The Effect of Pile-Head Boundary Conditions in Liquefiable Soil

*Pınar Sezin Öztürk Kardoğan and Nihat Sinan Işık* 

#### **Abstract**

The soil may lose its strength due to various reasons. Especially in saturated sandy soils, because of the increase in pore water pressure under dynamic loads, the soil loses its strength and behaves as a liquid. Liquefaction phenomenon, along with the occurrence of large soil displacements, can cause lateral spreading in sloping ground. These displacements cause some damage on pile foundations. It is still hard to determine the behavior of laterally loaded piles under dynamic loading with definite judgments. For this reason, some numerical and experimental methods are used. In this study, some centrifuge test results are compared with those obtained numerically with a beam on nonlinear Winkler foundation (BNWF) model.

**Keywords:** pile, pile-soil interaction, pile-head boundary conditions, soil liquefaction, BNWF

#### **1. Introduction**

The increase of pore water pressure during an earthquake triggers the liquefaction phenomenon, which in turn in the case of sloping ground conditions may trigger the lateral spreading phenomenon. Lateral spreading due to soil liquefaction in steep slopes causes the soil to lose its stiffness by being exposed to large lateral displacements. Both lateral spreading and liquefaction phenomenon occur due to the reduction of strength and stiffness of the soil. Therefore, in case of soil liquefaction and lateral spreading, large stresses may occur on pile foundation, resulting in considerable damage to the pile. In Turkey and in other countries severely exposed to large earthquakes, many studies have been carried out on pile damages due to liquefaction and lateral spreading, and their effects on the superstructure behavior have been evaluated. The 1964 Niigata (Mw = 7.5), 1964 Alaska (Mw = 9.2), 1971 San Fernando (Mw = 6.4), and 1995 Hyogoken-Nambu (Mw = 7.2) earthquakes caused liquefaction and lateral spreading and in literature is reported that piles damage due to these phenomena occurred. A semi-analytical analysis method was presented in [1] to capture the dynamic response of vertical floating pile groups. Numerical results suggested that this method gave better results than the finite element method [1]. In the past, the behavior of piles under seismic loads was studied with numerical methods (i.e., p-y curves and finite element method) [2], and the results obtained were compared with centrifuge test experimental data [3–6]. Other researchers have performed centrifuge tests to better understand the performance, in a liquefiable deposit, of piles beneath the bridge piers and then have attempted to reproduce these tests with the help of numerical tools based on the finite difference

method [7]. Especially in the case of pile foundation in liquefiable soil, bridge piers and structures were damaged due to large displacements caused by liquefaction and lateral spreading [8–12].

Particularly, static and dynamic analyses were used to determine the performance of an old bridge resting on a liquefiable soil deposit, where lateral spreading was observed, and it was found that the presence of a crust layer above a liquefiable soil can affect the pile foundation analysis results [13, 14]. In this work, a single pile subjected to liquefaction and lateral spreading effects is studied with a numerical model, and the results are compared with those obtained in centrifuge tests.

#### **2. Materials and methods**

#### **2.1 p-y curves for sandy soil in LPILE code**

 A proper assessment of lateral soil resistances is very important in case of deep foundation in liquefiable soils. Although reasonable and generally accepted methods have been developed to define p-y curves for unliquefied soils, a significant uncertainty remains about the evaluation of the lateral resistance that can be achieved in liquefied sands [15]. Sometimes liquefiable soils are thought to have no strength. In the code LPILE [16], soil strength reduction due to liquefaction is commonly introduced by applying reasonable reduction factors to the p-y curves or by entering a very low value for the angle of friction of the sand. Rollins et al. [17] developed p-y curves based on the results of load tests carried out on pile groups in liquefiable sandy soil. The p-y curves found in [17] revealed a concave upward shape as shown in **Figure 1**.

This shape is shown to result firstly from dilative behavior during shearing, although blank effects may also contribute to the monitored load transfer reply.

In addition, in [17] it was shown that these p-y curves increase their stiffness with depth. With increasing depth, a small displacement is required to develop significant resistance, and the rate at which resistance develops as a function of lateral pile displacement also increases. Rollins et al.'s model is shown to give results close to the actual liquefiable sandy soil behavior.

Reese et al. [18] developed p-y curves for laterally loaded piles, both under short-term static loads and under cyclic loads. In the latter work, some formulas were defined for the calculation of p-y curves of sandy materials under static and cyclic loads (**Figure 2**).

**Figure 1.**  *p-y curve in liquefiable sand as proposed in [17].* 

*The Effect of Pile-Head Boundary Conditions in Liquefiable Soil DOI: http://dx.doi.org/10.5772/intechopen.87836* 

#### **Figure 2.**

 *Characteristic shape of a set of p-y curves for static and cyclic loading in sand [18] (p = soil reaction; y = lateral displacement; b = pile diameter; x = depth).* 

#### **3. Case study**

Abdoun and Wang [19] carried out centrifuge test on a soil profile (**Figure 3**) consisting of an unliquefiable slightly cemented sand layer in the first 2 meters, a liquefiable loose Nevada sand layer (relative density, DR = 40%) in the next 6 meters, and a slightly cemented sand in the bottom layer thick 2 meters (dimensions above are in prototype scale). The centrifuge model was subjected to a centrifugal acceleration of 50 g.

 The soil in the centrifuge model was completely saturated with water. The model was adjusted to reach an angle of 4.8 degrees from the horizontal surface, thus allowing liquefaction-induced lateral spreading to occur during the experiments. Variable horizontal accelerations were applied at the base of the model at a frequency of 2 Hz in the form of a sine wave with a maximum value of 0.25 g. The diameter (D) and the flexural rigidity (EI) of the prototype pile used in these centrifuge tests are equal to 60 cm and 8000 kNm<sup>2</sup> , respectively. The box used for the centrifuge test is a laminar box with a size of 2 × 2.5 × 0.5 m (model scale).

**Figure 3.**  *Centrifuge model in prototype scale used for the verification analysis [19].* 

In the tests, the pile head was fixed against the rotation. The maximum bending moments reached during the tests at the upper and lower boundaries of the liquefiable layer were 270 and 305 kNm, respectively. It was also measured that the free-field ground surface moved 70 cm laterally, while the lateral displacement of the pile head was 85 cm.

#### **3.1 Analysis results using LPILE (beam on nonlinear Winkler foundation)**

The analyses simulating the centrifuge experiment were performed using LPILE, which is a computer code based on the BNWF model. The pile section properties (D and EI) were taken as reported above. A free-field ground surface displacement of 70 cm, which is the free-field ground surface displacement measured in the centrifuge test, was applied to the pile. A soil displacement of 70 cm was applied throughout the layer thickness in the unliquefiable upper layer, while a soil displacement profile, as shown in **Figure 4**, was applied in the liquefiable layer. The latter profile follows a cosine function.

The analyses were carried out considering the pile head both fixed against the rotation and free to rotate to evaluate the influence of the boundary condition at the pile head. p-y curves developed by Reese et al. [18] were applied at the top and bottom unliquefiable layers. The intermediate liquefiable layer was modeled in the following ways:


 When the p-y curve proposed by Rollins et al. [17] was adopted to model the liquefiable layer, the resulting bending moment at the transition zone (boundary line between liquefied and unliquefied soils) and the pile-head displacement were 265 kNm and 86.5 cm, respectively, in the case of fixed-head boundary conditions (also the centrifuge model pile was fixed-head).

In the case of free-head boundary condition, the pile bending at the transition zone and the pile-head displacement were 257 kNm and 101.5 cm, respectively. The pile bending and pile displacement profiles obtained when using the p-y curve proposed by Rollins et al. [17] are given in **Figure 5**.

When the p-y curve by Reese et al. [18] was used (to which a p-multiplier equal to 1/50 was applied to model the liquefiable soil), the pile bending at the transition

**Figure 4.**  *Model created with LPILE [16].* 

**Figure 5.** 

*Analysis results when using the p-y curve proposed by Rollins et al. [17]. (a) Pile bending vs. depth. (b) Pile deflection vs. depth.* 

zone and the pile-head displacement were 200 kNm and 33.5 cm, respectively, in the case of fixed-head conditions. In the case of free-head boundary condition, the pile bending at the transition zone and the pile-head displacement were 300 kNm and 90.5 cm, respectively. The results obtained using the p-y curve by Reese et al. [18] are shown in **Figure 6** in terms of pile bending and pile deflection profiles.

LPILE analysis results were compared to those obtained experimentally in centrifuge. It was found that the pile bending and the pile displacements inferred

#### **Figure 6.**

*Analysis results when using the p-y curve proposed by Reese et al. [18] scaled with a p-multiplier equal to 1/50. (a) Pile bending vs. depth. (b) Pile deflection vs. depth.* 

#### *ISBS 2019 - 4th International Sustainable Buildings Symposium*

by using the p-y curve developed by Rollins et al. [17] are closer to those measured in centrifuge than those inferred by using the p-curve by Reese et al. [18]. Nevertheless, especially in the case of pile buckling assessments, the use of p-y curve by Rollins et al. may lead to numerical stability problems. In fact, Ashford et al. [20] reported that convex p-y curves may cause numerical solution problems in some cases. The analysis results are summarized in **Table 1**.


#### **Table 1.**

*Comparison between LPILE and centrifuge test results.* 

#### **4. Conclusion**

In this work, the response of a single pile subjected to lateral spreading was studied by using a beam on nonlinear Winkler foundation (BNWF) program. Single-pile analyses were carried out considering both free-head and fixed-head boundary conditions. The analysis results were compared to centrifuge test data retrieved in literature.

To model the liquefiable soil layer, the p-y curves proposed by Rollins et al. [17] and by Reese et al. [18] were used. In the case of the p-y curve by Reese et al. [18], originally developed to model sandy soils under static/cyclic lateral loading, a p-multiplier set equal to 1/50 was applied to reproduce the stiffness and the strength reduction in the liquefiable layer.

 The analysis results by using the p-y curve developed by Rollins et al. [17] were successful in reproducing both the pile-head displacement and the pile bending at the transition zone depth obtained in centrifuge, when considering the pile head fixed against the rotation. On the contrary, the analysis results by using the p-y curve proposed by Reese et al. [18] were found to be significantly different compared to the experimental data. Thus, it was found that the p-y curve developed by Rollins et al. [17] to model liquefiable sands gave more reliable results than Reese et al. [18] springs.

*The Effect of Pile-Head Boundary Conditions in Liquefiable Soil DOI: http://dx.doi.org/10.5772/intechopen.87836* 

#### **Author details**

Pınar Sezin Öztürk Kardoğan\* and Nihat Sinan Işık Department of Civil Engineering, Gazi University, Ankara, Turkey

\*Address all correspondence to: sezinozturk@gazi.edu.tr

© 2019 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/ by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

### **References**

[1] Ettouney MM, Brennan JA, Forte MF. Dynamic behavior of pile groups. Journal of Geotechnical Engineering. 1983;**109**(3):301. DOI: 10.1061/ (ASCE)0733-9410(1983)109:3(301)

[2] Boulanger RW, Curras CJ, Kutter BL, Wilson DW, Abghari A. Seismic soilpile-structure interaction experiments and analyses. Journal of Geotechnical and Geoenvironmental Engineering. 1999;**125**(9):750-759. DOI: 10.1061/ (ASCE)1090-0241(1999)125:9(750)

 [3] Kimura M, Zhang F. Seismic evaluations of pile foundations with three different methods based on three-dimensional elasto-plastic finite element analysis. Soils and Foundations, The Japanese Geotechnical Society. 2000;**40**(5):113-132. DOI: 10.3208/ sandf.40.5\_113

[4] Brandenberg SJ, Boulanger RW, Kutter BL, Chang D. Static pushover analyses of pile groups in liquefied and laterally spreading ground in centrifuge tests. Journal of Geotechnical and Geoenvironmental Engineering. 2007a;**133**(9):1055-1066. DOI: 10.1061/ (ASCE)1090-0241(2007)133:9(1055)

[5] Brandenberg SJ, Boulanger RW, Kutter BL, Chang D. Liquefactioninduced softening of load transfer between pile groups and laterally spreading crusts. Journal of Geotechnical and Geoenvironmental Engineering. 2007b;**133**(1):91-103. DOI: 10.1061/ (ASCE)1090-0241(2007)133:1(91)

[6] Tasiopoulou P, Gerolymos N, Tazoh T, Gazetas G. Pile-group response to large soil displacements and liquefaction: Centrifuge experiments versus a physically simplified analysis. Journal of Geotechnical and Geoenvironmental Engineering.

2013;**139**(2):223-233. DOI: 10.1061/ (ASCE)GT.1943-5606.0000759

[7] Armstrong RJ, Boulanger RW, Beaty MH. Liquefaction effects on piled bridge abutments: Centrifuge tests and numerical analyses. Journal of Geotechnical and Geoenvironmental Engineering. 2013;**139**(3):433-443. DOI: 10.1061/(ASCE)GT.1943-5606.0000780

[8] Hwang J, Kim C, Chun C, Kim M. Viscous fluid characteristics of liquefied soils and behavior of piles subjected to flow of liquefied soils. Soil Dynamics and Earthquake Engineering. 2006;**26**(2-4):313-323. DOI: 10.1016/j. soildyn.2005.02.020

[9] Ledezma C, Bray JD. Factors that affect the performance of bridge foundations undergoing liquefaction-ınduced lateral spreading. In: Geotechnical Earthquake and Engineering and Soil Dynamics IV Congress; 18-22 May 2008; Sacramento, California. 2008. DOI: 10.1061/40975(318)161

[10] Shin H, Ardumio P, Kramer SL, Mackie K. Seismic Response of a Typical Highway Bridge in Liquefiable Soil. Geotechnical Earthquake and Engineering and Soil Dynamics IV Congress; 18-22 May 2008; Sacramento, California. 2008. DOI: 10.1061/40975(318)164

[11] Li Q , Zhang X, Yang Z. Analysis of Laterally Loaded Piles in Liquefiable Soils with a Frozen Crust. Oakland, California, United States: Geo Congress; 2012. pp. 25-29. DOI: 10.1061/9780784412121.022

[12] Turner BJ, Brandenberg SJ, Stewart JP. Case study of parallel bridges affected by liquefaction and lateral spreading. Journal of Geotechnical and Geoenvironmental Engineering.

*The Effect of Pile-Head Boundary Conditions in Liquefiable Soil DOI: http://dx.doi.org/10.5772/intechopen.87836* 

2016;**142**(7):1-12. DOI: 10.1061/(ASCE) GT.1943-5606.0001480

[13] Brandenberg SJ, Kashighandi P, Zhang J, Huo Y, Zhao M. Sensitivity Study of an Older-Vintage Bridge Subjected to Lateral Spreading. Geotechnical Earthquake and Engineering and Soil Dynamics IV Congress, 18-22 May 2008; Sacramento, California. DOI: 10.1061/40975(318)165

[14] Knappett JA, Mohammadi S, Griffin C. Lateral spreading forces on bridge piers and pile caps in laterally spreading soil: Effect of angle of incidence. Journal of Geotechnical and Geoenvironmental Engineering. 2010;**136**(12):1589-1599. DOI: 10.1061/ (ASCE)GT.1943-5606.0000387

[15] Isenhower WM, Wang S. Technical Manual for Lpile, Version 6. Austin: Ensoft Inc; 2011. p. 101

 [16] Wang S-T, Isenhower W. LPile A Program for the Analysis of Deep Foundations Under Lateral Loading. Austin, Texas: Ensoft Inc.; 2010. p. 87

[17] Rollins KM, Hales LJ, Ashford SA. P-y Curves for Large Diameter Shafts in Liquefied Sands from Blast Liquefaction Tests. Workshop on Seismic Performance and Simulation of Pile Foundations in Liquefied and Laterally Spreading Ground. 26-18 March 2005. United States: University of California; 2005. pp. 11-23. DOI: 10.1061/40822(184)2

[18] Reese LC, Cox WR, Koop FD. Analysis of Laterally Loaded Piles in Sand. In: Proceedings 6th Offshore Technology Conference, II, 6-8 May 1974 Houston, Texas; 1974. pp. 473-484. DOI: 10.4043/2080-MS

[19] Abdoun T, Wang Y. Performance of retrofitted pile foundations subjected to seismically induced lateral spreading. Pacific Conference on Earthquake

Engineering, University of Canterbury, Christchurch, New Zealand [Internet]. 2003. Available from: www.nzsee. org.nz/db/2003/View/Paper054s.pdf [Accessed: 20-09-2018]

[20] Ashford SA, Boulanger RW, Brandenberg SJ. Recommended Design Practice for Pile Foundations in Laterally Spreading Ground. USA: Pacific Earthquake Engineering Research Center College of Engineering, University of California, Berkeley, Rep; 2011

**719**

**Chapter 58**

**Abstract**

Sutcliffe 0.335 and RMSE 26.851 m3

boosting, data-driven models

tree-based model.

**1. Introduction**

able as inputs, Nash-Sutcliffe 0.671 and RMSE 18.879 m3

ments, and making decisions in irrigation management [5–7].

Prediction of Monthly Streamflow

Using Extreme Gradient Boosting

and Extreme Learning Machine

Streamflow is an essential part of the hydrologic cycle. The prediction of streamflow is important in most of the water resource management applications. In this study, the performance of two data-driven models, namely, extreme learning machine (ELM) and extreme gradient boosting (XGB), in predicting the streamflow 1 month ahead is evaluated. A basin located in the southeast of Turkey was selected as an application. Downstream flow is predicted 1 month ahead by using the optimum lags of downstream itself, upstream, rainfall, temperature, and potential evapotranspiration as input variables. Using these variables, several input combinations were developed to identify the best combination. The results showed that using the variables beside the lagged downstream flow increases the model performance. For example, using only lagged downstream flow, only Nash-

XGB. This study found that XGB outperformed the ELM although the former is a

**Keywords:** streamflow prediction, extreme learning machine, extreme gradient

One of the crucial components in the global- and regional-scale hydrologic cycle is streamflow [1–3]. The streamflow is highly related to the flood and drought disasters, and it is the main source of fresh water. Consequently, the highly accurate streamflow forecasting, especially in the regions that are vulnerable to floods and droughts, is very important for managing the water resources efficiently [1, 4]. Streamflow forecasting is done in two categories: short-term and long-term forecasts. Short-term forecast can be hourly or daily, and that is important in flood and warning system applications, while long-term forecast which can be monthly or annual is important for the operation of the reservoir, the transportation of sedi-

Hydrologic models in general are divided into two categories: physical-based model and data-driven model (DDM). A physical-based model formulation is

/s can be obtained, while using all the vari-

/s are obtained by using

*Sinan Jasim Hadi, Arkan J. Hadi, Kamaran S. Ismail,* 

*Mohammad Ali Ghorbani and Mustafa Tombul*

#### **Chapter 58**
