Numerical Assessment of Pipe Pile Axial Response under Seismic Excitation

I n engineering, the ground in seismically active places may be subjected to static and seismic stresses. To avoid bearing capacity collapse, increasing the system's dynamic rigidity, and/or reducing dynamic fluctuations, it may be required to employ deep foundations instead of shallow ones. The axial aptitude and pipe pile distribution of load under static conditions have been well reported, but more study is needed to understand the dynamic axial response. Therefore, this research discusses the outputs of the 3D finite element models on the soil-pile behavior under different acceleration intensities and soil states by using MIDAS GTS NX. The pipe pile was represented as a simple elastic, and a modified Mohr-Coulomb model was used to describe the surrounding soil layers. When low acceleration was introduced in the early stages, positive frictional resistance (i.e., in dry soil, the FR was about 1.61, 1.98, and 0.9 Mpa under Kobe, Halabja, and Ali Algharbi earthquakes, respectively) was recorded. However, as the acceleration increased (from PGA of 0.1 g and 0.102 g to 0.82 g), the resistance reduced and eventually turned negative. In this study, both internal and exterior frictional resistance were measured. It was found that the soil state and acceleration intensity both have a noticeable effect on the failure process, i.e., the maximum plug soil resistance decreased by about 55% by changing the soil condition from a dry to a saturated state under the recorded data of the Kobe earthquake. A rough estimation of the long-term settlements at the shaken soil surface is meant to be included in the results of this research.


INTRODUCTION
The earthquake, one of the worst natural disasters ever, caused significant damage and loss of life.Likewise, at this moment, earthquakes are unpredictable and unmanageable.The only possible choice available to engineers is to design and construct structures in a manner that eliminates the impact of earthquakes (Arora, 2004 to make an effort to mitigate seismic effects through soil treatment and mitigation.The axial force resulting from inertial and kinematic impacts, the inertial load attributed to the superstructure, and other loads associated with kinematic consequences (ground movement) are the major driving loads to a pile throughout such seismic situations.The 1-g shaking table test is widely employed to examine the ground's dynamic motion and earthquake-damaged soil structures' responses.A shaking table is used to replicate the earthquake-related shaking.The ultimate strength of long piles in frictional soil, frequently utilized in offshore projects, is strongly influenced by pile FR.According to (Vesic, 1970), the shaft resistance a long pile creates varies with pile depth.It usually increases with pile penetration depth till achieving a peak value a short distance above the pile tip, after which exterior skin resistance starts to drop dramatically at the zone above the pile tip, as seen in Fig. 1a.This was attributed to the 'Arching effect,' which restricted effective lateral stresses.The pattern of shaft resistance in sandy soil for various pile lengths is depicted in this diagram.) Vesic, 1970;Fellenius and Altaee, 1995;Klotz and Coop, 2001) proposed that the maximum possible FR was measured at a point above the pile tip.and 2022b).FR for piles built in homogeneous sand does not rise linearly with pile depth and is dependent on lateral stresses.Based on this pattern, the highest FR may be anticipated above the pile base.Owing to such an increase in shear displacement, (DeJong et al., 2003) reported that the radial distance of the shear zone surrounding the outer face of the pile is produced with a thickness ranging from (3.5 mm to 5.5 mm) for silica sand.The thin shear zone interfaces interpret the load formation and transmissions along the external shaft of the pile through the shear zone.Increased lateral pressure could be caused by an increase in limitation imposed by the surrounding soil, which could modify the volume of the thin shear zone.Thus, (Al-Soudani, 2020) stated that the pile FR distribution differs from the researchers' point of view and from one test to another, which led to high discrepancies in design approaches of friction distribution.Table 1 summarizes the allowable movement limits required for resistance mobilization according to the standardizations.Iraqi experts have been researching pile foundations in the sand under various shaking patterns of motion due to increased seismic activity in the last few years.Likewise, pipe piles are frequently used as they are simpler and more affordable to build, can be tested for safety before being established, and may be tailored to meet certain load specifications, thus saving money by reducing the need for extra reinforcing.Hence, in this study, using MIDAS GTS NX finite element software, the effects of static vertical and lateral stresses with three distinct seismic vibrations on the soil-pipe pile configuration about the kinematic interaction were investigated.The program findings were validated using (Hussein and Albusoda, 2021) laboratory findings.

NUMERICAL MODELING
Three-dimensional finite element models have been developed using the MIDAS GTS NX software, considering the shaking load nonlinearity.

MODEL DESCRIPTION
The current study adopted an open-ended aluminum pipe pile with three earthquake records (PGA = 0.82, 0.102, and 0.1 g) and pile L/D =25.The pipe piles are set in cohesionless two soil layers (Dr = 30% and 70% as the upper and bottom layers, respectively).The soil characteristics have been adopted from (Hussein and Albusoda, 2021), who performed physical laboratory tests using air-dried poorly graded sand (SP).These characteristics have been calibrated using (Beaty and Byrne, 2011) approach.The equivalent SPT blow count for clean sand (N1)60, obtained as per ASTM D 1586-99, is the fundamentals of the basic calibration formulas (Beaty and Byrne, 2011).The model has been exposed to the total allowable vertical load and 50% of the allowable lateral load, as shown in Fig. 1.These static loads were applied as point loads on the pile cap.The main input parameters for the soil and pile materials are presented in Tables 2 and 3.
The pile element's constitutive relationship was linear elastic.An extensive study uses an interconnection component with restricted shear resistance to demonstrate the relationship between the pile and the surrounding soil, employing the Modified Mohr-Coulomb failure criteria.The normal and shear coefficients of the interface elements were estimated.
Regarding the two-layer soil medium with an upper layer of sand that is 11.2 m thick and a bottom layer that is 28 m thick, a 0.56 m diameter pile is constructed.The core pile is assumed to have descended into the deep sand below after reaching the subsoil.2.7 x 10 -4 ξ (%) 5

PILE AXIAL RESPONSE
Numerous infrastructure and construction failures following earthquakes have been connected with notable settlements (Tokimatsu et al., 1998).Due to the increased pore pressure the loose sand creates, the soil shear stiffness and effective vertical stress are reduced when piles are installed in saturated sand layers.Additionally, there will be different effects on the end bearing and pile shaft's friction resistance.The higher-end bearing capacity could be activated due to the pile settling dramatically when additional pore pressure in the bearing stratum increases.Conversely, shaft friction reduces with the creation of excess pore pressure and may potentially vanish once liquefaction occurs.Yet, the nature of damage differs in dry-site piles.The pile may experience intense compression and tension stages, and the settlement may occur during rapid acceleration.However, pile settling is less than in saturated places, and bearing capacity failure is improbable because compaction may increase the soil's characteristics.Following is a discussion of how the acceleration intensity affects FR and open-ended pipe pile settlement for both saturated and dry models .Fig. 2 displays the time history of the OE Aluminum pipe pile's vertical movement in both wet and dry models.Unlike the preceding models, the saturated models showed noticeably higher settlements than the dry models, independent of the acceleration histories.Table 4 summarizes the modeling findings for the soil plug resistance and the highest value of adjacent soil resistance when static stress alone and mixed static-seismic load were applied, respectively.(i.e., the Kobe earthquake).After applying the combined staticdynamic load-at the final stage -14.6 13 -8 6.9 While seismic activation reduced the arching within the soil plug, a higher resistance than the maximum resistance of the soil next to it was found.It is crucial to emphasize that the soil plug's length was preserved at 4D since the author observed that when employing a plug length of 3D, the length is insufficient for plugging mode because the sand keeps sliding up within the pipe pile, as illustrated in Fig. 3a.As a result, the pile collapsed in the plugged mode (as a CE pipe pile) by using 4*Dinner, and no more soil was seen to go into the pipe pile, as demonstrated in Fig. 3b.This could have been explained by the existence of a loose soil layer (Dr of 30%); hence an additional soil plug was necessary to attain high soil particle density within the pipe pile.

CONCLUSIONS
This research investigates the acceleration intensity effects on the seismic response of an open-ended pipe pile set in dry and saturated sand soils.The main variables were adjusted depending on the modified Mohr-Coulomb model.The primary findings of the current study were verified with 1 g shaking table tests.The results of the present study indicate that the soil's shear strength is reduced with the release of excess pore pressure.The pile is anticipated to drop due to the development of high pore-water pressure (liquefaction commencement).At this point of the 1g laboratory tests, it was observed that soil particles had generally lost their shear strength and behaved roughly like a slurry.The pile frictional resistance decreased significantly (i.e., about 45% under the effect of the Kobe earthquake) when the ground condition shifted from dry to fully saturated.The computer simulations simulated this effect by considerably lowering the FR during the dynamic excitation until it reached the lowest possible value at the end.A significant mass circulates near the pile tip due to the soil plug arching and densification.The pile underwent noticeable deformations caused by a large drop in soil stiffness as the nearby loose sand soil exhibited liquefaction (excess pore water pressure ratio > 85%).Overall, the frictional resistance surrounding the pile body was less than that at the pile base, i.e., under the effect of the Kobe earthquake, the decreasing ratios were about 11% and 14% for the dry and saturated soil conditions, respectively.

Author
(Al-Jeznawi et al., 2022a) explained the laboratory experiments, meshing, static and dynamic boundary conditions, and validation of the constitutive models used in their latest work.The present computer analysis included a full-scale model and three recorded earthquake data (Kobe, Halabja, and Ali Algharbi).(Hussein and Albusoda, 2021) performed a 1g shaking table experiment to assess a closed-ended Aluminum pipe pile constructed in two different soil layers.The present numerical models were compared with the laboratory observations (Al-Jeznawi et al., 2022a; Al-Jeznawi et al., 2022c) and reported the validation outcomes.

Figure 1 .
Figure 1.The implemented finite element model in the current study (Al-Jeznawi et al., 2022a).

Figure 2 .
Pile settlement embedded in layered soil during the shaking of different acceleration histories.

Figure 3 .Figure 4 .
Figure 3. Plugging soil moving direction at the end of applying the static-seismic loading.

Table 1 .
The required Allowable movement limits for pile resistance mobilization.

Table 4 .
The numerical results of the plug and the surrounding soil resistance under static and Kobe earthquakes (PGA = 0.82g).