##
İnce daneli zeminlerin dinamik davranışı

İnce daneli zeminlerin dinamik davranışı

##### Dosyalar

##### Tarih

1995

##### Yazarlar

Sancar, Tufan

##### Süreli Yayın başlığı

##### Süreli Yayın ISSN

##### Cilt Başlığı

##### Yayınevi

Fen Bilimleri Enstitüsü

##### Özet

Bu araştırmada suya doygun, ince daneli zeminlerin dinamik davranış biçimleri ve bu davranış biçimi üzerinde dinamik gerilme seviyesinin, çevrim sayısının ve plastisitesinin etkisi incelenmiştir. Erzincan Ekşisu bölgesinde yapılan sondajlarda tüp ve pistonlu numune alıcılarla alınmış örselenmemiş plastik ve plastik olmayan siltli zemin numuneleri üzerinde, gerilme kontrolü olan dinamik basit kesme deney sisteminde, bir seri deney yapılmıştır. Zeminlerin dinamik mukavemetini etkileyen faktörlerden olan; uygulanan dinamik gerilme seviyesi, çevrim sayısı ve plastisitenin etkisi incelenmiştir. Deney sonuçlarına göre; uygulanan dinamik gerilme seviyesinin artması durumunda boşluk suyu basıncının ve birim kaymanın arttığı gözlenmiştir. Ayrıca, çevrim sayısının artması zemin numunelerindeki birim kayma değerlerini arttırmaktadır. Diğer bir sonuç ise, plastisitenin dinamik mukavemet üzerindeki etkisidir. Belli bir çevrim sayısı dikkate alındığında, plastik şiltlerin dinamik mukavemetinin, plastik olmayan şiltlere göre daha fazla olduğu gözlenmiştir.

It is inevitably necessary to determine the behaviour of cohesive soils while investigating the response of soil layers to the earthquake loads. In such soils there is no total loss of strength as in sands; therefore investigations should be directed toward stress-strain and porewater pressure behaviour. The problems encountered in cohesive soil layers during an earthquake can arise as instability at slopes where the soil surface is inclined. Also the cyclic properties of cohesive soils should be obtained in order to conduct earthquake analyses at soil structures such as an earthfill dam. Determination of the behaviour of soils under cyclic stresses, in other words the cyclic properties, can be divided into two sections: 1- Stress-strain properties 2- Strength properties The term stress-strain properties implies the determination of the cyclic shear modulus and the damping ratio values and their variation with unit strain. The shear stress amplitude that produces collapse or large deformations are used as strength parameters. Cyclic tests were conducted on properly prepared undisturbed soil samples and the factors affecting the cyclic behaviour were investigated. These factors have been thoroughly investigated by many researchers by tests done in the laboratory as well as in the field. Factors affecting the cyclic strength can be classified into two main groups according to their importance: Factors of primary importance: 1- Unit shear strain amplitude, y 2- Average effective confining pressure, oq 3- Void ratio, e 4- Number of cyclics, N 5- Saturation ratio, Sr 6- Overconsolidation ratio, OCR 7- Frequency, f Factors of secondary effect: 1- Testing methods xii 2- Loading type and direction Soils under the effect of cyclic loads are effected by cyclic shear stress at various amplitudes and frequencies. The laboratory tests developed nowadays to determine the strength properties of soils at such loading conditions are as follows: 1- Cyclic simple shear test 2- Cyclic triaxial test 3- Cyclic torsional simple shear 4- Shaking table 5- Resonant column test The cyclic simple shear testing system used in this study was a modified version of the Norwegian simple shear apparatus, developed by Prof. Ishihara and Prof. Silver. The cyclic shear stresses were controlled by a pneumatic system, applied at frequencies between 0,0001 Hz and 5 Hz. The horizontal shear stresses were applied as stress controlled at the top cap connected to a horizontally moveable shaft going through the cell. The test sample had a 70 mm diameter and a 30 mm height and it could be consolidated under isotropic and anisotropic stresses. The shear stresses were measured by a load cell located in the chamber. The porewater pressure transducer was connected to the bottom platen, axial and horizontal deformations were measured by sensitive displacement transducers located outside of the chamber. In this study a series of cyclic tests were conducted on plastic and non-plastic undisturbed silty soil samples obtained by Shelby tube and piston sampler from borings at the Erzincan Ekşisu region. Special care was given not to disturb the test samples and therefore sampler tubes of height 700-800 mm were divided into small pieces. The water content of carefully taken out samples was determined. Also other index properties (namely the liquid limit, plastic limit, natural unit weight, amount of coarse and fine graded particles) were determined. The soil sample was placed carefully on the cell base after it was brought to cylinders of height 30 and 70 mm. Afterwards a membrane of 0,3 mm thickness was fit onto the sample and the 0-rings were attached. The cell was filled with castor oil after all the connections of the device were completed. This filling up with castor oil was followed by the application of pressure. At this stage the initial height reading was taken on the deformation gauge. Then a 100 kPa confining pressure was applied to the sample. The connections were opened between the back pressure, axial pressure and confining pressure. The axial pressure increased equivalent to the confining pressure because the interconnections were free. The back pressure was slowly step-by-step increased to 400 kPa keeping the effective pressure at 100 kPa. Both of the pressures rose XIII equally to the back pressure to 500 kPa because the back pressure regulator controlled both the confining and the axial pressures. The soil sample was then left to consolidate for 24 hours at these pressure stages. The B-check was done after 24 hours. B value was obtained as 0,95 or greater after 24 hours. The sample was consolidated to the desired effective confining stress and it was cyclically loaded as the stress was controlled at a frequency of 1 Hz. Some of the silty soil samples used in the cyclic tests were non-plastic; some were plastic soils. The water content of non-plastic silty soil samples varied between 38% and 51% and those of plastic silty soil samples, from 36% to 50%. The liquid limit values of plastic silty samples varied between 36% and 60% and plastic limit had a span of 26% to 32%. The liquid limit of the non-plastic silty soils had range of 31% to 51?/o. Also the coarse portion of the non- plastic samples was between 2% to 17% and that among the plastic soil samples between 24% and 38%. The pore water pressures, applied cyclic loads, horizontal and vertical deformation readings were taken by computer during these tests. The applied cyclic load was recorded as T; therefore the force acting on the sample was ±T/2. Consequently, the applied stress to the sample was calculated as ±T/2A assuming that no longitudinal deformations developed and the sample area stayed constant during the testing. The deformation calculations were done in a similar manner and the unit shear strain values, y; obtained by dividing the deformation recordings Al from the computer by the sample height Ho. Pore water pressures were directly read from the data during the test. After the completion of the test these data were recorded with the help of floppy disk and processed for the calculations. It was crucial to carefully enter the necessary commands to the computer before starting the tests. In this study, a series of cyclic tests were conducted using the simple shear testing system. In all the tests the wave shapes were taken as sinusoidal and the effective confining pressure as 100 kPa. The cyclic loads were applied to the specimen until the porewater pressures were equal to the confining pressure or until the porewater were equal to the confining pressure or until the porewater pressure reached steady- state. The shear stress ratios were determined after each cyclic test. Besides, the porewater pressure ratio-time, shear strain-time and shear stress ratio-time relationships were shown after each cyclic tests. Whether the porewater pressure reached the 100 kPa value (whether there was liquefaction or not) was checked. The shear strain ratio was taken as y=±2,5% as the accepted failure criteria under cyclic loads after applying the unit shear stress and the cyclic numbers, N and the porewater pressure ratio, AuJcfc were determined corresponding to this level. XIV Finally, according to the results obtained from the tests, the following factors that affect cyclic strength were investigated in the study: 1- Effect of the cyclic stress ratio 2- Effect of the number of cyclics 3- Effect of plasticity 1. The effect of the cyclic stress ratio Cyclic loads at different stress ratios were applied to saturated high and low plasticity silty soil samples to determine the effect of cyclic stress ratio on cyclic behaviour. The deformation level, which was accepted as the failure criteria y=±2,5% was monitored for three different sets of tests and the cyclic number, and the porewater pressure ratio, Au/oJ. corresponding to this was determined. At soils not reaching the shear strain values at a certain stress level y=±2,5% was observed. For every soil at stresses exceeding the stress level the shear strain was monitored exceeding the y=±2,5% value. As the stress ratio increased the differences in the number of cycles, N also increased. It was found that the porewater pressures and the shear strain increased in proportion to the cyclic stress ratio. 2-The effect of number of cyclics Cyclic loads at different stress ratios were applied as well to saturated high and low plasticity silty soil samples to determine the effect of number of cycles on the cyclic behaviour. At different cyclics (N=l, N=10, N=20 and N=100) evaluatings were made at the same frequency and at a certain shear stress ratio and at the same shear stress ratio, the shear strain value at N=100 cycle obtained was higher than the shear strain value at N=l cycle. Three different sets of tests were conducted as well to determine the effect of the cyclic number on the cyclic strength. The increase in the number of the cycles resulted in an increase in the shear strain values in samples. This caused yield in soils. Certain repeated shear stress or shear strain amplitude had to be reached to obtain such a yield. The ties between the soil particles were then weakened and the strength was decreased due to the relative displacement of the particles. 3- The effect of plasticity Lastly, cyclic tests on two different soil groups of plastic and non-plastic (NP) behaviour at different shear stresses were conducted and the obtained results were compared to determine the effect of plasticity on the cyclic strength of fine graded soils. XV The results were evaluated at y=±\% deformation level as well as at y=±2,5% deformation level. There was a significant difference in their behaviour due to the effect of plasticity although both soils were classified as silt. The cyclic strength of plastic silt was observed to be higher than that of non-plastic (NP) silty soils taking a certain cyclic number into consideration. In this study, aim was to investigate the effects of cyclic shear stress level, the effect of the number of cycles and the effect of plasticity in fine graded soil samples. For every soil at stresses exceeding the stress level the shear strain was monitored exceeding the y=±2,5% value.

It is inevitably necessary to determine the behaviour of cohesive soils while investigating the response of soil layers to the earthquake loads. In such soils there is no total loss of strength as in sands; therefore investigations should be directed toward stress-strain and porewater pressure behaviour. The problems encountered in cohesive soil layers during an earthquake can arise as instability at slopes where the soil surface is inclined. Also the cyclic properties of cohesive soils should be obtained in order to conduct earthquake analyses at soil structures such as an earthfill dam. Determination of the behaviour of soils under cyclic stresses, in other words the cyclic properties, can be divided into two sections: 1- Stress-strain properties 2- Strength properties The term stress-strain properties implies the determination of the cyclic shear modulus and the damping ratio values and their variation with unit strain. The shear stress amplitude that produces collapse or large deformations are used as strength parameters. Cyclic tests were conducted on properly prepared undisturbed soil samples and the factors affecting the cyclic behaviour were investigated. These factors have been thoroughly investigated by many researchers by tests done in the laboratory as well as in the field. Factors affecting the cyclic strength can be classified into two main groups according to their importance: Factors of primary importance: 1- Unit shear strain amplitude, y 2- Average effective confining pressure, oq 3- Void ratio, e 4- Number of cyclics, N 5- Saturation ratio, Sr 6- Overconsolidation ratio, OCR 7- Frequency, f Factors of secondary effect: 1- Testing methods xii 2- Loading type and direction Soils under the effect of cyclic loads are effected by cyclic shear stress at various amplitudes and frequencies. The laboratory tests developed nowadays to determine the strength properties of soils at such loading conditions are as follows: 1- Cyclic simple shear test 2- Cyclic triaxial test 3- Cyclic torsional simple shear 4- Shaking table 5- Resonant column test The cyclic simple shear testing system used in this study was a modified version of the Norwegian simple shear apparatus, developed by Prof. Ishihara and Prof. Silver. The cyclic shear stresses were controlled by a pneumatic system, applied at frequencies between 0,0001 Hz and 5 Hz. The horizontal shear stresses were applied as stress controlled at the top cap connected to a horizontally moveable shaft going through the cell. The test sample had a 70 mm diameter and a 30 mm height and it could be consolidated under isotropic and anisotropic stresses. The shear stresses were measured by a load cell located in the chamber. The porewater pressure transducer was connected to the bottom platen, axial and horizontal deformations were measured by sensitive displacement transducers located outside of the chamber. In this study a series of cyclic tests were conducted on plastic and non-plastic undisturbed silty soil samples obtained by Shelby tube and piston sampler from borings at the Erzincan Ekşisu region. Special care was given not to disturb the test samples and therefore sampler tubes of height 700-800 mm were divided into small pieces. The water content of carefully taken out samples was determined. Also other index properties (namely the liquid limit, plastic limit, natural unit weight, amount of coarse and fine graded particles) were determined. The soil sample was placed carefully on the cell base after it was brought to cylinders of height 30 and 70 mm. Afterwards a membrane of 0,3 mm thickness was fit onto the sample and the 0-rings were attached. The cell was filled with castor oil after all the connections of the device were completed. This filling up with castor oil was followed by the application of pressure. At this stage the initial height reading was taken on the deformation gauge. Then a 100 kPa confining pressure was applied to the sample. The connections were opened between the back pressure, axial pressure and confining pressure. The axial pressure increased equivalent to the confining pressure because the interconnections were free. The back pressure was slowly step-by-step increased to 400 kPa keeping the effective pressure at 100 kPa. Both of the pressures rose XIII equally to the back pressure to 500 kPa because the back pressure regulator controlled both the confining and the axial pressures. The soil sample was then left to consolidate for 24 hours at these pressure stages. The B-check was done after 24 hours. B value was obtained as 0,95 or greater after 24 hours. The sample was consolidated to the desired effective confining stress and it was cyclically loaded as the stress was controlled at a frequency of 1 Hz. Some of the silty soil samples used in the cyclic tests were non-plastic; some were plastic soils. The water content of non-plastic silty soil samples varied between 38% and 51% and those of plastic silty soil samples, from 36% to 50%. The liquid limit values of plastic silty samples varied between 36% and 60% and plastic limit had a span of 26% to 32%. The liquid limit of the non-plastic silty soils had range of 31% to 51?/o. Also the coarse portion of the non- plastic samples was between 2% to 17% and that among the plastic soil samples between 24% and 38%. The pore water pressures, applied cyclic loads, horizontal and vertical deformation readings were taken by computer during these tests. The applied cyclic load was recorded as T; therefore the force acting on the sample was ±T/2. Consequently, the applied stress to the sample was calculated as ±T/2A assuming that no longitudinal deformations developed and the sample area stayed constant during the testing. The deformation calculations were done in a similar manner and the unit shear strain values, y; obtained by dividing the deformation recordings Al from the computer by the sample height Ho. Pore water pressures were directly read from the data during the test. After the completion of the test these data were recorded with the help of floppy disk and processed for the calculations. It was crucial to carefully enter the necessary commands to the computer before starting the tests. In this study, a series of cyclic tests were conducted using the simple shear testing system. In all the tests the wave shapes were taken as sinusoidal and the effective confining pressure as 100 kPa. The cyclic loads were applied to the specimen until the porewater pressures were equal to the confining pressure or until the porewater were equal to the confining pressure or until the porewater pressure reached steady- state. The shear stress ratios were determined after each cyclic test. Besides, the porewater pressure ratio-time, shear strain-time and shear stress ratio-time relationships were shown after each cyclic tests. Whether the porewater pressure reached the 100 kPa value (whether there was liquefaction or not) was checked. The shear strain ratio was taken as y=±2,5% as the accepted failure criteria under cyclic loads after applying the unit shear stress and the cyclic numbers, N and the porewater pressure ratio, AuJcfc were determined corresponding to this level. XIV Finally, according to the results obtained from the tests, the following factors that affect cyclic strength were investigated in the study: 1- Effect of the cyclic stress ratio 2- Effect of the number of cyclics 3- Effect of plasticity 1. The effect of the cyclic stress ratio Cyclic loads at different stress ratios were applied to saturated high and low plasticity silty soil samples to determine the effect of cyclic stress ratio on cyclic behaviour. The deformation level, which was accepted as the failure criteria y=±2,5% was monitored for three different sets of tests and the cyclic number, and the porewater pressure ratio, Au/oJ. corresponding to this was determined. At soils not reaching the shear strain values at a certain stress level y=±2,5% was observed. For every soil at stresses exceeding the stress level the shear strain was monitored exceeding the y=±2,5% value. As the stress ratio increased the differences in the number of cycles, N also increased. It was found that the porewater pressures and the shear strain increased in proportion to the cyclic stress ratio. 2-The effect of number of cyclics Cyclic loads at different stress ratios were applied as well to saturated high and low plasticity silty soil samples to determine the effect of number of cycles on the cyclic behaviour. At different cyclics (N=l, N=10, N=20 and N=100) evaluatings were made at the same frequency and at a certain shear stress ratio and at the same shear stress ratio, the shear strain value at N=100 cycle obtained was higher than the shear strain value at N=l cycle. Three different sets of tests were conducted as well to determine the effect of the cyclic number on the cyclic strength. The increase in the number of the cycles resulted in an increase in the shear strain values in samples. This caused yield in soils. Certain repeated shear stress or shear strain amplitude had to be reached to obtain such a yield. The ties between the soil particles were then weakened and the strength was decreased due to the relative displacement of the particles. 3- The effect of plasticity Lastly, cyclic tests on two different soil groups of plastic and non-plastic (NP) behaviour at different shear stresses were conducted and the obtained results were compared to determine the effect of plasticity on the cyclic strength of fine graded soils. XV The results were evaluated at y=±\% deformation level as well as at y=±2,5% deformation level. There was a significant difference in their behaviour due to the effect of plasticity although both soils were classified as silt. The cyclic strength of plastic silt was observed to be higher than that of non-plastic (NP) silty soils taking a certain cyclic number into consideration. In this study, aim was to investigate the effects of cyclic shear stress level, the effect of the number of cycles and the effect of plasticity in fine graded soil samples. For every soil at stresses exceeding the stress level the shear strain was monitored exceeding the y=±2,5% value.

##### Açıklama

Tez (Yüksek Lisans) -- İstanbul Teknik Üniversitesi, Fen Bilimleri Enstitüsü, 1995

##### Anahtar kelimeler

Dinamik davranış,
Zemin,
Dynamic behavior,
Soil