Civil Engineering

1 downloads 0 Views 13MB Size Report
by Amirhosein Yousefzadeh, Rassoul Ajalloeian, Meisam Kabiri ...... A. Yousefzadeh, R. Ajalloeian, M. Kabiri ...... Samir Bouharoun -Tizi Ouzou, (Algeria), 03.
International Review of

Civil Engineering (IRECE)

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

Contents Experimental Determination of the Liquefaction Potential of Sands Using Standard Geotechnical Laboratory Equipment by R. Porras-Soriano, S. López-Querol, R. Blázquez

259

Assessment of Modified Cam-Clay Theory in the Prediction of Settlement in Shallow Foundations by Amirhosein Yousefzadeh, Rassoul Ajalloeian, Meisam Kabiri

268

Analysis of Curved Prestressed Concrete Beams Under Short-Term and Long-Term Conditions by Using of Finite Element Method by M. B. Abdul Rahman, M. R. Abed

277

Influence of Water Content on the Tribological Behavior at Concrete/Wall Interface – Role of Cement Grains by S. Bouharoun

287

Uplift Behaviour of Plate Anchors Embedded in Cohesionless Soils by Baleshwar Singh, Birjukumar Mistri

294

Environmental Noise, Hold Body Vibration & Air Pollution Monitoring within Toll Booths of the Athens Ring Motorway Network: “ATTIKI ODOS” by K. Vogiatzis, N. Eliou

303

Tired Driver’s Behaviour Assessment Using Innovative Instrumentation by N. Eliou, K. Vogiatzis, F. Kehagia

311

Finite Element Analysis for Bearing Capacity of Circular Footing on Geogrid Reinforced Sand by Jawdat K. Abbas, Mohanad N. Al-Shindah

316

International Review of Civil Engineering (I.RE.C.E.), Vol. 2, N. 6 November 2011

Experimental Determination of the Liquefaction Potential of Sands Using Standard Geotechnical Laboratory Equipment R. Porras-Soriano1, S. López-Querol2, R. Blázquez2 Abstract – This paper is aimed to explain a new methodology for evaluating the behaviour of sandy soil under dynamic loadings. In order to do so, only standard soil mechanics laboratory equipment has been used. Currently, the liquefaction potential is determined in the laboratory by means of undrained and drained dynamic triaxial, resonant column or shaking table tests. However, these tests are far from being of generalized use in most soil mechanics laboratories. The equipment employed herein has been the conventional direct shear box, in which the cyclic loading is manually applied to different samples of quartzitic sand subjected to various vertical loading conditions. By correlating the stress distributions in the direct shear box and the simple shear apparatus, a recently proposed liquefaction model has been calibrated, and the sand liquefaction potential has been estimated. Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved.

Keywords: Densification, Cyclic Shear Tests, Liquefaction Potential

I.

The pioneer methods used for modeling the densification were semi-empirical approaches based on the above mentioned test results [6]-[8]. Nevertheless, the first rational analytical model was proposed by Cuéllar [9]-[10], based on the well known endochronic theory initially applied to metals [11]. This law takes into account the dependence of densification on the invariants of the stress and strain tensors, based on experimental findings. Recently, Blázquez and López-Querol [12] have expanded Cuéllar’s method and have developed a generalized densification endochronic law valid for quartzitic sands. Following this type of approach, earlier numerical models for predicting liquefaction potential were proposed by [1], [13]. These researchers were among the first ones who incorporated an inelastic constitutive law for densification in a coupled model for liquefaction, according to Eq. (1). After them, several investigations aiming the improve the numerical efficiency of these laws, and to consider further soil and loading characteristics, have been delivered in the literature [14], [15]. In this paper, the experimental results of a set of “cyclic” shear box tests on a quartzitic sand, with manually applied cyclic loading, are reported. By correlating the shear stress distribution in the shear box test and the simple shear apparatus, the generalized densification law proposed by [12] has been calibrated for a given sand. After obtaining the constrained modulus of the drained sand in unloading by means of oedometer tests, the liquefaction potential of the sand is investigated. The characteristics of the tested sand, as well as the experimental procedures carried out are described first in the paper. Then, the results derived

Introduction

Liquefaction potential of sands subjected to dynamics loadings may be determined by relating the evolution of excess pore water pressure, dpw, under undrained conditions to the volumetric strain under drained conditions (so called “densification”), dεy, through the unloading constrained modulus of the drained sand in unloading, Md [1]: dpw = − M d ⋅ d ε y

(1)

where dpw is taken positive in compression, and dεy is taken positive in extension. The constrained modulus is obtained experimentally in a standard oedometer apparatus, whereas, for evaluating the densification of the sand, cyclic testing devices – usually available only in highly specialized geotechnical laboratories – are required. Densification of sand takes place when the granular soil is subjected to dynamic loadings and it is either drained or not fully undrained. This phenomenon was first studied during the late 60’s and early 70’s, by establishing experimentally the influence of several sand properties on the final amount of plastic volumetric strain of the soil. As a result of these investigations, it was concluded that the key factors controlling densification are the amplitude of the dynamic shear strains, the initial relative density and the number of cycles of shear strain applied to the soil sample, while the dynamic volumetric strains, the frequency of loading and the confining stress seem to have small influence on the phenomenon [2]-[5].

Manuscript received and revised October 2011, accepted November 2011

259

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

R. Porras-Soriano, S. López-Querol, R. Blázquez

from these tests are summarized, and the calibration procedure of the densification law is provided. Finally, several numerical examples, both on dry and saturated sand, are shown in order to assess the validity of the methodology proposed in this paper.

II.

Experimental Procedures

The tested material is a quartzitic sand coming from Arija (Burgos, Spain). Its main properties, namely: specific weight of the solid skeleton (Gs), maximum and minimum particle sizes (Dmax and Dmin), sizes corresponding to percentiles 60, 30 and 10 in the grading curves (D60, D30 and D10, respectively), size distribution coefficients (Cu and Cc), maximum and minimum void ratios (emax and emin), and permeability (k), all of them derived from conventional soil characterization tests, are given in Table I. The grading curve of the sand is given in Fig. 1, where it is compared to the grain size distributions of other sands with various degrees of liquefaction potential [16].

Fig. 2. Oedometer tests results on Arija sand

The standard shear box apparatus has been used for applying cyclic loading to the samples of sand. In these tests, the maximum shear strain is computed (in percent) as the ratio between the maximum displacement of both parts of the box and the sample height, according to Fig. 3:

TABLE I CHARACTERISTICS OF THE ARIJA SAND Parameter Value Gs 2.65 Dmax 2 mm 0.08 mm Dmin D60 0.95 mm D30 0.77 mm D10 0.63 mm 0.99 Cc 1.51 Cu emin 0.57 0.76 emax

γ=

d ⋅100 h

(2)

Fig. 3. Sketch of the sample undergoing a shear box test

In the tests reported herein the maximum shear strain varied between 0.5 and 1.5%. Figures 4 show the shear box apparatus employed in this part of the research. The cyclic loading was applied manually, changing alternatively the shear direction, as indicated in the figure.

Fig. 1. Grading curves of the reported sands, compared to the conventional ranges of liquefiable sands

Standard oedometer tests were made on samples of sand at their maximum and minimum density conditions, subjected to different vertical loadings (100, 500, 800 and 1000 kPa). Figure 2 provides the complete set of results of both loading and unloading paths. It can be noticed in this graphs that, for the range of vertical stresses applied, the loading path does not converge towards a unique Normal Compression Line (NCL), a circumstance which is often observed in this kind of sands for stresses ranging between 10 and 20 MPa [17].

Figs. 4. Experimental procedures in the cyclic shear box test: a) Shear box apparatus. b) Detail of the manual change of direction for application of cyclic loading

The frequency of loading was about 0.01Hz for all tests because of the engine characteristics. Nevertheless an analysis of the sensitivity of the measurements to the frequency was performed for frequency values ranging between 0.01 and 0.05 Hz. The results of this analysis

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

260

R. Porras-Soriano, S. López-Querol, R. Blázquez

are given in Fig. 5, where it becomes quite evident the small influence that exerts the frequency on the final amount of the densification for the very low values used in these tests (50 cycles of vibration each).

noting that there are no differences between dry and saturated tests, for a frequency of loading of about 0.01Hz, as expected. On the basis of this finding, the whole set of tests in this apparatus was performed under dry conditions.

t (seg) 0

t (seg) 0

200

400

600

800

1000

1200

1400

1600

0

1800

0

1000

2000

3000

4000

5000

6000

-0.01

-0.01 -0.02

y (mm)

y (mm)

-0.02 -0.03

-0.04

-0.03

-0.04

-0.05

-0.05

-0.06

-0.07

-0.06

Dr=9.6%; f=0.01Hz

Dr=9.2%; f=0.04Hz

Dr=8.9%; f=0.02Hz

e11, dry, Dr = 8.95

Fig. 5. Comparison of cyclic shear box test results on samples under same initial conditions, using different frequencies of loading

e11, saturated, Dr = 21.1

Fig. 7. Comparison of densification results of dry and saturated (fully drained) samples under the same initial conditions

The horizontal and vertical displacement transducers originally used in the tests were of the type PY2 (GEFRAN), which are rectilinear displacement transducers with ball tip, and 10 to 100 mm stroke range. Since these transducers are only suitable for high strain levels, additional inductive transducers (Schreiber Messtechnik, serial MS223), capable of recording displacements higher than 0.0001 mm, were installed in the equipment (Fig. 6) and compared with the GEFRAN transducers, concluding that the later transducers are accurate enough for the horizontal displacements or the sample, but not for the settlements.

TABLE II INITIAL RELATIVE DENSITY FOR EACH TEST Test Dr % eI 2,90 eII 2,65 eIII 0,43 eIV 37,66 eV 45,67 eVI 42,47 eVII 81,59 eVIII 90,37 eIX 96,18

A sensitivity analysis of the influence of the main parameters controlling the densification on the final results was conducted. This study basically consisted of repeating several times (usually three) each test, under different initial conditions of vertical stress and relative density. Nine tests were carried out, the vertical stress ranging from 10 to 1000 kPa, and the initial relative density varying roughly between the maximal and the minimal relative densities of the sand (Table II). It was found from the experiments that the influence of the setting up process is smaller in the conditions of maximal initial relative density (three right graphs in Fig. 8), while for the other cases the final amount of densification seems to be less reliable. From the inspection of this figure, it also becomes clear that the vertical effective stress exerts a small influence on the final amount of densification. This assertion agrees well with a similar conclusion reached by [5], although the effective stresses applied in the latter research were much lower (zero to 16.8 kPa) than the ones used in the present investigation. In the following, all the experimental results reported in this investigation correspond to a maximal effective stress of the soil of 1000 kPa. In Figs. 9 and 10 the results of the densification strain against the number of cycles are given different initial conditions of the tests.

Fig. 6. Location of the vertical transducers on the shear box test

III. Experimental Results An analysis of the influence of saturated or dry conditions on the shear box tests has been carried out. Two tests were run using a vertical stress of 100 kPa, initial relative densities of 8.9 and 21.1 %, and dry and saturated conditions, respectively. The results in terms of densification against the number of cycles are shown in Fig. 7, where it is worth

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

261

R. Porras-Soriano, S. López-Querol, R. Blázquez

Fig. 8. Results of cyclic shear box tests

Similarly, Figs. 10(a), 10(b) and 10(c) show the variation of the densification against the number of cycles, for shear strain amplitudes of 0.5, 1 and 1.5 %, respectively, and different initial relative densities of the sand (maximum, mean and minimum). In most of the cases, it is found that the higher the initial relative density, the smaller the densification, as should be anticipated. This trend is not followed in Fig. 10(a). This result, however, is not very trustable, since represents the densification obtained for the lowest values of the shear strain amplitude, γ.

The shear strain ranged between 0.5 and 1.5%, while the initial relative density is the maximum, mean or minimum value of the soil sample. Figures 9(a), 9(b) and 9(c) show the temporal variation of the densification with the initial relative density of the soil. In each of these figures, the densification amount is plotted for shear strain amplitudes of 0.5, 1 y 1.5%, respectively. From the inspection of the graphs it is concluded that the larger the amplitude of the shear strain, the larger the densification.

(a) (a)

(b) (b)

(c) Figs. 9. Influence of the amplitude of the cyclic shear strain in the densification: (a) Maximum relative density, (b) Mean relative density, (c) Minimum relative density

(c) Figs. 10. Influence of the initial relative density in the densification: (a) Minimum amplitude of γ, (b) Mean amplitude of γ, (c) Minimum amplitude of γ

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

262

R. Porras-Soriano, S. López-Querol, R. Blázquez

IV.

of cyclic strain tests, are incrementally computed as follows:

Calibration of Constitutive Model

The shear stress distributions in the shear box and the simple shear stress apparatus have been compared, in order to correlate the experimental data reported in the present study with the ones corresponding to simple shear conditions. This comparison was made by means of a finite element code, in which elastic plane strain conditions were assumed at the central plane of the sample, since the strains measured in all the tests reported before are quite small. The final element mesh consists of linear triangles, while the boundary conditions differ for the simple shear and shear box tests. In both cases, the total applied force is the same, namely the force that generates a shear stress of 1Pa in the horizontal plane (Figs. 11). It should be noted that the boundary conditions for the shear box test are different from the ones considered in other studies, aimed to model the sample behavior for high strain levels, and therefore, imposing symmetry conditions to the model [18]. The non-symmetry hypothesis is only valid for very small strain ranges, like the ones dealt with in the present research. Figures 12 represent the shear stress distributions for the tests reported above. As expected, the shear stress distribution is more uniform in the case of simple shear conditions than in the shear box test.

dξ = d γ

( ( )) × dξ

d ζ = F1 J 2 d ε ij

(3) (4)

where F1 is an analytical function, J2 is the second invariant of the tensor in brackets, and εij is the strain tensor.

Figs. 12. Shear stress distribution within the samples: (a) in the simple shear test, (b) in the shear box test

The densification, ε ´y , is incrementally defined as

Figs. 11. Dimensions of the sample and boundary conditions, a) in the shear box test, b) in the simple shear test (dimensions in cm)

follows:

d ε ´y = − F2 (ζ ,Dr0 ) × d ζ

The correlation between both tests can also be established numerically. In order to do so, the absolute value of the mean normalized shear stress is computed in each case. Since this magnitude reaches values of 0.9543 and 0.292 for simple shear and shear box tests, respectively, the factor relating both tests is 0.9543/0.2920 = 3.268. Hence, the volumetric strains obtained in the shear box apparatus must be multiplied by this factor to be converted to equivalent simple shear strains. For calibration purposes, the densification law proposed by [9] and generalized by [12] for quartzitic sands is employed. This formulation is based on the numerical calculation of two endochronic monotonically increasing functions, named rearrangement measure, ξ, and densification measure, ζ, which, for the specific case

(5)

where F2 is an analytical function, and Dr0 is the initial relative density of the sand before the application of the dynamic loading. Functions F1 and F2 are defined as: F1 =

n ( n −1) × 100 × γ 4

F2 = 1 / (1 + α × ζ )

(6) (7)

where n and α are the densification law parameters, and γ is the shear strain. For the cyclic shear strain test:

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

263

R. Porras-Soriano, S. López-Querol, R. Blázquez

(

γ = 8 × J 2 ( ε ij )

)

1/ 2

V.

(8)

In order to assess the validity of the proposed methodology, two numerical examples have been analyzed, namely, densification and liquefaction cases of horizontal sandy soils under field conditions.

Rearranging the above equations, yields: d ε ´y = −

( n −1)

dζ n 100 × γ =− × (1 + α × ζ ) 4 1 + α × ζ

× dγ

(9)

V.1.

where: n n −1 ζ = ∫ d ζ = ∫ × 100 × γ ( ) × d γ

According to [12], for quartzitic sand the densification law parameters, n and α, can be related to several characteristics of the material (maximum and minimum void ratios, and initial relative density), as follows: n = A × ln ( N ) + B

(11)

α = C×ND

(12)

Densification of a Horizontal Layer of Dry Sand Under Seismic Loading

The first example is the case of a horizontal layer of 15 meters of dry sand resting over rock, and subjected to the San Fernando earthquake (1971; N69W component) at the rigid base. The geometry of this case is sketched in Fig. 14. This case has been reported previously in the literature by other researchers, using Crystal Silica No. 20 sand as ground soil [6], [8], [15], [12].

(10)

4

Validation of the Developed Methodology

where A, B and D are sand dependent values which can easily be computed by means of the maximum, minimum and current void ratios, and C is the only remaining densification law parameter which requires to be calibrated by means of cyclic shear strain tests, and it has been found that varies linearly with the initial relative density, according to: C = C* − 50 ⋅ Dr0 TABLE III CALIBRATION OF PARAMETER C * FOR CYCLIC SHEAR DENSIFICATION TESTS Dr C* Test γ % eI 0.03 0.50 0.50 eII 0.03 1.00 0.00 eIII 0.00 1.50 0.00 eIV 0.38 0.50 18.00 eV 0.46 1.00 18.00 eVI 0.42 1.50 18.00 eVII 0.82 0.50 18.00 eVIII 0.90 1.00 18.00 eIX 0.96 1.50 18.00

(13)

C 1.45 1.32 0.21 0.83 4.84 3.24 22.79 27.19 30.09

Fig. 13. Calibration process for parameter, C*, of the densification law

In turn, the densification law parameter, C*, is determined by fitting the numerical simulations to the experimental data obtained before. Figure 13 and Table III summarize this calibration. It is clear from this Table that, for the three smaller initial relative densities, the calibration gives a rather small value of C*. This is because the constitutive law used in this research is restricted to values of the initial relative density of the sand observed in field conditions, which are usually bigger than 0.40. Fig. 14. Sand deposit used in the validation problems

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

264

R. Porras-Soriano, S. López-Querol, R. Blázquez

V.2.

Before engaging in numerical computations, a comparison of the expected densification of Cristal Silica No. 20 sand with that of Arija sand, both under cyclic shear conditions, has been addressed. In order to do so, the experimental results of [6] are compared with the results of the tests described in this paper, modified by the correction factor derived in the last section. Figure 1 shows the grading curves of both sands. From the inspection of this figure, it is self evident that Arija sand is less liquefiable, attending to the range of grading curves for most liquefiable sands. Therefore, less densification is expected to occur. This prediction is confirmed by Fig. 15, where the densification of Arija sand (mean initial relative density of about 50%) and Crystal Silica No. 20 sand (initial relative densities of 45% and 60%) after 50 cycles of loading are compared. Figure 15 shows clearly that, as should be, for the same range of shear strain amplitudes, Crystal Silica sand densifies much more than Arija sand. The densification law previously described has been implemented in a simple one dimensional finite element code, the finite elements having two nodes each. The accelerogram applied at the rigid base of the field problem (Fig. 14) is given in Fig. 16. The computed settlement is clearly bigger for Crystal Silica sand (8 cm) than for Arija sand (5 cm). This result, at least qualitatively, proves the validity of the proposed methodology, because the numerical results fit very well the expected behaviour.

Liquefaction of a Horizontal Layer of Saturated Sand Under Seismic Loading

The performance of a soil profile similar to the one which liquefied during the Magnitude 7 Manica earthquake (February 2006), in Mozambique [19], [20] is analyzed next. The geometry is again a horizontal sandy layer, 15 meters deep, but the sand in this case is saturated instead of dry. The initial relative density at the soil surface is supposed to be of 45%. As in the former case, the grading curves of both sands are compared in Fig. 1. It can be conjectured from this figure that Arija sand is less liquefiable than Mozambique sand. The constitutive model for liquefaction analysis used in this research is the one developed by [21]. All the model parameters for Arija sand are calibrated using the experimental results for contractive behaviour (densification) previously described, as well as the oedometer tests results, except Bg and Ct, which control dilative sand response, and are of small influence in loose sand. For these parameters, typical values for quartzitic sand have been used. All the parameters, for both sands, are given in Table IV. TABLE IV LIQUEFACTION CONSTITUTIVE LAW PARAMETERS OF THE MOZAMBIQUE AND ARIJA SANDS Parameter Mozambique Arija emin 0.339 0.57 emax 0.839 0.76 Gs 2.64 2.65 C* 30 18 1.387×105 Kd 4.087×106 m 1.288 0.7 2.5 0.5 Ct 175 200 Bg 7.53×10-3 k (m/s) 2.12×10-4

In Fig. 17 the accelerogram used for numerical computations is plotted [20].

Fig. 15. Comparison of the experimental densification results of Arija and Crystal Silica sands

Fig. 17. Horizontal accelerogram sof the Irpinia earthquake (23rd November 1980) (Ambraseys et al. 2004)

The numerical results are presented in Figs. 18, where the isochrone lines (lines of excess pore water pressure at the same time) and the initial effective stresses are drawn, for both sands. While for Mozambique sand liquefaction is reached between t=7.5 and t=8.2 seconds (the corresponding isochrone lines touch the initial effective stress line), for Arija sand liquefaction does not

Fig. 16. Horizontal accelerogram (Component N69W) of the San Fernando earthquake (9th February 1971), recorded at Castaic station

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

265

R. Porras-Soriano, S. López-Querol, R. Blázquez

occur at all during the loading time, according to its lower liquefaction potential. Once again, this numerical example qualitatively shows the validity of the proposed methodology.

support the validity of the methodology proposed in this research. Further comparisons with the results of undrained cyclic tests are required in order to quantitatively complete this validation.

Acknowledgements This study has been partially supported by the projects “Nuevo método para evaluar el potencial de licuefacción de las arenas mediante equipos convencionales en laboratorios de mecánica de suelos” (University of Castilla La Mancha), “Diseño dinámico de geoestructuras viarias en Castilla La Mancha. Aspectos acústicos” (Regional Government of Castilla La Mancha) and “Presiones del terreno y movimiento de muros rígidos sometidos a acciones sísmicas” (Spanish Ministry of Public Works). The assistance in carrying out the laboratory tests, provided by the technician Óscar Merlo Espinosa, is gratefully appreciated by the authors.

References [1]

[2]

[3]

Figs. 18. Isochrone lines and initial effective stress lines for validation problems

VI.

Conclusion

[4]

This paper presents a new methodology for determining the liquefaction potential of sands using standard soil mechanics laboratory equipment. From the experimental and theoretical research conducted, the following conclusions can be advanced: - The densification of dry sand under cyclic loading increases with the number of cycles (faster during the first cycles), and also, with smaller initial relative densities and higher amplitudes of cyclic shear strains. In addition, for the low frequencies employed in the tests, it is concluded that the frequency of loading exerts a negligible influence in the final amount of densification. - The higher the initial relative density, the more reliable the experimental data. - The vertical stress applied to the sample has almost no effect on the densification of the sand for the range of 10 to 1000 kPa. - A corrective factor of 3.268 must be applied to the densification data derived from “cyclic” shear box tests for converting them to equivalent cyclic simple shear test data, on the basis of the assumption of very small strains taking place in the sample. - The numerical computations of densification and liquefaction of sand layers in the field using the generalized endochronic constitutive law seems to

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

G.R. Martin, W.D.L. Finn, and H.B. Seed, Fundamentals of liquefaction under cyclic loading, Journal of the Geotechnical Engineering Division, Vol. 101(GT5):423-438, 1975. H.Y. Ko, and R.F. Scott, Deformation of sand in hydrostatic compression, Journal of the Soil Mechanics and Foundations Division, Vol. 93(SM3):137-156, 1967. B.J. Greenfield, and E.T. Misiaszek, Vibration-settlement characteristics of four gradations of Ottawa sand, International Symposium on Wave Propagation and Dynamic Properties of Earth Materials, University of New Mexico, pp.787-795, Alburquerque, N.M., 1968. G.F. Anderson, An Earthquake Hazard Study of a Dune Sand Site in San Francisco, Ph.D. dissertation, San Jose State College, San Jose, California, 1969. M. Silver and H.B. Seed, Volume changes in sand during cyclic loadings, Journal of the Soil Mechanics and Foundations Division 1971; Vol. 97(SM9):1171-1182, 1971. H.B. Seed and M.L. Silver ML, Settlement of dry sands during earthquakes, Journal of the Soil Mechanics and Foundations Division, Vol. 98:381-396, 1972. T.L. Youd, Compaction of sands by repeated shear straining, Journal of the Soil Mechanics and Foundations Division, Vol. 98(SM7):709-725, 1972. K. Tokimatsu and H.B. Seed HB, Evaluation of settlements in sands due to earthquake shaking, Journal of Geotechnical Engineering, Vol. 113(8):861-878, 1987. V. Cuéllar, Rearrangement measure theory applied to dynamic behavior of sand, Ph.D. dissertation, Civil Engineering Department, Northwestern University, Evanston, Illinois, 1974. V. Cuéllar, Simple shear theory for the one-dimensional behavior of dry sand under cyclic loading, Proceedings of International Symposium on Dynamical Methods in Soil and Rock Mechanics, Karlsruhe, West Germany, 1977. K.C. Valanis, A theory of viscoplasticity with a yield surface; Part I. General theory; Part II. Application to mechanical behavior of metals, Archives of Mechanics (Archiwum Mechaniki Stosowanej) Vol. 23(4):517-555, 1971. R. Blázquez and S. López-Querol, Generalized densification law for dry sand subjected to dynamic loading, Soil Dynamics and Earthquake Engineering, Vol. 26:888-898, 2006. O.C. Zienkiewicz, C.T. Chang and E. Hilton, Non-linear seismic response and liquefaction, International Journal for Numerical and Analytical Methods in Geomechanic, Vol.2:381-404, 1978. J. Przewlocki and W. Knabe, Settlement of a soil stratum subjected to an earthquake, International Journal for Numerical

International Review of Civil Engineering, Vol. 2, N. 6

266

R. Porras-Soriano, S. López-Querol, R. Blázquez

[15]

[16]

[17]

[18]

[19]

[20]

[21]

and Analytical Methods in Geomechanics Vol.19(11):813821, 1995. E. Vincens, P. Labbé and B. Cambou, Simplified estimation of seismically induced setttlemens, International Journal for Numerical and Analytical Methods in Geomechanics, Vol.27:669683,2003. Japan Port and Harbour Association. Technical standards for port and harbour facilities in Japan 1999, Japan Port and Harbour Association, Yokosuka, 1999:281-288 (in Japanese). M.R. Coop, On the mechanics of reconstituted and natural sands. Keynote Lecture, in Deformation Characteristics of Geomaterials, Di Benedetto, H., Doanh, T., Geoffroy, H. and Sauéat, C. eds, Swets and Zeitlinger, Lisse, Vol.2, pp. 29-58,2003. D.M. Potts, G.T. Dounias and P.R. Vaughan, Finite element analysis of the direct shear box test, Geotechnique, Vol.37(1):1123,1987. C. Fenton, and J.J. Bommer, The Mw 7 Machaze, Mozambique, earthquake of 23 February 2006, Seismological Research Letters, Vol.77:426-439,2006. S. López-Querol, M.R. Coop, J.J. Bommer, C. Fenton and W.W. Sim, Back-analysis of liquefaction in the 2006 Mozambique earthquake. Georisk: Assessment and Management of Risk for Engineered Systems and Geohazards, Vol.1(2):89-101,2007. S. López-Querol and R. Blázquez, Liquefaction and cyclic mobility model for saturated granular media, International Journal for Numerical and Analytical Methods in Geomechanics, Vol.30(5):413-439,2006.

Rocio Porras-Soriano (Córdoba – Spain, 1982). Civil Engineer, University of Castilla-La Mancha (Spain), 2005. Master in Civil Engineering, University of Castilla-La Mancha (Spain), 2010. Ph.D., Civil Engineering, University of CastillaLa Mancha (Spain), 2011. Currently, she’s main research interests are Computational fracture mechanics, mechanical properties of quasi-brittle materials, material advanced characterization tests. Susana López-Querol (Madrid – Spain, 1972). Ms. Civil Engineering, Polytechnic University of Madrid (Spain), 1996. Ph.D., Civil Engineering, University of CastillaLa Mancha (Spain), 2006. She’s main research interests are soil dynamics, from both experimental and mathematical points of view. She also deals with advanced numerical methods in soil mechanics, and their application to design in geotechnical engineering. Rafael Blázquez (Madrid – Spain, 1946). Ms. Civil Engineering, Polytechnic University of Madrid (Spain), 1971. B.S. in Physisc, Universidad Complutense de Madrid (Spain), 1972. Ph.D., Northwestern University (USA), 1978. He’s main research interests are earthquake engineering and soil dynamics, soil-structure interaction, unsteady phenomena related to soils, numerical integration schemes in time-domain.

Authors’ information 1

Department of Applied Mechanics and Project Engineering, Univ. de Castilla – La Mancha, Avda. Camilo José Cela, s/n, 13071, Ciudad Real, Spain. E-mail: [email protected] 2 Department of Civil Engineering, Univ. de Castilla – La Mancha, Avda. Camilo José Cela, s/n, 13071, Ciudad Real, Spain, E-mails: [email protected] [email protected]

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

267

International Review of Civil Engineering (I.RE.C.E.), Vol. 2, N. 6 November 2011

Assessment of Modified Cam-Clay Theory in the Prediction of Settlement in Shallow Foundations Amirhosein Yousefzadeh1, Rassoul Ajalloeian2, Meisam Kabiri1 Abstract – Prediction and control of the excessive settlements are very important in the design of different structures. In recent decades, many ways of prediction of settlement have been developed by researchers. Using soil models via numerical methods for prediction of soil behavior is one of the newest of them. In this regard the critical state models have been used widely for this purpose. The Modified Cam-clay model is an elastic plastic strain hardening model that is based on Critical state theory. This model is used in geotechnical engineering practice. Nevertheless, due to some reasons using of this model is in doubt for practical issues. In this paper, a pair of circular and rectangular footings has been modeled via finite-element method and Modified Cam-clay model in a case study. In the next step, the predicted and monitored settlements of these footings were compared. Finally, relatively good and conservative performance of Modified Cam-clay theory is shown in practice, as the main aim of this paper. Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved. Keywords: Modified Cam-Clay Theory, Finite Element Method, Settlement

I.

This model is widely referenced and has been used in solving boundary value problems in geotechnical engineering practice (e.g., [11], [12], [13]). Nevertheless, this model was developed originally for reconstituted clays, and its usage was in doubt for practical purposes [14]. In “Iran” soft clays are fairly widespread, and some of these deposits exist extensively near capital cities. So using Modified Cam-clay model to predict the behavior of soil can be very beneficial in practice. Even so, this model is used with great caution by designers. In this regard, comparison of predicted and actual results in case studies is very important to give a good scale of the accuracy of predicted results. This paper presents a finite-element analysis of the settlement of a pair of the typical aeration tank and primary settling tank in the Shiraz wastewater treatment plant and compares, predicted and observed results. For this purpose, an extensive series of laboratory and in-situ tests have been performed to determine soil properties. In addition, the desired locations have been monitored during three years to measure the actual settlements.

Introduction

One of the most important aspects of foundation engineering deals with the estimation of settlement of foundations. In other words, settlement frequently controls the design of spread footings, and there are specifications for allowable levels of settlement especially when designing a wide footing. Geotechnical engineers usually face challenges to choose the most appropriate strategy to estimate the settlement of structures. During the past decades, many efforts have been conducted to develop methods to estimate the settlement of footings (e.g., [1], [2], [3], [4], [5], [6], [7], [8]). One of the earliest attempts to assess settlement potential in shallow foundations consisted of conducting the plate load tests. Nowadays, nearly all settlement analyses are based on the results of laboratory or in-situ testing. The laboratory methods are based on the results of consolidation tests, and the in-situ methods consist of standard penetration, cone penetration, or other in-situ tests. In recent decades, using soil models via numerical methods for prediction of soil behavior has been increased. For example, Cam-clay models and most models subsequently developed within the critical state framework have been used frequently in finite-element codes. The Modified Cam-clay model was proposed by Roscoe and Burland [9] and systematic study of the model can be found in the text by Muir Wood [10]. Formulation of this model is suitable for use in finite element analysis.

II.

Geotechnical Properties of Shiraz Industrial State

The location of the industrial state in the south west part of Shiraz is shown in Fig. 1. An aerial photograph of the wastewater treatment plant is shown in Fig. 2. Studies and surveys conducted at the waste water treatment through

Manuscript received and revised October 2011, accepted November 2011

268

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

A. Yousefzadeh, R. Ajalloeian, M. Kabiri

the drilling of three boreholes. The borehole positions are marked on Fig. 2. The Percussion bit drilling was used to excavate boreholes (e.g.; see [15]). Sampling was done at different depths. The sampling technique was used to obtain undisturbed and disturbed samples, according to the soil type. In addition during the drilling, standard penetration test (SPT) was performed according to ASTM D 158608a [16].

Fig. 2. Aerial photograph of the water treatment

Groundwater table was controlled during the drilling which it was about 6 m under ground level. Table I shows the soil profile properties as determined from the borehole logs at the sampling site in Shiraz industrial state. Almost all samples were classified CL (based on Unified soil classification system CL is clay with low plasticity).

Fig. 1. Location of sampling site

TABLE I SOIL PROFILE PROPERTIESAS DETERMINED FROM BOREHOLE LOGS BH1 Elevation (m)

S.P.T

PI (%)

BH2 LL (%)

MC (%)

S.P.T

1 8 29 2 4 24 5 3 11 30 4 6 25 4 5 11 32 6 11 28 10 7 12 33 8 13 36 25 9 13 32 10 15 34 9 11 15 34 12 14 37 14 13 17 30 14 13 33 13 15 15 36 16 16 34 18 17 13 32 18 13 33 17 Note: PI and LL are Plasticity Index and Liquid Limit, and MC is Moisture Content

BH3

PI (%)

LL (%)

8

28

9

32

8

27

9

29

MC (%)

S.P.T

20

5

25 28 32 10 9 11 14 12

32 28 32 33 29

32 35 34 33 35

PI (%)

LL (%)

10

32

9

30

10

29

13

30

11

31

11

32

14

28

13

31

14

33

24

4

23

4

27

6

38

12

35

16

35

17

37

16 22

MC (%)

34 35

Triaxial test is one of the most important geotechnical laboratory tests, and it is used extensively in the laboratory testing of cohesive soil. In this section, the results of consolidated undrained triaxial tests with considering pore water pressure are presented in Figs. 3. These tests were carried out in a triaxial cell according to ASTM D4767-02[17]. Two triaxial tests were performed on samples of boreholes 1 and 3. Properties of soil Samples used in the triaxial test are presented in Table III. It is possible to derive the Cam-clay parameters (λ, κ and M) from triaxial results. This is described in the next section.

III. Consolidation Characteristics Two types of consolidation tests, i.e., oedometer test and anisotropic consolidation test (triaxial test) were carried out in the present study. The oedometer test is popular because of its simplicity and its usage in modeling and predicting the behavior of the in-situ soil. Moreover, the oedometer can be used to study the consolidation of clays [15]. According to the results of oedometer tests on undisturbed and remoulded specimens, compression index (C ), swelling index (C ), and permeability index (k), were determined and are presented in Table II.

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

269

A. Yousefzadeh, R. Ajalloeian, M. Kabiri

TABLE II EXTRACTED PARAMETERS FROM OEDOMETER TEST BH2

BH1 SAMPLING

BH3

3

9

15

3

9

15

3

9

15

C

0.324

0.356

0.38

---

---

---

0.308

0.292

0.284

C

0.054

0.072

0.074

---

---

---

0.07

0.076

0.08

gr cm

1.61

1.68

1.66

1.48

1.53

1.56

1.62

1.69

1.66

cm s

10

---

---

----

DEPTH

5.9

Coordinates Of Sampling Place Ground Level Sampling Depth Sampling Level Test Type Back Pressure Cell Pressure Sample Diameter Sample L/D Ratio Soil Classification Soil Particles

10

3.5

10

Deviatoric stress q'(kPa)

2 1,5 (v)

10

6.7

10

5.9

10

TABLE III PROPERTIES OF SOIL SAMPLES USED IN THE TRIAXIAL TEST AND TEST CONDITION BH1 BH3 X=562456.8, X=562476.2, Y=3613553.3 m Y=3613576.9 m 1641.7 m 1641.5 m 9.3 9.5 1632.4 m 1632 m Consolidated Undrainde (CU) Consolidated Undrainde (CU) ----200 kPa 200 kPa 38 mm 38 mm 2 2 CL CL (< 60 m): 32.5% (< 60 m): 31.5%

2,5

1 =0.12 =0.04

0,5 0 0

2.4

2 Ln p' (kPa)

4

80 70 60 50 40 30 20 10 0

M=0.93 0

6

20

40

60

80

Mean effective stress P'(kPa)

(a)

(c) Deviatoric stress q'(kPa)

2,5

(v)

2 1,5 1 =0.14 =0.05

0,5 0 0

2

Ln p' (kPa)

4

80 60 40 20 M=0.99

0 0

6

20

40

60

80

100

120

Mean effective stress P'(kPa)

(b)

(d)

Figs. 3. Triaxial test results: (a), (b) Mean effective stress and specific volume relationship in BH1 and BH3; (c), (d) Mean effective stress and deviatoric strain relationship at failure

IV.

tests. In this research, all laboratory tests were performed by fully-trained, highly-skilled operators to obtain highquality information. This is particularly so when the program is to be used to predict soil behavior in a field situation.

Determining the Critical State Soil Parameters

All of the Modified Cam-clay parameters can be determined from the results of oedometer and triaxial

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

270

A. Yousefzadeh, R. Ajalloeian, M. Kabiri

The obtained values of λ and κ via different methods (oedometer and triaxial tests) in different depths are presented in Tables IV and V.

-The friction constant, M. Triaxial tests on isotropically consolidated samples can be used to obtain the frictional constant M. A number of tests need to be carried out with different consolidation pressures. The p q values should be plotted at failure. The slope of the best fitting straight line is taken as M (see Figs. 3). As a practical point it is necessary to continue these tests to large strains to ensure that the samples are close to the critical state [18]. -Slope of the normal consolidation and swelling lines ( These parameters can be obtained from both oedometer and triaxial tests on soil samples. From theoretical point of view, it is expected to obtain equal values of λ and κ from different sources. Nevertheless, in practice, the values of these parameters are not the same in oedometer and triaxial tests. Therefore, the analyst needs to obtain a ‘best fit’ between all the available data and the critical state parameters, bearing in mind the reliability of each piece of data. Indeed one of the strengths of the critical state theories is this ability to review data from different types of soil test [19]. In order to determine from oedometer results, it is common to plot the results of one-dimensional compression tests in terms of the void ratio (e) versus Log σ , where σ is the effective vertical stress. The slope C of the normally consolidated line is known as the ‘compression index’. The Slope of the normal consolidation line in the ν ln p diagram, i.e., is given by: 2.303

TABLE IV THE OBTAINED VALUES OF FROM TRIAXIAL AND OEDOMETERTEST AND PLASTICITY INDEX Triaxial Oedometer Plasticity test test index Depth(m) 9 3 6 9 3 6 9 0.12 0.14 0.154 0.165 0.29 0.32 0.31 BH1 --------------BH2

V.

0.127

0.123

0.38

0.34

0.37

Settlement Monitoring

In order to measure the settlement of tanks, one permanent survey mark (benchmark) and eight settlement plates were installed. Location of the plates is shown in Fig. 2. The benchmark was installed by building a concrete cube on a big rock. A 10 cm steel sleeve was placed on top of the concrete cube for more accuracy. The settlement plates were circular PVC plates with a diameter of 10 cm and 0.1 cm thick that were located on the edges of tanks. After installation of the benchmark and settlement plates, the height of the top of the settlement plates was surveyed by a standard surveyor (Leyica total station) over three years. The monitoring results are presented in Table VI.

(1)

VI.

Soil Model and Finite Element Program

In this study, CRISP, a critical state finite-element program is used to analyze technical problems [18]. This program has been developed in Cambridge University, Cambridge, England, that allows undrained, drained and consolidation analyses. The CRISP program has been rewritten in Najaf Abad Azad University for Windows operating system [20]. The modified cam clay model [9] was selected as the constitutive model, appropriate for Shiraz clay formations. In this study, both plane strain and axisymmetric conditions are used.

(2)

where, PI is percentage of the Plasticity index. The κ value can be determined in a way similar to λ. It is easy to show that: 2.303

0.133

TABLE V THE OBTAINED VALUES OF Κ FROM TRIAXIAL AND OEDOMETER Values of κ Triaxial test Oedometer test Depth(m) 9 3 6 9m BH1 0.04 0.023 0.031 0.032 BH2 --------BH3 0.05 0.030 0.033 0.035

2.303 10 . Alternatively λ can be directly determined from the slope of the compression (virgin consolidation) line in ν ln p diagram of a triaxial test (see Figs. 3). There is an approximate method to determine the value of λ via index tests: 160

0.14

BH3

(3)

Because in one dimensional swelling, the value of K (the coefficient of earth pressure at rest) is not constant, the value of κ, in previous relation is not very precise. In practice, κ values are often chosen in the range of one-fifth to one-third of λ. The data usually indicate a lower (stiffer) value on immediate unloading and a higher value at later stages of unloading [18].

VII.

FEM Analysis

Finite element simulation for a settling tank and an aeration tank is presented here. The cross sections of the settling and aeration tanks are shown in Figs. 4 and 5.

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

271

A. Yousefzadeh, R. Ajalloeian, M. Kabiri

TABLE VI MONITORING RESULTS OF THE POINTS

26/mar2006

22/oct2006

23/mar2007

25/oct2007

29/mar2008

24/oct2008

B.M

P1

P2

P3

P4

P5

P6

P7

P8

EL(m)

1648.334

1641.723

1641.740

1641.743

1641.733

1641.840

1641.821

1641.845

1641.837

SE(m)

---

0

0

0

0

0

0

0

0

EL(m)

1648.334

1641.685

1641.703

1641.708

1641.690

1641.792

1641.795

1641.806

1641.788

SE(m)

---

0.038

0.037

0.035

0.043

0.048

0.045

0.039

0.049

EL(m)

1648.334

1641.682

1641.701

1641.705

1641.688

1641.850

1641.769

1641.805

1641.784

SE(m)

---

0.041

0.039

0.038

0.045

0.055

0.052

0.040

0.053

EL(m)

1648.334

1641.682

1641.700

1641.703

1641.687

1641.776

1641.758

1641.803

1641.777

SE(m)

---

0.041

0.039

0.040

0.046

0.064

0.063

0.042

0.060

EL(m)

1648.334

1641.679

1641.699

1641.699

1641.685

1641.772

1641.756

1641.801

1641.775

SE(m)

---

0.044

0.041

0.044

0.048

0.068

0.065

0.044

0.062

EL(m)

1648.334

1641.678

1641.698

1641.698

1641.683

1641.767

1641.752

1641.798

1641.774

0.045

0.050

0.073

0.069

0.047

0.063

SE(m) --0.045 0.042 Note: EL and SE are Elevation and Settlement of plates in distinct times

520(m)

50(m)

1850(m)

107(m)

Fig. 4. Cross section of the settling tank

574(m)

3120(m)

Fig. 5. Cross section of the aeration tank

mass. Moreover, the initial state of consolidation must also be determined. The latter can be done in a straightforward manner by indicating the size of the current elliptical yield locus for each stress state. The parameter p (preconsolidation pressure that is a measure of the highest stress the soil have ever experienced) is used to define the size of the yield surface, as it is the maximum value of p on the current yield surface at the intersection of the yield ellipse and the isotropic stress axis [18]. The elliptical yield locus and p are shown in Fig. 6.

A foundation depth of 18 m was considered adequate for the analysis, because of existence of the bedrock beneath the clayey deposits. The lateral boundary of the finite-element model is defined by a distance of three times the vertical depth. The clay foundation is divided into three horizontal layers of elements, simulating the characteristic properties of the different soil deposits. The critical state parameters for the analyses are presented in the previous sections. In order to perform calculations with Modified Cam-clay model, initial effective stresses must be specified throughout the soil

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

272

A. Yousefzadeh, R. Ajalloeian, M. Kabiri

37 m. The rate of loading is rapid, and after loading a long term of consolidation occurs. In order to analyze this problem a consolidation analysis must be performed. Linear strain triangle (consolidating) elements have been used to represent the consolidating Modified Cam-clay material. In situ conditions are defined in terms of stresses at different elevations within the mesh. These stresses are derived from soil parameters that are obtained (almost) directly from lab testing of soil. In the present case, the magnitude of the uniform loading is equal to the weight of the primary settling tank, when it is full of wastewater. The finite-element advance during analysis is defined by units named Increment. In each increment, the effect of applied load and boundary conditions is calculated, and the accumulated strains and pore water pressures are updated. Increments are grouped together in blocks which then model particular events in the analysis. This analysis requires three increment blocks. The first block consists of only one increment and allows the water table to be defined. The next block of ten increments allows the loading process to be done and the last block of ten increments is related to the consolidation. The blocks and increments in this analysis are summarized in Table VII. The geometry and mesh generation for this problem is shown in Fig. 7.

CLS

q

Yield locus

P' Fig. 6. The elliptical yield locus of modified Cam-clay model

The size of the yield locus (i.e. P ) is calculated from the expression of the cam clay yield locus, since the stress state lies on the surface: (4) As Figs. 2 and 4 demonstrate, the primary settling tank is circular. Thus, for this tank, the analysis of a circular footing on a layer of clay is considered. Although the Cam-clay parameters have no significant changes with depth, but for more accurate modeling, the clay layer is subdivided to three layers. Therefore, the Cam-clay parameters for each layer of the clay were determined separately. In this case, a layer of clay is loaded by a uniform surface pressure applied over a circle of radius

Block No 1 2

TABLE VII THE BLOCKS AND INCREMENTS OF THE ANALYSIS Number of Total time-Step Description increments Water table definition 1 100 seconds Loading 10 1month Consolidation 10 3 years

Fig. 7. Mesh of 15-noded Triangles Used for settling tank

Fig. 8. Mesh of 15-noded Triangles Used for aeration tank

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

273

A. Yousefzadeh, R. Ajalloeian, M. Kabiri

As you can see in Figs. 2 and 5, the aeration tank is rectangular. This case is very similar to the settling tank, but there is only one difference. Unlike the previous analyses, this problem is one of plane strain. The other cases are exactly the same. The finite-element mesh used for aeration tank is shown in Fig. 8.

the curves from the previous in-situ measurements. Clearly the measured results have a lower value than calculated results. Even so, as expected, results are relatively close together. It seems that modified Cam-clay model has had a fairly good performance in determination of the settlement in this case study. Moreover, there are several error factors that maybe cause some differences between the determined values and measured values. For example, the variation of the ground water table level in different seasons is one of them. The other one is the errors that have arisen in determination of the Cam-clay parameters via laboratory tests.

VIII. Comparison of Calculated and Measured Results The predicted time versus vertical displacement curves for plane strain and axi-symmetry analyses are plotted in Figs. 9 and 10, where it can be compared with

Vertical displacements(m)

‐6,00E‐02 ‐5,00E‐02 ‐4,00E‐02 ‐3,00E‐02

Determined values in‐situ measurments

‐2,00E‐02 ‐1,00E‐02 0,00E+00 0

200

400

600

time(day)

800

1000

1200

Fig. 9. The predicted time versus vertical displacement curve for settling tank (axi-symmetric)

Vertical displacement(m)

‐1,20E‐01 ‐1,00E‐01 ‐8,00E‐02 ‐6,00E‐02 ‐4,00E‐02

Determined values In‐situ measurements

‐2,00E‐02 0,00E+00 0

200

400

600

Time(day)

800

1000

1200

Fig. 10. The predicted time versus vertical displacement curve for aeration tank (plane strain)

IX.

element program used to determine settlements via Modified Cam-clay model. Finally, the results of both methods were compared. Obtained results indicate that, the Modified Cam-clay model over estimates the settlement in both of these footings.

Conclusion

In this paper, settlement of a pair of tanks founded on soft clay was investigated via theoretical and practical methods. In practice the settlement of tanks were measured during three years. On the other hand, a finite Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

274

A. Yousefzadeh, R. Ajalloeian, M. Kabiri

[14] M. D. Liu, J. P. Carter, A structured Cam Clay model, Canadian Geotechnical journal, Vol. 39(Issue 6):1313–1332, 2002. [15] R. Day, Foundation Engineering Handbook: Design and Construction with 2006 International Building Code (McGrawHill Professional, 2006). [16] ASTM Standard D 1586-08a, Standard Test Method for Standard Penetration Test(SPT) and Split-Barrel Sampling of Soils, Annual Book of ASTM Standards, ASTM international, West Conshohocken, Pa, 2004. [17] ASTM Standard D 4767-02, Standard Test Method for Consolidated Undrained Triaxial Compression Test for Cohesive Soils, Annual Book of ASTM Standards, ASTM international, West Conshohocken, Pa, 2004. [18] A. M. Britto, M. J. Gunn, Critical state soil mechanics via finite elements (Ellis Horwood Ltd., Chichester 1987). [19] C. P. Wroth, The interpretation of in-situ soil tests, Geotechnique, Vol. 34 (Issue 4):449-489, 1984. [20] A. M. Yousefzadeh, M. MirmohamadSadeghi, H. Matinmanesh, Deformation and Pore Pressure Dissipation due to Excavation in Soft Clay, The new findings in the Civil Engineering Conference, Najaf Abad, Iran, 2-3 Mar. 2011.

In conclusion, no model exist as of yet that can simulate all of the inherent complexities and uncertainties of soil. However, there exist a large variety of models which have been recommended in recent years by researchers. All these models have certain advantages and limitations that depend on their application. Considering that the Modified Cam-clay model involves only four parameters that can be determined easily from standard test data (the convenience of parameters derivation) and the simplicity of computational application (the ease of implementing in computer calculations), it seems that, this model can be useful in practical and not very sensitive projects. In this case study, Modified Cam-clay model performed fairly suitable and conservative to describe deformation of normally consolidated soft soil. However one of the limitations of this method is that, it requires a series of accurate laboratory tests data. Finally, it is recommended to conduct more computation measurement along with additional fullscale experiments to ascertain the degree of realism in the Modified Cam-clay model in order to adjust and refine it.

Authors’ information 1

Department of Civil Engineering, Najafabad Branch, Islamic Azad University, Isfahan, Iran Address: No 2, Laleh Alley, 22bahman St., Isfahan, Iran. E-mails: [email protected], [email protected] (Corresponding Author); [email protected] [email protected]

References [1]

[2]

[3] [4]

[5]

[6]

[7]

[8]

[9]

[10] [11]

[12]

[13]

E. N. Fox, The mean elastic settlement of a uniformly loaded area at a depth below the ground surface, in Proc, 2nd International conference on Soil Mechanics and Foundation Engineering,Vol.1, pp. 129-132, Rotterdom, 1948. J. H. Schmertmann, Static cone to compute settlement over sand, J. Soil Mech. Found. Div., ASCE, Vol. 96:1011-1043, May/June 1970. G. G. Meyerhof, Shallow foundations, J. Soil Mech. Found. Div., ASCE, Vol. 91:21-31, 1965. D. J. D’Appolonia, H. G. Poulos, and C.C Ladd, Initial settlement of structures on clay, J. Soil Mech. Found .Div., ASCE, Vol. 97: 1359-1377, October 1971. E. Schultze, G. Sherif, Prediction of settlement from evaluation of settlement observations for sand, 8th Int. Conf. Soil Mech. Found. Eng., Moscow, U.S.S.R., Vol. 1.3, pp.225-230, 1973. A.W. Skempton, L. Bjerrum, A contribution to the settlement analysis of foundations on clay, Geotechnique, Vol. 7(Issue 1), 168-178, 1957. G. A. Leonards, Estimating Consolidation Settlements of Shallow Foundationson Over-consolidated Clay, Transportation Research Board, Special Report, (Issue163):13-16, Washington, D.C., October 1976. J. B. Burland, C. P. Worth, Allowable and differential settlement ofstructures, including damage and soil-structure interaction, Conf. On Settlement of Structures, pp.611-654, Cambridge University, U.K., 1970. K. H. Roscoe, J. B. Burland, On the generalised stress strain behaviour of wet clay, In Engineering plasticity. Edited by J. Heyman, and F.A. Leckie, Cambridge University Press, pp. 535– 609, 1968. D. Muir-Wood, Soil Behaviour and Critical State Soil Mechanics (Cambridge University Press, 1990). A. Gens, D. M. Potts, Critical state models in computational geomechanics, Engineering Computations, Vol. 5(Issue 1):178 – 197,1988 H. S. Yu, A unified state parameter model for clay and sand, International journal for Numerical and Analytical Methods in Geomechanics, Vol. 22 (Issue 8):621-653, 1998. L. Zdravkovic, D. M. Potts, H. D. St John, Modelling of a 3D excavation in finite element analysis, Geotechnique, vol. 55(issue 7):497-513, 1999.

2 Geological Science Department, Faculty of Science, The University of Isfahan, Isfahan, Iran. E-mail: [email protected]

Amirhosein Yousefzadeh (Isfahan, Iran, 1983). B.SC. in Civil Engineering, Islamic Azad University, Najafabad Branch, Najafabad, Isfahan, Iran, 2007.M.sc. in Geotechnical Engineering, Islamic Azad University, Najafabad Branch, Isfahan, Iran, 2011. He has published 10 refereed conference papers (in Persian and English) in different subjects (e.g., numerical stability analysis of different structures and bearing capacity of foundations). His main research interestsare finite element analysis and development of numerical methods in geotechnical engineering,using laboratory testing of geomaterials. Mr. Yousefzadeh is a member of “Joint Working Group on Geotechnical Engineering for Disaster Mitigation and Rehabilitation (JWG-DMR)”, “Iranian Geotechnical Society”, “Young Researchers Club, Islamic Azad University” and “Isfahan Organization of Civil Engineers. Rassoul Ajalloeian (Isfahan, Iran, 1959). B.Sc. in Geology, The University of Isfahan, Isfahan, Iran, 1984. Ms.sc. in engineering Geology, Tarbiat Modares University, Tehran, Iran, 1995. Ph.D. in Geotechnic and Soil Mechanics, The University of Newcastle, Newcastle, Australia, 1995. He has published four books (in Persian) including “Applied Engineering Geology of Tunnels”, “Principles of Geotechnical Engineering”, “Experimental Criteria for Rocks Failure” and “Rock Mass Classification”. He also has published more than 15 international journal papers, 14 national journal articles (in Persian), more than 50 international conference papers and more than 90 national conference papers. Some of his works published in international journals include: “Determination of silica sand stiffness” published in “Engineering Geology”; “Geochemistry of the Gavkhoni Playa Lake Brine” in “Carbonates and Evaporites”; “Investigation on causes of the

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

275

A. Yousefzadeh, R. Ajalloeian, M. Kabiri

Siruyeh Landslide, West Semirom(Iran)” in Landslides; “Application of Rock Mass Characterization for Determining the Mechanical Properties of Rock Mass: a Comparative Study” in Rock Mechanics and Rock Engineering; “The effect of Scale Direct Shear Tests on The Strength parameters of Clayey Sand in Isfahan city, Iran” in journal of Applied Sciences; “Chamber studies of the Effects of Pressuremeter Geometry in Test Results in Sand” in Geotechnique; “The effect of water content on the mechanical behavior of fine grained sedimentary rocks” in Science Journal, Isfahan University; “Strength anisotropies in mud rocks” in Bulletin of Engineering Geology and the Environment; “Rocks mass characterization for an underground excavation support system: the sabzkuh water con” in Int. J. of Rock mech. & Mining sciences; “The provenance of the varzaneh Aeolian sand field regarding to lithology and tenture, Sw Esfahan, Iran” and “Evaluation of damsite grout ability using secondary permeability index rock classification (case study)” in American Journal of applied sciences. His main research interests are laboratory testing of geomaterials, soil characterization, ground improvement, problematic soils, rock mechanics, tunnel engineering and site investigation and in situ testing. Dr. Ajalloeian is currently Associate professor in Geological Science department of University of Isfahan and member of Iranian Association of Rock Mechanics, Iranian Earthquake Engineering Association, Editorial Board of the Journal of Geological Society of Iran, Iranian Construction Engineering Organization and International Association for Engineering Geology (IAEG).

Meisam Kabiri (Isfahan, Iran 1984). BSc in Civil Engineering, Najafabad Azad University, Isfahan, Iran, 2008.He has been a member of Civil and Geotechnical Engineering Department of Pars Arian Ab Engineering Consultants since 2010 and head of the department since early 2011. His specialty is in designing liquid concrete tank structures and foundations. He has been supervisor of field exploration of south-east sewer network of city of Isfahan. His main research interests are settlement of foundation, soil-structure interaction, liquefaction analysis and retaining wall analysis for earthquakes. He is a member of “Isfahan Organization of Civil Engineers” and has a third grade certificate for supervising building construction.

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

276

International Review of Civil Engineering (I.RE.C.E.), Vol. 2, N. 6 November 2011

Analysis of Curved Prestressed Concrete Beams Under Short-Term and Long-Term Conditions by Using of Finite Element Method M. B. Abdul Rahman, M. R. Abed Abstract – In this study, the theoretical behavior of curved prestressed concrete beams have been investigated by using of three dimensional finite element method, in order to understand the behavior of this beam under incremental loads up to failure and also under long-term time conditions. Therefore, the analytical study have been done by parts: First part deals with shortterm analysis, and other part deals with long-term analysis. This analysis is done by depending on package programs (ANSYS+CivilFEM V12.0 version 2009). Willam and Warnke models have used to represent the nonlinear behavior for concrete. The curved prestressed concrete beam represents by three dimensional model include of Isoparametric 8-Node Brick Element known as SOLID65 which is used in short-term analysis, while the Element known as SOLID185 have been used in long-term analysis. Due to the time-dependent effect on ultimate load value for concrete structures, both of the creep and the shrinkage effects on properties and behavior for concrete have been taken into consideration using finite elements technology dependence on recommendation committee of American Concrete Institute (ACI 209). Also, the effective modulus approaches used in the representation of creep effect. Also, all the prestressed loss for prestress concrete members have been included . Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved. Keywords: Curved Beam, ANSYS+CivilFEM, SOLID65, Creep, Long-Term, Shrinkage, Prestressed Losses

to u, v, w Wc x, y, z α ∆PCR ∆PEL ∆PFR ∆PSH ∆PSL ∆PSR εc εcr εo εps

Nomenclature ξ, η, ζ ϕcr φcr A a, b Aps Ec Eceff Eps fc' fpi

fps fpy fr ft' K L Po r Rt s t Tc

Local coordinate system Creep coefficient Ultimate creep coefficient Anchorage slip Constants of cement type and curing method Area of the prestress steel Modulus of elasticity of concrete Effective modulus Modulus of elasticity of the prestress steel Compressive strength of concrete Prestress stress after discounting the immediate losses Prestress stress Yield stress of prestressed tendon. Direct tensile strength Tensile strength of concrete Unintentional angular Length of prestress tendon. Initial prestressed force Vector radius expresses this curved line Secant modulus Length of the arc Time (day) Stiffness multiplier for cracked tensile

εsh εsu

θ µ ρ σa, τa

σc

σcr

σs

τ

Manuscript received and revised October 2011, accepted November 2011

277

Curing time Displacement components Density of concrete Cartesian coordinates Sum of prestress tendon angles along X Creep loss Elastic shortening loss Friction loss Shrinkage loss Anchorage slip loss Relaxation loss Strain of concrete Cracked strain. Strain of concrete at fc' Prestress strain. Shrinkage stain Ultimate shrinkage strain Angle of failure surface Friction coefficient, Apex of the surface Average stress components Stress of concrete at any strain Concrete stress duo to prestress and loads applied Stress of concrete at prestress steel location Time of stress applied

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

M. B. Abdul Rahman, M. R. Abed

I.

elements of a high degree of accuracy that may be combined to form macro-elements having the form of transversal slices of a box-girder bridge . Include the effect of the prestressing in the element formulation. The program BOXGBR (BOX-Girder BRidge) have been developed for static analysis of curved box-girder bridges with variable cross section, the program also is used for test the accuracy and practical suitability of the proposed approach. Where, to simplify the calculations, the tension in prestressing cables was assumed to be independent of the deformation of the structure. The numerical tests conducted were concluded that, the inplane distortion of the cross sections of the assembly elements proved to be of little practical importance. Campbell, Lee, and Chan (19800 [5], tested six twospan continuous horizontally curved post-tensioned beams up to failure, in order to study three regions of behavior, namely, uncracked, cracked and collapse behaviors, and each is studied analytically. The crosssection of this curved beam is rectangular, (127 mm) wide by (203 mm) deep, with (6520 mm) of radius and (40.16o) of angle. During the test it is explained that the first cracking observe in the top surface of the beam at the central support. Subsequently, cracks developed in the bottom surface in the vicinity of the load points. With the addition of further load, spiral cracking occurred in the bottom surface and on each side of the beam between the load points and the end supports, as well, the collapse load of a continuous curved prestressed concrete beam can be predicted by methods of plastic analysis provided the cross sections possess sufficient ductility to allow development of the collapse mechanism. Ebana, Hoshino, Kutuzawa, and Sugimoto (1991) [6], showed the design and the construction of the Yokomuki bridge. This bridge is 9 span prestressed concrete curved continuous box girder bridge, both ends of the bridge are fixed in the horizontal plane. Thus the bridge acts as a horizontal arch for seismic force and temperature. It change was constructed by the incremental launching method. In (1997), Debaiky [7], developed a numerical procedures and a computer program for the analysis of the time-dependent behaviour of segmentally erected curved prestressed concrete box-girder bridges. The analysis gives the instantaneous and time-dependent changes in the displacements, in the support reactions and in the internal forces from which the strains and stresses at various sections of the structure can be calculated. The analysis accounts for the effects of sequence of construction, loading and prestressing and changes in the statically system and support conditions. Prediction of the time-dependent properties of concrete, namely, the modulus of elasticity, creep coefficient and shrinkage strain, is based on the recommendations of either the CEB-FIP mode1 Code (1990) or the AC1 Cornmittee 209 (1992). The numerical applications conducted were concluded that, Prediction of stresses and deformations in segmental erected curved bridges can be considerably in error if the effects of creep,

Introduction

Curved prestressed concrete beams are frequently employed in structures such as: highway bridges to provide a smooth traffic flow, interchanges in large urban areas, containment structures, ring beams, balconies of the spatial buildings and other structures. Where, a curved beam is generated by a plane cross-section which centroid sweeps perpendicularly through all the points of an axis line. The vector radius r=r(s) expresses this curved line, where (s) length of the arc [1]. Prestressed concrete is a particular form of reinforced concrete. Prestressing involves the application of an initial compressive load on a structure to reduce or eliminate the internal tensile forces and thereby control or eliminate cracking. The initial compressive load is imposed and sustained by highly tensioned steel reinforcement reacting on the concrete. With cracking reduced or eliminated, a prestressed section is considerably stiffer than the equivalent (usually cracked) reinforced section. Prestressing may also impose internal forces which are of opposite sign to the external loads and may therefore significantly reduce or even eliminate deflection [2]. A prestressed concrete structure has many advantages, such as delaying cracks, saving materials, reducing deflection, and has been widely or increasingly used in long-span structures, shells, and nuclear containment vessels [3]. The FEM (Finite Element Method) is a numerical technique. In this method all the complexities of the problems, like varying shape, boundary conditions and loads are maintained as they are but the solutions obtained are approximate. Because of its diversity and flexibility as an analysis tool, it is receiving much attention in engineering. The fast improvements in computer hardware technology and slashing of cost of computers have boosted this method, since the computer is the basic need for the application of this method. A number of popular brand of finite element analysis packages are now available commercially. Some of the popular packages are STAAD-PRO, GT-STRUDEL, NASTRAN, NISA and ANSYS. Using these packages one can analyze several complex structures. In this study, a three dimensional finite elements will be displayed, which are used for represent of the concrete and reinforcing steel, then, the material properties of concrete, reinforcing steel and prestressed steel will be expressed by ACI and PCI codes . Also, the present study is studied same parametric as the ratio of stress applied on prestress tendons.

II.

Previous Researches

Few Experimental and analytical studies have been published on the behaviour of this curved prestressed concrete beams, where, most that studies deal with curved prestressed concrete bridges. In (1979), Jirousek, Boubergu, and Saygun[4], developed an appropriate system of special purpose finite

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

278

M. B. Abdul Rahman, M. R. Abed

orthogonal directions), crushing, plastic deformation, while the rebar capability is available for modeling reinforcement behavior. In long-term analysis, SOLID185 is used for 3-D modeling of solid structures, as shown in Fig. 2(b). The element has plasticity, hyperelasticity, stress stiffening, creep, large deflection, and large strain capabilities. It also has mixed formulation capability for simulating deformations of nearly incompressible elasto-plastic materials, and fully incompressible hyperelastic materials [11].

shrinkage and relaxation are ignored. Choi, Kim, and Hong (2002) [8], introduced the FSM (Finite Strip Method) for the analysis of the some numerical applications of the simple and continuous straight box bridges, other curved continuous supported. As well, losses of the anchorage slips and the friction have been included in the analysis. The aim of FSM is not to replace entirely the FEM but compliment it where FSM is more effectively used for the analysis of structures with geometry of a large length in the longitudinal direction and relatively short transverse length such as box-girder bridges. Kholoo and Kafimosavi (2007) [9], studied the flexural behavior of horizontally curved prestressed (posttensioned) box bridges by using of three dimensional and refined finite element modeling and analysis. Bridge length, section geometry, and material properties are the same in all the models, while angle of curvature varies from 0 to 90°. The results of analysis show that in curved bridges, stress distribution is significantly different in comparison to straight bridges. Also, the level of stresses at some locations of section width is considerably high. Results show that by proper redistribution of prestressing in section width, significant reduction in resultant stress is possible.

Fig. 1. Three-dimensional 8-node element (Brick Element) [11]

III. Finite Element Model The ultimate purpose of a finite element analysis is to recreate mathematically the behavior of an actual engineering system. In other words, the analysis must use an accurate mathematical model of the physical prototype. In the broadest sense, this model comprises all the nodes, elements, material properties, real constants, boundary conditions, and other features that are used to represent the physical system [10].

(a)

III.1. Concrete In the present study, three-dimensional 8-node elements are used to model the concrete, which known as the brick element. This element has eight corner nodes, each has three degrees of freedom of movement (u, v and w) along the cartesian coordinates (x, y and z) respectively. The element has its local coordinate system ξ, η, ζ, as shown in Fig. 1, with the origin at the center of the element such that each local coordinate range from (-1) to (+1). The geometry of the element can be defined in terms of the shape functions. ANSYS program have more of the elements, which, such as: SOLID65 and SOLID185 will have been used in this study for concrete modeling. In short-term analysis, SOLID65 (or 3-D reinforced concrete solid) is used for the 3-D modeling of solids with or without reinforcing bars (rebar), as shown in Fig. 2(a). The solid is capable of cracking in tension and crushing in compression The most important aspect of this element is the treatment of nonlinear material properties, the concrete is capable of cracking (in three

(b) Figs. 2. ANSYS Brick Element, (a)SOLID65, (b) SOLID185 [11]

III.2. Reinforcing Steel Bars In this study, three-dimensional 2-node truss element is used to model the reinforcing steel bars, which known as the LINK8 element, as shown in Fig. 3. LINK8 is a spar which may be used in a variety of engineering applications. This element can be used to model trusses, sagging cables, links, springs, etc. The 3-D spar element is a uniaxial tension-compression element with three degrees of freedom at each node: translations in the nodal x, y, and z directions. As in a pin-jointed structure, no bending of the element is considered. plasticity, creep, swelling, stress stiffening, and large deflection capabilities are included [11].

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

279

M. B. Abdul Rahman, M. R. Abed

⎧ ⎪⎪0.85 f c' σc = ⎨ ⎪ ⎪⎩0.85 f c'

0 ≤ εc ≤ εo

(1)

εo ≤ εc ≤ εu

where σc is a stress of concrete at any strain of it εc, fc' is a compressive strength of concrete, εo is a strain at fc', Ec is modulus of elasticity of concrete (Young's Modulus), and Wc is density of concrete:

Fig. 3. Three-dimensional 2-node truss element (Link8) [11]

⎛ 0.85 f c' ⎞ ⎟⎟ ⎝ Ec ⎠

(2)

Ec = 0.043 Wc3 f c'

(3)

εo = 2 ⎜ ⎜

III.3. Prestressed Steel In ANSYS+CivilFEM programs, there are two possible ways to taking into account the prestress actions in finite element models. The first one is to model the prestressing tendons, using beam or link elements with an initial strain, inside the model it have been already created. To do this must adapt the original finite element mesh, to locate the nodes where the tendons will be, to be able to connect the beam elements to the solid elements. If the mesh cannot be changed, then coupling equations can be used to blend the tendons movements to those of the solid elements, putting together in the same model elements with such a different nature (solids and beams), with different degrees of freedom, etc. can easily lead to errors in the results, meshing problems, peaks of stresses on certain nodes, etc. and therefore results should be checked very carefully. The second way of considering prestress actions is to create a group of loads equivalent to the action of the prestressing tendon which create on the model. This loads will be put directly in the model’s nodes, without having to change its geometry or mesh . This method, which is the one implemented by CivilFEM program [12], has the advantage that it can be used on any kind of model, or mesh. It can also be applied on a model made up of beam elements [13].

IV.

⎡ 2ε ⎛ ε ⎞2 ⎤ ⎢ c −⎜ c ⎟ ⎥ ⎢ εo ⎝ εo ⎠ ⎥ ⎣ ⎦

Fig. 4. Uniaxial compressive stress-strain relationship of concrete [14]

The Willam–Warnke failure model is used for modeling the failure collapsing surface of concrete without reinforcement under stress with three axis. If the calculated principle stress is more than threshold stress, its behavior is considered to be nonlinear. In this case, the calculated principle stresses are used to determine the failure situation using the Willam–Warnke model. If Equation (4) is obtained using these principals, it means that stresses occur on the failure surface:

Modelling of Material Properties

1 σa 1 τa + =1 ρ f c r (θ ) f c'

During recent years, interest in nonlinear analysis of concrete structures has increased steadily, because of the wide use of plain, reinforced and prestressed concrete as structural materials, and because of the development of the relatively powerful analysis techniques implemented on computers.

(4)

where σa and τa are average stress components, ρ is the apex of the surface and fc' is the uniaxial compressive strength, r is the position vector locating the failure surface with angle θ. The use of the Willam–Warnke mathematical model of the failure surface for the concrete has the following advantages: close fit of experimental data in the operating range, simple identification of model parameters from standard test data, smoothness (continuous surface with continuously varying tangent planes) and convexity (monotonically curved surface without infection points) [15]. The cracked concrete is generally modeled by a linear-elastic-fracture

IV.1. Concrete The concrete is assumed to be homogeneous and initially isotropic. The compressive uniaxial stress-strain relationship for the concrete model is suggested by ACI code is adopted in this study. It is composed of a parabolic ascending branch and a linear up to crushing , as shown in Fig. 4, indicated by Equation (1) [14]:

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

280

M. B. Abdul Rahman, M. R. Abed

committee [16] is adopted in this study, indicated by Equation (6), as shown in Fig. 8:

relationship. When a principal stress exceeds its limited value, a crack is assumed to occur in a plane normal to the direction of the offending principal stress. The cracking of concrete in the present study is modeled as smeared-cracking model, as shown in Fig. 5.

f ps

ε ps ≤ 0.0086 ⎧193060 ε ps ⎪ =⎨ 0.2758 ⎪1861.65 − ε − 0.007 ε ps > 0.0086 ps ⎩

(6)

where fps is prestress stress, and εps is prestress strain:

Fig. 5. Smeared-cracking model [10]

The tensile stress normal to the crack dose not drop to zero immediately, when the crack is formed, it decreases with increasing crack width. The secant modulus of elasticity post-cracking is represented by an equation (5), as shown in Fig. 6: Rt =

Tc ft' 5

⎛6 1 ⎞ ⎜ − ⎟ ε cr ≤ ε i ≤ 6ε cr ⎝ ε i ε cr ⎠

Fig. 7. Uniaxial stress-strain relationship of reinforcing steel [14]

(5)

where Rt is secant modulus, Tc is stiffness multiplier for cracked tensile, ft' is tensile strength of concrete, and εcr is cracked strain:

Fig. 8. Uniaxial stress-strain relationship of prestressed steel [16]

V.

Time – Dependent Analysis

Mechanisms of creep and shrinkage in concrete are not fully understood and the prediction of creep and shrinkage behavior in concrete is not precise at best. Creep and shrinkage are often treated as separate and independent phenomena. Actually, the effect of creep is significantly greater when accompanied by shrinkage e.g., drying creep is the additional creep resulting from drying of concrete. Over the last two decades, number of predictive models have been developed to take into account the interdependence of creep and shrinkage. The ACI committee 209 [17], proposed the following form of equations for predicting the compressive strength (fc'), direct tensile strength (fr) and the modulus of elasticity (Ec) at any time (t):

Fig. 6. Post-cracking model of concrete [11]

IV.2. Reinforcing Steel Bars In this study, the idealization of steel bars neglected the strength increase due to strain hardening and the reinforcing steel is modeled as a linear, perfectly plastic material (bilinear isotropic) as shown in Fig. 7. Bilinear isotropic material is also based on the von Mises failure criteria is adopted. This assumption underlies the design equations of the ACI code [14]. IV.3. Prestressed Steel The uniaxial stress-strain relationship for the prestressed steel model is suggested by ASTM

f c' ( t ) =

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

t f c' a + b × t ( 28)

(7)

International Review of Civil Engineering, Vol. 2, N. 6

281

M. B. Abdul Rahman, M. R. Abed

f r ( t ) = 0.012 Wc f c' ( t )

(8)

Er ( t ) = 0.043 Wc3 f c' ( t )

(9)

V.3.

Prestress losses are the reduction in the initial prestressing force in the strands (the jacking force) and can be grouped into two general categories, immediate losses and long-term losses. immediate losses occur quickly upon release of the tendons and include anchorage slip, elastic shortening, and friction. Timedependent losses occur more slowly over the life of the member and include steel relaxation and concrete creep and shrinkage [19]. Immediately after transfer, the change in strain in the prestressing steel caused by elastic shortening of the concrete is equal to the strain in the concrete at the steel level. This loss may be obtained by Equation (14):

where a and b are constants depending on type of cement and type of curing. Shrinkage

V.1.

Shrinkage, after hardening of concrete, is the decrease with time of concrete volume. The ACI committee 209 [17], suggested the use of the following empirical equation to calculate shrinkage strain for ''standard condition'' and the use of correction factors for condition other than the standard case:

ε sh ( t ,to ) = ε su f ( t,to ) ⎧ ( t − to ) ⎪ ⎪ 35 + ( t − to ) f ( t ,to ) = ⎨ ⎪ ( t − to ) ⎪ 55 + ( t − t ) o ⎩

∆PEL =

(10)

after 7 day for moist curing

(11) after 1 - 3 day for steam curing

The time-dependent increase of strain in hardened concrete subjected to sustained stress is defined as creep. The ACI committee 209 [17], suggests the use of the following empirical equation to calculate creep for ''standard condition'' and use of correction factors for condition other than the standard case:

∆PSL =

0.6

(12)

Ec (τ ) Ec (τ ) 1+ ϕcr ( t,τ ) Ec ( 28 )

Apsσ s

(14)

A E ps Aps L

(16)

where  is anchorage slip, and L is length of prestress tendon. In long-term losses, the loss of stress in a tendon due to shrinkage of concrete may be obtained:

where ϕcr is creep coefficient, φcr is ultimate creep coefficient, and τ is time of stress applied. In creep analysis, several methods to analysis of the creep, but the effective modulus method, proposed by McMillan and Faber [18]. Is the oldest and simplest method to analyze the time dependent effects in concrete structures. This method involves a reduction of the modulus of elasticity to account for creep in concrete. To obtain the effective modulus Eceff elastic modulus, Ec(τ) is reduced by a factor which incorporates the creep coefficient ϕcr ( t ,τ ) defined in Equation (12): Eceff ( t,τ ) =

Ec

where Po is initial prestressed force,  µ  is friction coefficient, and α is sum of prestress tendon angles along X, and K is unintentional angular. The slip at the anchorage occurs when the prestressing force is transferred from the jack to the anchorage. This causes an additional loss of prestress. This loss may be approximated by Equation (16):

Creep

(t −τ ) ϕcr ( t,τ ) = φcr 0.6 10 + ( t − τ )

E ps

where Eps is modulus of elasticity of the prestress steel Aps is area of the prestress steel (tendon), and σs is stress of concrete at prestress steel location. In post-tensioned members, friction losses occur along the tendon during the stressing operation. A reliable estimate of friction losses may be obtained by Equation (15): − µ ⋅α + K ⋅ X ) ⎤ ∆PFR = Po ⎡1 − e ( (15) ⎣ ⎦

where εsh is shrinkage stain, εsu is ultimate shrinkage strain, and to is curing time. V.2.

Prestressed Losses

∆PSH = ε ( t,to ) E ps Aps

(17)

The creep loss in the tendon may be approximated by Equation (18): ∆PCR = ϕcr ( t ,τ ) σ cr Aps

E ps Ec

(18)

where σcr is concrete stress duo to prestress force and loads applied (CivilFEM is suggested its value 5 MPa). Relaxation can be described as “the loss of stress in a stressed material held at constant length”. Although changes in applied loads along with concrete creep and

(13)

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

282

M. B. Abdul Rahman, M. R. Abed

shrinkage produce variations in strand length in prestressed concrete elements, tendon length is assumed to remain constant for the purposes of calculating loss of force due to steel relaxation [18]. After discounting the immediate losses, the relaxation loss for low-relaxation prestressing steel type can be calculated by means of the expression (19): (a)

∆PSR

⎡ ⎞⎤ log ( 24t ) ⎛ f pi = Aps ⎢ f pi − 0.55 ⎟ ⎥ ⎜ ⎜ ⎟⎥ 45 ⎝ f py ⎢⎣ ⎠⎦

(19)

where fpi prestress stress after discounting the immediate losses, and fpy is yield stress of prestressed tendon.

VI.

Theoretical Applications

In theoretical example, the two-span curved prestressed concrete have been constructed by posttension technology, the cross-section of this curved beam is rectangular, (200 mm) wide, (300 mm) deep , a ratio of a span length to it’s a radius is (0.3) and (3000 mm) it’s of span length. The variable eccentricity of three tendons ( 12.7 mm) form curve along the curved beam, this curved beam is reinforced by four reinforcing steel bars ( 12 mm) and stirrups ( 10 mm), as shown in Figs. 9. Material properties of this curved beam, as shown in Table I.

fc' Ec ft' υ εu βt βc Tc

φcr εsu

fy Es υ fpy fpu Hp Eps υ K µ A

TABLE I MATERIAL PROPERTIES OF THIS CURVED BEAM Prestressing Technology (Post-Tension) Concrete 52.0 Compressive Strength 34290 Young's Modulus 4.48 Tension Strength 0.20 Poisson's Ratio 0.003 Ultimate Compressive Strain Shear transfer coefficients for an 0.20 open crack Shear transfer coefficients for an 0.80 closed crack Stiffness multiplier for cracked 0.60 tensile 2.350 Ultimate creep coefficient 780×10-6 Ultimate shrinkage strain III Cement type Steam Curing technologic Reinforcing Steel Yield Stress 414 Young's Modulus 200000 Poisson's Ratio 0.30 Prestressed Steel (Low – Relaxation) Yield Stress 1674 Ultimate Stress 1860 Elastic / Plastic Modulus Ratio 27.70 Young's Modulus 195000 Poisson's Ratio 0.30 Unintentional angular 0.01 Friction coefficient 0.20 Anchorage slip 2.0

(b)

(c)

Mpa Mpa Mpa (d) Figs. 9. Two-span curved prestressed concrete beam, (a) the crosssection, (b) finite element representation, (c) details of reinforcing, (d) details of prestressed tendons

VI.1. Short-Term Analysis of the Example Under applied the incremental load at center each span, the results of analyzing long-term conditions have been included study of (load-deflection) curve at midspan of this curved prestressed concrete beam. The results to clarified that the deflection increase with increase of the ratio of stress applied on prestress tendons (fpi/fpu) as shown in Fig. 10. The ratio of cracking load to the failure load (Pcr/Pu) have been reached to about (66%) for value of (fpi/fpu) reach to (0.65 to 0.75) as shown in Table II. Fig. 11(a) shows that the first crack occurs gradually begin from top of the cross-section at mid support due to negative moment effect on this curved beam at that location, while Fig. 11(b) clarifies the cracks at failure of the curved prestressed concrete beam.

Mpa Mpa

Mpa Mpa Mpa m-1 mm

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

283

M. B. Abdul Rahman, M. R. Abed

Figs. 13 show that the immediate and long-term total losses at edge and center supports, as well as, the ratio of immediate and long-term total losses (IL / FL) at edge and center supports have been reached about: (68%) and (60%) respectively, as clarified in Table IV and Table V respectively. Figs. 14 show that total losses diagram along of the two-span curved prestressed concrete decrease with increase of (fpi / fpu) at any time.

Fig. 10. Load-Deflection curve of two-span curved prestressed concrete beam at mid-span TABLE II VALUES OF EQUILIBRIUM, CRACKING AND FAILURE LOADS Failure Equilibrium Cracking f pi Pcr Load Pu load Load Pcr f pu Pu (kN) (kN) (kN) 0.60 90.8 225.83 350.0 0.645 0.65 100.0 239.17 360.0 0.664 0.70 110.6 252.00 380.5 0.662 0.75 120.0 265.00 397.0 0.668 Fig. 12. Load-Time curve of two-span curved prestressed concrete beam

(a)

(a) (b) Figs. 11. Cracks apparition, (a) first crack, (b) cracks at failure the curved prestressed concrete beam

VI.2. Long-Term Analysis of the Example The analyzing of the long-term conditions have been included study of (load-time) curve at mid-span for the curved prestressed concrete beam of the same beam adopted in short-term analysis example. A static load of (P =200 kN) is applied at center of each span of beam along life's service. The results to clarified that long-term deflection decrease with increase (fpi/fpu), as shown in Fig. 10, as well as the ratio of short-term deflection to the long-term deflection (∆28 day/∆10000 day) have been reached to about (30%), as shown in Table III.

(b) Figs. 13. Total losses-Time curve of two-span curved prestressed concrete beam, (a) at edge supports, (b) at centre supports

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

284

M. B. Abdul Rahman, M. R. Abed

TABLE III VALUES OF SHORT-TERM AND LONG-TERM DEFLECTIONS f pi f pu

Initial ∆28 day (mm)

Final ∆ 10000 day (mm)

∆ 28 ∆10000

0.60 0.65 0.70 0.75

0.881 0.821 0.756 0.697

2.880 2.714 2.536 2.376

0.3057 0.3024 0.2981 0.2936

TABLE IV VALUES OF IMMEDIATE AND LONG-TERM TOTAL LOSSES AT EDGE SUPPORTS Final Loss 10000 Immediate Loss f pi IL day 28 day f pu FL FL (%) IL (%) 0.60 30.09 44.32 0.679 0.65 28.59 41.72 0.685 0.70 27.50 39.92 0.689 0.75 26.90 38.71 0.695

(a)

TABLE V VALUES OF IMMEDIATE AND LONG-TERM TOTAL LOSSES AT CENTER SUPPORT Immediate Loss Final Loss 10000 f pi IL 28 day day f pu FL IL (%) FL (%) 0.60 19.83 34.15 0.581 0.65 19.68 33.13 0.594 0.70 19.60 32.35 0.606 0.75 19.56 31.73 0.617 (b)

VII.

Conclusion

(c)

The results of this study shows that the characters of the curved prestressed concrete beams, in general, are constant with respect to the ratio of stress applied on prestress tendons (fpi / fpu). The deflection of the curved prestressed concrete beam increases with time, and it decreases with increase the ratio of stress applied on prestress tendons. Therefore, the suitable ratio of stress applied on prestress tendons in curved prestressed concrete beam is about (0.65-0.70) . The creep and shrinkage have a principal effects on long-term behavior of the curved prestressed concrete beams. While the long-term deflection reached about (70%) from the short-term deflection at age (10000 day), which is proved that the time parameter have very important on long-term behavior of this the curved beam. Additional numerical studies are needed to analyze of the curved partial-prestressed concrete beams in order to show the effect a reinforcing on behaviour of this curved beam. Also, more studies are required to show the effect more parameters on the behavior of examined structure, like: the ratio of the span length for curved beam to its radius , the ratio of the stress applied on prestress tendons , amount of prestress tendons , the effect of type of curing on concrete and cement type.

(d)

References [1]

Figs. 14. Total losses diagram along of the two-span curved prestressed concrete, (a) ( fpi/fpu=0.60), (b) ( fpi/fpu=0.65), (c) ( fpi/fpu=0.70), (d) ( fpi/fpu=0.75)

[2]

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

L. Gimena, F. N. Gimena, P. Gonzaga, Structural Analysis of a Curved Beam Element defined in Global Coordinates, Engineering Structures 30, (2008), pp. 3355-3364. Hurst M. K., Prestressed Concrete Design (1st. Edition, Chapman

International Review of Civil Engineering, Vol. 2, N. 6

285

M. B. Abdul Rahman, M. R. Abed

[3] [4] [5] [6] [7] [8] [9]

[10] [11] [12] [13] [14] [15] [16] [17] [18] [19]

and Hall, USA, 1988). X. H. Wu, S. Otani, H. Shiohara, Tendon Model for Nonlinear Analysis of Prestressed Concrete Structures, ASCE Journal of Structural Engineering Vol. 127, No. 4, (2001), pp. 398-405. J. Jirousek, G. A. Boubergu, A. Saygun, A Macro–Element Analysis of Prestressed Curved Box – Girder Bridges, Computer and Structures Journal 10, (1979), pp. 467-482. T. I. Campbell, E. Y. Lee, K. S. Chan, Strength of Continuous horizontally Curved Post-Tensioned Beams, PCI Journal, (1980), pp. 118-145. R. Ebana, K. Hoshino, K. Kutuzawa, T. Sugimoto, Prestressed Concrete Horizontal Arch Bridge, IABSE Journal, (1991), pp. 259-264. A. S. Debaiky, Analysis of Time-Dependent effects on Segmental Prestressed Concrete Curved Box – Girder Bridges, MSc. thesis, Civil Eng., Concordia Univ., Montreal, 1997. C. K. Choi, K. H. Kim, H. S. Hong, Spline Finite Strip Analysis of Prestressed Concrete Curved Box-Girder Bridges, Engineering structure Journal 24, (2002), pp. 1575-1586. A. R. Kholoo, M. Kafimosavi, Enhancement of Flexural Design of Horizontal curved Prestressed Concrete Bridges, ASCE Journal of Bridge Engineering, Vol. 12, No. 5, (2007), pp. 585590. S. S. Kadhim, Finite Element Analysis of Composite ConcreteSteel Arches up to Failure, MSc. thesis, Tikrit Univ., Tikrit, 2007. ANSYS, Theory Manual Release 12.0 (SAS IP, 2009). CivilFEM, Theory Manual Release 12.0, (Ingeciber, S.A, 2009). J. Aparicio, I. Maia, E. Salete, ANSYS Customization for Bridges and Prestressed Concrete Structures Analysis and Design, Ingeciber, S.A. (2000), pp.1-11. ACI Committee 318, Building Code Requirements for Structural Concrete and Commentary, ACI 318M-02, American Concrete Institute. (2002). J. W. Anthony, Flexural Behavior of Reinforced and Prestressed Concrete Beams using Finite Element Analysis, MSc. thesis, Graduate School, Marquette Univ., Wisconsin, 2004. PCI Industry Handbook Committee, PCI Design Handbook Precast and Prestressed Concrete (6nd edition, U.S.A., 2004). ACI Committee 209, Prediction of Creep, Shrinkage and Temperature Effects in Concrete Structures, ACI 209 R-92, American Concrete Institute, Detroit, (1992). PCI Committee on Prestress Losses, Recommendations for Estimating Prestress Losses, PCI Journal, (1975), pp. 43-56. S. E. Bowers, Recommendations for Longitudinal PostTensioning in Full-Depth Precast Concrete Bridge Deck Panels, M.Sc. thesis, Civil Eng., Virginia Polytechnic Institute and State Univ., Virginia, 2007.

Authors’ information Mazin Burhan Adeen Abdul Rahman birth in Iraq, Kirkuk, October-1973, associate professor, Ph.D. in structural Engineering, University of Technology, Baghdad-Iraq, Jully-2007. Areas of Expertise :structural engineering, composite beams and long term effects concrete. E-mail: [email protected] Munther Rashed Abed birth in Iraq, Tikrit, January-1988, Postgraduate of Structures Engineering, B.Sc. in Civil Engineering, University of Tikrit, Salah ad Din-Iraq, 2009. Areas of Expertise: structural engineering, ANSYS+CivilFEM programs. E-mail: [email protected]

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

286

International Review of Civil Engineering (I.RE.C.E.), Vol. 2, N. 6 November2011

Influence of Water Content on the Tribological Behavior at Concrete/Wall Interface – Role of Cement Grains S. Bouharoun

Abstract – Each concrete component has a very important role on the tribological behavior in the vicinity of wall surface. The objective of this paper is to understand the influence of the paste volume on the mechanisms intervening at the concrete/formwork. Two concrete with 28 and 34% of paste were formulated in order to study the influence of fines content at the concrete/wall interface. Friction tests were carried out using an apparatus which can reproduce the same conditions of jobsites. Then, three types of mixtures fine elements and the aggregates forming the granular skeleton were prepared to study the effect of water content and aggregates on the interface behavior of fresh concrete. The results show that the friction is governed by the cement grains and its water content. Hypotheses were proposed to explain the different mechanisms occurring at the interface. Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved. Keywords: Cement Grains, Tribology, Friction, Interface, Fresh concrete, Formwork

I.

Vanhove [7] also analyzed the effect of formwork friction based on the friction coefficient. They measured the friction by means of the tribometer specially designed for a complex medium such as fresh concrete [8], in which the friction coefficient is known to be dependent on the roughness of the formwork surface. There are many factors influencing the behavior at the interface, such as mixture proportions, admixture, temperature, casting rate, height of formwork. Those factors can be classified into two categories: intrinsic and extrinsic. The intrinsic factors are related only to the material characteristics, while the extrinsic factors include the external contributions such as formwork flexibility, wall roughness, and external temperature. The frictions at concrete/wall can be related to the compactness of the granular skeleton and the fines content of the mixture. It can occur in the case of concrete mixtures with a very compact granular skeleton and a low or high content of binder [9]. The ratio is a factor which modifies the rheological properties of fresh concrete and affects the interface properties. The ability of concrete to flow is improved when this ratio increases. But the concretes at very high w/c ratio are more exposed to the segregation phenomenon, with a possible blockage of mixture in the pumping pipe and an increase of fiction at the concrete/formwork interface by aggregate/aggregate contact. The existing studies have some limitation in defining the role of fine elements and water content of the interface layer at the concrete/wall interface. This work presents the influence of the water content and the concrete constituents on the friction at concrete/wall interface. Friction tests were carried out using a plan/plan tribometer specially designed to study such frictions due

Introduction

The flexibility of concrete formulations allowed to this day the realization of highly developed structures by giving them a complex architecture, a high mechanical strength and improved durability. This progress has been made by the mean of extensive research on fresh concrete behavior. Knowledge of the rheological and tribological properties of fresh concrete is indeed essential because they determine the final quality of the structures. The paste is a unique and active element of concrete coating aggregates and filling the gaps existing in the granular skeleton [1]. In terms of rheological behavior, it acts as a lubricant, reducing the intergranular forces. Thus, this paste can facilitate the work in jobsite, offering sufficient fluidity of concrete in the formwork to flow under the effect of vibration and fill perfectly the formwork. Currently, there are many lacunas to understand the phenomena involved during the concrete casting in the vicinity of the formwork. Some of the phenomena are directly related to a friction between the materials in contact. This friction at the concrete/formwork interface can be considered as a favorable factor to reduce lateral pressure exerted on the formwork, but it can also be unfavorably to the quality of concrete surface [2]-[3]. In the case of concrete pumping, the excessive friction is a problem representing a lack of mobility or stability which causes blockage in the pumping pipes in jobsites [4]-[5]. Some studies have reported on the effect of friction on formwork pressure. Proske [6] modeled the friction effect based on the friction coefficient increasing over time.

Manuscript received and revised October 2011, accepted November 2011

287

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

S. Bouharoun

to the placement of concrete in formwork. Firstly, two ordinary concretes with 28 and 34% of paste volume were studied at different pressures. Secondly, the friction stress was determined for several mixtures of paste and of sand at deferent water content. The results highlight the role of fine elements on the tribological behavior of fresh concrete. The results showed that the friction stress at the concrete/formwork interface depends on the volume of paste and the water content of the interface layer.

100 Final step (Vertical pressure P = 90 ± 2 kPa)

Vertical pressure (kPa)

II.

displacement of the concrete after placing in real formwork.

Experimental Method

A testing apparatus was developed to analyze the friction and deformability of the fresh concrete. The principle of this device was inspired by the box shear apparatus used in soil mechanics (see Fig. 1). It can reproduce the conditions encountered by manufacturers of concrete walls and precast elements. In particular, it can create sliding contacts between concrete, release agent and formwork [2]-[8]. Two 120 mm diameter cylinders, with the concrete inside, were placed either side of a metal plate. The sample-holders were fitted with a gasket system to prevent water egress. The plate was set in motion using a motor coupled to an endless screw. Plate travel was 800 mm. The concrete was pressured against the plate by a jack.

80 60

33 s

40 20

First step (Vertical pressure P = 30 ± 2 kPa)

0 0

50

100

150

Times (s) Fig. 2. Experimental protocol

The frictional, or tangential, stress was calculated by the following equation for each step of load: τf =

Fmes − Fpar

(1)

Sc

Fpar is the resultant of the parasitic frictional forces due to the watertight system against the plate. The area in contact between the concrete and the plate is calculated from the diameter of the sample-holder. In our case, this area is Sc = 113.1 cm². The characterization of the tribological behavior of the concrete/formwork interface is based on the study of the evolution of the tangential stress over time. II.1.

Water Content of the Interface Layer

The knowledge of the amount of water present at the concrete/wall interface is an important parameter to explain the mechanisms that occur in the vicinity of the formwork during friction. The volume of water mobilized by the material close to the wall can define the interface friction. An interface layer can be formed in the vicinity of wall surfaces in contact with fresh concrete during its implementation in formwork. It is composed of fine particles and water [8]. The experimental procedure is consists to take five samples in the surface of interface layer to determine the water content after friction at different contact pressure (0, 30, 50, and 90 kPa). This paste recovered is weighed in a cup (previously tarred) and then subjected to a drying process. The drying was performed in an oven for 24 hours at a temperature of 105°C. The water content can be calculated by the formula:

Fig. 1. Testing apparatus

For this experimental campaign, the study parameters were chosen according to the jobsite conditions. After mixing, the concrete was placed in the sample-holders. The normal pressure was increased step by step until the final setting state. The size of the load increment depends on the casting rate. In the friction tests, the applied vertical stress was normally 30 ± 2, 50 ± 2, 70 ± 2, 90 ± 2 kPa which corresponding to a lateral pressure applied by the concrete on the formworks of 1.2 and 3.6 m in height (see Fig. 2). A displacement of 2.5 mm with pull out velocity of 5 mm/min was applied to the metallic plate during every load step. The selected velocity simulates the vertical

ω=

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

M f − Ms Ms

(2)

International Review of Civil Engineering, Vol. 2, N. 6

288

S. Bouharoun

TABLE I COMPOSITION AND PROPERTIES OF CONCRETES

After the determination of its water content, a sieving in water was carried out to identify the maximum diameter of the grains that compose this mixture of the interface layer.

II.2.

Concretes Paste volume (%) Cement CEM I 52.5 CP2 (kg/m3) Limestone filler (kg/m3) Sand 0/4 (kg/m3) Crushed aggregate 4/8 (kg/m3) Crushed aggregate 8/12.5 (kg/m3) Water (kg/m3) Water/(cement+ Limestone filler) Crushed aggregate/Sand Slump (mm) Yield stress Plastic viscosity

Wall Roughness

For the tests, the tribometer plate was taken from the wall of formwork in order to reproduce the same jobsite conditions. Its surface roughness was measured using a portable roughness meter with graphic display. This apparatus can determine the following parameters: Ra (arithmetic average), Rv (maximum valley depth), Rq (root mean squared), Rp (maximum peak height) and Rt (Maximum Height of the profile). Fig. 3(a) shows the roughness parameters.

C1 28 232 77 838 287 778 176 0.57 1.27 12 549.1 30.1

C2 34 282 94 768 263 712 214 0.57 1.27 15 355.4 16.7

30 s Aggregates + Sand + 1/3 Water

Stop 10 min

Binder 1 min

2/3 Water

End of mixing

1 min

Fig. 3(b). Mixing procedure

The rest time of 10 min is the time of absorption of water by the aggregates. This precaution allows preventing an alteration of the rheological properties of concrete few minutes after mixing. The mixer used to prepare of concrete is DZ120V DIEM. It is equipped with several blades to optimize the mixing. A slump test was conducted after each mix. The yield stress and the plastic viscosity of each concrete were measured using the ICAR rheometer. This device is portable on jobsite; it was developed at the University of Texas [10]-[11]. For each concrete mixture, the protocol test begins with a pre-sheared, followed by a rest period so that the concrete can be completely restructured. The concrete is then sheared at a velocity of 3.14 rad/s during 20 s to obtain a complete breakdown of the material. Then, a ramped down from 3.14 rad/s to 0.314 rad/s, by steps of 0.0684 rad/s is applied. When increasing the volume of paste the yield stress and plastic viscosity decrease (see Table I). The paste gives to the concrete a better deformability. The yield stress depends on the friction between the grains forming the granular skeleton of concrete [12]. For concretes tested, the paste volume was increased by keeping the w/b and a/s ratios constant in the mixtures. This increase contributes to reduce the intergranular frictions by lubricating the contacts aggregate/aggregate and therefore a decrease in the yield stress and plastic viscosity.

Fig. 3(a). Roughness parameters

The two parameters Ra and Rt, will be measured in this study. The knowledge of the height of maximum peak to valley profile (Rt) allows to evaluate the ability of fine grains to be trapped in the roughness of metallic plate during the friction process. Several measurements were carried out on both sides of the plate in order to determine the average roughness. The obtained results indicate that the height of the maximum peak to valley profile (Rt) ranges 4 to 19 µm. For the arithmetic mean of the profile deviations from the mean (Ra), it is comprised between 0.7 and 1.7 µm. The average of these parameters is Ra = 1 µm and Rt = 13 µm.

III. Materials III.1. Concretes Composition The effect of paste volume was studied by formulating two concretes without superplasticizer at 28 and 34% of paste, named, C1 and C2 respectively. The water/binder ratio (w/b) and the Aggregate/sand ratio (a/s) are keeping constant. Table I shows the formulation of concretes used for the study. In this work, the French standard NF P 18-404 (entitled “Concretes – Design, suitability and inspection testing – Specimen production and conservation) was taken into account. The concrete was mixed from dry materials, using the following procedure (see Fig. 3(b)).

III.2. Binder Characteristics The cementitious material used in all mixtures consisted of a CEM I 52.5 CP2 Type Portland cement with a Blaine fineness of 410 m²/kg, a density of 3100

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

289

S. Bouharoun

TABLE II AGGREGATE CHARACTERISTICS

kg/m3, and a ground granulated blast furnace slag. The maximum size of grains is 60 µm with 80% of the grains are smaller than 20 µm. The limestone filler used in this study is BETOCARB-MQ which is essentially composed of carbonate (99.3%). The maximum size of grains is 80 µm with 63% of the grains are smaller than 20 µm in studied volume. In order to better observe the percentage of fine elements in a fresh concrete volume, the cumulative proportion of cement and limestone filler grains was illustrated as a function of their diameter.

Sand

Gravel 1

Gravel 2

0/4 1,49 2,57 0,8 97,38

4/8 1,42 2,72 -

8/12.5 1,45 2,6 -

100 Cumulative passing (%)

Cumulative proportion (%)

100

Aggregates Granular class Apparent volumetric mass (t/m3) Absolute volumetric mass (t/m3) Absorption coefficient SE (%)

Cement Limestone filler

80 60 40

80

Sand 0/4 Gravel 4/8 Gravel 4/12.5

60 40 20 0 0.01

0.1

20

1 Grains diameter (mm)

10

100

Fig. 5. Distribution of the grain size

0 125 100 80

63

50

40

20

10

6

2

1

The fineness modulus (Fm) of sand is derived from the granulometric curve. Its value is 2.51. This sand includes 0.6% of sand grains less than 80 µm. In order to understand the role of this grains class, all elements less than 80 µm were separated from the sand and studied as a function of water content. The different water contents which are chosen for the tribological tests are 0, 5, 10, 15, 20, 19 and 23%. The friction tests were also performed for the aggregates forming the granular skeleton with identical proportion of those used in the formulation of concretes (without binder and water).

Grains diameter (µm) Fig. 4. Distribution of the grain size of cement and limestone filler

In the concrete formulation, the amount of fine elements consists of 25% of limestone filler and 75% of cement. In addition, 69% of cement and 12% of limestone filler have a diameter less than 20 µm. This information will be used to explain some phenomena at the interface. To understand the mechanisms at the interface linked to the fluidity of paste, several mixtures were prepared with different water content. A friction tests were carried out using plan/plan tribometer to identify the influence of fine element on the friction stress. The mixtures studied were prepared with water content ranging between 0 and 36%.

IV.

Results and Discussion

IV.1. Friction Stress Evolution To understand the mechanisms at the concrete/formwork interface linked to the paste volume, friction tests were conducted using plan/plan tribometer. Fig. 6 illustrates an example of the tangential force recording as a function of time pour the concrete C1. These graphs can be decomposed into two zones [7]: - Zone I presents an increase of friction stress due to the start-up time of engine to catching up the gaps and the elastic response of the mechanical system. - Zone II reflects a stationary regime. Friction is substantially constant during the test. The stresses of static and dynamic friction are similar for this concrete/wall interface. This will be the values of friction that will be taken into account. Fig. 7 shows the evolution of the friction stress as a function of contact pressure for the both concretes and the aggregates (gravel + sand) forming the granular

III.3. Aggregate Properties Sand and two types of gravels were used in the concrete manufacturing. To obtain a formulation corresponding to the specifications, characterization tests were performed in the laboratory. These tests concern the absolute and apparent densities, cleanliness (sand equivalent) and the absorption coefficient and the sand granulometric analysis. The characteristics of the aggregates used are given in Table II. The results obtained show that the sand is very clean and can be used without risk in the composition of concrete. Fig. 5 shows the size and the percentage of different grains diameter making up the concrete.

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

290

S. Bouharoun

skeleton. Each point on this graph represents an average of five points in the dynamic regime (zone II).

with contact pressure which is rigidifying the medium at the interface and requires a more important shear force to create a movement close to the wall. In addition, during the displacement of the plate, the friction between particles acts as a resistance force. Indeed, the movement of plate leads to break the network formed by fine particles. This process indicates that the material undergoes an internal shear near the wall (see Fig. 8).

18 P = 90 kPa

15

P = 70 kPa

Tangential force (daN)

P = 50 kPa

12

P = 30 kPa

9 Zone II

6 Zone I

3 0 0

50

100

150

Times (s) Fig. 6. Evolution of the friction stress a function of time

Friction stress (kPa)

30

C1 C2 Aggregates (Gravel + Sand)

25

Fig. 8. Physical interpretation of friction at concrete/wall interface – Influence of the contact pressure

20

Fig. 7 also shows that the friction stress recorded for the aggregates are lower than those of concretes. This phenomenon confirms that the friction is not governed by the granular skeleton. So, the mechanisms of friction at the concrete/wall interface are probably linked to the fine elements of mixtures.

15 10 5 0 0

20

40 60 Contact pressure (kPa)

80

100

IV.2. Water Content of the Interface Layer

Fig. 7. Evolution of the friction stress as a function of contact pressure

The determination of water content and the role of fine elements at the concrete/wall interface allow explaining the influence of paste volume on the evolution of friction stress. Fig. 9 shows the obtained results of water content for the both concretes/wall interfaces.

The evolution of the curve is linear and follows a friction law of Coulomb type in the range of pressure studied. In addition, an increase in friction stress is observed when the contact pressure and the paste volume increase. The normal stress applied to the material is transmitted to the interface layer by a chain force. This phenomenon implies the contacts grain-grain, which plays an important role in the stress distribution at the interface and increased intergranular friction [13]. When the contact pressure increases, the intensity of these contacts becomes more important between grains forming the mixture (see Fig. 8). The medium becomes more rigid at the interface, which requires a more important shear force to create a movement close to the wall. This configuration generates an increase of shear stresses at the interface concrete/formwork. The metallic plate has a maximum peak-to-valley height of the roughness profile (Rt) that can reach up 19 µm, allowing fine particles (cement, filler) to be trapped in the roughness. Grains trapped cannot move in this configuration, they tend to lead other grains near the wall creating a mechanical adherence between the interface layer and the wall of formwork. This adherence increases

25

C1 C2

Water content (%)

23 21 19 17 15 0

20

40 60 Contact pressure (kPa)

80

100

Fig. 9. Water content as a function of paste volume

A slight increase of water content is observed when the volume of paste decreases and the contact pressure increases. It is possible to note that the concrete C1 has

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

291

S. Bouharoun

16

20

Friction stress (kPa)

an increase about 4% of water content at the interface between 0 and 90 kPa. In this same pressure range, the concrete C2 has meanwhile increased approximately 3% of its water content. In the case of zero pressure, the interface is formed during the implementation of concrete. At the approach of the wall, the aggregates of large size give the way to the smallest grain size to fill the spaces in the granular skeleton. However, voids are present despite the spread granular distribution in size. These spaces can be finally filled completely or partially by the free water of the concrete. With increasing of paste volume, the amount of fine elements becomes more important, allowing a better filling of the spaces between the gravel and sand close to the wall. Consequently, fewer spaces remain unoccupied and this small amount of free water is more distributed in the mixture. Under the influence of the confining pressure, a small amount of free water can be directed to these small cavities at the interface that are difficult to fill with a simple stitching of concrete. This phenomenon can explain the high water content observed for concrete with low paste volume (C1). After a sieving in water of the mixture collected at the concrete/wall interface, it appears that the interface layer is constituted of grains less than 80 µm. To understand the role of water and all fine element in the vicinity of the formwork surface, friction tests were conducted using the tribometer (see Fig. 1) in order to analysis the influence of water content. Figs. 10 and 11 illustrate the evolution of friction stress as a function of water content for the limestone filler and the fines of sand. The results show that the lubricant role of water at the interface is occurs when the water content exceeds 20% for limestone filler and 10% for fines of sand. The results of these figures also show that these components can’t be the origin of the mechanisms of increase of friction stress. Fig. 12 shows the evolution of friction stress as a function of water content for the cement. The frictions stress are similar than those obtained for the two concretes. This information confirms that the cement is the component that governs the friction behavior at the interface.

30

Friction stress (kPa)

10 15 20 Water content (%)

P = 90 kPa 25

30

Friction stress (kPa)

20 15 10 5 P = 30 kPa 0

10

P = 50 kPa

P = 70 kPa

20 Water content (%)

P = 90 kPa 30

40

Fig. 12. Evolution of friction stress as a function of water content for the cement

The cement grains in contact with water tend to agglomerate in flocs form in the presence of water by the electrical charges on the grain surface. They tend to trap a certain amount of water inside the flocs [14]. When these flocs are trapping in the roughness, it acts as a resistance force. With an increase of water content, the number of these flocs increase which causes an increase in friction stress. From a water content of 30%, the friction stress decreases. This phenomenon is probably related to the saturation of the cement. So, the lubricant role of cement begins from 30% of water content. In the concretes composition, the paste volume was increased by keeping the w/b constant and by reducing the amount of aggregates. The amount of fine in the vicinity of the formwork surface is greater for the concrete with higher paste

P = 90 kPa 30

P = 70 kPa

Cement

0

0 20 Water content (%)

5

P = 50 kPa

25

4

10

P = 30 kPa

At a pressure of 30 kPa, the friction stress is stable whatever the water content. At this load, the material is not sufficiently solicited to allow the trapping of cement grains in the roughness of the metal plate. From a pressure of 50 kPa, the results show that the friction stress is constant for water contents ranging between 0 and 20%. In this range of pressure, the tribological behavior of materials is identical. From 20 to 30% of water content, an increase of friction stress is observed in this range for three contact pressure (50, 70, 90 kPa).

8

0

4

Fig. 11. Evolution of friction stress as a function of water content for the fines of sand

12

P = 70 kPa

8

0

16

P = 50 kPa

12

0

Limestone filler

P = 30 kPa

Fines of sand

40

Fig. 10. Evolution of friction stress as a function of water content for the limestone filler

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

292

S. Bouharoun

volume which increases the thickness of the interface layer. This phenomenon also increases the possibility to trap the fine particles in the formwork roughness. It is important to remember that in concrete composition, 69% of cement is less than the maximum peak-to-valley height of the roughness profile (Rt = 19 µm). This information confirms that the trapping of the cement particles in the wall roughness is a key element in explaining the mechanisms of interface.

V.

[6]

[7]

[8]

[9] [10]

Conclusion

This work allowed identifying the influence of fines content in a concrete and water content at the interface on the tribological behavior. Rheologically, the increase of the paste quantity fluidizes the concrete and reduces the rheological parameters but also an increase in the friction stress at the concrete/wall interface. This surprising result could be explained by the role of cement grains in the interface layer which generates in an increase of friction. This augmentation is linked to the enrichment of the interface layer of fine elements. The thickness of this layer becomes greater which increases the probability to trap the cement grains in the roughness of the metal plate. The water content plays an important role in reducing friction at the interface. An increase in the volume of paste produces a decrease in water content near the wall which reduces the friction stress. The results presented in this study allow to comprehending the influence of the paste volume the tribological behavior of traditional concrete without superplasticizer. The important element that could complete this work is the reconstruction of the interface layer for each concrete. This layer will be the subject of tribological and rheological study to determine the real properties of concrete in the vicinity of formwork. This study can also be completed by an investigation on the influence of superplasticizer dosage on the behavior of the interface.

[11]

[12] [13]

[14]

la constante visqueuse, Ph.D. dissertation in French, University of Cergy Pontoise, 2009. T. Proske, Frischbetondruck bei Verwendung von Selbstverdichtendem Beton - Ein wirklichkeitsnahes Modell zur Bestimmung der Einwirkungen auf Schalung und Rüstung, Ph.D. dissertation in German, Technical University of Darmstadt, 2007. Y. Vanhove, Contribution à l’étude du frottement d’un béton autoplaçant contre une surface métallique – Application aux poussées contre les coffrages, Ph.D. dissertation in French, University of Artois, Béthune, 2001. Y. Vanhove, C. Djelal, A. Magnin, Device for studying fresh concrete friction, Cement Concrete and Aggregates, vol. 26: 3541, 2004. A. M. Neville, Properties of Concrete (Wiley; 4 edition, 1995). E. P. Koehler, Development of a Portable Rheometer for Portland Cement Concrete, Ph.D. dissertation, University of Texas, Austin, 2004. E. P. Koehler, D. W. Fowler, C. F. Ferraris, S. Amziane, A New Portable Rheometer for Fresh Self-Consolidating Concrete, Proc. of session ACI , New York, 2005. F. De Larrard, Concrete mixture proportioning - Ascientific approach (Spon Press, 1999). L. E. Silbert, G. S. Grest, G. W. Landry, Statistics of the contact network in frictional and frictionless granular packing, The American Physical Society, vol. 66:1-9, 2002. M. R. Rixom, N. P. Mailvaganam, Chemical admixtures for concrete (Spon Press, 1986).

Authors’ information Laboratoire Génie Civil et géo-Environnement (LGCgE) - Lille Nord de France (EA 4515), FSA, TechnoParc Futura, F-62400 Béthune, France. Samir Bouharoun -Tizi Ouzou, (Algeria), 03 June 1982. Ph.D. Dissertation in civil engineering, University of Artois, Béthune, France, 2011. Research Master's degrees in Mechanics, Materials, Structure, Process, Ecole Nationale Supérieure d’Arts et Métiers, Metz, France, 2007. Engineer in Civil Engineering, Ecole Nationale des Travaux Publics, Algiers, Algeria, 2006. Research interests: - Tribological and rheological behavior of fresh concrete, - Physicochemical of the concrete/release agent/formwork interface. - Quality of concrete surface

References [1]

[2]

[3]

[4]

[5]

K. Salhi, B .Mezghiche, "Effects of Sand of Dune and Granulated Slag on the Properties of Cement, International Review of Civil Engineering, Vol. 1: 165-169, 2010. C. Djelal, Y. Vanhove, and A. Magnin, Tribological behaviour of self-compacting concrete, Cement and Concrete Research, vol. 34: 821-828, 2004. L. Libessart, Influence de la composition des agents de démoulage à l’interface coffrage/béton - Impact sur l’esthétique des parements en béton, Ph.D. dissertation in French, University of Artois, Béthune, 2006. T. T. Ngo, E. H. Kadri, R. Bennacer, F. Cussigh, Use of tribometer to estimate interface friction and concrete next term boundary layer composition during the fluid “concrete pumping”, Construction and Building Materials, Vol. 24: 1253-1261, July 2010. T. T. Ngo, Influence de la composition des bétons sur les paramètres de pompage et validation d'un modèle de prévision de

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

293

International Review of Civil Engineering (I.RE.C.E.), Vol. 2, N. 6 November2011

Uplift Behaviour of Plate Anchors Embedded in Cohesionless Soils Baleshwar Singh, Birjukumar Mistri Abstract – Plate anchors are being increasingly used as a foundation system to provide uplift resistance for several types of structures constructed both on land and offshore sites. Accurate estimation of the ultimate uplift capacity is necessary in the design to ensure the safety and stability of the supported structures. A parametric study has been carried out with finite element modeling using PLAXIS to estimate the ultimate uplift capacity of horizontal anchor plates embedded in cohesionless soils by varying soil relative density and embedment ratio. The results have been compared with those obtained from various methods and approaches which are based on theoretical, computational, physical and experimental studies. Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved. Keywords: Plate Anchors, Uplift Capacity, Finite Element Analysis, Parametric Study

Nomenclature A A1 & A2 B c E F1, F2 Fr h H H/h, H/B K Ko Kp L Nu Nr Qu R Rψ RR RK t W Z δ γ γdry γsat

ν ψ

I.

Area of plate anchor Functions of soil friction angle Width of plate anchor Soil cohesion Soil modulus of elasticity Functions dependent on soil friction angle Capacity factor for the basic case of a smooth anchor Diameter of plate anchor Embedment depth of plate anchor Embedment ratio of plate anchor Coefficient of earth pressure Coefficient of earth pressure at rest Coefficient of passive earth pressure Length of plate anchor Breakout factor Anchor capacity factor Ultimate uplift capacity of plate anchor Radius of circular failure surface Correction factor for dilatancy Correction factor for anchor roughness Correction factor for initial stress state Thickness of plate anchor Weight of lifted soil mass Depth from ground surface Angle of earth pressure reaction with normal Unit weight of soil Dry unit weight of soil Saturated unit weight of soil Soil friction angle Soil critical state friction angle in plane strain Peak soil friction angle Soil friction angle in plane strain Poisson’s ratio of soil Dilation angle of soil

Introduction

The design of many structures requires the foundation system to resist vertical uplift forces. Different systems have been developed in order to take care of such forces. As ocean operations and construction have expanded and moved into deeper waters, the need for the development of high capacity, reliable anchor systems for long-term moorings has also emerged. Plate anchors find their use in providing tie-back resistance for transmission towers, utility poles, submerged pipelines, offshore platforms, wind turbines etc., and represent an alternative to gravity and other embedded anchors in both land and marine environments. In the design of plate anchors, it is necessary to determine the ultimate uplift resistance and the influence of size and shape on this resistance. It is also necessary to obtain the displacement at which the ultimate resistance is mobilized as well as the resistance under working loads. Several theoretical and experimental studies on plate anchors embedded in homogeneous cohesionless soils have been performed in the past. A plate anchor is classified as shallow or deep depending on its mode of failure. It is termed as shallow when a definite slip surface is visible on the soil surface at failure. For a deep anchor plate, no such surface is observed. Various researchers have confirmed this dual mode of failure phenomenon. The depth at which the transition from a shallow to a deep anchor takes place will depend on the properties of the particular soil. The present study is related to the uplift behavior horizontal plate anchors embedded in sandy soils. The load-displacement response and ultimate uplift capacity have been determined from finite element analysis using PLAXIS, and the computed capacity is compared with those obtained from various approaches and methods found in the literature.

Manuscript received and revised October 2011, accepted November 2011

294

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

B. Singh, B. Mistri

II.

Literature Review

·

·

·

(3)

1

The ultimate uplift capacity of a horizontal plate anchor in cohesionless soil can be expressed as:

2

2

(1)

The magnitude of this dimensionless breakout factor Nu is influenced by the geometry of the failure surface developed within the soil mass, and is dependent on a number of factors that include soil properties, relative depth and anchor dimensions. The analysis of anchor behavior can be divided into two distinct categories, with the soil-anchor interface being capable of sustaining tension or not. If there is uncertainty surrounding the actual magnitude of any suction force, it should not be considered in the analysis. II.1.

Theoretical Investigations

Fig. 2. Vesic’s failure mechanism

The uplift capacity of plate anchors can be determined by a number of approaches. Majer [1] adopted the simplest approach in which the assumed failure mechanism is a vertical slip surface as shown in Fig. 1. According to this approach, the failure is assumed to take place along the surface of a cylinder of soil above the plate. The capacity is computed from the weight of soil within the cylindrical failure surface directly above the anchor and the frictional resistance along this surface. For predicting the ultimate uplift capacity of a strip plate anchor of width B, unit length and thickness t, he proposed the following relation for breakout factor: 1

 

Meyerhof & Adams [3] conducted a number of model and full-scale uplift tests on footings with special reference for transmission towers. They found that at ultimate load, the failure surface for shallow depths makes an angle with the ground surface as shown in Fig. 3, and that the magnitude of this angle depends on the relative density and angle of internal friction of the soil. With increasing depth of the footing, the compressibility and deformation of the soil mass above the footing prevent the failure surface from reaching the ground surface. The extent of this local shear fai1ure can be included in the analysis by limiting the vertical extent H of the failure surface and utilizing the surcharge pressure above the level of the failure surface. They proposed a semi-theoretical relationship for the estimation of uplift capacity for square, circular and rectangular plate anchors, with the use of shape factors. It was concluded that this theory is good for dense sand but overestimates in loose to medium dense sand. They gave the expression for ultimate uplift capacity of strip plate anchor embedded at shallow depths in c- soil as:

(2)

2

(4)

Rowe & Davis [4] considered the effects of soil dilatancy, initial stress state and anchor roughness on the uplift capacity of plate anchors, and presented numerical solutions that were obtained from an elasto-plastic finite element analysis. They found that soil dilatancy has a significant effect on ultimate uplift capacity. The effect of initial stress state decreases ultimate uplift capacity by about 10%, whereas roughness has a negligible effect. The following relation was proposed for the ultimate uplift capacity of plate anchors in sand:

Fig. 1. Majer’s vertical failure surface

Vesic [2] considered the problem of an explosive point charge expanding a spherical cavity close to the surface of semi-infinite, homogeneous isotropic solid. If the distance H is small enough in a soil mass as shown in Fig. 2, there will be an ultimate pressure that will shear away the soil located above the cavity. At that time, the diameter of the spherical cavity is equal to h. Based on these, he proposed the following expression for the breakout factor of a circular plate anchor of diameter h:

(5)

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

295

B. Singh, B. Mistri

where Nr = Fr .Rψ .RR .RK . Influence charts were presented which can be used in hand calculations to obtain an estimate of anchor capacity for a wide range of geometries and soil types.

complete load-deformation characteristics of horizontal plate anchors by assuming a truncated conical failure surface and by using hyperbolic stress-strain curves of cohesive-frictional soils as the constitutive law. The parameters of the model are computed from the stressstrain relationship of the soil obtained from triaxial tests. The analysis incorporates the effect of the shape of the anchor for strip, square and circular anchors. Merifield et al. [8] applied three-dimensional numerical limit analysis and axisymmetrical displacement finite element analysis to evaluate the effect of anchor shape on the uplift capacity of horizontal anchors in sand. The anchor is idealized as either square or circular in shape. Results are presented in the familiar form of break-out factors based on various anchor shapes and embedment depths. Hanna et al. [9] presented an analytical model to predict the uplift capacity and the load-displacement relationship for circular plate anchors in sand. Design charts were presented for practical purposes. The model was based on the failure mechanism observed in laboratory testing and utilized the limit equilibrium technique. Expression was given to estimate the critical depth which separates deep from shallow anchors. The critical depth depends upon the diameter of anchor and angle of internal friction. The radius of influence of an individual anchor on the ground surface was established, and accordingly the spacing between anchors can be determined to avoid overlap of the stress zones between anchors. Kumar & Kouzer [10] examined the uplift capacity of rigid horizontal strip anchors placed in sand on the basis of an upper bound limit analysis in combination with finite elements and linear programming. Even though the analysis considered the development of plastic strains within elements in all cases, the soil mass lying above the anchor was noticed to remain rigid and a planar rupture surface emanated from the anchor edge making an angle with the vertical. The magnitude of the breakout factor was found to increase substantially with an increase in the embedment ratio of the anchor and the friction angle of the soil mass. The influence of friction angle on the uplift resistance was found to be greater at higher embedment ratios. White et al. [11] presented a limit equilibrium solution for the uplift resistance of plate anchors buried in sand. The geometry of this solution reflected observations from model tests. Peak angles of friction and dilation were found using established correlations that captured the influence of stress level and density. These angles governed the geometry of the failure mechanism and the mobilised resistance. The solution was validated using a database assembled from the published literature. Simple charts for the prediction of ultimate uplift capacity from critical state friction angle, relative density and normalised burial depth were presented. They proposed the breakout factor for a strip plate anchor of unit length as:

Fig. 3. Failure surface of Meyerhof & Adams

Vermeer & Sutjiadi [5] considered straight rupture surfaces at an inclination to the vertical equal to the soil dilatancy angle, and proposed the following relation for breakout factor: 1

 

(6)

Chattopadhyay & Pise [6] proposed a theoretical model for evaluating the ultimate vertical breakout resistance of horizontal plate anchors embedded in sand, by assuming an exponential equation for the curved axisymmetric failure surface through the surrounding soil as shown in Fig. 4. It indicated the existence of a characteristic relative depth dependent on the soil friction angle, and beyond which breakout factor approaches a constant value. It was capable of predicting the breakout factors for a wide range of values of angle of shearing resistance of sand.

1

Fig. 4. Failure surface of Chattopadhyay & Pise

(7)

Saran et al. [7] proposed a new method for obtaining Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

296

B. Singh, B. Mistri

where: tan

tan

tan

2

cos 2

1

only static loading may be misleading. Charts were provided for use in design. The design approach involves the assessment of both anchor capacity and displacement with reference to the useful life of the system. Murray & Geddes [14] described an investigation into the vertical uplift of anchor plates in a cohesionless soil. The factors investigated in relation to the loaddisplacement response were the size and shape of plate, depth of embedment, sand density, and plate surface roughness. The results of laboratory tests were presented together with equilibrium analysis and limit analysis methods of predicting the ultimate resistance of strip, circular and rectangular anchors. A curved failure surface was used in the equilibrium analysis whereas a plane rupture surface was considered for the limit equilibrium analysis. The proposed breakout factors were as follows: For strip anchors from equilibrium analysis:

1

2 II.2.

Experimental Investigations

Several design methods are based on observed failure surfaces in small-scale model tests at unit gravity, and usually a limit equilibrium approach is employed. Balla [12] conducted several model and field tests on shallow circular under-reamed mushroom foundations in dense sand. The tests revealed a vertical rupture surface at the upper surface of the foundation, curving outwards and intersecting the ground surface at approximately 45° – Ø/2. He simplified this surface to a circular arc as shown in Fig. 5, having the following radius:

1   45

2

For circular anchors from equilibrium analysis:

Kotter's equation was used to describe the distribution of shearing resistance on this surface. In examining the vertical equilibrium of the system, he ignored the presence of normal stresses on the failure surface and only resolved the shear stress components in the vertical direction. The ultimate uplift capacity was expressed as:  

,

(10)

2

(8)

,

1

2 2 3

2

Ø 2 2

1

(11)

For rectangular anchors from limit analysis:

(9)

1

1

(12)

3

Frydman & Shaham [15] reanalyzed published experimental data and results of uplift tests on small models and prototypes of slab anchors. Based on these, they presented a simple theoretical expression to reasonably predict the uplift capacity of a continuous, horizontal slab as a function of depth-towidth ratio. Factors to account for shape and inclination were also established leading to expressions for the estimation of uplift capacity of any slab anchor. The resulting expressions were used to predict the uplift capacity of nine full-scale tests on prototype slab anchors placed at various inclinations and depths in a dense sand profile, and satisfactory agreement was obtained with measured values. The proposed expressions for the breakout factor of horizontal plate anchors were: For dense sand:

Fig. 5. Balla’s failure surface

Andreadis & Harvey [13] proposed a procedure for the estimation of uplift capacity of circular plate anchors based on medium scale laboratory tests and a few prototype tests. In addition to the influence of embedment ratio and anchor size, the uplift capacity was found to be dependent on the degree of soil disturbance adjacent to the anchor during installation. They introduced the effect of cyclic loading and showed that the analysis based on

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

1

Ø 1

· 0.51

2.35

0.15 1

0.15

·

(13)

International Review of Civil Engineering, Vol. 2, N. 6

297

B. Singh, B. Mistri

the plate. For the soil model, coarse element is adopted as the basic element type. During the generation of the mesh, 15-node triangular elements have been selected in preference to the alternative 6-noded versions in order to provide greater accuracy in the determination of stresses. The analyses were carried out using a plane strain model for loose, medium dense and dense sands.

For loose sand:

1

Ø 1

0.5

0.15 1

(14)

0.15

III. Present Study The anchor used in the study is a strip plate anchor of unit length connected to a rigid tie shaft. PLAXIS 2D foundation has been used in the analysis. Mohr-Coulomb soil model was adopted to predict soil behavior. It involves five input parameters: E and γ for soil elasticity, Ø and c for soil plasticity and ψ for angle of dilatancy. The initial stresses were generated by using Jaky’s formula which gives the at rest earth pressure coefficient Ko = 1−sinØ, where Ø is the angle of internal friction in terms of effective stresses. The soil properties used in the analysis are shown in Table I. TABLE I PROPERTIES OF SANDY SOILS Medium Loose Property Dense Sand Sand Angle of internal friction, Ø 30° 35° Angle of dilation, ψ 0° 12° 14 17 Dry unit weight of soil, γdry (kN/m3) 17 20 Saturated unit weight of soil, γsat (kN/m3) Poisson’s ratio, ν 0.3 0.35 20,000 25,000 Modulus of elasticity, E (kN/m2)

Dense Sand

Fig. 6. Geometrical arrangement of plate anchor in PLAXIS analysis

40° 15° 20

IV.

Figures 7 and 8 show the displacement contours obtained from the PLAXIS analysis for both shallow and deep anchors in loose, medium dense and dense sands.

23 0.4 30,000

Three sand soils of different relative densities, namely loose sand, medium dense sand and dense sand were selected. The plate having width B of 2m, unit length and thickness of 0.1m is analyzed by varying its embedment ratio as 2, 4, 6, 8 and 10. For increasing strength analysis, the modulus of elasticity at any depth is computed from the following equation: 1

Results and Discussion

Fig. 7. Displacement contours of shallow plate anchors (H/B = 2) in loose, medium dense and dense sands

(15)

where E0 = Initial modulus of elasticity at 1 m depth, Z = depth in m, and n is exponent assumed as 0.5. The typical model geometry adopted is shown in Fig. 6. The plate has width B of 2 m with unit length and thickness of 0.1 m. In elevation, the distance between the plate and bottom of soil boundary is adopted as 4B. In plan, the distance between the plate and the mesh boundary is adopted as 2B. The anchor plate is modeled by using floor elements. Floor element is used to model thin structural element in ground with suitable flexural rigidity. It is composed of 6-noded triangular plate element with six degree of freedom per node. As soil behavior under uplift is being studied, the flexural rigidity of plate has been assumed to be very high to avoid buckling of

Fig. 8. Displacement contours of deep plate anchors (H/B = 6) in loose, medium dense and dense sand

These figures show that for an anchor placed at a

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

298

B. Singh, B. Mistri

shallow depth, the soil displacements and the shear stresses extend to the soil surface. On the other hand, the mechanism for deep embedded anchors is more localized in nature, predominating in the region above the anchor. Another observation is that as the soil density increases, the volume of soil involved in failure is larger leading to an increase in the uplift capacity. Ilamparuthi et al. [16] observed similar failure mechanisms from tests on half-cut models of circular plate anchors embedded in dense sand. For shallow anchors, the failure pattern was in the form of a cone extending from the top edge of the plate to the soil surface with sides inclined at Ø/2 with vertical as shown in Fig. 9. For deep anchors, the failure surface was of balloon shape above the plate as shown in Fig. 10.

The plots show a rapid increase in uplift capacity within a small range of deformation followed by smaller increments in uplift capacity with wide range of deformation. It is observed that with the increase in density of sand and embedment ratio, the uplift resistance increases significantly. From serviceability point of view, 2.5% of embedment depth is adopted as the limit. The uplift resistance at that point is adopted as the ultimate uplift capacity for that particular embedment depth. As per this criterion limit, the maximum displacement limit ranges from 100 to 500 mm for embedment ratio ranging from 2 to 10. 8000

Uplift Load (kN)

7000 6000 5000 H/B = 2

4000

H/B=4 3000

H/B=6 H/B=8

2000

H/B=10

1000 0 0

0,2

0,4

0,6

Displacement (m) Fig. 11. Influence of embedment ratio on load-displacement response in loose sand

Fig. 9. Failure surface in model tests on shallow anchor in dense sand

14000

Uplift Load (kN)

12000 10000 8000

H/B = 2 H/B=4

6000

H/B=6 H/B=8

4000

H/B=10 2000 0 0

0,2

0,4

0,6

Displacement (m) Fig. 10. Failure surface in model tests on deep anchor in dense sand

Fig. 12. Influence of embedment ratio on load-displacement response in medium dense sand

The uplift resistance of plate anchors with displacement in sand of various densities is presented in Figs. 11 to 13.

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

299

B. Singh, B. Mistri

Majer (1955) Vesic (1965) Meyerhof & Adams (1968) Rowe & Davis (1982) Vermeer & Sutjiadi (1985) Chattopadhyay & Pise (1986) Saran et al. (1986) Merifield et al. (2006) Hanna et al. (2007) Kumar & Kouzer (2007) White et al. (2008) PLAXIS Analysis  PLAXIS Analysis (Increasing Strength)

20000

30000

12000

Ultimate Capacity  (kN)

Uplift Load (kN)

16000

H/B=2 H/B=4

8000

H/B=6 H/B=8

4000

20000

10000

H/B=10 0 2

0 0

0,2

0,4

8

Ultimate Capacity  (kN)

Ultimate Capacity  (kN)

10000

4

6

8

10

Embedment Ratio Fig. 16. Comparison of uplift capacity in dense sand from PLAXIS analysis with those of theoretical studies TABLE II ULTIMATE UPLIFT CAPACITY (kN) FOR SHALLOW ANCHORS FROM THEORETICAL STUDIES & PLAXIS ANALYSIS Method Embedment Loose Medium Dense ratio Sand Dense Sand Sand Majer (1955) 4 288 366 453 Vesic (1965) 4 1,680 2,338 3,200 Meyerhof & Adams 4 1,600 2,927 4,934 (1968) Rowe & Davis (1982) 4 806 809 980 Vermeer & Sutjiadi 4 707 984 980 (1985) Chattopadhyay & Pise 4 1,120 1,992 2,176 (1986) Saran et al. (1986) 4 245 330 432 Merifield et al. (2006) 4 2,240 3,360 5,440 Hanna et al. (2007) 4 2,912 5,440 7,360 Kumar & Kouzer 4 336 347 358 (2007) White et al. (2008) 4 741 1,033 1,393 PLAXIS analysis 4 1,971 3,948 5,964

0 6

20000

2

5000

4

30000

0

Majer (1955) Vesic (1965) Meyerhof & Adams (1968) Rowe & Davis (1982) Vermeer & Sutjiadi (1985) Chattopadhyay & Pise (1986) Saran et al. (1986) Merifield et al. (2006) Hanna et al. (2007) Kumar & Kouzer (2007) White et al. (2008) PLAXIS Analysis  PLAXIS Analysis (Increasing Strength)

2

10

Majer (1955) Vesic (1965) Meyerhof & Adams (1968) Rowe & Davis (1982) Vermeer & Sutjiadi (1985) Chattopadhyay & Pise (1986) Saran et al. (1986) Merifield et al. (2006) Hanna et al. (2007) Kumar & Kouzer (2007) White et al. (2008) PLAXIS Analysis  PLAXIS Analysis (Increasing Strength)

40000

Figures 14 to 19 show the comparison of results obtained from PLAXIS analysis with the various approaches and methods found in the literature. For greater clarity, they are separated in two parts: comparison with the theoretical studies, and then with the experimental studies. The comparative study has been carried out for the sands of different relative densities. Table II and Table III respectively show the ultimate uplift capacities of shallow plate anchors (at embedment ratio of 4) and deep plate anchors (at embedment ratio of 10) from the theoretical methods and PLAXIS analysis. Table IV and Table V respectively present the ultimate uplift capacity values of shallow plate anchors and deep plate anchors obtained from the experimental studies and PLAXIS analysis.

10000

8

Fig. 15. Comparison of uplift capacity in medium dense sand from PLAXIS analysis with those of theoretical studies

Fig. 13. Influence of embedment ratio on load-displacement response in dense sand

15000

6

Embedment Ratio

0,6

Displacement (m)

20000

4

10

Embedment Ratio Fig. 14. Comparison of uplift capacity in loose sand from PLAXIS analysis with those of theoretical studies

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

300

B. Singh, B. Mistri

TABLE III ULTIMATE UPLIFT CAPACITY (KN) FOR DEEP ANCHORS FROM THEORETICAL STUDIES & PLAXIS ANALYSIS Method Embedment Loose Medium Dense ratio Sand Dense Sand Sand Majer (1955) Vesic (1965) Meyerhof & Adams (1968) Rowe & Davis (1982) Vermeer & Sutjiadi (1985) Chattopadhyay & Pise (1986) Saran et al. (1986) Merifield et al. (2006) Hanna et al. (2007) Kumar & Kouzer (2007) White et al. (2008) PLAXIS analysis

10 8 10

721 8,511 19,876

917 13,055 31,676

1134 23,039 57,393

8

2,886

3,168

4,842

10

3,621

5,189

7,156

10

5,240

14,280

32,640

10 8

1,097 10,304

1,540 16,128

2086 32,640

8 10

15,680 1,120

27,200 1,142

39,680 1,164

10 10

3,791 6,720

5,438 12,132

7,507 18,528

Vermeer & Sutjiadi, Kumar & Kouzer, and Saran et al. predict results on comparatively lower side. The method of Chattopadhaya & Pise is based on slipline approach with an assumed curved failure surface, and the results from this method are also falling in middle range. The method of Hanna et al. predicts very high values as embedment depth increases. At shallow embedment depths, results from all the methods are varying within a close range. But as the embedment depth becomes greater, the variation in results becomes larger. Methods that show either very high or low uplift capacities are mainly based on slip-line approach or limit equilibrium approach. Such a variation is found to occur because all parameters such as soil dilation, initial stress state, soil rigidity, overburden pressure, inhomogeneous soil strata, anchor roughness, and suction beneath the plate, which affect the uplift capacity may not be included in the analysis. Methods based on finite element analysis are giving results in the middle range as most of the parameters which affect the uplift capacity have been included in the analysis, and hence more reliable results can be obtained. From Figs. 17 to 19, it is observed that the ultimate uplift capacities from all the experimental studies are showing values on the lower side of the present PLAXIS analysis. Only Balla’s approach shows higher values in loose and medium dense sands as experiments were mainly carried out for dense sand. The method of Frydman & Shaham shows results on the higher side in case of dense sand with deep embedment.

TABLE IV ULTIMATE UPLIFT CAPACITY (KN) FOR SHALLOW ANCHORS FROM EXPERIMENTAL STUDIES & PLAXIS ANALYSIS Method Embedment Loose Medium Dense ratio Sand Dense Sand Sand Balla (1961) 4 5,158 6,257 7,324 Andreadis & Harvey 4 560 951 1,600 (1981) Murray & Geddes 4 903 1,222 1,580 (1987) Frydman & Shaham 4 1,111 1,550 4,075 (1989) PLAXIS analysis 4 1,971 3,948 5,964 TABLE V ULTIMATE UPLIFT CAPACITY (KN) FOR DEEP ANCHORS FROM EXPERIMENTAL STUDIES & PLAXIS ANALYSIS Method Embedment Loose Medium Dense ratio Sand Dense Sand Sand 10

2,127

3,535

Andreadis & Harvey (1981) Murray & Geddes (1987)

8000

Ultimate Capacity (kN)

Andreadis & Harvey (1981) Murray & Geddes (1987) Frydman & Shaham (1989) PLAXIS analysis

6,400

10

4,807

6,622

8,675

10

5,686

8,157

28,980

10

6,720

12,132

18,528

Balla (1961)

10000

From Figs. 14 to 16, it is observed that the ultimate uplift capacities from PLAXIS analysis are falling in the middle of all other theoretical studies. With increase in modulus of elasticity with depth, not much variation in uplift capacity is observed. Methods of Hanna et al., Merifield et al., Meyerhof & Adams, and Vesic give results on the higher side of present analysis. The present analysis shows fair amount of correlation with the methods of Chattopadhyay & Pise, Rowe & Davis, and White et al., whereas the methods of Majer,

Frydman & Shaham (1989) PLAXIS Analysis 

6000

PLAXIS Analysis (Increasing  Strength)

4000

2000

0 2

4

6

8

10

Embedment Ratio Fig. 17. Comparison of uplift capacity in loose sand from PLAXIS analysis with those of experimental studies

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

301

B. Singh, B. Mistri

increase in embedment depth, the range increases. As the soil relative density increases, variation in ultimate uplift capacity also increases.

Balla (1961)

14000

Andreadis & Harvey (1981)

Ultimate Capacity (kN)

12000

Murray & Geddes (1987) Frydman & Shaham (1989)

10000

References

PLAXIS Analysis 

8000

[1]

PLAXIS Analysis (Increasing  Strength)

[2]

6000 [3] 4000 [4] 2000 [5] 0 2

4

6

8

10

[6]

Embedment Ratio

[7]

Fig. 18. Comparison of uplift capacity in medium dense sand from PLAXIS analysis with those of experimental studies

[8]

Ultimate Capacity (kN)

40000

Balla (1961)

35000

Andreadis & Harvey (1981)

30000

Murray & Geddes (1987)

[9] [10]

Frydman & Shaham (1989)

25000 20000

PLAXIS Analysis 

[11]

PLAXIS Analysis (Increasing  Strength)

[12]

15000 [13]

10000

[14] 5000 [15]

0 2

4

6

Embedment Ratio

8

10

[16]

Fig. 19. Comparison of uplift capacity in dense sand from PLAXIS analysis with those of experimental studies

V.

J. Majer, Zur Berechnung von Zugfundamenten (in German), Osterreichische Bauzeitschrift, 10(5): 85-90 1955. A. S. Vesic, Cratering by Explosives as an Earth Pressure Problem, Proc. of VI International Symposium on Soil Structure Interaction, Roorkee, India, 1: 389-397, 1965. G.G. Meyerhof, J.I. Adams, The Ultimate Uplift Capacity of Foundations, Canadian Geotechnical Journal, 24(4): 589-592, 1968. R. K. Rowe, E. H. Davis, The Behavior of Anchor Plates in Sand, Geotechnique, 32(1): 25-41, 1982. P. A. Vermeer, W. Sutjiadi, The Uplift Resistance of Shallow Embedded Anchors, Proc. of 11th International conference on Soil Mechanics and Foundation Engineering, 3, 1635-1638, San Francisco, California,1985. B.C. Chattopadhyay, P. J. Pise, Breakout Resistance of Horizontal Anchors in Sand, Soils and Foundations, 26(4): 16-22, 1986. S. Saran, G. Ranjan, A. S. Nene, Soil Anchors and Constitutive Laws, Journal of Geotechnical Engineering, ASCE, 12(12): 10841100, 1986. R. S. Merifield, A. V. Lyamin, S. W. Sloan, Three-Dimensional Lower-Bound Solutions for the Stability of Plate Anchors in Sand, Geotechnique, 56(2): 123–132, 2006. A. Hanna, T. Ayadat, M. Sabry, Pullout Resistance of Single Vertical Shallow Helical and Plate Anchors in Sand, Geotechnical and Geological Engineering, 25(4): 559-573, 2007. J. Kumar, K. M. Kouzer, Vertical Uplift Capacity of Horizontal Anchors Using Upper Bound Limit Analysis and Finite Elements, Canadian Geotechnical Journal, 44, 698-704, 2007. D. J. White, C. Y. Cheuk, M. D. Bolton, The Uplift Resistance of Pipes and Plate Anchors Buried in Sand, Geotechnique, 58(10): 771–779, 2008. A. Balla, The Resistance to Breaking-Out of Mushroom Foundations for Pylons, Proc. of 5th Intl. Conf. on Soil Mechanics & Foundation Engineering, Paris, France, 1, 569-576, 1961. A. Andreadis, R. C. Harvey, A Design Procedure for Embedded Anchors, Applied Ocean Research, 3(4): 177-182, 1981. E. J. Murray, J. D. Geddes, Uplift of Anchor Plates in Sand, Journal of Geotechnical and Geoenvironmental Engineering, ASCE, 113(3): 202-214, 1987. S. Frydman, I. Shamam, Pullout Capacity of Slab Anchors in Sand, Canadian Geotechnical Journal, 114(11): 1300-1317, 1989. K. Ilamparuthi, E. A. Dickin, K. Muthukrisnaiah, Experimental Investigation of Uplift Behaviour of Circular Plate Anchor Embedded In Sand, Canadian Geotechnical Journal, 39(3): 648– 664, 2002.

Authors’ information

Conclusion

Dr. Baleshwar Singh is an Associate Professor in the Department of Civil Engineering, Indian Institute of Technology, Guwahati, India. His research interests are in offshore foundations, soil-structure interaction, soil stabilization & ground modification. He has published several research papers in many international and national journals.

The design of buried anchors requires the assessment of the peak uplift capacity. Various studies on plate anchors embedded in sand have been reviewed. Finite element analysis results for plate anchors have been presented in the form of ultimate uplift capacity. Comparison has been made with all the design approaches in terms of uplift capacity by using common soil data. Consideration has also been given to increasing modulus of elasticity with depth. The influence of embedment ratio on ultimate uplift capacity has been brought out. The variation in ultimate uplift capacities is in close range for shallow embedment depth. With

Birjukumar Mistri is a post-graduate in Geotechnical Engineering from Indian Institute of Technology, Guwahati, India India. His areas of research include foundation design of offshore and onshore structures, and foundations systems for wind turbines.

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

302

International Review of Civil Engineering (I.RE.C.E.), Vol. 2, N. 6 November 2011

Environmental Noise, Hold Body Vibration & Air Pollution Monitoring within Toll Booths of the Athens Ring Motorway Network: “ATTIKI ODOS” K. Vogiatzis, N. Eliou Abstract – Attikes Diadromes S.A (the company ensuring the “Maintenance – Operation and Exploitation of the motorway Elefsina – Stavros – Spata Airport & West Peripheral Ymittos Avenue”), in cooperation with the University of Thessaly - Faculty of Civil Engineering (Laboratory of Environmental Transportation Acoustics-LTEA) has executed during 2010 (an extended environmental monitoring & research program [1] in, selected toll booths of “Attiki Odos” network (Athens Ring Road) & also to selected maintenance machinery equipment, aiming at the evaluation of environmental noise, body vibration & air pollution generated from the road operation aiming at the evaluation of the relevant impact at the employees exposed during working periods. This article presents and evaluates the relevant full typical week (24/7) monitoring measurements in the selected booths in comparison with the existing criteria & limits. Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved. Keywords: Air Pollution, Attiki Odos Motorway, Body Vibration, Environmental Noise, Toll Booths

Being a closed motorway, it has controlled access points and consists of two sections, which are perpendicular to one another (Figure 1): • The Elefsina - Stavros - Spata A/P motorway (ESSM), extending along approximately 52 km, and • The Imittos Western Peripheral Motorway (IWPM), extending along approximately 13 km.

I. The Athens Ring Motorway Monitoring Program of Environmental Parameters in the Tool Booths Attica Tollway (Athens Ring Road or "Attiki Odos") is a pioneering road project constructed on a concession basis and constitutes one of the biggest co-financed road projects in Europe's urban mega region. It belongs to the first generation of co-financed projects awarded in Greece during the '90s and essentially paved the way and laid the foundations for the execution of future successful concession contracts, in Greece and in other European countries. This modern motorway is extending along 65 km, and constitutes the ring road of the greater metropolitan area of Athens and the backbone of the road network of the whole Attica Prefecture. It is an urban motorway, with two separate directional carriageways, each consisting of 3 lanes and an emergency lane (hard shoulder). The suburban railway of Athens has been constructed in the central reservation of the motorway. Attica Tollway constitutes a unique piece of infrastructure, even in European terms, since it is essentially a closed toll motorway, within a metropolitan capital, where the problem of traffic congestion is acute. Attica Tollway is part of the PATHE road axis (Patra - Athens Thessaloniki - Evzoni) and connects the Athens - Lamia Highway with the Athens - Κorinthοs Highway (currently under upgrade), by-passing the centre of Athens [1], [2].

Fig. 1. Attiki Odos motorway (Attica Tollway)

The harmonious co-existence of the motorway with the environment has been a difficult task, but successfully achieved by Attica Tollway regarding environmental noise through the implementation of comprehensive management, including monitoring,

Manuscript received and revised October 2011, accepted November 2011

303

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

K. Vogiatzis, N. Eliou

TRAFFIC DATA FOR ALL BOOTH LOCATIONS Average Average Booth location mopeds/24hrs PC/24hrs

assessment & implementation of the appropriate mitigation measures i.e. noise barriers & land use organization. The monitoring program includes a complete series of measurements of the following environmental parameters inside the toll booths in 4 separate toll locations for a typical week per location [3]: • Environmental Noise according to the Greek Presidential Decree 85/18-3-91/FEK38 (as per the relevant 2003/10/EC directive of 6/02/2003), • Atmospheric Pollutants: Carbon monoxide – CO / Nitrogen monoxide – NO & Nitrogen Dioxide – NO2 / Total volatile organic combinations – TVOC /Particulate materials (PM10) as per the relevant greek legislation, • Ηand-arm & Whole body vibration according to the Greek Presidential Decree 176/2006 (as per the relevant 2002/44/EC directive of 25 June 2002) for several motorway maintenance equipment in normal operation conditions. The relevant road traffic data for both ESSM & IWPM) are presented in the table hereafter. For 2011 a further reduction of some 10% was estimated due mainly to the well known financial crisis. TABLE I ATTIKI ODOS OVERALL TRAFFIC DATA Year Total passages (veh.) Heavy vehicles % 2008 300.993 5,08 2009 307.300 4,46 2011(est.) 281.329 4,29

II.

Average HV/24hrs

KATEHAKI TOLL AREA (IWPM)

1149

32426

627

METAMORPHOSSI TOLL AREA (ESSM)

1439

39349

1884

KIFISSIAS TOLL AREA (ESSM)

97

4435

97

PENTELI TOLL AREA (ESSM)

244

6914

193

Fig. 2. Typical booth

The important parameters that affect the formation of all environmental / occupational noise and air pollutants are, obviously, the traffic load, the synthesis of the vehicle’s flow using Attiki Odos and, also the typical remaining (background) air pollution in the immediate region. Moreover, the wind meteorological conditions had an important role, since they influence the diffusion of air pollutants. From the respective analysis of detected recordings it became clear that both, concentrations of all air pollutants and occupational noise levels, inside the working areas (booths) were found to be lower compared to the relevant permissible limits for the workers’ safety and health. The following parameters were monitored in all locations working environment using the respective instrumentation: (a) for the atmospheric pollutants: The use of special mobile monitoring stations was secured ensuring continuous measurement 24/7 of the following air pollutants : • Carbon monoxide – CO; • Nitrogen Dioxide- NO2; • Total volatile organic compound – TVOC; • Dust (PM10) . The following instrumentation was used in a mobile station (Fig. 3): • CO Analyser 48C (Thermo Environmental Instruments - USA); • NOx Analyser 42C (Thermo Environmental Instruments - USA);

Monitoring Setup

The monitoring program has been executed as follows: • Both Environmental Noise & Atmospheric Pollutants monitoring measurements were executed in 4 typical weeks (24/7) at the most important four selected toll areas as follows : 1. KATEHAKI TOLL AREA (IWPM), from 04/10/2010 to 11/10/2010, 2. METAMORPHOSSI TOLL AREA (ESSM) from 11/10/2010 to 18/10/2010, 3. KIFISSIAS TOLL AREA (ESSM) from 18/10/2010 to 25/10/2010, 4. PENTELI TOLL AREA (ESSM) from 25/10/2010 to 11/10/2010. The relevant traffic data (private cars & Heavy vehicles>3,5t diesel) for the above locations & designated time periods (24/7 per booth location) are presented in the table hereafter. In Figure 2 a typical booth formation is given. From all the above locations, Metamorphosis tolls were comparatively characterized by a considerable vast traffic flow – both private and heavy vehicles - passages on a daily basis. The relative measurement’s program was carried out in each location for a whole typical week during typical after summer traffic & climatic conditions. TABLE II

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

304

K. Vogiatzis, N. Eliou

• T-VOC Analyser ppbRAE (RAE Systems

(c) For the Ηand-arm & Whole body vibration measurements were conducted according to the EU directive 44/EC for selected machinery & construction vehicles as per figure hereafter.

USA); • Dust Analyser 224-PCEX (SKC – USA); • Personal Data-logging Real-time Aerosol Monitor PDR-1200 (Thermo ELECTRON –USA); • TSP analyser PARTISOL 2000 (Thermo ELECTRON –USA).

Fig. 3. Mobile Air pollution mobile monitoring station

(b) for the environmental noise in the working environment inside the booths the physical parameters used as risk predictors are defined as follows: • peak sound pressure (ppeak): maximum value of the ‘C’- frequency weighted instantaneous noise pressure; • daily noise exposure level (LEX,8h) (dB(A) re. 20 µPa): timeweighted average of the noise exposure levels for a nominal eight-hour working day as defined by international standard ISO 1999: 1990. It covers the environmental noise in the working environment, including impulsive noise; • weekly noise exposure level (LEX,8h): timeweighted average of the daily noise exposure levels for a nominal week of eight-hour working days as defined by international standard ISO 1999:1990. Measurements were executed inside the working environment (sensors placed adjacent to the booth employee), and both Leq(100ms) index and the daily noise exposure level (LEX,8h) in dB(A) ref.20µPa: timeweighted average of the noise exposure levels were calculated. The integrated sonometer/noise statistical analyzer SOLO Master, (as per 651 and 804 standards I.E.C. PUBLICATIONS 651-1979 and 804-1985) and IEC 1260 and IEC 61672-1. (Fig. 3), was used.

Fig. 5. Equipment tested for environmental vibration

The relevant instrumentation used was «Maestro» (01 dB France) with 4 available channels for both noise & vibration monitoring. In the following figures both “hand arm” & “whole body” measurements setup are presented. It is important to notify that according to both Greek & EU legislation: • “hand-arm vibration”: is the mechanical vibration that, when transmitted to the human hand-arm system, entails risks to the health and safety of workers, in particular vascular, bone or joint, neurological or muscular disorders; • “whole-body vibration”: is the mechanical vibration that, when transmitted to the whole body, entails risks to the health and safety of workers, in particular lower-back morbidity and trauma.

Vibration sensor

Vibration Data

Fig. 4. Occupational noise monitoring inside the working environment (booth)

Fig. 6. “Hand arm & Whole body vibration setup”

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

305

K. Vogiatzis, N. Eliou

III.

Criteria – Indices

For the atmospheric pollution the following criteria for the working environment, according to Greek [4] were assessed (see table hereafter). TABLE III ATMOSPHERIC POLLUTION CRITERIA & LIMITS Air pollutant

Limit value for 8 hrs

CO ΝΟ2

50 ppm 5 ppm

Max value for 15 min exposure 300 ppm 5 ppm

TVOC

----

----

PM10

10 mg/m³

----

Add. limit ------< 200 µg/m³ (comfort value for 8hrs) ----

Fig. 7. Katehaki Toll area: Average concentration for CO (ppm) & NO2 (ppb) air pollutants. Time history in 15 min intervals 1,800.0 1,600.0 1,400.0 1,200.0 1,000.0 800.0 600.0 400.0 200.0 0.0

ƒ

ƒ

for hand-arm vibration: the daily exposure limit value standardised to an eight-hour reference period shall be 5 m/s2; and the daily exposure action value standardized to an 8 hour reference period shall be 2,5m/s2, for whole-body vibration: the daily exposure limit value standardised to an eight-hour reference period shall be 1,15m/s2 or, at the choice of the EU Member State concerned, a vibration dose value of 21 m/s1,75; (b) the daily exposure action value standardized to an 8thour reference period shall be 0,5 m/s2 or, at the choice of the Member State concerned, a vibration dose value of 9,1 m/s1,75.

IV.

14:30 15:15 16:00 16:45 17:30 18:15 19:00 19:45 20:30 21:15 22:00 22:45 23:30 0:15 1:00 1:45 2:30 3:15 4:00 4:45 5:30 6:15 7:00 7:45 8:30 9:15 10:00 10:45 11:30 12:15 13:00 13:45

• As for the environmental noise, the max value of the noise index LEX,8h ≤ 87(dB(A) was assessed [3]. • As per the relevant exposure limit values and action values according to European directive 44/EC/2005 we underline the following [5]:

4 ‐ 5 Οκτωβρίου

5 ‐ 6 Οκτωβρίου

6 ‐ 7 Οκτωβρίου

8 ‐ 9 Οκτωβρίου

9 ‐ 10 Οκτωβρίου

10 ‐ 11 Οκτωβρίου

7 ‐ 8 Οκτωβρίου

Fig. 8. Katehaki Toll area: TVOC µg/m3 -8hours averages

2. METAMORPHOSSI TOLL AREA (ESSM)

Day PM10 µg/m³

TABLE V METAMORPHOSSI: PM10 MAX DAILY VALUES µg/m3 12/10 13/10 14/10 15/10 16/10 17/10 18/10 146

187

167

271

146

31

31

Max. 271

Results

The relevant results of the monitoring program per booth location and environmental parameter are as follows:

Fig. 9. Metamorphossi Toll area : Average concentration for CO (ppm) & NO2 (ppb) air pollutants. Time history in 15 min intervals

• For the atmospheric pollution the results are presented in the graphs (x axis representing the time history) and tables hereafter per booth location:

1,800.0 1,600.0 1,400.0 1,200.0 1,000.0 800.0 600.0 400.0 200.0 0.0

1. KATEHAKI TOLL AREA (IWPM)

14:30 15:15 16:00 16:45 17:30 18:15 19:00 19:45 20:30 21:15 22:00 22:45 23:30 0:15 1:00 1:45 2:30 3:15 4:00 4:45 5:30 6:15 7:00 7:45 8:30 9:15 10:00 10:45 11:30 12:15 13:00 13:45

TABLE IV KATEHAKI TOLL AREA: PM10 MAX DAILY VALUES mg/m3 5/10 6/10 7/10 8/10 9/10 10/10 11/10 Max. Day PM10 167 115 83 73 177 104 52 177 µg/m³

11‐12 Οκτωβρίου

12‐13 Οκτωβρίου

13‐14 Οκτωβρίου

15‐16  Οκτωβρίου

16‐17 Οκτωβρίου

17‐18  Οκτωβρίου

14‐15 Οκτωβρίου

Fig. 10. Metamorphossi Toll area: TVOC µg/m3 -8hours averages

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

306

K. Vogiatzis, N. Eliou

25‐26 Οκτωβρίου

26‐27 Οκτωβρίου

27‐28 Οκτωβρίου

28‐29 Οκτωβρίου

29‐30 Οκτωβρίου

30‐31 Οκτωβρίου

13:30

12:30

11:30

9:30

10:30

8:30

135

7:30

73

6:30

73

5:30

73

4:30

52

3:30

83

2:30

31

1:30

135

0:30

Max.

23:30

25/10

22:30

24/10

21:30

23/10

20:30

22/10

19:30

21/10

18:30

20/10

17:30

19/10

14:30

Day PM10 µg/m³

16:30

TABLE VI KIFISSIAS: PM10 MAX DAILY VALUES µg/m3

1,800.0 1,600.0 1,400.0 1,200.0 1,000.0 800.0 600.0 400.0 200.0 0.0 15:30

3. KIFISSIAS TOLL AREA (ESSM)

31 Οκτ. ‐ 1 Νοεμβρίου

Fig. 14. Pentelis Toll area: TVOC µg/m3 -8hours averages

• For the Environmental noise in the LEX,8h index fluctuation in all booths location and relevant working shifts are presented in the following graphs in Figure 16. In Figs. 15, in particular, some representative time history data for the noise descriptor Leq (x axis in dB(A), for a selected morning/typical day shift, in all locations for comparison reasons are also presented. • For the “hand-arm vibration" the representative measurements results are shown in the Figs. 17. • For the "whole-body vibration" the representative measurements results are shown in the Figs. 18.

Fig. 11. Kifissias Toll area: Average concentration for CO (ppm) & NO2 (ppb) air pollutants. Time history in 15 min intervals 1,800.0 1,600.0 1,400.0 1,200.0 1,000.0 800.0 600.0 400.0 200.0 0.0

14:30 15:15 16:00 16:45 17:30 18:15 19:00 19:45 20:30 21:15 22:00 22:45 23:30 0:15 1:00 1:45 2:30 3:15 4:00 4:45 5:30 6:15 7:00 7:45 8:30 9:15 10:00 10:45 11:30 12:15 13:00 13:45

V.

18‐19 Οκτωβρίου

19‐20 Οκτωβρίου

20‐21 Οκτωβρίου

22‐23 Οκτωβρίου

23‐24 Οκτωβρίου

24‐25 Οκτωβρίου

21‐22 Οκτωβρίου

Fig. 12. Kifissias Toll area: TVOC µg/m3 -8hours averages

4. PENTELI TOLL AREA (ESSM)

Day PM10 µg/m³

TABLE VII PENTELIS: PM10 MAX DAILY VALUES µg/m3 26/10 27/10 28/10 29/10 30/10 31/10 1/11 187

198

21

31

94

42

21

Max. 198

Fig. 13. Pentelis Toll area: Average concentration for CO (ppm) & NO2 (ppb) air pollutants. Time history in 15 min intervals

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

Discussion

Based on the relevant results and in accordance with last 5 year previous monitoring programs in the same selected toll locations [6] we conclude the following: • The executed monitoring program succesfully covers all aspects of environmental parameters inside the booths of "Attica Tollway" network and is projected to be continued also for 2011 up 2014 in order to produce comparative data, especially under the estimated assumption if further reduction of road traffic flows in the motorway. • Regarding the Environmentla Noise inside all respected both areas the relevant results for all measurements suggests that the exposure to noise index LEX,8h is less than the given criterion of 87 dB(A), and it fluctuates in quite acceptable levels for an adequate working environment. • In the respected major both areas all atmospheric pollutants measurements (i.e: Carbon monoxide – CO / Nitrogen Dioxide – NO2 / Total volatile organic combinations – TVOC /Dust - PM10) were also found within the respected criteria in all cases, especially in rather busy days prior - for the ruch morning hours especially for a time period when flows were not severely affected by the financial crisis. • Regarding Ηand-arm & Whole body vibration measurements it was found that the relevant values for the most important and signifiquant machinery were within the given criteria. However its usefull to underline that drivers/operators behaviour found to influence all relevant measurements in a signifiquant level, bur alway not esceeding the given criteria.

International Review of Civil Engineering, Vol. 2, N. 6

307

K. Vogiatzis, N. Eliou

105

100

95

90

85

80

75

70

65

60

55

50

45 40 07h

08h

09h

10h

11h

12h

13h

14h

Fig. 15(a). Katehaki Toll area: 06:30-14:30 shift for a typical Tuesday 105

100

95

90

85

80

75

70

65

60

55

50

45 40 07h

08h

09h

10h

11h

12h

13h

14h

Fig. 15(b). Kifissias Toll area : 06:30-14:30 shift for a typical Tuesday q 90

,

,

85

80

75

70

65

60

55

50 07h

08h

09h

10h

11h

12h

13h

14h

Fig. 15(c). Pentelis Toll area: 06:30-14:30 shift for a typical Tuesday 90

85

80

75

70

65

60

55

50 07h

08h

09h

10h

11h

12h

13h

14h

Fig. 15(d). Metamorphossi Toll area: 06:30-14:30 shift for a typical Tuesday Figs. 15. Representative measurements results for for the noise descriptor Leq (x axis in dB(A), for a selected morning/typical day shift, in all toll areas

Figs. 16. Environmental noise index inside booth fluctuation LEX,8h

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

308

K. Vogiatzis, N. Eliou

X Session #8 aw 1s Wd

THU 10/12/09 11h01m35 9.001e-02m/s²

THU 10/12/09 11h32m40 3.400e-01m/s²

Y Session #8 aw 1s Wd

THU 10/12/09 11h01m35 1.100e-01m/s²

THU 10/12/09 11h32m40 1.400e-01m/s²

Z Session #8 aw 1s Wk

THU 10/12/09 11h01m35 5.100e-01m/s²

THU 10/12/09 11h32m40 2.100e-01m/s²

1.6

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0 11h05

11h10

11h15

11h20

11h25

11h30

Figs. 17. Representative measurements results for “hand-arm” vibration in the representative machinery equipments

Fig. 18. Representative measurements results for "whole-body” vibration in a representative machinery equipments

Monitoring programs of all signifiquant environmental parameters in the working environment consists a major factor of ensuring acceptable levels of comfort during work and contributes in the project's sustainability, therefore their implementation should be a major concern for motorway & highway operators especially in urban mega regions in europe.

risks) in respected both areas arising from physical agents (noise) (Seventeenth individual Directive within the meaning of Article 16(1) of Directive 89/391/EEC). [6] EC directive 44/EC of 25 June 2002 on the minimum health and safety requirements regarding the exposure of workers to the risks arising from physical agents (vibration) within the meaning of Article 16(1) of Directive 89/391/EEC). [7] Vogiatzis K., Eliou N., Occupational Noise Vibration & Air Pollutants Fluctuation inside the Toll Booths in the Athens Ring Motorway Network: “Attiki Odos”, 14th International Congress on Sound and Vibration ICSV 14, 9-12 July 2007, Cairns Australia

References [1] http://www.aodos.gr [2] Vogiatzis K. "Strategic Environmental Noise Mapping & Action Plans In Athens Ring Road (Atiiki Odos) - Greece" Wseas Transactions On Environment and DEVELOPMENT, ISSN: 17905079 Issue 9, Volume 8, October 2011. [3] Vogiatzis K. "Research Program “Attiki Odos Occupational Noise Vibration & Atmospheric pollutants Monitoring Program” (University of Thessaly, Faculty of Civil Engineers-2010). [4] Greek legislation PD 90/1999 & PD 307/1986. [5] EC directive 2003/10 of 6 February 2003 on the minimum health and safety requirements regarding the exposure of workers to the

Authors’ information Prof. Konstantinos Vogiatzis, University of Thessaly- Faculty of Civil Engineers, Laboratory of Environmental Transportation Acoustics-LTEA Volos - Greece, Born: 6th February 1956 Degrees: 1. Rural and Surveying Engineering, National Technical University of Athens (U.N.T.A.) July 1977.

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

309

K. Vogiatzis, N. Eliou

2. 3. 4.

Civil Engineering (Transportation planning), National Technical University of Athens (U.N.T.A.) December, 1980.. Urban and Rural Planning : Diplômé de Formation Interdisciplinaire en Aménagement et en Urbanisme (D.F.I.A.U.) Centre of Research in Urban Planning of Paris, November 1980. PhD. in REGIONAL ANALYSIS AND RURAL PLANNING (Doctorat en Analyse Regionale et Amenagement du Territoire) OPTION: REGIONAL PLANNING. University of Paris 1 PANTHEON-SORBONNE, January 1981. Nikolaos Eliou - Associate ProfessorUniversity of ThessalyFaculty of Civil Engineering, Volos - Greece, Email:[email protected] Born: 26th May 1961Courses Taught: Applications of Transportation Systems Simulation: Infrastructure Education: 1. Civil Engineering. Aristotle University of

Thessaloniki 2. Ph.D., Civil Engineering, Aristotle University of Thessaloniki Professional Organizations Membership: 1. Technical Chamber of Greece 2. Union of Civil Engineers of Greece 3. Union of Transportation Engineers of Greece 4. World Road Association (PIARC) Technical Committees Membership: 1. Committee of Infrastruction Projects of the Technical Chamber of Greece 2. Committee of the Technical Chamber of Greece for the Local Administration 3. Committee of Enquiry of the Technical Chamber of Greece 4. Committee of Transportation Affairs of the Technical Chamber of Greece 5. Senate of University of Thessaly 6. Technical Board of University of Thessaly (Chairman) 7. Technical Board of Region of Thessaly-GREECE (Prof. expert in transportation projects)

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

310

International Review of Civil Engineering (I.RE.C.E.), Vol. 2, N. 6 November 2011

Tired Driver’s Behaviour Assessment Using Innovative Instrumentation N. Eliou, K. Vogiatzis, F. Kehagia

Abstract – Driver’s fatigue has received considerable attention as a road safety issue, worldwide, for all kind of vehicles (light, heavy etc)and all kind of drivers(ordinary, professional etc). The deteriorating driver performance associated with driver fatigue presents a serious safety risk. In the present paper, the relationship between driving performance and fatigue is examined using an innovative, specially instrumented, vehicle. According to a pilot study that was conducted, the investigation whether the driving behaviour, as it can be interpreted through speeding and accelerating profiles analysis, is directly related to drive’s fatigue. A deep analysis of the variation of the trajectories followed in different driver’s fatigue levels combined with the relevant speeding and accelerating profiles is very revealing, in order to assess the severe impact of the tired driver’s behaviour and attitude. Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved.

Keywords: Driver’s Behaviour, Fatigue, Speed, Safety, Accidents

I.

Introduction

I.1.

Driver’s Fatigue

asleep is a widespread phenomenon. Among the 2783 private drivers and 2854 professional drivers who are the sample base, the 44,8% of the private drivers respondents and the 36% of the professional drivers respondents have experienced falling asleep while driving one time or more, [12]. Fatigue or sleepiness or drowsiness are concepts with the same meaning and they are often used interchangeable. Fatigue is associated with physiological components or physical aspects (changes in brain wave activity, eye movement, muscle tone, heart rate) and psychological components (changes in mood, motivation or mental aspect). The underlying causes of fatigue and the consideration of the level of fatigue are difficult to be examined, since there is no reliable device for detecting the level of fatigue. Generally, the most general factors that cause fatigue are lack of sleep or bad sleep quality, prolonged and monotonous driving (time-on-task) and individual characteristics including medical conditions. In the literature, it is mentioned that a variety of psychophysiological parameters have been used as indicators of fatigue, with electroencephalography perhaps being the most promising. Most researches found changes in theta and delta activity to be strongly linked to transition to fatigue. The review also identified anxiety and mood states as factors that may possibly affect driver fatigue. Furthermore, personality and temperament may also influence fatigue, [10].

Road Safety is a priority for the European Transport Policy [1]-[15]. An implementation of a great variety of policies and safety measures is necessary in order to satisfy the objective of halving the number of fatalities. It is well known that the humans in general are the ones who mostly contribute to the occurrence of accidents. Speeding, alcohol, inattention and driver fatigue are the most important causative factors in road accidents. Driver fatigue has received considerable attention as road safety issue, worldwide, for light and heavy vehicle drivers. Fatigue has been identified as contributing factor in up to 20% of vehicle accidents. In Australia, statistics from 1998 indicate that there were 16.6% of total roads fatalities due to fatigue related accidents. From 1990 to 1998, the proportion of fatal accidents involving driver fatigue increased from 14.9% to 18.0%, [2]. A pilot study in New Zealand estimated that up to 18% of truck crashes may involve driver fatigue, [5]. Moreover, it is estimated that injury crash rates could be reduced by as much as 19%, by reducing the number of drivers who drive sleepy, or with 5h or less of sleep, or who drive between 2 and 5 am, [1]. The US National Transportation Safety Board has identified driver fatigue as being implicated in 30-40% of all truck accidents and as the most common cause (31%) of fatal to the driver truck accidents, [11]. According to the results of a study conducted by Institute of Transport Economics of Norway, falling

Manuscript received and revised October 2011, accepted November 2011

311

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

N. Eliou, K. Vogiatzis, F. Kehagia

I.2.

Driving and Fatigue

accident causation. For this reason, when the police officers evaluate and record an accident according a specific checklist of accident causes, the choice of fatigue as an accident cause is depended solely on police officers' subjective opinions. Conversely, in the United Kingdom, fatigue related crashes have been identified using some criteria as for example, the absence of skid marks or braking or the possibility that the driver could see the point of run-off or the object hit prior to the crash, [6]. Moreover, Finnish police and the courts punish a significant number of drivers every year on the basis of fatigue. Some drivers are even punished for "just" the drifting on a road, [13].

Driving is simultaneously an intellectual, cognitive and emotional experience. In general, increased fatigue causes a decrease in mental and physical functioning. As for driving task for the driver, fatigue deteriorates the level of driving performance by impairing alertness, reducing vigilance and reaction times. Moreover, the motivation to carry out a task diminishes, the communication and interaction with the surroundings deteriorates and the amount of efficient processed information is quite reduced. Williamson et al [14] found, in their research, that a performance of a driver after 17–19 hours without sleep, on some tests was equivalent or worse than those with 0.05%blood alcohol concentration. Response speeds were up to 50% slower for some tests and accuracy measures were significantly poorer than at this level of alcohol. After longer periods without sleep, performance reached levels equivalent to the maximum alcohol dose given to subjects (BAC of 0.1%). Horne [7] has noted that drivers typically underestimate the impact of driver fatigue, ignore feelings of drowsiness and continue driving when they become sleepy. These findings reinforce the evidence that the fatigue of sleep deprivation is an important factor likely to compromise performance of speed and accuracy that is needed for safety on the road.

II.

III. Research Set Up III.1. Design Based on Lewin's (1951) equation, B = f (P,E) a person behavior (B) is a function of the person (P) and the environment (E) and the interaction between them. Driving behavior is an effect of the drivers' basic emotional process. Moreover, is related to the physical environment, to the other users, to the driving task, to individual factors and own abilities and to the interaction between them, [3]:

(

B= f E

Road Safety Policy in Greece

road

,E

users

,A

task

,P

person

)

where: B = the adapted driving behavior as a consequence of the drivers' basic emotional process; E road = the physical environment of the road-traffic; E users = the social environment including other users; A task = the driving task; P person = the individual factors and own abilities. The basic emotional process that affects the driving behaviour is expressed in terms of activation, the drivers' preparedness to establish safe behaviour, orientation, the ability to distinguish the traffic situation, evaluation, the relevance of perceived information and control, the chosen speed behaviour. Being a first step, this pilot study investigated whether the driving behaviour, specifically, the factor of control is related to driver fatigue as a specific driver individual factor. The factor of control is expressed by the variation of speed and acceleration. A 2-lane rural road, with length 3.5 km, between two villages in the area of Prefecture of Thessaloniki was used for the study (Figure 1). The landscape of the road is mountainous, with tight radius curves in horizontal alignment and steep gradient in vertical alignment. The traffic volume of the road is quite small. The choice of this examined route gives the opportunity to examine the influence of factor of road geometry, which affects the value of speed, and diminish the effect of the other factors that affect driver's behaviour.

Greece keeps one of the last places among the Europe’s 27 members, according to road safety statistics (the number of fatalities was 1550 during 2008 compared to 1463 fatalities during 2009). During the last years many efforts have been made attempting to reduce road traffic accidents in Greece. The implementation of 1st National Strategic Road Safety Plan 2001-2005 for the improvement of road safety, had the main objective to reduce the number of injured persons in road accidents by 20% up to the year 2005 compared to the base year 2000. The main target of the 2nd National Strategic Road Safety Plan 2006-2010 is an another attempt that can lead to achieving the European road safety target of decreasing the number of fatalities in road accidents in 2010 by 50%, compared to 2000. Despite of the implementation of the two plans, registered number of fatalities was 1593 in 2008, showing deviation of 50% of the European target for 2010. The two Strategic Plans have not been implemented efficiently due to lack of reliable data on road accidents and scientific approach to road safety issues and lack of accountability by the ambiguity of duties and responsibilities, [9]. There are no official statistics in Greece regarding the fatigue-related accidents. The contribution of sleepiness and fatigue in road crashes is still underestimated in police reports as there is a lack of identification or documentation. There are no specific criteria for the recognizing of fatigue as major or contributing factor in Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

312

N. Eliou, K. Vogiatzis, F. Kehagia

drive in the specific road segment, both directions (straight and return) twice. The two routes, according to the experiment, were firstly in the morning when the drivers feel refreshed after a period of adequate hours of sleep and secondly when the drivers feel fatigued and tired after a period of 8-hour work. The measuring data represented speed, traveled distance and acceleration were stored on ASCII files and processed in excel programs.

IV.

Results

IV.1. Speed Distribution The results from the study provide significant differences of the speed behavior among tired and rest drivers (Figure 3). The difference value of speed was about 3-6 km/h. The factor of feeling of fatigue, caused after a period of 8-hour work, seems to influence driver's speed behavior since their operational speed is smaller. In Figure 4, the variation of operational speed from all drivers depending on the factor of experience and the feeling of fatigue is presented. The results provide interesting information about the relationship of the factor of experience and the driver fatigue. An experience driver seems to manage better difficult situations.

Fig. 1. Satellite image of the selected route

III.2. Instrumentation A device called Video VBOX was used to assess the relationship between driving behaviour and fatigue. Video VBOX combines a powerful GPS data logger with a high quality solid-state video recorder, which takes two cameras and combines them with a graphical overlay, with the resulting video streamed onto an SD card as a DVD quality MPEG4 file. (Figure 2). VBOX records the following parameters as standard along with the video file: satellites, time, latitude, longitude, velocity, heading, height, vertical velocity, and avi_sync_time [15]. The VBOX Tools Software provided with the unit allows you to view the recorded video and analyse all of the parameters which have been logged. Video VBOX is compatible with the display which can be used to display Speed, Max Speed, Lap-time, and Lateral and Longitudinal acceleration data from the Video.

All the drivers 90

80

Speed (km/h)

70

tired rest

60

50

40

30 0

500

1000

1500

2000

2500

3000

3500

Distance (m)

Fig. 3. The variation of operational speed between tired and rest drivers Fig. 2. Video VBOX All the drivers 90

III.3. Participants

80 70

S peed (km/h)

Ten volunteer participants with driving licenses were recruited for participation (5 male, 5 female). Half of the participants had high driving experience and the other half of the participants had low driving experience (less than 3 years from the acquisition of driving license).

60 no-experience and tired 50

experince and tired

40

no-experience and rest experience and rest

30 20 10 0 0

500

1000

1500

2000

2500

3000

3500

Distance (m)

III.4. Procedure Fig. 4. The variation of operational speed from all drivers depending on the factor of experience and the feeling of fatigue

The imposed requirement of each participant was to

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

313

N. Eliou, K. Vogiatzis, F. Kehagia

IV.2. Acceleration Distribution The variation of acceleration values determines how smooth the variation of driver speed was. The finding of results indicates that there are no significant differences between rest and tired drivers. All of them had the same upper and lower limits (Figure 5). Conversely, the factor of experience seems that affects the range of the values of acceleration. An experienced driver drives with greater changes in acceleration values (Figure 6). All the drivers 3

2

Acceleration (m/s2)

1

0 0

500

1000

1500

2000

2500

3000

3500

tired rest

-1

-2

-3

Fig. 7. Typical Trajectory (driving path)

-4 Distance (m)

The number of cross sections points of two consecutive driving paths either return or transition are examined. The number of these points are more when the drivers were tired than rest because as driving progressed, steering performance became less flexible and the amplitude of steering corrections increased. The number of the intersection points between the trajectories (transition and return) of some drivers is presented (Figure 8). Although a GPS based plot of the trajectory includes an error ±0,6m regarding the absolute position of the vehicle, the relative position that is used for Speeding, Accelerating and Trajectory plot, is very accurate (±0,02m) and reliable.

Fig. 5. The variation of acceleration between tired and rest drivers All the drivers 3

2

Acceleration

1

0 0

500

1000

1500

2000

2500

3000

3500

-1

no-experience and tired experience and tired no-experience and rest experience and rest

-2

-3

-4 Distance (m)

Fig. 6. The variation of acceleration from all drivers depending on the factor of experience and the feeling of fatigue

IV.3.

Steering Performance

According to studies, steering performance gradually deteriorates as the feeling of driving fatigue increases. Van der Hulst et al. found that after a 2.5 hour drive in a simulator, steering performance deteriorated. In the study presented in this paper, the form of driving paths was recorded by help of the Video VBox. This track map is a plot of the latitude and longitude data of a circuit combined with a vehicle position indicator (Figure 7). It is found that there is difference between two consecutive paths, transition and return in all drivers. The difference is more significant when the drivers were tired.

Fig. 8. The number of intersection points between transition and return paths

V.

Discussions

The findings of this study provide interesting

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

314

N. Eliou, K. Vogiatzis, F. Kehagia

information about fatigue in terms of variation of speed driving. Still, some other variables might also influence driver's behavior (mood, mentality, stress, travel purpose etc). It is difficult to isolate the effect of the factor of fatigue from other factors. For example, it is hard to know how determinant parameter is the driver experience. Whatever the multitude of the immeasurable competitive variables that might negatively influence driver's behavior, one reflection is that fatigue made drivers less alert, traveling with lower speed.

Authors’ information Nikolaos Eliou - Associate Professor University of Thessaly - Faculty of Civil Engineering, Volos – Greece. E-mail: [email protected] Born: 26th May 1961 Courses Taught: Applications of Transportation Systems Simulation: Infrastructure Education: 1. Civil Engineering. Aristotle University of Thessaloniki 2. Ph.D., Civil Engineering, Aristotle University of Thessaloniki Professional Organizations Membership: 1. Technical Chamber of Greece 2. Union of Civil Engineers of Greece 3. Union of Transportation Engineers of Greece 4. World Road Association (PIARC) Technical Committees Membership: 1. Committee of Infrastruction Projects of the Technical Chamber of Greece 2. Committee of the Technical Chamber of Greece for the Local Administration 3. Committee of Enquiry of the Technical Chamber of Greece 4. Committee of Transportation Affairs of the Technical Chamber of Greece 5. Senate of University of Thessaly 6. Technical Board of University of Thessaly (Chairman) 7. Technical Board of Region of Thessaly-GREECE (Prof. expert in transportation projects)

References [1]

[2]

[3] [4] [5]

[6]

[7] [8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

Connor J., Norton R., Ameratunga S., Robinson E., Civil I., Dunn R., Bailey J., Jackson R. (2002), “Driver sleepiness and risk of serious injury to car occupants: Population based case control study”, British Medical Journal, 324, pp.1125-1130. Dobbie K., (2002). “Fatigue related crashes: an analysis of fatigue related crashes on Australian roads using an operational definition of fatigue”, Australian Transport Safety Bureau. Drottenborg H., (2002). “Are Beautiful traffic environments safer than ugly traffic environments?” Lund Institute of Technology. European Road Safety Observatory Roads, www.erso.eu. Gander P. H., LeQuesne L., Marsall MS (2003). “How important is driver fatigue in truck crashes?” Wellington: Sleep/Wake Research Centre. Horne J., Reyner L., (2001). “Vehicle accidents related to sleep; a review”, Occupational and Environmental Medicine, 56(5), pp.289-294. Horne J., Baulk S. D. (2004). “Awareness of sleepiness when driving”. Psychophysiology 41, pp.161-5. Hulst van der Meijman, Rothengatter, (2001). “Maintaining task set under fatigue: a study of time-on-task effects in simulated driving”, Transportation Research Part F, 4(2), pp.103-118. Kanellaidis G. et al, (2005). “Development of a Strategic Plan for the improvement of Road Safety in Greece, (2006-2010)”, in: Proceeding of 3rd Hellenic Conference on Road Safety, Patra, Greece. Lal S. K. L., Craig A., (2001). “A critical review of the psychophysiology of driver fatigue”, Biological Psychology, 55, pp. 428-436. National Transportation Safety Board, (1995). “Factors that affect fatigue in heavy truck accidents”, Safety Study 95/01, Washington, DC: National Transportation Board. Nordbakke S., (2004). “Driver fatigue and falling asleepexperience, knowledge and conduct among private drivers and professional drivers”, TOI report 706. Radun I., Radun J., (2009). “Convicted of fatigued driving: Who, why and how?”, Accident Analysis and Prevention 41, pp. 869875. Williamson A. M., Feyer A., (2000). “Moderate sleep deprivation produces impairments in cognitive and motor performance equivalent to legally prescribed levels of alcohol intoxication”. Occupational and Environmental Medicine, 57, pp. 649-655. Nikolaos Katsianis, Nikolaos Eliou, Evdokia Iliadou, Konstantinos E. Vogiatzis, (2011). "Effect of Driver’s Roadway Familiarity on Risk Perception & Driving Behaviour", International Review of Civil Engineering (I.RE.C.E.), Vol.2, n.2., pp. 107-112.

Prof. Konstantinos Vogiatzis - University of Thessaly - Faculty of Civil Engineers Volos – Greece. E-mail: [email protected] Born: 6th February 1956 Degrees: 1. Rural and Surveying Engineering, National Technical University of Athens (U.N.T.A.) July 1977. 2. Civil Engineering (Transportation planning), National Technical University of Athens (U.N.T.A.) December, 1980. 3. Urban and Rural Planning : Diplome de Formation Interdisciplinaire en Amenagement et en Urbanisme (D.F.I.A.U.) Centre of Research in Urban Planning of Paris, November 1980. 4. PhD. in Regional Analysis And Rural Planning (Doctorat en Analyse Regionale et Amenagement du Territoire) Option: Regional Planning. University of Paris 1, Pantheon-Sorbonne, January 1981. Fotini Kehagia - Lecturer of Highway Engineering - University of Thessaloniki, Greece. Born: 5th March 1966 Degrees: 1. Aristotle University of Thessaloniki, Dipl. of Civil Engineering (1988). 2. Aristotle University of Thessaloniki, Dipl. of Civil Engineering Master Diploma: “Environmental Protection and Sustainable. 3. Aristotle University of Thessaloniki, Dipl. of Civil Engineering Diploma of Doctor in Engineering, PhD, Doctoral Research Work : “Road design through environmental criteria”. Applied Research: • 2002-2011: Research work in 15 projects in "Sustainable Design and Alternative Materials in Highway Engineering", "Road Safety Audits", Pavement Monitoring and Condition Assessment in Motorway " and "Road Asset Management". • 2007-2009:National Delegate in COST research Action 356.

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

315

International Review of Civil Engineering (I.RE.C.E.), Vol. 2, N. 6 November 2011

Finite Element Analysis for Bearing Capacity of Circular Footing on Geogrid Reinforced Sand Jawdat K. Abbas, Mohanad N. Al-Shindah Abstract – Finite element method is used to determine the ultimate bearing capacity of Circular Footing resting on geogrid reinforced sand. The effect of each of the depth ratio of the topmost layer of geogrid (u/D), the vertical distance ratio between consecutive layers (h/D), number of geogrid layers (N), and the effective depth ratio of reinforcement (d/ D) on the bearing capacity were studied, where (D) is the footing diameter. In addition, for all these parameters, depth of embedment ratio of footing (D / D f ) and the angle of internal friction ( φ )on the ultimate bearing capacity were studied as well. In general, the results showed that by increasing the number of reinforcement layers (N), the bearing capacity increased. The optimum value of reinforcement layers was (N=3-5). The optimum depth ratio of the topmost layer geogrid was between (0.25D0.3D-0.35D) depending on the angle of internal friction (30˚,35˚,40˚) respectively. The optimum vertical distance ratio between consecutive layers was between (0.325D-0.35D),(0.225D-0.25D) and (0.3D-0.375D). As a result, the optimum effective depth ratio of reinforcement was between (0.95D-1.55D), (0.8D-1.2D) and (0.95D-1.1D) depending on the angle of internal friction (30˚,35˚,40˚) respectively. Also, the depth of embedment ratio of footing was (0D,0.25D,0.5D) respectively. Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved.

Keywords: Bearing Capacity, Finite element, Geogrid, Reinforced Sand, Circular Footing

Nomenclature BCR c d D Df E50 EA EI h

N quR qu Rint u

φ

ψ γ

µ

I.

Bearing capacity ratio Soil cohesion Effective depth of reinforcement Diameter of the footing Depth of footing base below the ground surface Modulus of elasticity Normal stiffness Flexural rigidity Vertical distance between consecutive layers of reinforcement Number of reinforcement layers Ultimate bearing capacity of circular footing on reinforced sand Ultimate bearing capacity of circular footing on unreinforced sand Interface reduction factor Depth of first layer of reinforcement below the footing base Angle of internal friction of sand Angle of dilatancy Unit weight of the soil Poisson's ratio

Introduction

Foundation is that part of the substructure which provides interface between the substructure and the supporting earth in general. A substructure may be defined as that part of a structure which helps in transferring the load of the superstructur and its own load on to the supporting soil. A substructure may be partly or wholly embedded inside the earth. It is, therefore, proper to concentrate on the aspects of economy and structural efficiency in the design of such foundations. Circular foundation is that type of shallow foundation, which has been used in many structures such as (tank, chimneys, towers, etc.). This was achieved by removing the existing weak soil up to a certain depth and then replacing the soil or filling the same soil back with the inclusion of horizontal layers of geosynthetics at different depths under the footing. Therefore, with the potential benefit of using soil reinforcement, both the type and the size of foundation may be changed leading to an economic design. Reinforced soil, or mechanically stabilized soil, is a construction technique that consists of soil that has been strengthened by tensile elements such as metal strips, geotextiles, or geogrids. In the 1960's, the French Road Research Laboratory conducted extensive research to evaluate the walls and embankments were constructed all over the world using reinforced soil, and they have performed efficiently [1].

Manuscript received and revised October 2011, accepted November 2011

316

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

J. K. Abbas, M. N. Al-Shindah

ΙΙ.

TABLE IΙ FOOTING AND GEOGRID PARAMETER Parameter Footing ** Geogrid*** 1500 EI ( kN ⋅ m 2 / m )

Finite Element Formulation and Material Modeling

The nonlinear behavior of sand was modeled using Mohr-Coulomb soil model, which is an elastic perfectplastic stress-strain model, the footing was treated as elastic beam elements with significant flexural rigidity (EI) and normal stiffness (EA). Interface elements have been used to model both the interaction between the footing base and the soil and the interaction between the geogrid and soil from both sides. The interface elements allow for the specification of a reduced wall beneficial effects of using reinforced soil as a construction technique. During the last thirty years many retaining The Bearing capacity analysis of soils represents an important step for evaluation of the stability and economy of shallow foundations. Along with the analysis of settlements which is made to ascertain whether the foundations can fulfill their intended function from the structural and utilitarian point of view, the computation of ultimate loads constitutes the primary framework for design. Sometimes it is required to construct footing on weak soil with low bearing capacity. This might be avoided either by constructing the footing with larger dimensions to reduce the contact pressure which leads to uneconomical design or by increasing the bearing capacity of the supporting soil. Several studies have been reported on the successful use of soil reinforcement as a cost-effective method to increase the ultimate bearing capacity at a given settlement under shallow foundations. compared to the friction of the soil. Automatic generation of (15 node) triangle plane strain elements for the soil, (5 node) beam elements for the footing and (5 node) elastic line elements for the geogrid were used. The soil parameters used in the present study are listed in Table Ι the footing and geogrid parameters are listed in Table ΙΙ. TABLE I SOIL PARAMETERS USED IN THE PRESENT STUDY Medium Medium Parameter Loose dense sand Sand 1 1 Cohesion (c) ( kN ⋅ m 2 / m ) Angle of internal friction

ϕ

()

EA ( kN / m )

For more details about the elements types, formulation and the parameters used refer to [2]. For the present study, a general computer program called (PLAXIS 2D Professional Version 8.2) is used. (PLAXIS) is a finite element program for geotechnical applications in which soil models are used to simulate the non-linear behavior of soil. The name (PLAXIS) came from first parts of the two phrases (Plane strain and Axisymmetric).

ΙΙΙ. Parametric Studied The bearing capacity of circular footing on geogrid reinforced sand was studied. The effect of the number of geogrid layers (N), the depth of the topmost layer of geogrid (u), the distance between consecutive layers (h), and the effective depth of reinforcement (d) on the bearing capacity were studied. In addition, for all these parameters, depth of embedment of footing ( D f ) and the angle of internal friction ( φ ) on each of them are studied too. After that optimum values are obtained. Fig. 1 shows the major reinforcement parameters of circular footing on geogrid reinforced sand.

Dense sand 1

30

35

40

0.25

0.3

0.35

25000

35000

50000

0

5

10

Fig. 1. Circular footing on geogrid - reinforced sand

Poisson's ratio (µ)* E50

The study depends on the varied parameters, the depth ratio of footing embedment ( D f /D) was varied (from 0

( kN ⋅ m 2 / m ))* Angle of dilatancy

ψ

2000

* [3] ** [4] *** [5]

D

Modulus of elasticity

200000

( ) **

to 0.5) and the angles of internal friction ( φ ) were chosen (30, 35D and 40D ) to represent loose, medium and dense sand respectively. The depth ratio of topmost layer (u/D) varied (0.1, 0.15, 0.2, 0.25,0.3,0.35,0.4,0.45 and 0.5), the vertical distance ratio between consecutive layers of geogrid (h/D) is varied (0.2, 0.25, 0.3, 0.35and 0.4), and the number of geogrid layers is varied (from 1 to 6).

D

Soil unit weight γ

15.5

16.5

17.5

( kN ⋅ m 2 / m )) Interface reduction factor ( Rint )**

1.0

1.0

1.0

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

317

J. K. Abbas, M. N. Al-Shindah

The term bearing capacity ratio (BCR) is used to express the combined effect of soil reinforcement the bearing capacity and it can be written as: BCR =

quR qu

(1)

ΙV. Optimum Number of Geogrid Layers

Fig. 4. (B C R) - ( D f /D) relationship for different (N)

Figs. 2, 3 and 4 show the relationship between the depth ratio of footing ( D f /D) and the bearing capacity

(D=4, u/D=0.3, h/D=0.35, φ = 40D )

ratio (BCR) for different the number of geogrid layers (N) for varies value of the angles of internal friction ( φ ) (30˚, 35˚, 40˚) respectively. It can be seen that, increasing (BCR) with increasing ( D f /D) because the over bearding pressure for soil increasing, increasing ( D f /D) the optimum number of (N) decreased because the soil has been improve already, for the ( φ =40˚) the optimum value of number of reinforcement layers is (N=3) for all depth ratio of footing ( D f /D). Also Fig. 5,

Fig. 5. (B C R) - (Ф) relationship for different (N) (D=4, u/D=0.25, h/D=0.35, D f /D=0.5)

shows the relationship between the angles of internal friction ( φ ) and the bearing capacity ratio (BCR) for different the number of geogrid layers (N). It can be seen that increasing the angle of internal friction ( φ ) increases the bearing capacity ratio (BCR). But it has no effect on the optimum value of (N). Fig. 6 shows the relationship between the depth ratio of topmost layer (u/D) and the bearing capacity ratio (BCR) for different the number of geogrid layers (N).

Fig. 6. (B C R) - (U/D) relationship for different (N) (D=4, D f /D=0.25, φ = 35D )

It can be seen that there is an optimum value of (u/D) which gives the higher values of (BCR), and as a result, gives the optimum number of geogrid layers (N=4) where the optimum value of (h/D). As increasing the value of (u/D), the value of (BCR) for the optimum number of geogrid is decreased. Fig. 7 shows the relationship between the vertical distance ratio between consecutive layers of geogrid (h/D) and the bearing capacity ratio (BCR) for different number of geogrid layers (N). It can be seen that increasing the number of geogrid layers (N) more than three layers (the optimum value), decreases the optimum value of (h/D). This can be attributed to the effective reinforcement zone. Furthermore, increasing of the value of (h/D) larger than the optimum, is significantly decreases the value of (BCR). Fig. 8 show the relationship between the diameter (D) and the bearing capacity ratio (BCR) for different the number of geogrid layers (N). It can be seen that increasing of the diameter (D), decreases the optimum number of geogrid layers (N), because increasing the diameter, decreases the soil

Fig. 2. (B C R) - ( D f /D) relationship for different (N) (D=4, u/D=0.3, h/D=0.35, φ = 30D )

Fig. 3. (B C R) - ( D f /D) relationship for different (N) (D=4, u/D=0.3, h/D=0.35, φ = 35D )

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

318

J. K. Abbas, M. N. Al-Shindah

required improvement.

Fig. 11. (B C R) - (U/D) relationship for different (N) (D=4, D f /D=0.5, φ = 30D )

For varies value of the depth ratio of the footing ( D f /D) (0,0.25,0.5) respectively. It can be seen that,

Fig. 7. (B C R) - (h/D) relationship for different (N) (D=4, D f /D=0.25, φ = 40D )

increased the depth ratio of the footing ( D f /D), that the optimum value of (u/D) is constant because is independent to ( D f /D). Figs. 11, 12 and 13 show the relationship between the depth ratio of topmost layer (u/D) and the bearing capacity ratio (BCR) for different the number of geogrid layers (N), for varies value of the angles of internal friction ( φ ) (30˚, 35˚, 40˚) respectively, the optimum value of (u/D) (0.25,0.3,0.35) respectively. It can be seen that, increasing the value of ( φ ), the value of (u/D), increased.

Fig. 8. (B C R) - (D) relationship for different (N) ( D f /D=0,U/D=0.3,h/D=0.35, φ = 40D )

V.

Optimum Depth of Topmost Layer

Several tests were conducted to investigate the effect of the depth ratio of the topmost layer of reinforcement (u/D) and to obtain its optimum values. Figs. 9, 10 and 11 show the relationship between the depth ratio of topmost layer (u/D) and the bearing capacity ratio (BCR) for different the number of geogrid layers (N). Fig. 12. (B C R) - (U/D) relationship for different (N) (D=4, D f /D=0.5, φ = 35D )

Fig. 9. (B C R) - ( U/D) relationship for different (N) (D=4, D f /D=0, φ = 30D ) Fig. 13. (B C R) - ( U/D) relationship for different (N) (D=4, D f /D=0.5, φ = 35D )

VΙ. Optimum Vertical Distance Between Geogrid Layers Figs. 14, 15 and 16 show the relationship between the vertical distance ratio between consecutive layers of geogrid (h/D)and the bearing capacity ratio (BCR) for different the number of geogrid layers (N). For varies

Fig. 10. (B C R) - (U/D) relationship for different (N) (D=4, D f /D=0.25, φ = 30D )

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

319

J. K. Abbas, M. N. Al-Shindah

value of the depth ratio of the footing ( D f /D) (0,0.25,0.5) respectively, the optimum value of (h/D) (0.325,0.33,0.35) respectively. It can be seen that, when the value of depth ratio of the footing ( D f /D) increased, the value of (h/ D), it also increased depended on the value of ( D f /D). Figs. 16, 17 and 18 show the relationship between the vertical distance ratio between consecutive layers of geogrid (h/D) and the bearing capacity ratio (BCR) for different number of geogrid layers (N).

Fig. 18. (B C R) - (h/D) relationship for different (N) (D=4,U/D=0.25, D f /D=0.5, φ = 40D )

For value of the angles of internal friction ( φ ) (30˚,35˚,40˚) respectively, the optimum value of (h/D) (0. 25,0.35,0.375). It can be seen that, increasing the angle of internal friction ( φ ), increasing the value of (h/D). It is depended on the value of internal friction ( φ ). Fig. 14. (B C R) - (h/D) relationship for different (N) (D=4,U/D=0.25, D f /D=0, φ = 30D )

VΙΙ. Effective Depth Zone of Reinforcement The effective depth zone of the reinforcement (d) is the depth beneath the footing base, under which no longer effect of the reinforcement on the bearing capacity is observed. This depth could be calculated as follow: d = u + ( N − 1) h

(2)

Since the optimum values of ( u/D, h/ D and N) for case D=4, are ( φ =30), D f /D=0 (0.25, 0.325, 5)

Fig. 15. (B C R) - (h/D) relationship for different (N) (D=4,U/D=0.25, D f /D=0.25, φ = 30D )

respectively, the effective zone will be(d ≈ 1,55D) the optimum values of case ( u/D, h/ D and N) D=4, φ =35, D f /D=0(0.3, 0.225, 5 )respectively, the effective zone will be(d ≈ 1.2D), .and the optimum values of case ( u/D, h/ D and N) D=4, φ =40, D f /D=0(0.35, 0.3, 3

)respectively. the effective zone will be(d ≈ 0.95D). It is noticed. That when the angle of internal friction ( φ )increases, the effective zone(d) decreased.

Fig. 16. (B C R) - (h/D) relationship for different (N) (D=4,U/D=0.25, D f /D=0.5, φ = 30D )

VΙΙΙ. Conclusions From the numerical results and their discussion presented in this Research, the major conclusions that could be drawn on the behavior of circular footing resting on geogrid reinforced sand are outlined below: (1) The results show that increasing the number of geogrid layers (N) significantly increases the ultimate bearing capacity, but there is an optimum value after which little effect is observed. This optimum value is varied (N=3-5) depending on the value of the depth ratio of footing ( D f /D) and the angle of internal

Fig. 17. (B C R) - (h/D) relationship for different (N) (D=4,U/D=0.25, D f /D=0.5, φ = 35D )

friction ( φ ).

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

320

J. K. Abbas, M. N. Al-Shindah

[11] Lee J., Salgado R., Kim S., Bearing Capacity of Circular Footings under Surcharge using State-dependent Finite Element Analysis, Journal of Computers and Geotechnics.,(32), pp. 445-457, 2005. [11] Loukidis D., Salgado R. (2009), Bearing Capacity of Strip and Circular Footings in Sand using Finite Elements, Journal of Computers and Geotechnics, (36), pp.871-879. [12] Lovisa J., Shukla S. K., Sivakugan N., Behaviour of prestressed Geotextile-Reinforced Sand bed Supporting a loaded Circular Footing, Journal of Geotextiles and Geomembranes, (28) pp. 2332, 2010. [13] Sitharam T. G., Sireesh S., Model Studies of embedded Circular Footing on Geogrid-reinforced Sand beds, Ground Improvement, (8), No.(2), pp.69-75.2004. [14] Boushehrian J. H., Hataf N., Experimental and Numerical Investigation of the Bearing Capacity of Model Circular and Ring Footings on Reinforced Sand, Journal of Geotextiles and Geomembranes, (21), pp. 241-256.2003.

(2) Increasing the depth of footing ( D f /D), decreases optimum value of (N), and increases the ultimate bearing capacity. (3) Increasing the angle of internal friction ( φ ) increases the ultimate bearing capacity. But it has no effect on the optimum value of (N) or the optimum values of (u/D) and (h/D). (4) For the dense sand ( φ =40˚), the optimum value of number of reinforcement layers is (N=3) and it is independent from the value of the depth ratio of footing ( D f /D). (5) The depth of first layer (u/D) and the vertical distance between consecutive layers (h/D) have no direct effect on the optimum number of geogrid layers (N). (6) Increasing of the diameter (D),decreases each of the optimum number of geogrid layers (N), the optimum values of (u/D) and the optimum values of (h/D). (7) The optimum value of (u/D) varies (0.25-0.35) depending on the angle of internal friction ( φ ). (8) The optimum value of the vertical distance between layers (h/D) varies (0.225-0.375) depending on the depth of footing ( D f /D), and the angle of internal

Authors’ information Dr. Jawdat K. Abbas, Assist Prof in soil mechanics- Tikrit University, Engineering College, Civil Dept.

friction ( φ ). (9)The effective depth zone of reinforcement is varied (d/D=0.8-1.55) depending on the value of the angle of internal friction ( φ ) and the value of depth ratio of footing ( D f /D).

Mohanad N. Al-Shindah, B.Sc. in Civil Engineering. Tikrit University, Engineering College, Civil Dept.

References [1]

Das B. M., Shin E. C., Omar M. T., The Bearing Capacity of Surface Strip Foundation on Geogrid-Reinforced Sand and ClayA Comparative Study, Geotechnical and Geological Engineering, 12, 1-141994. [2] Al-Shindah I., N Finite Element Analysis for Bearing Capacity of Circular Footing on Geogrid Reinforced Sand, M.Sc. Thesis, Tikrit University, Iraq. 2011. [3] Bowles J. E., Foundation Analysis and Design, (McGraw-Hill Book Co. New York 1996). [4] PLAXIS b.v. Software manual. Scientific Manual, A. A. Balkema Publishers. Tokyo.2002. [5] El Sawwaf M., Experimental and Numerical Study of Eccentrically Loaded Strip Footings Resting on Reinforced Sand, Journal of Geotech. and Geoenv. Engrg., Vol. 135, No. 10, pp. 1509-1518, 2009. [6] Adams M. T., Collins J. G., Large Model Spread Footing Load Tests on Geosynthetic Reinforced Soil Foundations, Journal of Geotech. and Geoenv. Engrg. ASCE, Vol. 123, No.1, pp. 6672, 1997. [7] Al-Taay A. H., Bearing Capacity of Eccentrically Loaded Strip Footing on Geogrid Reinforced Sand. M.Sc. Thesis, Tikrit University, Iraq, 2010. [8] Al-Ramadan A. K., Assessing The Behavior of Excavation Using the Finite Element Method., M.sc. Thesis, Building and Construction Engineering Department, University of Technology . 2001. [9] Al-Juboory F. M., Finite Element Analysis of Conical Footings, M.Sc. Thesis, Tikrit University, Iraq. 1998. [10] Aswad M. F., The Effect of Reinforcement Material-Nylon on Sandy Soil Behavior under Foundation, M.Sc. Thesis, University of Al- Kufa, Iraq. 1989.

Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Civil Engineering, Vol. 2, N. 6

321

2036-9921(201111)2:6;1-5 Copyright © 2011 Praise Worthy Prize S.r.l. - All rights reserved