Probe for site investigation in soft soils. Quarterly Journal Engineering Geology and Hydrology 39, 189-196. [6] Lee, C., Kim, K.S., Woo, W., Lee, W. 2014. Soil Stiffness Gauge (SSG) and Dynamic Cone Penetrometer (DCP) tests for estimating engineering properties of weathered sandy soils in Korea. Engineering Geology 169, 91-99. [7] Mohammadi, S.D., Nikudel, M.R., Rahimi, H., Khamehchiyan, M. 2008. Application of the Dynamic Cone Penetrometer (DCP) for determination of the engineering parameters of sandy soils. Engineering Geology 101, 195-203. [8] Mohammadi, S.D., Nikudel, M.R., Rahimi, H., Khamehchiyan, M. 2012. Unit Skin Friction from the Extended Dynamic Cone Penetrometer (EDCP) Test Supplemented by Measurement of Torque within testing wells. Iranian Journal of Sciences and Technology Transaction Civil Engineering 36, 115-119. [9] Mohammadi, S.D., Nikudel, M.R., Rahimi, H. 2009. Assesment of the effect of DCP tip angle on its results repeatability. Journal of Iranian Association of Engineering Geology 1, 2, 31-41 (in Persian). 10] Baecher, B.G., Christian, J.T. 2003. Reliability and statistics in geotechnical engineering. John Wiley & Sons Ltd, England. 11] Esmailian, M. 2006. SPSS 14 guide book. Tehran Dibagaran, Tehran, Iran (in Persian). 12] Khodaparast, M. 2006. Expanding of soil geotechnical identification techniques using Dynamic probes. Dissertation, Tehran University, (in Persian). 13] Lee, I.K., White, W., Ingles, O.G. 1983. Geotechnical Engineering. Copp Clark Pitman; Incorporated. 14] Abu-farsakh, M., Khalid Alshibi, P.E., Nazzal, M., Seyman, E. 2004. Assessment of in-situ test technology for construction control of base courses and embankments. Report No, FHWA/LA.04/385, Louisiana Transportation Research Center. 15] Bowles, J.E. 1997. Foundation analysis an design. 5th edition, McGraw-Hill International Editions. 16] ASTM 1998. Standard test method for repetitive static plate load tests of soils and flexible pavement components, for use in evaluation and design of airport and highway pavements (D1195-93). Annual Book of ASTM Standards 04.08, pp. 110-113. 17] Timoshenko, S.P., Goodier, J.N. 1970. Theory of Elasticity. 3rd Edition, Mc Graw-Hill Book company, New York. 18] American Society of Testing and Materials. 2000. Standard test method for direct shear test of under drained conditions (D3080-98). Annual book of ASTM standards 04.08, pp. 894-904. 19] Coduto, P.D. 2004. Foundation design, principal and practices. Second Edition, Prentice Hall, New Jersey. 78. Acta Geotechnica Slovenica, 2016/1 NAVODILA AVTORJEM NAVODILA AVTORJEM Vsebina članka Članek naj bo napisan v naslednji obliki: - Naslov, ki primerno opisuje vsebino članka in ne presega 80 znakov. - Izvleček, ki naj bo skrajšana oblika članka in naj ne presega 250 besed. Izvleček mora vsebovati osnove, jedro in cilje raziskave, uporabljeno metodologijo dela, povzetek izidov in osnovne sklepe. - Največ 6 ključnih besed, ki bi morale biti napisane takoj po izvlečku. - Uvod, v katerem naj bo pregled novejšega stanja in zadostne informacije za razumevanje ter pregled izidov dela, predstavljenih v članku. - Teorija. - Eksperimentalni del, ki naj vsebuje podatke o postavitvi preiskusa in metode, uporabljene pri pridobitvi izidov. - Izidi, ki naj bodo jasno prikazani, po potrebi v obliki slik in preglednic. - Razprava, v kateri naj bodo prikazane povezave in posplošitve, uporabljene za pridobitev izidov. Prikazana naj bo tudi pomembnost izidov in primerjava s poprej objavljenimi deli. - Sklepi, v katerih naj bo prikazan en ali več sklepov, ki izhajajo iz izidov in razprave. - Vse navedbe v besedilu morajo biti na koncu zbrane v seznamu literature, in obratno. Dodatne zahteve - Vrstice morajo biti zaporedno oštevilčene. - Predložen članek ne sme imeti več kot 18 strani (brez tabel, legend in literature); velikost črk 12, dvojni razmik med vrsticami. V članek je lahko vključenih največ 10 slik. Isti rezultati so lahko prikazani v tabelah ali na slikah, ne pa na oba načina. - Potrebno je priložiti imena, naslove in elektronske naslove štirih potencialnih recenzentov članka. Urednik ima izključno pravico do odločitve, ali bo te predloge upošteval. Enote in okrajšave V besedilu, preglednicah in slikah uporabljajte le standardne označbe in okrajšave SI. Simbole fizikalnih veličin v besedilu pišite poševno (npr. v, T itn.). Simbole enot, ki so sestavljene iz črk, pa pokončno (npr. Pa, m itn.). Vse okrajšave naj bodo, ko se prvič pojavijo, izpisane v celoti. Slike Slike morajo biti zaporedno oštevilčene in označene, v besedilu in podnaslovu, kot sl. 1, sl. 2 itn. Posnete naj bodo v katerem koli od razširjenih formatov, npr. BMP, JPG, GIF. Za pripravo diagramov in risb priporočamo CDR format (CorelDraw), saj so slike v njem vektorske in jih lahko pri končni obdelavi preprosto povečujemo ali pomanjšujemo. Pri označevanju osi v diagramih, kadar je le mogoče, uporabite označbe veličin (npr. v, T itn.). V diagramih z več krivuljami mora biti vsaka krivulja označena. Pomen oznake mora biti razložen v podnapisu slike. Za vse slike po fotografskih posnetkih je treba priložiti izvirne fotografije ali kakovostno narejen posnetek. Preglednice Preglednice morajo biti zaporedno oštevilčene in označene, v besedilu in podnaslovu, kot preglednica 1, preglednica 2 itn. V preglednicah ne uporabljajte izpisanih imen veličin, ampak samo ustrezne simbole. K fizikalnim količinam, npr. t (pisano poševno), pripišite enote (pisano pokončno) v novo vrsto brez oklepajev. Vse opombe naj bodo označene z uporabo dvignjene številke1. Seznam literature Navedba v besedilu Vsaka navedba, na katero se sklicujete v besedilu, mora biti v seznamu literature (in obratno). Neobjavljeni rezultati in osebne komunikacije se ne priporočajo v seznamu literature, navedejo pa se lahko v besedilu, če je nujno potrebno. Oblika navajanja literature V besedilu: Navedite reference zaporedno po številkah v oglatih oklepajih v skladu z besedilom. Dejanski avtorji so lahko navedeni, vendar mora obvezno biti podana referenčna številka. Primer: ».....kot je razvidno [1,2]. Brandl and Blovsky [4], sta pridobila drugačen rezultat...« V seznamu: Literaturni viri so oštevilčeni po vrstnem redu, kakor se pojavijo v članku. Označimo jih s številkami v oglatih oklepajih. Sklicevanje na objave v revijah: [1] Jelušič, P., Žlender, B. 2013. Soil-nail wall stability analysis using ANFIS. Acta Geotechnica Slovenica 10(1), 61-73. 88. Acta Geotechnica Slovenica, 2016/1 INSTRUCTIONS FOR AUTHORS Sklicevanje na knjigo: [2] Šuklje, L. 1969. Rheological aspects of soil mechanics. Wiley-Interscience, London Sklicevanje na poglavje v monografiji: [3] Mitchel, J.K. 1992. Characteristics and mechanisms of clay creep and creep rupture, in N. Guven, R.M. Pollastro (eds.), Clay-Water Interface and Its Rheological Implications, CMS Workshop Lectures, Vol. 4, The clay minerals Society, USA, pp. 212-244.. Sklicevanje na objave v zbornikih konferenc: [4] Brandl, H., Blovsky, S. 2005. Slope stabilization with socket walls using the observational method. Proc. Int. conf. on Soil Mechanics and Geotechnical Engineering, Bratislava, pp. 2485-2488. Sklicevanje na spletne objave: [5] Kot najmanj, je potrebno podati celoten URL. Če so poznani drugi podatki (DOI, imena avtorjev, datumi, sklicevanje na izvorno literaturo), se naj prav tako dodajo. INSTRUCTIONS FOR AUTHORS Format of the paper The paper should have the following structure: - A Title, which adequately describes the content of the paper and should not exceed 80 characters; - An Abstract, which should be viewed as a mini version of the paper and should not exceed 250 words. The Abstract should state the principal objectives and the scope of the investigation and the methodology employed; it should also summarise the results and state the principal conclusions; - Immediately after the abstract, provide a maximum of 6 keywords; - An Introduction, which should provide a review of recent literature and sufficient background information to allow the results of the paper to be understood and evaluated; - A Theoretical section; - An Experimental section, which should provide details of the experimental set-up and the methods used to obtain the results; - A Results section, which should clearly and concisely present the data, using figures and tables where appropriate; - A Discussion section, which should describe the relationships shown and the generalisations made possible by the results and discuss the significance Podatki o avtorjih Članku priložite tudi podatke o avtorjih: imena, nazive, popolne poštne naslove, številke telefona in faksa, naslove elektronske pošte. Navedite kontaktno osebo. Sprejem člankov in avtorske pravIce Uredništvo si pridržuje pravico do odločanja o sprejemu članka za objavo, strokovno oceno mednarodnih recenzentov in morebitnem predlogu za krajšanje ali izpopolnitev ter terminološke in jezikovne korekture. Z objavo preidejo avtorske pravice na revijo ACTA GEOTECHNICA SLOVENICA. Pri morebitnih kasnejših objavah mora biti AGS navedena kot vir. Vsa nadaljnja pojasnila daje: Uredništvo ACTA GEOTECHNICA SLOVENICA Univerza v Mariboru, Fakulteta za gradbeništvo, prometno inženirstvo in arhitekturo Smetanova ulica 17, 2000 Maribor, Slovenija E-pošta: ags@uni-mb.si of the results, making comparisons with previously published work; - Conclusions, which should present one or more conclusions that have been drawn from the results and subsequent discussion; - A list of References, which comprises all the references cited in the text, and vice versa. Additional Requirements for Manuscripts - Use double line-spacing. - Insert continuous line numbering. - The submitted text of Research Papers should cover no more than 18 pages (without Tables, Legends, and References, style: font size 12, double line spacing). The number of illustrations should not exceed 10. Results may be shown in tables or figures, but not in both of them. - Please submit, with the manuscript, the names, addresses and e-mail addresses of four potential referees. Note that the editor retains the sole right to decide whether or not the suggested reviewers are used. Units and abbreviations Only standard SI symbols and abbreviations should be used in the text, tables and figures. Symbols for physical quantities in the text should be written in Italics (e.g. v, T, etc.). Symbols for units that consist of letters should Acta Geotechnica Slovenica, 2016/1 89. INSTRUCTIONS FOR AUTHORS be in plain text (e.g. Pa, m, etc.). All abbreviations should be spelt out in full on first appearance. Figures Figures must be cited in consecutive numerical order in the text and referred to in both the text and the caption as Fig. 1, Fig. 2, etc. Figures may be saved in any common format, e.g. BMP, JPG, GIF. However, the use of CDR format (CorelDraw) is recommended for graphs and line drawings, since vector images can be easily reduced or enlarged during final processing of the paper. When labelling axes, physical quantities (e.g. v, T, etc.) should be used whenever possible. Multi-curve graphs should have individual curves marked with a symbol; the meaning of the symbol should be explained in the figure caption. Good quality black-and-white photographs or scanned images should be supplied for the illustrations. Tables Tables must be cited in consecutive numerical order in the text and referred to in both the text and the caption as Table 1, Table 2, etc. The use of names for quantities in tables should be avoided if possible: corresponding symbols are preferred. In addition to the physical quantity, e.g. t (in Italics), units (normal text), should be added on a new line without brackets. Any footnotes should be indicated by the use of the superscript1. LIST OF references Citation in text Please ensure that every reference cited in the text is also present in the reference list (and vice versa). Any references cited in the abstract must be given in full. Unpublished results and personal communications are not recommended in the reference list, but may be mentioned in the text, if necessary. Reference style Text: Indicate references by number(s) in square brackets consecutively in line with the text. The actual authors can be referred to, but the reference number(s) must always be given: Example: "... as demonstrated [1,2]. Brandl and Blovsky [4] obtained a different result ..." List: Number the references (numbers in square brackets) in the list in the order in which they appear in the text. Reference to a journal publication: [1] Jelusic, P., Zlender, B. 2013. Soil-nail wall stability analysis using ANFIS. Acta Geotechnica Slovenica 10(1), 61-73. Reference to a book: [2] Suklje, L. 1969. Rheological aspects of soil mechanics. Wiley-Interscience, London Reference to a chapter in an edited book: [3] Mitchel, J.K. 1992. Characteristics and mechanisms of clay creep and creep rupture, in N. Guven, R.M. Pollastro (eds.), Clay-Water Interface and Its Rheological Implications, CMS Workshop Lectures, Vol. 4, The clay minerals Society, USA, pp. 212-244. Conference proceedings: [4] Brandl, H., Blovsky, S. 2005. Slope stabilization with socket walls using the observational method. Proc. Int. conf. on Soil Mechanics and Geotechni-cal Engineering, Bratislava, pp. 2485-2488. Web references: [5] As a minimum, the full URL should be given and the date when the reference was last accessed. Any further information, if known (DOI, author names, dates, reference to a source publication, etc.), should also be given. Author information The following information about the authors should be enclosed with the paper: names, complete postal addresses, telephone and fax numbers and E-mail addresses. Indicate the name of the corresponding author. Acceptance of papers and copyright The Editorial Committee of the Slovenian Geotechnical Review reserves the right to decide whether a paper is acceptable for publication, to obtain peer reviews for the submitted papers, and if necessary, to require changes in the content, length or language. On publication, copyright for the paper shall pass to the ACTA GEOTECHNICA SLOVENICA. The AGS must be stated as a source in all later publication. For further information contact: Editorial Board ACTA GEOTECHNICA SLOVENICA University of Maribor, Faculty of Civil Engineering, Transportation Engineering and Architecture Smetanova ulica 17, 2000 Maribor, Slovenia E-mail: ags@uni-mb.si 90. Acta Geotechnica Slovenica, 2016/1 89. NAMEN REVIJE AIMS AND SCOPE Namen revije ACTA GEOTECHNICA SLOVENICA je objavljanje kakovostnih teoretičnih člankov z novih pomembnih področij geomehanike in geotehnike, ki bodo dolgoročno vplivali na temeljne in praktične vidike teh področij. ACTA GEOTECHNICA SLOVENICA objavlja članke s področij: mehanika zemljin in kamnin, inženirska geologija, okoljska geotehnika, geosintetika, geotehnične konstrukcije, numerične in analitične metode, računalniško modeliranje, optimizacija geotehničnih konstrukcij, terenske in laboratorijske preiskave. ACTA GEOTECHNICA SLOVENICA aims to play an important role in publishing high-quality, theoretical papers from important and emerging areas that will have a lasting impact on fundamental and practical aspects of geomechanics and geotechnical engineering. ACTA GEOTECHNICA SLOVENICA publishes papers from the following areas: soil and rock mechanics, engineering geology, environmental geotechnics, geosynthetic, geotechnical structures, numerical and analytical methods, computer modelling, optimization of geotechnical structures, field and laboratory testing. Revija redno izhaja dvakrat letno. The journal is published twice a year. AVTORSKE PRAVICE Ko uredništvo prejme članek v objavo, prosi avtorja(je), da prenese(jo) avtorske pravice za članek na izdajatelja, da bi zagotovili kar se da obsežno razširjanje informacij. Naša revija in posamezni prispevki so zaščiteni z avtorskimi pravicami izdajatelja in zanje veljajo naslednji pogoji: Fotokopiranje V skladu z našimi zakoni o zaščiti avtorskih pravic je dovoljeno narediti eno kopijo posameznega članka za osebno uporabo. Za naslednje fotokopije, vključno z večkratnim fotokopiranjem, sistematičnim fotokopiranjem, kopiranjem za reklamne ali predstavitvene namene, nadaljnjo prodajo in vsemi oblikami nedobičk-onosne uporabe je treba pridobiti dovoljenje izdajatelja in plačati določen znesek. Naročniki revije smejo kopirati kazalo z vsebino revije ali pripraviti seznam člankov z izvlečki za rabo v svojih ustanovah. COPYRIGHT Upon acceptance of an article by the Editorial Board, the author(s) will be asked to transfer copyright for the article to the publisher. This transfer will ensure the widest possible dissemination of information. This review and the individual contributions contained in it are protected by publisher's copyright, and the following terms and conditions apply to their use: Photocopying Single photocopies of single articles may be made for personal use, as allowed by national copyright laws. Permission of the publisher and payment of a fee are required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Subscribers may reproduce tables of contents or prepare lists of papers, including abstracts for internal circulation, within their institutions. Elektronsko shranjevanje Za elektronsko shranjevanje vsakršnega gradiva iz revije, vključno z vsemi članki ali deli članka, je potrebno dovoljenje izdajatelja. Electronic Storage Permission of the publisher is required to store electronically any material contained in this review, including any paper or part of the paper. ODGOVORNOST Revija ne prevzame nobene odgovornosti za poškodbe in/ali škodo na osebah in na lastnini na podlagi odgovornosti za izdelke, zaradi malomarnosti ali drugače, ali zaradi uporabe kakršnekoli metode, izdelka, navodil ali zamisli, ki so opisani v njej. RESPONSIBILITY No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of product liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein.