1D ground response analyses of compacted soil with dynamic properties measured in laboratory

Earthquakes could be expressed as the vibrations caused by the ruptures in the inner structure of the earth. Even though it is known that the properties of waves reached ground surface could be affected by the earthquake source characteristics such as magnitude, type and length of the rapture they were also influenced by soil layers which they travel through. Therefore two seismic waves which are created by the same source with the same characteristics could have completely different shapes when they arrive ground surface due to the distinct properties of soil layers they pass through. This effect, which is called as a site effect, causes the earthquake waves to change differently for each medium depending on the properties of sites under dynamic loading such as stress-strain behavior, natural period and topography. Since our country has a high seismic activity, the characteristic and destructiveness of the waves at the ground surface are one of the most important topics in the evaluation of earthquake effects in terms of loss of life and property. For this reason, the effect of dynamic loads such as earthquake is vital on stress-strain, bearing capacity and settlement behavior of soils. The dynamic properties which determine the stress-strain the behavior of soils under cyclic loading called shear modulus and damping ratio are determined by both laboratory and field experiments. Since the laboratory experiments such as dynamic simple shear test have the ability to create various loading, stress and drainage conditions comes to the forefront when field tests are considered. Dynamic analyzes have an important role in determining the behavior of the soil subjected to dynamic load. Horizontal displacements, settlements, response spectra, amplification, shear deformations that may occur on the ground surface by earthquake loading could be obtained from the numerical analyzes that can be performed in one, two or three dimensional according to their sufficiency to modeling of real site conditions. With the results obtained, the planned construction can be designed as earthquake resistant. Among the analyzes methods, one-dimensional analyzes, in which the soil model is formed as layered horizontally comes forward in terms of practicality. The methods used to perform one-dimensional analysis are divided into three as linear, equivalent linear and non-linear procedure in which the dynamic properties that define the stress-strain behavior of the soil under repetitive loads is based on different assumptions. The linear method which dynamic properties are considered constant during the earthquake loading for each layer is not preferred since it is not able to model the soil stiffness and damping ratio that varies under repetitive load. The nonlinear method, which takes into account the change in shear modulus and the damping ratio due to time and deformation is a complicated method because it requires many parameters. The equivalent linear method, in which the dynamic properties are obtained as a results of iterative procedure of initial shear modulus and damping ratio according to the shear strain induced in soil is widely used in terms of practicality. In this study, firstly the geotechnical and dynamic properties of the soil samples which obtained from the boring in the engineering fill which was constructed with high compaction energy was determined by laboratory test at Prof. Dr. Kutay Özaydın Soil Mechanics Laboratory. After the determination of the physical and dynamic properties of soil samples, engineering fill was modeled with the usage of the data obtained from test results by Deepsoil V6.1 in order to perform ground response analyzes. In line with the thesis purposes, sieve and hydrometer analyzes were conducted to determine the granulometry of the compacted soil. From the results of the experiments, it was obtained that soil sample has coarse grain material varying between 7-28% and fine grain material changing between 72-93% by sieve analysis. It was stated that soil samples attained from drilling highly consist of fine grain material. After the determination of the percentage of fine grain material, Atterberg Limits tests were carried out on soil samples to know the relationship between soil's water content and consistency. From the result of Casagrande Liquid Limit tests, it was observed that liquid limit (LL) is changing between 38-80% for compacted soil while the results of plasticity limit tests indicate soil's plasticity limit (PL) is between 22-26%. With the evaluation of the test results, soil classification has been made according to the Unified Soil Classification System (USCS). It was concluded that the soil samples used in this study were low plasticity clay (CL), high plasticity clay (CH) and low plasticity silt (ML). To better understanding of the consistency of samples liquidity index (LI) and consistency index (CI) were calculated according to experimental results and the natural water content (wn). It was found that liquidity index (LI) is between -1 and 0 while consistency index (CI) is between 0.9 – 2. Both of the indexes indicated that soil samples are in the hard-stiff state. Pycnometer tests were carried out to define specific gravity. It was seen that values are varying between 2.66 and 2.75 compatible with fine-grained soil's specific gravity. In order to determine the shear strength of soil samples, unconfined compression tests were performed. It was calculated that unconfined compressive strength (qu) for soil samples except T5 was bigger than 400 kPa which is convenient with the liquidity and consistency index results since they correspond hard consistency. From the results of consolidation test the preconsolidation pressure (Pc') was determined between 200 - 400 kPa. It was also observed during the experiment that heave pressure (Pheave) that soil sample had varying approximately between 50-200 kPa. The compression index was defined for soil samples in a range of 0.0685 -0.0959. After the determination of the geotechnical and physical properties of samples dynamic simple shear tests were conducted to define the shear modulus (G) and damping ratio (ξ) change. The experiments were carried out with sinusoidal waveforms with 1 Hz loading frequency. Before starting to dynamic experiment program, with the tentative test it was seen that calibration of the test apparatus should be done since each type of materials such as soft clay, loose sand, dense sand, stiff clay requires different calibration factors suggested by test apparatus. Following the tentative tests the experiment device has operationalized. Since the aim of this thesis was to perform ground response analyzes with the dynamic properties of soil samples, engineering fill was divided into six layers which are called T1 for the 0-12m, T2 for 12-24 m, T3 for 24-36 m, T4 for 36-48 m, T5 for 48-60 m and T6 for 60-75 m. Stress-controlled cyclic simple shear tests were conducted with a different cyclic stress ratio (CSR) values to define the relationship between shear strain induced in the soil sample and the shear modulus for soil layers consisted of different elevation depth. With the evaluation of the hysteretic curves obtained from test results the variation of the damping ratio with shear strain also determined by soil samples at different depth. Since the materials investigated through thesis were clayey soils it was observed that the change in the dynamic properties is highly dependent on plasticity characteristic of soils rather than normal stress applied to it. As a result of tests and calculations, six different shear modulus and damping ratio curves were obtained. By the evaluation of the dynamic simple shear and consistency tests, empirical correlations have found between the shear modulus, consistency limit and plasticity index for shear strain values bigger than 10-3. Since the numerical analyzes require normalized shear modulus degradation (G/Gmaks) to determine the dynamic response, shear wave velocity has been predicted by the correlations in the literature and the Earthquake Regulation for Turkey. Obtained shear wave velocities were used to calculate the maximum shear modulus (Gmaks) which corresponds small strain range. Thereafter one-dimensional ground response analyzes were performed by Deepsoil with the using normalized shear modulus- shear strain, damping ratio- shear strain curves procured by dynamic simple shear tests. Within the scope of analyzes eighteen strong ground motion data with different PGA values (0.1g, 0.2g, 0.3g, 0.4g) and various Arias Şiddeti values were used to obtain horizontal displacements, shear strains, amplifications, peak ground accelerations and response spectrum of the soil profile by dynamic loading. It was concluded that the peak ground acceleration through the soil profile was changing slightly for the various PGA values of input motion. As a reason for this slight change the uniformity of soil profile (except T5) in terms of shear wave velocities and dynamic properties could be pronounced with the impedance ratio which is very close to one. From the results of analyzes for earthquakes with 0.1g input PGA, the surface PGA values for Coyoto Lake, Darfield New Zealand(14W), Hector Mine(HBS180), Hector Mine (JTN90), Hollister 04, Landers, Sitka Alaska, Darfield New Zealand(76W) were found as 0.15g, 0.12g, 0.13g, 0.12g, 0.13g, 0.12g, 0.11g, 0.11g respectively. Earthquakes with 0.2g PGA the surface PGA values for Big Bear, Basso Tirreno, Anza, Northridge, Parkfield-02 CA(36529360) were found as 0.21g, 0.22g, 0.22g, 0.24g, 0.32g respectively. Earthquakes with 0.3g PGA the surface PGA values for Iwate, ParkDonna -90 were found as 0.30g and 0.32g respectively. Then the Iwate and ParkDonna – 90 earthquakes was scaled to 0.4g to observe the effect of higher PGA values and the 0.31g and 0.38g were found respectively. For Kocaeli earthquake with 0.4g the surface PGA was found as 0.24g. The horizontal displacements were found as 0.01m for Big Bear, Anza, Northridge, 0.02m for Coyoto Lake, Basso Tirreno, Darfield New Zealand(14W), Hector Mine(HBS180), Hector Mine(JTN90), Landers, 0.03m for Hollister 04, Darfield New Zealand(76W), Parkfield-02 CA(36529360), Parkfield-02 CA(Donna90), 0.07m for Iwate. For the 0.4g - scaled Iwate and ParkDonna-90 earthquakes the displacements at surface was found as 0.10m and 0.04m respectively. From the analysis of Kocaeli the displacement was found as 0.13m at ground surface. It can be said that considering the shear wave velocities of the input motions, the earthquakes with small PGA and Arias Şiddeti values also have less displacement values on the ground surface in comparison with the earthquakes with higher PGA and Arias Şiddeti. However the input motions with same PGA values could generate various displacement values regarding to their Arias Şiddeti (While ParkDonna-90 with 0.4g generated 0.04m displacement, Kocaeli with same PGA and higher Arias Şiddeti generated 0.13m displacement at ground surface. Shear strains occurred in the soil profile were evaluated through the results of analyzes. As it was expected the maximum shear strains were generated between 48-60 m which corresponds to T5 layer with the lowest shear wave velocity for each strong ground motion record. The amount of shear strains induced in the soil profile at 5 m depth (approximately the base of elevation for shallow foundations) were found as 0.0045% for 0.1g input motion, 0.008% for 0.2g input motions, 0.011% for 0.3g input motion. Shear strains generated by Kocaeli, Iwate, Parkdonna – 90 were found 0.014 %, 0.018%, 0.022% respectively.From the results of analysis the spectral amplifications was found to be varying between 2.15 and 2.59 at predominant periods between 0.96-1.69 sn. In the scope of the thesis, the ground response analyzes also performed to base on Turkish Building Earthquake Regulation (TBDY). The soil classification was made according to TBDY regarding shear wave velocity for upper 30 m of the soil profile and determined as ZC ((Vs)30 = 360-760 m/s) type of soil with average shear wave velocity of 383 m/s. Then a response spectrum was drawn according to the TBDY for ZC type of soil and earthquake level for DD-2 (with the probability of exceeding the spectral aspects over 50 years is 10% and corresponding repetition period is 475 years.). It was seen that the response spectrum obtained from the analyzes and calculated from the earthquake regulation has dominant periods between 0.09-0.40 sec for both. The maximum spectral acceleration values were 0.74g for the results obtained from the regulation and 0.68g for the ground response analyses. The different peak values of the spectrum drawn using two methods indicate the importance of site-specific analyzes in determining the response spectrum to be used in the design phase. In the study, the borehole in the engineering fill field has been utilized. Since the number of boreholes and the type of engineering fill used is the only one the data obtained from them remained limited in order to obtain dynamic properties of compacted soils. By using the samples obtained from the borehole to be carried out in the various engineering fill areas, the correlation and the dynamic property curves suggested by this study are open to be developed.