Development Of A Model For Estimation Of Buried Large Diameter Thin-walled Steel Pipe Deflection Due To External Loads

Abstract

Design of buried pipeline systems involves solution of geotechnical and structural problems in addition to the hydraulics and mechanical issues. Just like any buried structure, it is of utmost importance to understand how the pipe interacts with the soil when subjected to external and internal loads. Based on the mode of withstanding loads, pipes are classified into two major categories, which are rigid and flexible pipes. Pipe material is the major factor governing the classification of a pipe being rigid or flexible. Rigid pipe is a pipe which is designed to withstand external dead and live loads and internal pressure loads without deformation. Flexible pipe on the other hand is designed with allowance to deform within a specified limit depending upon the pipe material and type of coatings and linings on the pipe. Designs of flexible pipes are generally based on hydraulic criteria of the pipeline, also known as Hydraulic Design Basis (HDB). Side soil column plays a pivotal role in flexible pipe's ability to withstand external loads.Pipe diameters and pipe wall thicknesses of flexible pipes are usually designed as per hydraulic requirements, such as, flow capacity, internal fluid pressure, pipe material strength and elasticity, and so on. Analysis of flexible pipe for response to external loads is commonly carried out with proper embedment rather than to increase pipe structural capacity. This approach is rightly adopted because it is much more economical to provide good embedment rather than increasing stiffness of the pipe with increased thickness. Most common methods for flexible pipe analyses to predict pipe deflecions include the Modified Iowa and the Bureau of Reclamation equations.The Modified Iowa formula and the Bureau of Reclamation equations are semi-empirical methods to predict flexible pipe deflections. The pipe material properties used in these equations are engineering properties. However, the Modulus of soil reaction (E') which is a key property in determining the predicted long term deflection of pipe is an empirical value.One of the key assumptions in Spangler's (1941) soil pipe interaction model is that the passive soil resistances offered by embedment soil above and below the pipe springline are symmetric. This assumption is addressed in this dissertation, especially for the case of large diameter pipes. It is a widely accepted principle in geotechnical engineering that lateral pressure (active, at-rest or passive) from soil is dependent on depth, with deeper soils with higher lateral forces potential due to greater overburden pressures and also in cases where two different embedment materials are used. The Spangler's model does not consider peaking behavior (increase of vertical diameter) of pipe during embedment construction. There is a need to develop a model to predict pipe behavior due to embedment construction. This model needs to consider the cycle that embedment soil goes through from at-rest conditions (at the time of placement of layer), to active conditions (during peaking deflection), and finally to passive conditions (due to deflection of pipe).The objectives of this research are to consider engineering properties of embedment soils in analysis of flexible pipe-soil system for external load conditions and develop a new model for prediction of deflection of flexible steel pipe. Full scale laboratory tests were perfomed to develop the new model and finite element models were analysed to validate the test results. In this research, finite element method was effectively used to model the soil pipe interaction for five full scale laboratory tests conducted on a steel pipe. Such models can be used for analysis of flexible pipe embedment design for layered embedment conditions. The results of finite element analysis showed that the squaring of the pipe occurs when haunch soil is weak compared to the side column. Another critical observations made during the tests were stresses at the bottom of pipe and bedding angle. It is desirable that the stress due to surcharge load on top of the pipe, weight of the pipe, and water inside the pipe be distributed uniformly across width of the bedding.Best results against peaking deflection were obtained with crushed limestone (Test 3) due to lesser lateral earth pressure coefficient and lesser energy required for compaction. Perhaps, that is the reason why peaking deflections in flexible pipe have not been studied extensively in the past. However, if clayey materials are considered, peaking deflections need to be examined closely.Best results against deflection due to surcharge load were obtained in Test 4 with mixed embedment of crushed limestone and native clay. This was the only case when horizontal deflection due to surcharge load was observed to be approximately equal to vertical deflection in magnitude. This only echoes the importance of haunch area in behavior of pipe. The haunch area consisted of flow-able crushed limestone which was also subjected to compaction energy from compaction of clay embedment above 0.3 diameter. Also, the bedding angle for Test 4 was highest of all tests. The stress at top of pipe was well distributed along the bedding of pipe which is a favorable condition for integrity of bedding.