Numerical Analysis on Effects of Soil Improvement on Pile Forces on Existing High-Rise Building

Abstract

Nowadays, seismic codes are regularly updated with new knowledge and a better understanding of the earthquake phenomenon. With these updates, existing buildings require a reevaluation of their stability and a process of reinforcement and/or retrofitting. This study investigated the effects of two types of ground improvement which use cement-mixing soil surrounding the foundation structure to reduce and redistribute forces acting on piles. This is especially important when the reevaluation of high-rise buildings leads to increased forces in the piles. Typically, buildings are designed while assuming fixed base boundary conditions at the foundation level, without considering soil–pile–structure interaction (SPSI). SPSI significantly influences the response of high-rise buildings supported by soft soil. Increasing the lateral resistance of the surrounding soil can reduce the influence of SPSI. In this study, a detailed dynamic numerical analysis was used to investigate the dynamic response of an SPSI system of a high-rise building under seismic load. A dynamic analysis was conducted on a modified layout of a real building, using real-time earthquake motion. The finite element program DIANA FEA was used to perform nonlinear 3D FEM numerical simulations, taking into account the essential SPSI phenomena, gap-slip between the piles and the soil, and free-field boundary conditions. A comparison of the data suggests that the bending moment and shear forces in the piles are reduced in magnitude and evenly distributed along the upper part of the pile, which reduces the stress concentration of the bending moment and shear forces at the contact between the piles and the pile cap.

1. Introduction

Nowadays, the number of high-rise buildings in Thailand is increasing dramatically, especially in Bangkok, the capital of Thailand. Most of the design methods in Thailand are based on the assumption that the foundations of the building are assumed to be fixed or hinged. This simplification can reduce the complexity of the problem and the time required to analyze the problem by analyzing the superstructure and pile foundation separately. Thus, this simplification can help engineers work more efficiently and save the cost of designing the building. However, this simplification excludes soil–structure interaction, and the flexibility of the soil and foundation will be ignored.

Many building codes, including the National Building Code of Thailand, suggest neglecting the soil–structure interaction [1]. However, SSI is usually considered to be beneficial to the structure because SSI can provide damping to dissipate energy, which may help the structural system. In addition, several studies have indicated that SSI is reasonably neglected for low-rise structures and structures supported by very stiff soil. However, SSI will significantly influence the response of a structure supported by a soft soil [2], especially in the case of high-rise buildings [3,4,5,6,7].

Even though the Bangkok area is considered to be located in a low seismic hazard zone, the soil profiles at a shallow depth are very soft soils that are potentially influenced by far-distance earthquakes. High-rise buildings are more likely to be susceptible to the SSI effect when subjected to seismic loads. Past earthquakes, such as the 1985 Mexico City, 1994 Northridge, and 1995 Kobe earthquakes, have provided sufficient evidence to prove that the SSI effects should be investigated [8,9,10,11]. Additionally, SSI can influence the natural frequency [12], increase the total displacement [13,14,15], and amplify the floor drifts of the superstructure [16,17]. Consequently, neglecting the kinematic effects from SSI may lead to underestimating the maximum bending moment of the piles [18]. This can result in piles and foundations experiencing rocking effects and potentially failing during earthquake loading [19,20,21].

Cement mixing is one of the most cost-effective ground-improvement techniques that is widely used in Thailand. This ground-improvement technique can improve the strength and stiffness of the weak soil area [22,23]. The improved area of the soil around the building can potentially reduce the negative effect of SSI by increasing the horizontal stiffness and strength [24,25]. This ground-improvement technique improves the soil strength in the shallow zone around the foundation’s periphery. The lateral resistance can be further increased by extending the area of ground improvement to include the soil beneath the foundation. However, additional numerical analyses are needed to validate the capability of these approaches.

In the past, most of the numerical research about SSI was simplified to be linear elastic to reduce the complexity of the problem. Even though some of the previous research implements a nonlinear soil model with nonlinear Winkler spring models for the SSI problem, the effect of SSI is still not clearly understood. The Winkler spring model is the spring model based on the Winkler spring theory, which uses beam elements to model the pile and uses the spring element to model the soil along the embedded pile surface [26]. This model will consider only the load and displacement of the soil, and springs are independent of each other, which excludes shear coupling between the springs. A three-dimensional analysis of the soil–pile–structure interaction problem using an equivalent linear model performed by Lu et al. [27] indicates that the computational analysis shows a significant error. The equivalent linear model does not include the important behavior of SPSI, such as the nonlinearity of the surrounding soil and the SPSI phenomena of separation, closing, and sliding between the pile foundation and the surrounding soil.

This study utilized a powerful civil engineering numerical simulation program based on the finite element method DIANA to simulate the 3D soil–structure interaction (SSI) behavior of pile foundations subjected to seismic load, using the continuum SSI model. This model considers the normal pressure, shear drag, and gap/slap mechanism, which provide separation, closing, and sliding between the soil and pile. The 3D finite element analysis, which includes most of the essential SSI behavior, can provide reliable responses when compared to other approaches [28]. The responses of the superstructure and pile foundations are then observed.

2. Materials and Methods

2.1. Approach for Soil–Pile–Structure Interaction in Numerical Analysis

The interface properties between the soil and pile foundation for SPSI should be able to satisfy the combination of normal behavior and tangential behavior [28,29]. In the past, most SSI models based on the Winkler spring theory simulated the surrounding soil as springs with dashpots, but recently, an SSI model was developed to be compatible with the continuum model that can simulate the separation, closing, and sliding. In this paper, the interaction between two parts of a structure is governed by a structural interface element in terms of the Coulomb frictional behavior with a brittle gap criterion. The load–displacement curve of the interface stiffness value should relatively closely resemble that without the interface [30]. The normal stiffness should be relatively high to prevent the overlapping between the structure and subsoil. However, substantial stiffness values will result in a numerically ill condition and stress oscillation [31]. The interface stiffness values are defined according to the following equation [32]:

$$D_{ss}=\frac{A^2}{t}\frac{E_{soil}}{2\left(1+v_{soil}\right)}$$

(1)

$$D_{nn}=f\times D_{ss}$$

(2)

where D_nn is the normal stiffness; D_ss is the shear stiffness; E_soil is the Young’s modulus of the surrounding soil; v_soil is the soil Poisson’s Ratio; A is the reduction factor; t is a small length representing the virtual thickness of interfaces; and f is the multiplication factor. The Coulomb’s friction properties of the interface element between the soil and pile are based on the following equation [32]:

$$c=A{c}_{soil}$$

(3)

$$\tan \varnothing =A\tan {\varnothing }_{soil}$$

(4)

where $c$ is the interface cohesion, $c_{soil}$ is the soil cohesion; $\varnothing$ is the internal friction angle, and ${\varnothing }_{soil}$ is the soil friction angle.

The behavior of SPSI was modeled as a surface-to-surface interaction between the subsoil and the pile. Therefore, the source and target interaction surfaces are assigned to pile and soil, respectively. This type of surface-to-surface interface allows for gap forming, which can prevent the tensile-stress transfer from the structure to the surrounding soil through interface elements [33]. In addition, the same interface principles based on the Mohr Coulomb model can be seen in another powerful FEA program, ABAQUS [34].

A numerical model of a single pile in soft soil with selected interface properties is used to verify the shape of the p-y curve. The pile head is subjected to a prescribed cyclic displacement. Figure 1 shows the shape of the p-y curve from the Coulomb frictional interface, showing a good correlation with the p-y curve obtained from the experimental observations [35]. The SPSI phenomena of separation, closing, and sliding between the pile and the surrounding soil are clearly observed, as shown in Figure 2.

2.2. Numerical Investigation of SPSI with Ground Improvement

High-rise buildings on soft soil are susceptible to the SPSI effect under seismic load. Thus, increasing the lateral resistance of the surrounding soil is beneficial. One of the ground improvement methods that is widely used in Thailand is the cement-mixing ground improvement method [36,37]. The cement-mixing ground improvement can increase the lateral pile resistance, resulting in reducing the effect of SPSI and minimizing the gap separation between the pile and the surrounding soil [38,39].

Increasing lateral resistance by using the cement-mixing ground-improvement technique for the whole area around the foundation of the existing building is very hard. Thus, the alternative to this technique should be investigated. The alternative technique that is proven to increase lateral resistance is a cement-mixing wall where the soft soil around the perimeter of an existing foundation is improved by using the cement-mixing technique [40].

The numerical models are divided into three parts, namely the 18-story superstructure, pile foundations, and soil body, as shown in Figure 3a. In order to simulate the effect of SPSI on a high-rise building, the superstructure in this analysis was modeled as an 18-story building. To avoid the copyright issue, the 18-story superstructure was modified from an existing 18-story building in Thailand. The floor-to-floor height of the superstructure was 3 m, with 1.5 m of clearance between the first floor and ground level. The superstructure is composed of columns, beams, slabs, elevator shafts, and footings. The columns have two different sizes of cross-sections, namely 0.4 × 0.8 m and 0.4 × 0.15 m, and the slab thickness is 0.25 m. The beam’s cross-section, including the slab thickness, is 0.5 × 0.25 m. The total height of the superstructure is 55.5 m, with a width and length of 18.8 m and 40.4 m, respectively.

The pile foundations for the superstructure consist of several types of configurations, from single piles to a pile raft, as shown in Figure 4. The pile length for all piles is 40 m, with a pile-cap thickness of 1.7 m. The pile diameter for the single piles, three-piles with a rectangular-shaped pile cap, and pile raft is 1.0 m, while the pile diameter for two piles with a rectangular-shaped pile cap and three piles with a triangular-shaped pile cap is 0.8 m. The piles are divided into seven groups according to their foundation type, as shown in

Table 1. Pile group sorted by type of footing.

Pile Groups	Number of Piles	Pile Number
Group 1, corner	4	1, 5, 34, 38
Group 2, 3 × 1, center	2	3, 36
Group 3, 3 × 1, side	4	2, 4, 35, 37
Group 4, 2 × 1, X direction	8	6, 7, 12, 13, 26, 27, 32, 33
Group 5, 2 × 1, Y direction	8	8, 9, 10, 11, 28, 29, 30, 31
Group 6, 3 piles group	12	14 through 25
Group 7, pile raft	20	39 through 58

. The soil layers are modelled as a cylindrical container with a diameter of 100 m and a depth of 235 m, where the shear wave velocity of the soil layer is considered to be engineering bedrock.

Figure 3b shows the finite element model of the high-rise building on soft soil. The meshing method, number of elements, and node position are the same for all cases. In this study, two types of ground improvement were considered: cement-mixing wall and whole-area cement mixing, as shown in Figure 3a. The whole-area cement mixing improves the soil in the area around and under the building, while the cement-mixing wall improves the soil only around the perimeter of the foundation. The depth of the ground-improvement zone for both ground-improvement methods is 6 m, while the thickness of the cement-mixing wall is 2 m, starting from the outer perimeter of the foundation. The overall dimensions of the improved area are 42 m × 24 m × 6 m.

Bangkok is situated on the lower Central Plain of Thailand, which is known as the Chao Phraya Basin. It is well-known that this basin is filled with sedimentary soil deposits. These soil deposits form alternated layers of sand and clay with a thick, soft clay layer deposited at the top [41]. The soil parameters of Bangkok’s soft soil with the Mohr–Coulomb (MCM) soil model at a depth of 0 to 60 m were calibrated by Likitlersuang [42]. The soil parameters between 60 and 235 m were adopted from the microtremor array analysis in the work of Arai [43]. The material properties and soil parameters are shown in

Table 2. Material properties and Bangkok’s soft-soil parameter [42,43].

Material	Depth (m)	Model	γ	Su	Friction Angle	Eu, E′	v	Rayleigh Damping
			(kN/m3)	(kPa)	(ϕ)	(mPa)
Soil
Soft Clay1	0–7.5	MCM *	16.5	20	-	10	0.5	5%
Soft Clay2	7.5–12	MCM *	16.5	39	-	20.5	0.5	5%
Medium Clay	12–14	MCM *	17.5	55	-	27.5	0.5	5%
Stiff Clay1	14–20	MCM *	19.5	80	-	40	0.5	5%
Sand	20–21.5	MCM *	19	-	27	53	0.5	5%
Stiff Clay2	21.5–26	MCM *	20	120	-	72	0.5	5%
Hard Clay1	26–60	MCM *	20	240	-	240	0.5	5%
Hard Clay2	60–235	MCM *	20	400	-	1350	0.5	5%
Cement Mixing	-	MCM *	16	300	-	221.9	0.5	5%
Concrete		LEM **	23			35,000	0.2	5%

* MCM = Mohr–Coulomb model. ** LEM = Linear Elastic Model.

.

2.2.1. Boundary Conditions

The realistic effect of the earthquake wave propagation at the edge of the soil layer model was simulated with a free-field boundary condition as a lateral boundary connection at the side of the model. A free-field element represents the infinite soil domain, as shown in Figure 5. This type of element involves one-way coupling, where the free-field element has dashpots to absorb any outgoing waves [44]. Thus, the free-field element allows the earthquake wave to propagate from the bedrock to the soil surface without causing unrealistic wave reflections. Each analysis case is subjected to earthquake motion in all three directions as base excitation at the bottom boundary of the model.

2.2.2. Analysis Procedure

Before the dynamic analysis is executed, initial static stresses in the soil layers and superstructure are calculated for the geostatic and superstructure loads. The analysis step is then followed by a nonlinear time history analysis where the selected earthquake motions are applied as base excitation in all three directions. The analysis considers three cases, which are one soil structure interaction case and two SPSIs with ground-improvement cases, namely cement-mixing wall and whole-area cement mixing, as shown in Figure 6.

2.2.3. Input Motion

Thailand’s national code does not consider SPSI; therefore, the superstructure is analyzed separately from the pile foundation. The earthquake motions provided by Thailand’s national code are free-field motion, which are amplified by the soil site conditions. The free-field motions provided by the code are not suitable for this study, where the input motions are applied at the base of the model. Thus, the input motion was obtained from the Pacific Earthquake Engineering Research Center (PEER). The selected earthquake motions matched the spectral acceleration from the codes and the input earthquake motion given from the codes. The detailed list of each selected earthquake motion is shown in

Table 3. Selected strong motion for SPSI analysis.

Strong Motion Parameter	Earthquake
	Tottori	Chuetsu-Oki	Niigata	Tottori
Record Sequence Number	3893	5013	6526	3920
Year	2000	2007	2004	2000
Station Name	HYG004	FKSH15	FKSH15	OKYH02
Magnitude	6.61	6.8	6.63	6.61
Scale Factor	1.7194	3.5902	5.8522	1.071
5–95% Duration (s)	29.1	18	20.4	19.5
Rjb (km)	108.34	125.46	110.08	70.52
Rrup (km)	108.34	126.64	110.16	70.52
Vs30 (m/s)	834.56	803.57	803.57	1047.01

. All selected earthquake motions have a free-field PGA of around 0.02 g, which is equivalent to the motion that Thailand’s national code provided.

3. Results

The primary objective of this study was to investigate the SSI effect of soil improvement on pile forces. Additionally, a brief discussion is provided on the impact of soil improvement on the superstructure, focusing only on the seismic response at the top of the superstructure.

The comparison of relative displacement at the top of the superstructure during the earthquake in both directions is shown in Figure 7. It is evident that in the earthquake records TOTTORI_HYG004 (Figure 7a), CHUETSU_FKSH15 (Figure 7b), and NIIGATA_FKSH15 (Figure 7c), the main impulse is directed along the y-axis, while in the case of the earthquake record TOTTORI_OKYH02 (Figure 7d), the main impulse exhibits directionality along a 45-degree line. The overall relative displacement of the case without ground improvement is significantly higher than the case with ground improvement. The relative displacement of the structure tends to have a longer vibration period, which is one of the effects of SPSI, where the gap-slap mechanism will increase the unconfined area of the superstructure. The ground improvement in both methods can improve horizontal stiffness, which minimizes the effect of the gap-slap mechanism. Figure 8 shows the relative displacement in the Y direction of the superstructure subjected to earthquake motion RSN3920_TOTTORI_OKYH02. The relative displacement, in this case, is the highest among all cases. The gap opening is formed as the superstructure vibrates. After the superstructure passes the highest relative displacement point, the changes in the vibration period are clearly visible.

After the first strong motion impulse, the vibration period starts to show deviation among each analysis case. The normal case is highly susceptible to gap separation due to the low stiffness, which is characteristic of the soft soil layer. On the other hand, the cement-mixing wall is only susceptible to gap separation around the soft soil area inside the cement-mixing wall. Thus, after the point of peak acceleration, the vibration period of the normal case is noticeably higher than the case with ground improvement. The influence of SPSI on the vibration period decreases as the lateral stiffness of the foundation system increases.

Figure 9 shows the spectral acceleration at the top of the superstructure. The natural frequency of the superstructure is 0.74 Hz and 0.82 Hz for the X direction and Y direction, respectively. At the current level of acceleration and Bangkok’s soft soil condition, the kinematic interaction affects the superstructure significantly, especially when additional damping from SPSI due to plastic deformation is small. In addition, the ground improvement method increases the overall stiffness of the foundation system, resulting in a slightly higher response amplitude in the first mode when compared to the case without ground improvement.

The case with a cement-mixing wall shows a higher amplitude at a higher frequency when compared to other cases, especially in the Y direction. The observed amplitude increases could potentially be attributed to the reflection of the surface wave. Since the size of the pile-raft and footing is large and the base excitation of this analysis also includes the vertical acceleration time history, the surface wave was created as the superstructure vibrated. As the surface wave spreads out radially toward the model’s boundary, the energy will continuously dissipate, and the surface wave will be completed and absorbed by the free-field boundary when the wave reaches the model’s boundary. However, in the case of the cement-mixing wall, the shear wave velocity of the cement-mixing wall is significantly higher than that of the Bangkok soft soil. Thus, the surface wave is reflected toward the center of the model. As a result, the surface wave is trapped inside the soft soil area until the wave energy is dissipated. This kind of behavior is expectable; however, the amplitude is very low. Thus, the surface wave in the case of the cement-mixing wall is insignificant when compared to the overall benefit of the ground improvement.

Due to the high number of piles analyzed in this study, it was necessary to categorize them into seven groups based on their foundation types, as presented in Table 1. The pile cross-section in this study was uniformly circular, resulting in an axis-symmetric pile strength. Thus, the bending moments in both the X and Y directions were combined into the total bending moment. The average maximum and minimum total bending moments were calculated for each pile group as an average value from all piles in the group and all four earthquake records at a given depth of the pile. The total bending moment was averaged for the maximum and minimum observed values throughout the entire duration of the earthquake records. The standard deviation and observed extremes at piles were also calculated. Figure 10 illustrates the average total bending moment for each analysis case, considering all earthquake motions and time increments.

From the analysis results, both the normal case and the cement-wall case exhibited a similar distribution for the total bending moment along the depth, characterized by a comparable shape, for all pile groups. However, it was noticeable that the cement-wall case demonstrated a relatively smaller total bending moment compared to the normal case. In the cement-mixing case, it is noteworthy that the formation of a stiff zone between the piles resulted in a reduction in the magnitude of the total bending moment, which exhibited a uniform distribution along the depth of the cement mixing. This effect was particularly pronounced in the pile raft zone. Moreover, for single-pile configurations and pile groups consisting of two and three piles, the location of the maximum total bending moment shifted toward the end depth of the cement mixing.

It was observed that Pile Group 7 exhibited the highest bending moment due to the superstructure configuration, with two elevator shafts located directly above the pile raft. Given that the elevator shaft is the stiffest component of the superstructure, a significant portion of the bending moment was distributed into the pile raft.

Comparing the cases with and without ground improvement, it was noted that the bending moment in the ground-improvement case was more uniformly distributed. The improved area displayed higher stiffness than the surrounding soils, reducing the excessive displacement of the piles caused by the soil–structure interaction. Furthermore, the whole-area cement-mixing case exhibited similar behavior to that of a large pile raft, with both the axial stress and bending moment evenly distributed across the improved area. Therefore, the maximum bending moment in the whole-area cement-mixing case was located underneath the improved area.

The maximum bending moment in Pile Group 7 was significantly reduced due to the improved stress distribution, resulting in a proportional increase of bending moment in piles with lower bending moments in the normal case (i.e., without ground improvement). The summary of the bending moment at the pile cap in Figure 10 is shown in

Table 4. Summary of pile total bending moment at the pile cap (−1.7 m).

Case		Total Bending Moment (Nm)
		Group 1	Group 2	Group 3	Group 4	Group 5	Group 6	Group 7
Normal	Mean	111,522	113,862	121,256	55,446	52,029	51,476	182,163
	SD	34,863	14,012	19,150	11,475	7451	9558	36,921
	High	177,149	135,269	152,339	75,598	63,902	70,874	247,784
	Low	73,886	93,137	94,644	33,578	41,348	36,209	113,873
Cement wall	Mean	96,987	95,504	102,927	49,276	46,738	43,228	166,106
	SD	21,413	20,261	19,268	11,169	7072	9317	35,769
	High	134,562	124,757	137,465	72,010	60,130	63,977	238,189
	Low	69,474	74,728	79,768	30,853	35,123	27,536	104,294
Cement mixing	Mean	62,367	63,401	65,949	32,183	27,360	37,555	108,460
	SD	14,526	12,482	10,997	8911	5881	6920	29,250
	High	92,734	84,431	86,087	45,868	38,618	51,877	161,207
	Low	44,443	46,446	52,408	19,657	17,482	27,684	64,644

Note: Data for all points on the pile are available upon request.

.

Similarly, as in the case of the total bending moment, shear forces were calculated for both the X and Y directions and then combined into a total shear force. The total shear forces are presented in Figure 11 for each analysis case and averaged for all earthquake motions, and the summary for shear forces at the pile cap in Figure 11 is shown in

Table 5. Summary of pile total shear forces at the pile cap (−1.7 m).

Case		Total Shear Force (N)
		Group 1	Group 2	Group 3	Group 4	Group 5	Group 6	Group 7
Normal	Mean	805,554	944,009	1,028,693	729,069	680,295	677,659	1,525,612
	SD	244,687	113,713	149,743	157,680	111,755	137,286	317,270
	High	1,285,038	1,140,997	1,285,279	972,544	896,355	941,972	2,111,692
	Low	517,881	811,316	789,664	454,486	491,230	472,291	935,039
Cement wall	Mean	739,773	748,211	805,757	619,633	601,831	548,193	1,378,372
	SD	158,678	139,052	156,113	148,659	93,705	122,286	305,049
	High	1,009,736	951,149	1,066,179	946,867	761,228	829,351	1,953,279
	Low	530,827	603,251	627,590	366,518	470,467	350,769	837,257
Cement mixing	Mean	422,221	426,601	431,321	347,745	304,286	423,874	860,034
	SD	111,327	72,321	68,152	104,679	67,417	82,773	243,605
	High	651,981	542,812	551,320	505,495	415,028	596,759	1,268,231
	Low	239,930	340,844	350,343	211,291	192,943	301,101	511,695

Note: Data for all points on the pile are available upon request.

.

In the normal case, it is observed that there is a concentration of shear forces at the contact between the pile and the pile cap, which can be critical during pile design. In the cases of improved soil, a reduction in shear forces was observed compared to the normal case. The reduction is significant in the case of the cement-mixing improvement, especially at the contact between the pile and the pile cap. The reduction of total shear forces was observed across all groups. This behavior can be beneficial in cases where the shear capacity of the pile is limited.

A comparison was conducted between the normal cases and the soil improvement cases regarding the increase in the total bending moment and shear forces. This comparison involved assessing the change in the total bending moment and shear forces on the piles relative to the mean maximum observed in the normal case (Figure 12). Notably, at the interface between the piles and the pile cap, a reduction in both the total bending moment and shear forces was observed for both soil improvement methods. It was found that the method involving cement mixing exhibited a larger reduction in magnitude compared to the other method. In the case of the cement-wall-method soil improvement, the forces acting on the piles are reduced; however, the level of reduction is negligible, and therefore, there are limited benefits to the soil improvement. However, with the cement-mixing method for soil improvement, the reduction of the forces in the piles was observed at the levels where the bending moment and shear forces reached their highest magnitude. Although the total bending moment near the top of the pile was higher than the reference normal case, it should be noted that this magnitude gradually decreases with depth, ultimately becoming lower than the maximum observed on the piles (Figure 10). Based on these findings, it is evident that the cement-mixing method for soil improvement can provide significant benefits in terms of reducing the total bending moment and shear forces.

4. Conclusions

The study aimed to propose a method to reduce the forces acting on the foundations by reducing the effect of soil–pile–structure interaction. Two methods are proposed for improving the soft foundation soil, one by improving the soil around the foundations of the buildings, cement-mixing wall with a depth of 5 m, and cement-mixing between the pile foundations. In both cases, no strengthening or retrofitting is performed on the superstructure so that it can be usable while the soil improvement is performed.

The seismic response of the superstructure shows that, in the case of Bangkok’s soft soil condition, the kinematic interaction affects the superstructure significantly, while the ground improvement method increases the overall stiffness of the foundation system, resulting in a reduction of the seismic response at the top of the superstructure when the soil is in the nonlinear state.

The analysis of pile forces, bending moments, and shear forces was performed by grouping the piles into seven groups based on the type and number of piles in the pile cap. The effect of soil improvement with a cement wall showed a small reduction in the bending moment and the shear forces in the piles. A significant reduction in the bending moment was observed for the single piles at the corners of the building, limited only to the depth of the cement wall improvement. It is possible that increasing the depth of the cement wall may further reduce the bending moment to the required depth. In the case of cement mixing, a significant reduction of the bending moment is observed for the piles in the pile raft, Group 7, and a better distribution among all pile groups is observed. In the case of the total shear force, a significant reduction is observed for all pile groups, and this may be due to the better transfer of the base shear to the foundation soil through the cement-mixing area and the reduced effect of the kinematic interaction. The results of the analysis suggest that this type of improvement can be beneficial in cases where the piles do not have a great enough bending moment or shear capacity.

The data obtained from the analyses suggest that the width of the cement wall and the depth of soil improvement influence both the pile forces and possibly the response of the superstructure. Further investigation is required to explore the effects of widening the cement wall and adjusting the dimensions of the soil-improved area.

The effect of the soil improvement on the superstructure can be further investigated, considering that it showed that the effect of the kinematic interaction can be reduced. This method can be applied to both existing and new structures, considering its cost-effectiveness and ease of construction.