Determination of soil displacements depending on the soil behaviour during earthquakes
The August 17, 1999 Kocaeli Earthquake had a moment magnitude of Mw = 7.4 with a focal depth of 17 km. Several cases of liquefaction in silt and sand layers and bearing capacity loss in fine-grained soils have been observed in Adapazarı City that is located on alluvial soil deposits. Such phenomenon resulted in tilting of buildings, excessive settlements and lateral displacements and caused several thousands of casualties and billions of dollars of losses consequently.
For the mitigation of such hazards, the determination of the cyclic behaviour of the soils in the region, the mode and magnitude of displacements under cyclic loading conditions and ascertaining the displacements or deformations being whether within tolerable limits or not become very important.
According to the existing liquefaction analysis methods, it is necessary to determine the shear wave velocity values, maximum ground acceleration value and soil index properties for the determination of liquefaction potentials of soils. But in spite of such a study, still accurate results for liquefaction potential may not be obtained. Therefore, it becomes very important to utilize laboratory cyclic tests for the investigation of liquefaction potential and bearing capacity loss of soils.
Within the context of this study, the cyclic behaviour of Adapazarı soils with reference to August 17, 1999 Kocaeli Earthquake have been investigated both in the field and laboratory by studying on soil samples taken from 10 borings loacted at 8 different sections of Adapazarı. By focusing on the lateral spreading of soils caused by the Kocaeli Earthquake and studying different methods proposed by various researchers, a new approach for the determination of lateral soil spreading under cyclic loadings has been developed and this new method was applied to a couple of cases.
The first part of this study covers a literature survey on current methods developed for the determination of liquefaction potential of coarsed and fine grained soils, and a detailed information on the definition of lateral spreading of soils, its effects on engineering structures and current methods for the determination of the intensity of lateral spreading. The second part involves the laboratory determination of dynamic soil properties and also ascertaining the values of soil parameters for liquefaction analyses. As being one of these parameters, shear wave velocity of the soil layers were calculated by using emprical relationships while the maximum ground surface acceleration values were obtained with the help of attenuation relationships and by running the software ProSHAKE. The liquefaction analyses were conducted by using these parameters and the possible liquefaction depths were calculated accordingly.
The results of different methods of liquefaction analysis performed for fine grained soils have shown that most of the silt-clay soils do not liquefy. But with regard to the experimental results obtained within this study, it was determined that bearing capacity loss and liquefaction may occur for soils having fine-grained particles more than 15%. This important finding is in harmony with the studies of Seed et al. (2001), Bray and Stewart (2000), Bray et al. (2001) and Sancio et al. (2002;2003;2003b).
In order to determine the behaviour of Adapazarı soils under cyclic loading conditions and the reduction in post-cyclic static strength of these soils, two different sets of cyclic triaxial tests were conducted on undisturbed soil samples taken from borings made in Adapazarı. The first set includes the tests where cyclic loading was continued until liquefaction occurred (=±2.5 %) while the second set involves the tests with a constant cyclic stress ratio applied with a loading frequency of f=0.1 Hz for a specific number of cycles, i.e. N=20 was chosen to represent August 17, 1999 Earthquake with Mw=7.4. Strain controlled monotonic tests (20 mm/min) were performed at the end of cyclic loading application stage. Due to the limitations of the loading system, the monotonic tests were carried on until an axial strain value of 10% or 20 %.
Undisturbed soil samples were taken from 10 borings located at 8 different districts of Adapazarı. The soil samples have moisture content values varying between wn=22-50%, fines content values changing between FC = 1-100 %, liquid limit values between wL = 26-74 % and plasticity index values change between Ip=0-44% while uncorrected SPT-N values vary between N30=5-44. For most of the soils investigated, fines content is greater than 35% and SPT-N values change between N30=4-9.
According to the results of oedometer tests conducted on soil samples taken from near surface between 2.0 m to 4.5 m depths, overconsolidation ratio values vary between OCR=2.75-3.85 which means that soils near to the soil surface had become overconsolidated due to seasonal variations in ground water level and dessication phenomena.
The first set of tests involving cyclic loadings being applied regarding the failure criterion of the axial strains reaching = 2.5 %, were conducted on non-plastic silty sands or very sandy silts and soils having plasticity index values of Ip=%15-22. According to the strength curves obtained by using the results of these tests, the cyclic strength of silty sands-sands (FC=39-46) is 45% less than the cyclic strength of high plasticity soils (MH, IP=15-22%). It has also been determined that cyclic strength of soils with a plastisity index of Ip=15-16 is less than that of high plasticity
According to the strength curves obtained by using the results of the first set of cyclic triaxial tests performed within this study obeying the failure criteria of reaching an axial strain of =±%2.5, the cyclic stress ratio values obtained for a number of cycles of N=20 which represents an earthquake with magnitude Mw=7.4, are determined to be as CSR=0.490 for MH soils with plasticity index values changing between Ip = 27-31 % and CSR=0.360 for MH soils with plasticity index values between Ip = 15-22 %, CSR=0.430 for CH soils, CSR=0.325 for CL soils, CSR=0.240 for ML soils and CSR=0.220 for SM soils.
The second set of tests involve a cyclic loading stage for a constant stress ratio with a specific number of cycles of N=20 and then a monotonic loading stage with a loading speed of 0.20 mm/min It was determined that soils reaching excessive axial strain levels during cyclic loading stage have a lower post-cyclic monotonic strength when compared with their initial strengths. But for some samples, the negative pore water pressure build-up due to dilatation gave rise to higher strength values.
For the second set of tests, the maximum elasticity modulus values were determined at very low strain levels prior to the loading levels causing plastic deformations. For the later stages of cyclic loading continued for number of cycles of N=20, the stiffness of soils decrease as the number of applied cycles increase. Elasticity modulus values were determined also for successive loading stages where plastic deformations encountered and shear modulus values were calculated accordingly. Consequently the shear modulus ratio (G/Gmax) variation with the strain () have been determined. According to the graphical plots obtained from the results of the tests conducted with this study, the minimum value of shear modulus was attained for non-plastic soils while soils with IP=40% have the highest shear modulus values.
In the proceeding stage, the existing methods for the determination of the lateral spreading of soils were applied to cases those being a lateral spreading case for the region where an indoor swimming pool facility is located along the sea-shore in Golcük after the 1999 Kocaeli Earthquake and some cases where lateral spreading due to liquefaction had been encountered in some soil sections in Kobe City after the 1995 Kobe Earthquake. In addition tho this, the lateral soil spreading values for 5 borings among 10 located at 8 different districts of Adapazarı City have been determined by utilising some current methods for lateral soil spreading determination. Cyclic 1D software and shear strain curves obtained in a manner similar to the method proposed by Wu (2002) by using the results of the tests conducted within this study have been used for the calculation of the lateral soil spreading values. These values have been compared with the observed values deduced by using 1/1000 scaled maps of the investigated region belonging to the periods of before and after the 1999 Kocaeli Earthquake. It has been shown that the calculated lateral soil spreading values determined by using the curve obtained by limited number of tests conducted are generally in harmony with the observed ones. On the other hand, it has been deduced that this curve should be refined by using more experimental data.
Zülküf Kaya
Depremler sırasında zeminlerin davranışına bağlı olarak meydana gelen zemin deplasmanlarının belirlenmesi · 2008 · 425 sayfa.
Danışman: Prof. Dr. Ayfer Erken
Bu araştırmanı devamına: http://tez2.yok.gov.tr/tez.htm adresinden ulaşabilirsiniz.