Predicted and Measured Response of Precast Concrete Piles under Lateral Load
Assistant Professor, Department of Civil and Environmental Engineering, Southern Illinois University Carbondale, Carbondale, Illinois
Professor, Department of Civil and Environmental Engineering, Southern Illinois University Carbondale, Carbondale, Illinois
Precast concrete piles are often used to support heavy structures. In addition to vertical compressive loads, these piles are subjected to significant lateral loads. Therefore, prediction of lateral load-deflection response of precast concrete piles is of significant interest to structural and geotechnical engineers. Prediction of lateral behavior of pile foundation is complicated due to the fact that the soil reaction is dependent on the pile movement and the pile movement is dependent on the soil response. Non-linear behavior of soils and concrete makes the problem even more complicated. Several methods and computer programs are available to analyze the piles under lateral loads. This paper presents the results of predictions made on four, precast reinforced concrete piles which were constructed using coal combustion products. Predictions are compared with the measured data on all four piles. All piles analyzed are 12x12 inch square piles. The results presented show that assumption of completely uncracked section or cracked section may not be realistic.
Keywords: Precast Concrete, Pile Foundations, Nonlinear Behavior, Lateral Load Prediction
Pile foundations of various types, sizes, shapes, and materials are part of every other structure in the world. Piles made of pre-cast reinforced concrete and pre-stressed concrete have also gained popularity due to several advantages. Pile foundations are often subject to both axial and lateral loads. Piles that sustain lateral loads of significant magnitude occur in offshore structures, waterfront structures, bridges, buildings, industrial plants, locks and dams, and retaining walls. Piles used to stabilize slopes are also subjected to lateral loading. The lateral loads on piles are derived from earth pressures, wind pressures, current forces from flowing water, earthquakes, impact loads from barges and other vessels, hurricanes, and moving vehicles. Even if the above loads are not present, lateral load on piles can result from eccentric application of vertical load (Kumar et al., 2005).
It is a well known fact that concrete is very strong in compression compared to soil and therefore, under axial compression loads on concrete piles, soils reach their ultimate capacity much before the concrete. However, due to weak tensile and flexural strength of concrete, lateral load-deflection behavior of concrete piles depends significantly on material properties of the piles. The most precise approach to determine the capacity of pile foundations is to install a full-size prototype piles at the site and load it to failure (Coduto, 1994). However, because of the cost to perform field tests on full-size piles, these tests are not performed routinely. Therefore, behavior of pile foundations is predicted using load test data and design methods available in literature.
Prediction of lateral behavior of pile foundations is complicated due to the fact that the soil reaction is dependent on the pile movement and the pile movement is dependent on the soil response. The fact that the behavior of soils is nonlinear even at small strains makes the problem even more complicated. In order to incorporate the nonlinear behavior of soil and interaction between a pile and soil for prediction of lateral-load deflection response of piles, a discrete model is commonly used (Reese et al. 2000).
Many of the developing and developed countries use coal to generate energy. Use of coal in generation of electricity has resulted in production and accumulation of large quantities of coal combustion products (CCPs). Within the past 50 years, the concrete industry has given special attention to the safe and economical utilization of these residues (Helmuth 1987). Current research on the beneficial use of CCPs as building and highway construction materials has identified several promising uses for these materials. In addition to use in concrete, CCPs have been successfully used in the agricultural industry, blasting grit and roofing material, cement clinker raw feed, flowable fill, grout, mineral filler, mining applications, snow and ice control, wallboard, roller compacted concrete, structural fill, embankments, and soil stabilization (Kalyoncu, 2003). Several case histories of utilization of coal combustion products in construction projects are available (ACAA, 2001; GAI, 1988; Golden, 1986; Hosin, 2001; Korcak, 1998; Kumar et al., 2004, 2001; Kumar and Stewart, 2003a and b; Kumar and Vaddu, 2003 and 2004; Lovell et al. 1997; Naik et. al., 1997; Ng, 2001; Schroeder, 1994; Seals et al. 1972; Tikalsky and Carrasquillo, 1989). Use of coal combustions products in the construction of reinforced precast or prestressed concrete piles is still very limited, if any.
Kumar et al. (2005) conducted four full size pile load tests on precast reinforced concrete piles. The piles were made using coal combustion by-products. They showed that the field performance of the piles constructed using concrete composites containing bottom ash would be similar to that of piles constructed using an equivalent conventional concrete, provided they are constructed, installed, and tested in a similar manner. This paper presents the results of lateral load-deflection response predictions made and the data measured from the full-scale tests.
CHARACTERISTICS OF THE PILES ANALYZED
All piles analyzed in this study were square piles 12 x 12 inches in cross-section and were 21 feet long. Three piles were made from coal combustion products and one pile was made from an equivalent conventional concrete. Table 1 shows the compressive strength of the concrete composites and an equivalent conventional concrete at various curing ages. All piles were appropriately reinforced as shown in Figure 1. Piles were driven at the site located in the Carterville, Illinois campus of SIUC. The site was selected due to its proximity to the CCP research lab. A single-acting diesel hammer, Delmag D15, with a rated energy of approximately 27,100 ft-lbs was used to drive the piles. The piles were driven into the ground till it encountered the specified resistance. All piles, except PB_RP50, were equipped with inclinometer pile to measure deflections along the pile length. Pile PB_RP50 was used as reaction pile and, therefore, only head deflection was measured for this pile.
Table 1. Compressive Strength of the Concretes used for Piles Analyzed
Figure 1. Shape and reinforcement details of the piles tested
The soil stratigraphy at the site consisted of medium stiff to stiff, brown silty clay to depths of approximately 21 feet. The silty clay is underlain by very stiff to hard, sandy clay shale to the maximum depths explored, i.e., 25.5 feet. The results of field and laboratory testing indicated that the compressive strength of the silty clay generally ranged between 0.75 and 1.75 ton per square foot. The moisture content of the silty clay generally varied between 15 and 29 percent. Groundwater at the site was measured by installing a piezometer. The groundwater table was observed to be at depths between 4 and 6 ft below the ground surface. An artificial pond existed within 100 feet of the test site. The water level in the pond was observed to be 4 feet below the bank surface which generally corresponded with the groundwater level at the test site.
MEASURED LOAD-DEFLECTION RESPONSE
Figure 2 shows the results of lateral load tests on piles analyzed. The piles were tested in two groups. Reaction was provided by either a reaction pile (a pile without instrumentation and inclinometer pipe) or by another test pile. Testing continued until the pre-determined maximum load was reached. The load was subsequently removed in four decrements, according to the standard loading schedule. For both tests, the load was applied using a 450-kip O-cell (Hayes et al. 2004) and the loads were measured using pressure-load calibration of the hydraulic jack. A vibrating wire load cell was also used as a check on the applied load.
It is clear from Figure 2 that the lateral load – deflection response of all piles in nonlinear. Figure 2 also shows that the performance of pile made with concrete composite PB100 is similar to that of pile made with an equivalent concrete CM whereas, the performance of the pile made with concrete composite PB50 appears to be better than the performance of piles made with concrete composite PB100 and an equivalent conventional concrete CM. However, Kumar et al. (2005) concluded that the difference in the response was due to the way piles were installed. They also concluded that if all the piles are constructed, installed, and tested in a similar manner, the performance of piles constructed with concrete composites made with PCC bottom ash and an equivalent conventional concrete are expected to perform in a similar manner.
Figure 2. Measured load-deflection response of four piles analyzed
The data from the pile load tests were analyzed by using subgrade reaction approach and the theoretical expressions and non-dimensional coefficients recommended by Reese and Matlock (1956) and Matlock and Reese (1960). Subgrade reaction approach treats a laterally loaded pile as a beam on elastic foundation. It is assumed that the beam is supported by a Winkler soil model according to which the soil is replaced by a series of infinitely closely spaced independent and elastic springs. For more information on this topic refer to any textbook on pile foundations, e.g., Prakash and Sharma (1990).
According to the subgrade reaction approach and non-dimensional coefficients suggested by Matlock and Reese (1960), when a load is applied at the ground surface, the deflection in the free-head pile at the ground surface can be determined by using (1).
Qg = Load at the ground surface
T = Relative stiffness factor
E = Modulus of elasticity of the pile material
I = Moment of inertia of the pile.
Since the load and displacement during the tests were measured and EI of the piles are known, the relative stiffness factor, T, can be calculated from (1) as follows:
Also, the relative stiffness factor can be expressed in terms of coefficient of subgrade reaction, nh , as,
Values of the coefficient of subgrade reaction, nh, can be computed for the test piles from the following equation using load and displacement data from the test piles.
To predict the lateral-load deflection of piles in this study, a computer program Lpile Plus was used. Lpile plus program is a special purpose program based on rational procedures for analyzing piles or drilled shafts under lateral loads (Reese et al., 2000). This program is capable of computing deflection, shear, bending moments, and soil response considering nonlinear behavior of soils. Soil behavior is modeled with p-y curves. In this study, the option of generating the p-y curves internally was used.
RESULTS AND DISCUSSION
The lateral load-deflection response of a particular pile is dependent of its material properties. For concrete piles, pile stiffness, EI, is a function of applied loading. Since concrete is weak in tension and flexure, there is a possibility of cracking in the pile due to application of lateral load which can cause reduction in pile stiffness. In most of the analysis the pile stiffness is assumed to be constant throughout the loading range, i.e. it is assumed that the pile section remains uncracked throughout the loading range.
Figure 3 shows the measured and predicted response of a precast reinforced concrete pile made with concrete composite PB50. In predicting the behavior, it was assumed that the flexural stiffness of the piles did not change due to application of the lateral load (Type 1_Uncracked Section). It is clear from Figure 3 that the predicted response is much higher than the measured response. Similar observation was made from the response of other piles. Further analysis was done by considering nonlinear variation in the flexural stiffness of pile (Type 3_Cracked Section). This is modeled in the program by calculating the ultimate bending moment and corresponding flexural stiffness of the pile first and then calculating the response by using the nonlinear stiffness of the pile. Figure 4 shows the measured and predicted response of the same pile, PB50, considering the nonlinear variation of flexural stiffness of the pile. Response predicted for uncracked section is also presented on the same figure. Similar responses of other piles analyzed are presented in Figures 5 through 7.
Predicted response shown in Figures 4 through 7 for precast concrete piles made from concrete composites and equivalent conventional concrete show that the measured response was somewhere in between the responses predicted by assuming uncracked section and cracked section, i.e. considering nonlinear variation of flexural stiffness of piles. Therefore, it was concluded that assumption of completely uncracked section of cracked section may not be appropriate for all types of reinforced precast concrete piles.
Figure 3. Predicted and measured response of precast concrete pile made with concrete composite PB50 assuming uncracked section
Figure 4. Predicted and measured response of precast concrete pile made with concrete composite PB50 assuming uncracked and cracked sections
Figure 5. Predicted and measured response of precast concrete pile made with concrete composite PB_RP50 assuming uncracked and cracked sections
Figure 6. Predicted and measured response of precast concrete pile made with concrete composite PB100 assuming uncracked and cracked sections
Figure 7. Predicted and measured response of precast concrete pile made with concrete composite CM assuming uncracked and cracked sections
Results of lateral-load deflection response predictions made on four, precast reinforced concrete piles, constructed using coal combustion products and conventional concrete and tested in the field, are presented. The predictions were made assuming the uncracked sections of the piles and nonlinear variation of flexural stiffness, i.e., cracked sections. Predictions are compared with the measured data on all four piles. All piles analyzed are 12x12 inch square piles. Measured response of all four piles was similar to each other indicating that utilization of coal combustion products did not have significant effect on the lateral-load response of piles. Results of predictions show that assumption of completely uncracked section or cracked section may not predict the response realistically.
|© 2005 ejge|