یزدان شمس ملکی

In this research, the effect of soil stabilization with cement at the same time as its reinforcement with fibers has been studied on the shear strength of sandy soil exposed to freeze-thaw cycles. In order to achieve this goal, laboratory studies were carried out with the help of unconfined compressive strength tests (UCS tests) on different compounds obtained from mixing cement, fibers, and sandy soil. More than 336 cylindrical laboratory models with dimensions including 3.6cm in diameter and 8cm in length have been made. Various modes have been observed during the failure of the samples, including shear, tensile, plastic yielding, and composite failure modes. The fibers used in the present research are waste products of tire factories known as DTY. Percentages of 2, 4, and 6 for cement and 0, 0.5, and 1 for fibers with lengths of 0.5, 1, and 1.5 cm were used relative to the weight of dry sandy soil in making the samples. Uniaxial cylindrical samples were tested for unconfined compressive strength after 7 and 28 days of curing time and under 0, 1, 2, and 3 freeze-thaw cycles. The results show that the act of stabilizing the soil with cement, along with reinforcing it to a certain amount of fibers, improves the uniaxial compressive strength before and after freezing and thawing cycles. This amount depends on the percentage of cement and the curing period. Also, adding cement in a certain curing time increases the unconfined compressive strength before and after applying the cycle, increases stiffness, reduces the ductility and toughness of the sample, and brittle failure when breaking occurs in the soil. Also, the addition of fibers, to some extent, improves the weaknesses caused by soil stabilization, such as reducing the failure axial strain, decreasing the residual strength, and the toughness of the materials in the conditions before and after freezing and thawing.

To study the dynamics of embankment models, which is the basis for the construction of many engineering structures, physical modeling with different scales can be performed. The pre-seismic behavior of physical models has a great impact on the subsequent results of their seismic modeling. Appropriate experiments in this field can be used for pre-seismic examination of physical models. Free vibration and impact tests are among these tests (Dezi et al., 2012; Capatti et al., 2018; Kramer, 1996). Impact pulse tests (IPT) can be used to determine the natural frequency characteristics of physical models. Interpretation of frequency responses provides researchers with appropriate pre-seismic parameters. According to the evaluation of these findings in the first step, a suitable geometric scale can be selected for physical modeling. In addition, the relationship between soil conditions in terms of density and particle size and frequency response quantities of physical models is also determined. In this paper pre-seismic simulations of a small-scale physical model of a granular soil layer inside a rigid physical modeling box are developed.

The occurrence of ice lenses and the subsequent melting of the ice causes a lot of damage to the beds consisting of fine sand soils every year. Volumetric changes of soil during freezing-thawing is a factor that reduces soil strength and increases deformations. In this study, several laboratory measurements are presented to investigate the effects of freeze-thaw cycles on the behavior of cylindrical sand-cement-fiber specimens. Fine-grained sand has been chosen to investigate the effects of freeze-thaw. At the same time, chemical stabilization methods of soil by mixing with cement with amounts of 2, 4, and 6% by weight of dry soil and mechanical reinforcement by adding 0, 0.5, and 1% by weight of recycled nylon fibers have been used. This study shows that the presence of fibers next to cement causes obvious changes in the stiffness, strength, and durability characteristics of the samples under the effect of freeze-thaw cycles. From the findings of this study, it can be concluded that in samples without fibers, distinct and wide cracks are observed. While, in samples armed with fibers, the cracks are smaller and distributed in a wider width. The results related to the behavior of the samples during loading showed that in the samples reinforced with fibers, failure occurred due to the pull-out of the fibers. In 7-day dry samples (without the freezing-thawing cycle effect), the compressive strength of the samples increases with the addition of fibers. In the 28-day samples, with an increase of only 0.5% of fibers with a length of 0.5cm, the unconfined compressive strength increased , and its decrease was observed after that. In all 7d and 28d dry samples, with the increase of fiber size from 0.5cm to 1 and 1.5cm, the compressive strength of the samples has a decreasing trend. Also, by adding the percentage of fibers from 0.5 to 1%, the trend of decreasing strength of the specimens can be seen.

In this research, the free vibration or natural frequency analyzes have been performed with the help of small-scale physical models. Laboratory modeling in the geotechnical engineering can be performed in the acceleration field of 1g.  In each of the physical modeling modes, the relationship between the model and prototype frequencies is very essential. In this paper, with the help of hammer impact pulse tests (HIPTs) -dynamic experiments- the optimal frequency ranges and the best geometric scales for physical modeling are investigated by a strongbox. The frequency range studied has been selected according to the study of shaking table models between 0.001Hz and 150Hz. To perform impact pulse tests, the physical models of dry sandy slope with different inclination angles from 25 to 60 degrees (and a constatnt slope height) have been instrumented by the piezoelectric acceleration sensors.  The relative density of the sandy slope models is medium dense and about 50% to 52%. In addition to 8 physical models of sandy slopes, two models of level-ground and empty box have also been investigated. The time-history of the acceleration function of the input excitation shock at the slope floor (base point) and the response acceleration at the slope crest are recorded by the acceleration sensors.  These acceleration time responses last for a short stroke (short impact) of less than 1.0 second duration. After extracting temporal responses, the frequency analyzes including transfer function (TF), Fourier response spectrum ratio (RFRS), and spectral energy density function (PSD-function) are derived from the temporal results. Using the transfer function or RFRS, quantitative values of natural frequencies of the physical model of the sandy slope and the storngbox are extracted in different vibration modes. According to the findings of the present study, for a constant slope model the frequencies at which the maximum seismic or dynamic energy is emitted are quite different from the frequencies with the maximum magnified response amplitude. The results of the present study prove the existence of a logic relationship between the sandy slope inclination angle (physical model natural frequencies) and the model response amplification frequency. So that by increasing the angle of inclination of the model slopes at a constant height, the magnification values of the impact acceleration response decrease. Because in general, the amount of sandy materials magnifies or weakens the amplitude of frequency responses. The presence of low sandy materials (on steep slope models) reduces the magnification range of the acceleration response and high sandy materials (on gentle slopes) increase the response range. Optimal frequencies in strong box modeling in the 1g acceleration field are frequencies that do not interfere with acceleration magnifications before or during seismic excitation (pre-seismic mode). Acceleration magnification causes resonance and premature failure in the physical model, which is generally undesirable and unmeasurable in laboratory studies. In this research, the optimal frequency range according to the measurements is proposed for the physical modeling of the 1g acceleration field. These ranges and frequency values are presented according to the various constraints such as the type of strong box, slope angle, relative density of sand, the actual frequency effect of the horizontal components of earthquakes, and so on.

In this paper, the irregularity effects in the initial directions of the seismic loading on the nonlinear time-history responses of the Shohada dam have been investigated. The maximum cross-section of the dam has been simulated by the 2D finite element method under the normal lake level conditions. The near-fault acceleration record of the Tabas earthquake has been applied as dynamic input motion to the two-dimensional numerical models. Numerical analysis has been conducted in the Newmark explicit time integration scheme framework. The main three patterns of horizontal, vertical, and oblique directional seismic loading are considered to investigate the effects of the initial directions of seismic motion propagation. In each case, the seismic responses are compared to the conventional seismic loading responses in the horizontal direction from the upstream (reservoir) to the downstream of the dam. The hardening soil model with small strain (HS-small) and Mohr-Coulomb model was used to model the dam's body and the foundation materials. In most 2D simulations, the initial direction of the input seismic movements has been only in one or two dimensions and in the horizontal or vertical orientation. This study has attempted to compare these traditional seismic loading patterns with other possible states in 2D numerical models. The present study results show that the seismic loading mode with the initial direction of inclination of 45 degrees has the worst and most significant effect on the seismic response of the dam compared to the traditional method of horizontal loading.

The tunnel overburden height has considerable effect on stability or instability of surface tunnels. In this research, effects of overburden surface tunnel in strong rocks in both static and pseudo-static states investigated. In this paper, the results of pseudo-static analysis comparing to static state analysis are outstanding. Due to superficiality of tunnel, the water table line considered lower than tunnel bottom. The static analysis performed by Finite Element Method (FEM) and Finite Difference Method (FDM) and both results are close and acceptable. Results of static analysis indicated that by increasing in overburden height, horizontal displacement in sidewalls of tunnel and vertical deformation of tunnel crest would decrease. The stress amounts in sidewalls and crest of tunnel in pseudo-static state are higher than static condition.

Using piles row is a well-known method to stabilize the soil slopes in the both static and seismic conditions. The load transfer mechanism of installed piles row may be as end- bearing or floating. In this study the behavior of floating piles row with circular cross- section were installed inside the dry sandy slope by help of three- dimensional numerical analyses and physical modeling have been simultaneously studied. The three- dimensional numerical modeling was used for conducting the parametric studies about effects of directions of imposition of near-field and far- field earthquakes loading on the main structural and geotechnical interaction parameters of floating piles row-sandy slope problem (Al-Defae and Knappett, 2015; Ashour and Ardalan, 2012).

Considering countless repetitions of Gaussian-shape bending moment-depth curves, M(z)-z, in the results of several experimental/numerical and static-seismic physical modeling, suggestion of a new relationship for using this opportunity appears to be suitable. This relationship is created a connection between high-used p-y curves and the structural internal efforts of piles in the form of bending moment along the piles lengths. In this study, the bending moment equation along the depth for the issue of floating pile row under near-fault earthquake lateral loading in the dry sandy slope is obtained by using the results of the three-dimensional numerical models and physical models. Therefore, in the next step, the values of bending moment by the suggested new relationships transform to the proportional p-y curves. The main abilities and advantages of the new suggested relationships include generation of relationships compatible for different soil types such as clayey soils (i.e., cohesive soils), mixed cohesive-granular soils, sandy soil and some weak rocks. Furthermore, other advantages of new relationships include (1) the compatibility of relationships with the experimental tests and the static field tests, cyclic, and seismic modeling, (2) have no need to calculate slip depth in the sliding sand mass in the sandy slope failure problem and the fully analytical form of the relationships. In addition, all the conducted mathematical calculations are done in the parametric form; therefore, all other kinds of similar problems can be totally computed by the suggested new analytical relationships. In this study, the mathematical-analytical relationships among monotonic static p-y curves, tangent hyperbolic cyclic p-y curves, and Gaussian bending moment curves were presented. The Gaussian-shape bending-moment curves have been calculated for dynamic loading of floating pile row in the dry sandy slope under different combinations of near-fault earthquakes. According to the findings of the present paper, the values of Gaussian bending moment curves can be simply transformed to the soil pressure, p, and relative pile-soil deflection, y. Essentially, by using this strategy, the quantity of generated stresses within the soil due to pile lateral loading and soil yielding or soil plasticity can be controlled. On the other hand, by understanding the deflection of pile, y, the values of pile lateral deflection can be compared with allowable deflections in each project; the surviving of the superstructure can be judged based on the pile in the safe zone. In each arbitrary depth, the values of two parameters pu and Ki are different, and the values of these two parameters generally depend on the depth change; in addition, these differences there are between p-y curves in the different depths. The slope of the initial portion of both static and cyclic p-y curves at points p=0 (equivalent to the earth surface i.e., z=0 in the sandy soils) and y=0, according to the present article calculations is always Ki, which is a stress-kind parameter and has a stress unit. The findings of the paper show that the shape of the bending moment curves obtained from the double integration of static and cyclic p-y curves, p(z) component, similar to the shape of numerical bending moment curves, is Gaussian. Moreover, the sign and depth-pattern of the obtained bending moment curves are completely similar to the considered predictions. The depth location of the maximum bending moment from double integration of static monotonic p-y curves is in a depth that is close to the depth of the maximum bending moment of numerical results, while the maximum bending moment from double integration of cyclic p-y curves occurs in the shallow depth (i.e., at the zone of failure surface of the slope).

In this study the behavior of floating piles row with circular cross-section were installed inside the dry sandy slope by help of three-dimensional numerical analyses and physical modeling have been simultaneously studied. The three-dimensional numerical modeling was used for conducting the parametric studies about effects of directions of imposition of harmonic seismic loading on the main geotechnical parameters of floating piles row-sandy slope problem. The seismic loading of harmonic sinusoidal waves in the form of seismic motions in the in-plane and out-of-plane directions along the longitudinal and transverse directions of slope model and both of them were imposed on the slope physical and numerical models. Moreover, the physical modeling of the investigating problem was implemented for validation of numerical results by imposing the sinusoidal harmonic loading in the longitudinal and transverse horizontal directions of micro-scale slope model by help of small-scale geotechnical shaking table. Reinforcing of a dry sandy slope by a row of floating piles (similar to reinforcing of slope by end-bearing piles row) results in significant decrease (to about more than 50 percent) in slope’s vertical displacements. Out-of-plane components of seismic loading such as transverse component of earthquake, T component, (productive of horizontal shear waves, i.e., SH waves) also in the presence of site’s effects such as “directivity effects” can produce the responses as large as the in-plane motion components such as earthquake longitudinal contained component, L (productive of P and SV seismic waves). The motion of slope sliding wedge in the strong ground motions is a rigid block motion while the failure wedge displacement under weak ground motions is a negligible motion and occur in a flexible block manner. Simultaneous seismic loading along two-axes of three coordinate axes in contrast to the current slope seismic loading that the seismic loading are imposed along one axis and in the longitudinal direction of slope failures surface, have great effects on the slope displacements values and internal efforts generated in the reinforcing piles row. Studying and solving the classic problem of in-plane and out-of-plane seismic-waves propagations in the combination with the existence of slope in the ground and piles row interaction (i.e., adding the piles row-slope seismic interaction to the initial classic problem) by help of present available analytical and mathematical solutions will be a very difficult problem. By combining of the in-plane and out-of-plane seismic motions the complexity of the piles row-sandy slope dynamic interaction problem will increase and in the some cases presenting analytical and closed-form solutions can be impossilble. The alternating solutions for solving these complex problems proposed by the present paper are the simultaneous using of numerical and shaking table physical modeling for understanding the precise details and ambiguities of the problem in the complete scientific and practical-empirical frameworks. The results of present study show that installing a row of floating piles similar to the end-bearing piles row can reduce the displacements of loose dry sandy slope and through this manner the seismic stability of slope against the local and general failures increased. In the present paper despite of increasing the directions of seismic loading from one-direction to the two-directions the installed floating piles row sufficiently played their roles in the reducing seismic displacements of slope. Indeed, according to the experimental-numerical results of the present paper, decreasing of slope crest settlements by installing a row of floating piles in the seismic loading cases are more than 50 percent. The results of small-scale 2DOF geotechnical shaking table physical model were used to verification of the obtained 3D numerical results. There is a good agreement between the numerical and physical models results.

Predicting the behavior of urban tunnels in weak rocks with a behavior similar to soil's or rock's by swelling and squeezing ability is one of the important challenges in tunnel engineering and rock mechanics. In this research plane strain two-dimensional model of a real urban tunnel with horseshoe cross section is modeled through 2D nite element method. The aim of modeling is the investigation of the eects of considering or ignoring time factor in the constitutive model of materials on the interaction behavior of tunnel support system and also stress-strain and de ection behavior of surrounding rock masses.
For numerical modeling of tunnel-support system interaction two constitutive models including Mohr-Coulomb model(time-independentmodel),MC,andsoftsoilcreep model (time-dependent model), SSC, similar and constant rock mass properties were considered. The results of these two models in the framework of structural behavior of tunnel support system and also stress-strain and de ection behaviors of its circumference rock mass were compared. The results of study show that considering time eect on some cases caused an approximately 63 percent dierence between two model's results. For verication of research problem, a tunnel with similar conditions and jointed rock mass constitutive model in the frameworks of 2D and 3D nite-element analyses were used. The results of verication show good agreement between the obtained results by numerical models of the present paper and numerical models of the reference paper.
Mohr-Coulomb constitutive model is a model independent of time, stress history conditions, consolidation state, creep, hardening and softening, in order that no change could occur in the results obtained with this model in plastic numerical analyses by changing time intervals of numerical computations. However, considering creep of materials and time factor will generally increase the values of structural interaction parameters (secondary parameters such as bending moments, shear forces and axial forces in shotcrete and rock bolts) and geotechnical interaction parameters(initial parameters such as stressstrain relationships parameters, displacements and de ections of rock materials) of SSC model rather than MC model in the excavation of tunnel and installation of its support system. Impressibility of secondary parameters (i.e., structural parameters) in support system of tunnel compared to initial parameters (i.e. geotechnical parameters) is half by changing time intervals of implementationofcreepplasticanalyses. Inotherwords,with respect to the obtained results by changing real time in numerical analysis of geotechnical parameters (initial parameters) is approximately 2.30 times greater than structural parameters (secondary parameters), because the occurrence of geotechnical parameters will generate and mobilize secondary parameters (structural parameters) in the interaction of support system of tunnel-rock.

Investigation of lateral bearing capacity of piles under lateral loading in the vicinity of soil slopes and level ground is one of the interesting issues in the engineering of the deep foundations in recent decades. In the meantime, the estimation of lateral stiffness of soil layers, coefficient of soil reaction, ks, is an old challenge on the basis of the creation of empirical (testing) and theoretical p-y curves. In this paper, we attempt to estimate lateral soil stiffness in the length of concrete piles in various depths in sloped ground and level ground and in the all type of soils, including cohesive soil, frictional (granular) soil and mixed soil, i.e. c-φ soil with the aid of numerical analysis in three dimensional finite difference software, Flac3D. For the purpose of estimating lateral soil stiffness, Ki, in different depths by means of this numerical method, the p-y curves have been derived in several depths in the length of concrete pile with circular section under static lateral loading up to the threshold of the yielding of soil materials. The empirical relationship established between soil lateral stiffness Ki and the effective dimensionless parameters on the Ki in general is calculated as a six order polynomial, and this six order polynomial reduces to a fourth order polynomial for the cohesion component of the soil. At the end of the paper, some well documented lateral loading tests by the present numerical method have been investigated for validation of the numerical method of derivation of the p-y curves and pile-head load deflection curves, H-y curves have been calculated and are compared with other results obtained by other researchers and methods. There are good agreements between test results and the results obtained from the numerical method presented in this paper.

In the present paper, derivation of the yield function of the Hoek-Brown model in stress invariant space is investigated. This model is an empirical model that is dened for estimating the bearing capacity of rocks. The main equations of this model, according to principal stresses and in 3D stress space, are dened here . By dening of the model in the stress invariant space, the yield criterion will be independent from the coordinate directions and the rotation of stress axes in general static loadings. By denition of the Lode angle presented in certain papers and books, basically one cannot derive the relationship between the yield function of the model in the principal stress space and stress invariants space. This problem is because of ignoring the 30 degrees difference between the Lodes angle and the used angle in the denition of the models. After denition of the yield function, the plastic potential function of the model is also investigated. In this study, Hessian matrices are obtained by means of the chain rule in dierentiating stress invariants. Basically, any arbitrary constitutive model that is expressed by principal stresses can transform to the stress invariant space via the basic relationships presented here. After this step, by computation of the elasto-plastic constitutive matrix, we can estimate the stress-strain behavior of material using that arbitrary constitutive model. This paper focuses on the elasto-plastic behavior and corresponding elasto-plastic relationships of the Hoek-Brown generalized criterion in three dimensional principal stress space and introduces a simple way to convert these relationships to three dimensional stress invariant space.

Behavior of braced excavation in a cohesive-frictional soil has been evaluated by present paper. Two groups distinct analyses based on plastic calculations and consolidation calculations were implemented by considering time intervals of staged excavation. In order to investigation of time effect and pore water pressure impact on the staged excavation phases, numerical modeling has been conducted by help of consolidation and plastic analyses. In major of former analysis and before present study numerical analyses of staged excavation procedure have selected without considering the effect of time and in form of plastic analyses, while pore water pressures were also ignored. In this, paper thereto the time effect other effects such as length of time interval, kind of analysis depending on drainage conditions, constitutive modeling for soil and location of ground water table were considered. Two dimensional finite element analyses in PLAXIS 2D software are the basis of the numerical calculations of present study. Excavation bracing selected as a kind of concrete facing wall and grouted soil nailing. The results of this research show that the values of excavation wall lateral displacement and soil heave in bottom of the excavation in consolidation analysis by considering time effect in comparison with plastic analysis often reduced approximately 20%.
It seems that effect of time and staged excavation just with implementation of numerical deflection analysis depending on the time such as consolidation analysis with creep models can be evaluated and these conditions in plastic numerical analysis with staged excavation without creep (time-depended) models is meaningless. Soft soil creep model (i.e. SSC model) for considering time effect in plastic and consolidation analyses has been used by authors. The results of present paper show that neither consideration of analyses that consider time interval nor analysis such as consolidation analysis that considers time are not adequate, but soil constitutive model that defines the material behavior must be contains time and also in their mathematic equations structure parameters such as time, strain rate and stress rate that vary with time must be taken into account. Present paper analyses show that consolidation analysis by considering time effect obtain less wall deflections by comparison with plastic analyses.
However, from present study outcomes can conclude that the plastic analyses also by considering constitutive model that contain time can take into account time effects in stress-strain calculations. On the other hand, responses of plastic analyses by comparison with consolidation analyses always are preservative and show more values. Therefore, in structural designing of bracing of an excavation, reliance on results obtained from plastic analysis is preservative and real values of time-dependent deflections of wall and bottom of excavation via consolidation analyses are obtainable. This paper has recommended that both plastic and consolidation analyses for designing of braced cut were considered by engineers and optimum system between those according to the economically advantages and disadvantages be selected. Because, occasionally reliance on plastic preservative analyses lead to imposition of high values of design and construction costs on a certain project that is revealed by implementation of consolidation analyses that those are not necessary. At the end of the paper, verifications and comparisons are related to the topic of present study have been carried out by authors and the obtained results have been compared with together and then are investigated with the obtained results by present study.