|M.Sc Student||Bar Alexander|
|Subject||Prediction of Pile Behavior Using a Rehological Model|
|Department||Department of Civil and Environmental Engineering||Supervisors||Professor Emeritus Sam Frydman|
|Professor Shmuel Paikowsky|
|Full Thesis text - in Hebrew|
Use of pile foundations is a very common, and sometimes the only option for building foundations. Their use requires the ability to predict the relation between the load and the settlement of the pile, sometimes as a function of time.
There are essentially two groups of methods for pile behavior determination under a vertical load: analytical and empirical methods and interpretation of the results of load tests. Because of high price and duration of load tests, it is desirable to use a rheological model which will allow the prediction of pile behavior based on existing data base of model parameters and in a cheapest way and in a short time.
In this work the rheological model first suggested by Operstein (1986) has been further developed. The model consists of a combination of perfectly elastic linear springs, perfectly viscous elements and yield stress elements. The model of Operstein is based on interpretation of static- cyclic loading tests and allows description of load- settlement- time relations for vertically loaded and unloaded piles. In this work, it was demonstrated that the proposed rheological model can also be used for description of conventional static pile loading methods.
For determination of model parameters, 44 conventional loading tests on piles at two sites in Canada and USA were analyzed in this work.
The tests from both sites were analyzed using pile-top movement equations developed in this research.
In order to develop a relationship between the model parameters and the soil and pile types, the stiffness parameters were normalized by pile stiffness, and the viscous parameters by the friction area between pile and soil. Relations between the model parameters and the stiffness of the pile or the area of side surface of the piles are presented.
Because of the limited number of soil types at the test sites studied in the work, actual relations between the model parameters and soil type were not found. In spite of this, it was found out that the absolute values and variation of model parameters in clay soils are greater than in sand or silt soils.
It was found out that the model parameters may not be constant during loading. The time periods between measurements of pile top movement during conventional loading tests must be as short as possible, allowing collection of adequate information for successful analyses using the proposed model, and determination of model parameters.