Influence Analysis of Underground Excavation on the Adjacent Buildings and Surrounding Soil Based on Scale Model Test

In urban areas, it is common to construct underground structure nearby existing buildings. To investigate the impact of excavation construction on the adjacent existing buildings and surrounding soil, eight parallel scale model tests that considered the process of cut and cover construction are carried out with two kinds of the diameter of support piles, two kinds of the adjacent structure, and two kinds of the relative horizontal distance from the excavation in the laboratory. And, the variations of horizontal and vertical displacements of adjacent buildings and the soil pressure surrounding the excavation and the foundations of existing buildings with different parameters are presented and discussed. )en, the experimental data and the results of eight prototype finite element models are compared and analyzed.


Introduction
Urbanization rates continue to climb steadily throughout China.Nowadays, in the quest to maintain rapid economic growth, Chinese cities are confronted with many pressing problems which greatly threaten the urban living environment, such as overpopulation, land resource shortage, worsening urban traffic congestion, and environmental degradation.In the process of urban development, underground space, as an important component of the urban land space resource, has been increasingly utilized to effectively solve urban problems like municipal transportation, disaster prevention, environmental protection, and land scarcity [1][2][3][4].It became a trend in development to exploit underground space as the land resource in cities decreased.
erefore, it became a more and more commonplace to dig underground structures close to buildings that had already been constructed [5].
Open cuts and underground excavations are gradually increasing in frequency because of the development and upgrade of infrastructures and the construction of new buildings.In urban areas, there are many situations, where basement construction or underground facilities such as underground civil air defense project are proposed to be constructed adjacent to existing buildings.In this case, it is important to evaluate the influence of excavation construction on surrounding soil and adjacent buildings and to estimate settlement of the ground for proper designing of the underground structure.
e construction of the underground structure often causes ground settlement ranging from regional subsidence to small-scale collapses leading to many problems, such as building damage and structural failure [6].At the same time, public concerns have risen over the effects of underground excavation on adjacent structures and utilities [7].Nearby structures can be distorted and damaged by excavation construction, causing many problems such as loss of invaluable historic property, third party impact, construction delay, and substantial increase of project cost.
Similarity theorem is a kind of principle studying similar phenomenon and similarity principles.It is the basic of the model test and feasible to apply this theorem in the new model test design.e theory of the scale model similitude explains the relationship between the scale model and the behavior of the corresponding prototype.Rocha [8] systematically described scale modeling for problems in soil mechanics in a 1 g gravitation field.Moncarz and Krawinkler [9] considered that if an adequate model correctly scales the primary features of the problem, the scaling relations between the prototype and model are not significantly affected.In the previous research, Sawwaf and Nazir [10] have analyzed the results of laboratory model tests on the influence of deep excavation-induced lateral soil movements on the behavior of a model strip footing adjacent to the excavation and discussed the variation of the footing measured vertical settlements with different parameters.Fang et al. [11] conducted physical model tests of highway tunnel construction and examined the stability of the surrounding rocks regarding different caved zone-tunnel distance and dip angles of the coal seam.Yang et al. [12] have investigated the mechanical behavior of a typical jointed rock block located adjacent to an underground excavation by comparing the failure modes of the numerical simulations with the corresponding experimental physical jointed block test results.Shahin et al. [13,14] studied about the influence of existing building load on the deformation and Earth pressure of the ground in shallow tunneling by two-dimensional model tests and numerical analysis.Kusui et al. [15] conducted a large number of scaled-down tunnel experiments to investigate the response of unsupported walls to an increased stress field.Xu et al. [16] have carried out a series of threedimensional shaking table tests to investigate the mechanism and effect of seismic measures of the mountain tunnel using a scaled model based on a real tunnel.Furthermore, the centrifuge test was often used to study the effect of excavation construction, and Liyanapathirana and Nishanthan [17] and Zhang et al. [18,19] analyzed the excavation-induced pile behavior using the finite element method and the centrifuge test.In terms of influence parameters of excavation construction, Yashiro et al. [20] conducted parameter analysis focusing on the overburden and the stiffness coefficient of the ground by anecdotal surveying and numerical analysis.Boone et al. [21] provided a modified approach for estimating potential damage and compared to case histories using construction data from a large braced excavation.Hsiao et al. [22][23][24] considered the calculated settlement as the load in the context of reliability analysis and proposed a simplified model for evaluating the damage potential of a building adjacent to a braced excavation.Long [25] summarized the general trends and patterns based on some 300 case histories of wall and ground movements due to deep excavations worldwide.
In this research, some new aspects and mechanism of the influence due to underground excavation are illustrated using scale model tests and finite element analysis.e objective of this study is the impact of underground excavation on adjacent existing buildings and surrounding soil in the cases of different influence parameters, such as the size of support piles, the load and layer of adjacent existing buildings, and the relative horizontal distance from the excavation.
e process of excavation construction is simulated by the way of the indoor scale model test based on the similarity principle.is paper compares and analyzes the difference in changes of the horizontal and vertical displacements of the existing buildings and the soil pressure near the foundation of the existing building models according to eight parallel tests of different influence parameters and numerical simulate analysis.

Scale Model Test Setup
e foundation of the similarity theorem is the three similarity principles, the science of the conditions under which physical phenomena are similar.It is related to dimensional analysis and provides the basis for physical modeling.Similarity theory establishes similarity criteria for different physical phenomena and studies the properties of the phenomena by means of these criteria.Physical similarity is a generalization of the elementary and intuitively obvious concept of geometric similarity.In physics, a calculation is said to be from first principles, if it starts directly at the level of established laws of physics and does not make assumptions such as empirical model and fitting parameters.Two geometrical objects are called similar if they both have the same shape, or one has the same shape as the mirror image of the other.More precisely, one can be obtained from the other by uniformly scaling (enlarging or reducing), possibly with additional translation, rotation and reflection. is means that object can be rescaled, repositioned, or reflected, so as to coincide precisely with the other object.If two objects are similar, each is congruent to the result of a particular uniform scaling of the other.A modern and novel perspective of similarity is to consider geometrical objects similar if one appears congruent to the other when zoomed in or out at some level.
In geometry, if corresponding geometrical elements of two figures or solids are in proportion, the figures are said to be similar.Based on the first principle, the size of the prototype is reduced (or enlarged) by a certain proportion to make the model, and the similarity constants of parameters are as follows: where C l is a geometrical similarity parameter, l is the length, and the subscripts p and m stand for the prototype and model, respectively.e purpose of the scale test is to investigate the soil pressure and the displacement of adjacent buildings during the process of underground excavation construction.In general, the larger the scale model, the better the actual situation of the prototype model can be reflected.1.
Considering the most unfavorable situation in the underground excavation construction, the soil used in this scale test is ne sand, the density of the sand is found to be 1.56 kg/m 3 , the void ratios is 0.43, and the moisture content is 3.43%.e speci c gravity of the soil particles was measured by the pycnometer method.
ree tests were carried out producing an average value of speci c gravity of 2.63.Furthermore, the optimum moisture content is 10.1%, and the maximum dry density is 1.875 kg/m 3 by ve compaction tests.e particle size distribution was determined using the dry sieving method, and the results are presented in Figures 2  and 3.
e uniformity coe cient (C u ) and coe cient of curvature (C c ) for the sand were 2.74 and 0.89, respectively.e triaxial compression test was conducted as shown in Figure 4, and the internal friction angle and the cohesive force were 30.54 °and 17 kPa, respectively.And the average compression modulus of the sand soil is 35.53MPa according to three compression tests.
e sand was evenly poured into the test box three times to ensure the uniformity of the sand density and 72 hours after each pouring.As shown in Figures 5 and 6, the soil pressure sensors (range 0-100 kPa, accuracy 10 −5 kPa, China) were buried horizontally in the sand after lling, which were used to measure the change of soil pressure.And, the data of sensors were collected by static strain collection instrument (Donghua DH3816, China).Furthermore, the soil pressure sensors were calibrated before being buried by the liquid water pressure test, to determine the rate K of the soil pressure sensors.
As  Advances in Civil Engineering accuracy 0.01 mm, China) were installed by magnetic bases, respectively, to monitor the settlement response of the adjacent building models during excavation construction simulation, as seen in Figure 8.In addition to this, the adjacent frame structure model was made by polyvinyl chloride material which considered linear deformation and stiffness, and the frame structure models were heaped load with some sand boxes based on the geometric similar parameter C l .
Table 1 illustrates the influence parameters of each parallel scale test.A load means that the left frame structure was a four-storey model (stiffness 37.268 N/mm) with strip foundation and the right frame structure was a two-storey model (stiffness 97.3 N/mm) with column foundation, and B load means that the left frame structure was an eight-storey model (stiffness 49.86 N/mm) with strip foundation and the right frame structure was a four-storey (stiffness 115.9 N/ mm) model with column foundation.
ese structure models with load blocks are similar to the prototype concrete frame structures, as shown in Figure 9.
e cut-and-cover method is a traditional construction method of the underground structure: firstly, the cover soil is excavated, and then the support structure (contiguous piles) is constructed on both side of the excavation.Secondly, the soil is excavated while the temporary support structure is strengthened.irdly, the bottom plate, side walls columns, and roof of the underground structure are poured in turn.Finally, the underground excavation construction is completed after the backfilling of cover soil.Each parallel scale model test simulated the construction process of the underground excavation used cut-and-cover method.e main test steps are as follows: the initial state, excavation of cover soil, construction of the support structure, excavation of soil, construction of the underground structure, and backfill of cover soil, as shown in Figure 10.Furthermore, the simulation process of each scale model test is manual operation, and the support structure is simulated by the polyvinyl chloride pipe which is like the contiguous piles.
e test data of soil pressure sensors and dial indicators were collected at the time of two hours, six hours, ten hours, and twenty hours after each test step.Figure 11 is the rendering of the scale test model.

Displacement of Adjacent Building Model
In the process of underground structures construction, the adjacent buildings are affected by the soil excavation unloading, support piles squeezing soil, and so on, which could cause horizontal tilt and vertical settlement.2. e contrast variations of horizontal displacement measured from each parallel test on the top and bottom of the adjacent buildings are shown in Figure 12. e curves show that the horizontal displacement changes obviously during the construction of soil excavation and support piles which are affected by soil-squeezing action especially, and the impact on the existing buildings are relatively small in other construction stages.
Additionally, the amplitude of change and value of the horizontal displacement are relatively larger in the model test, in which both the number of structural layers and the load are less, but the diameter of support pile is larger.e horizontal displacement of the left adjacent structure model is not obvious because the relative horizontal distance from the excavation is farther, and the value is within a very small range.Moreover, it should be mentioned that the peak value of the horizontal displacement on the bottom of the frame structure model exists in the test II, with the largest diameter of support piles and the least relative horizontal distance, and the impact is evident considerably as the support piles construction.However, the magnitude and the peak value of the horizontal displacement of the right frame structure are obviously larger than that of the left frame structure.
e vertical displacement variations of the adjacent frame structure model from each parallel scale model test are shown in Figure 13.Similar to the horizontal displacement, the peak value of vertical displacement also appears in the test II.It can be verified that the overall value of vertical displacement at the closest point to the excavation is relatively large, and the displacement change is significantly obvious due to the proximity to support piles during the construction of row piles.However, the vertical displacement changes are relatively insignificant when the load and structural layers of the adjacent building model increase.

Advances in Civil Engineering
Furthermore, the vertical displacement of the existing frame structure model has a small overall value within 0.5 mm at the farthest distance from the excavation, and the reverse tilt occurs in some model tests.
To investigate the impact of excavation construction on adjacent buildings, the horizontal displacements measured from each scale model test at the top and bottom of the left and right frame models in the scale model tests are subtracted to obtain the relative values as shown in Figure 14.
e curves show that the horizontal tilt is signi cant as the support piles construction especially in test II and test IV in which the horizontal distance is farther and the diameter of support piles is larger relatively.
In the same way, the relative vertical displacement is obtained by subtracting the vertical displacements of the nearest and farthest distances from the excavation in each scale model test, as shown in Figure 15.Di erent from the relative horizontal displacement, the relative vertical displacement of the right frame structure model in test II and test IV increases signi cantly as the support piles construction, and the adjacent building models subsidence obviously, due to relatively small distance from the excavation.

Soil Pressure
Excavation-induced ground movements and their impact on the adjacent existing buildings is a source of signi cant concern for the underground structure.is paper analyses the change of soil pressure surrounding the excavation which could re ect the disturbance condition of the Earth by excavation construction.During each parallel scale model test, the steps of soil pressure sensors data acquisition are demonstrated in Table 3.
e soil pressure sensor nos.112, 118, and 122 were embedded under the middle of excavation every 0.1 m directly, which was measured from each scale model test and are shown in Figure 16.It can also be seen that the soil pressure under excavation displays a uctuation signi cantly because of an unloading e ect as excavation of soil, and the curves change slightly during excavation of cover soil and construction of the support structure as compared to other construction stages.Furthermore, the decrease in soil pressure is signi cant with increasing distance from the bottom of excavation, and the curves reach the maximum value due to the soil excavation.However, it is important to mention that the in uence of the existing building load is relatively large near the bottom of excavation.
To investigate the e ect of relative vertical distance on the soil below the underground structure, the numerical relationship between the relative distance from the bottom of excavation and the soil pressure from each scale model test is tted in Figure 17.
where l r is the relative distance from the bottom of the excavation and p 1 , p 2 , p 3 , p 4 , p 5 , p 6 , p 7 , and p 8 are the soil pressures under excavation from the scale model test I, II, III, IV, V, VI, VII, and VIII, respectively.Figure 18 shows the variation of soil pressure at di erent horizontal distances from the middle of the excavation, but the same vertical distance from the bottom of the excavation.It can be seen that the soil pressure demonstrates uctuations more obviously near the existing buildings (nos.107 and 114), when compared with under the midline of the excavation (no.110) during the excavation of the cover soil and support structure construction, which is mainly a ected by the number of layers and the load of the adjacent buildings and relative horizontal distance, and the soil pressure near the existing building with bigger load is relatively large.
In each scale model test, the soil pressure sensors are buried 0.1 m below the foundation of the existing frame structure model on the left and right sides.Figures 19 and 20 show the change of soil pressure below the foundation of the left and right existing structure model, respectively, farthest from each parallel test.In contrast, the soil pressure below the foundation of existing buildings is lower than that of the overall value below excavation, and the curves of soil pressure uctuate obviously near the support piles especially.Furthermore, the overall numerical change of soil pressure below the right foundation is relatively obvious due to the reduction of the number of structural layers and loads.

Advances in Civil Engineering
It can be claimed that the stability of the soil under the adjacent building foundation has a signi cant relationship with the load and foundation form of structure.However, the soil is not as a ected as the soil under the excavation during excavation construction, and the overall value is relatively small.In conclusion, the soil pressure displays di erent changes from each parallel scale model test because of 10 Advances in Civil Engineering different size of support piles, the load and layer of adjacent existing buildings, and the relative horizontal distance during excavation construction.e size of support piles is obvious comparatively to the disturbance of the surrounding soil, and the soil pressure under the excavation is affected by the unloading of the excavated soil and the loading of the underground structure.Furthermore, the peak value of the fluctuation is greater than the soil pressure under the adjacent building foundation.Additionally, due to the discreteness of the sand used in the model test and the influence of the similarity principle of the scale model, there are certain differences in the data results of some soil pressure sensors, but it does not affect the overall trend of soil pressure around underground structure during excavation construction.e measured values of the soil pressure sensors in each of the scale model test could reflect the changes of surrounding soil caused by the influence parameters in approaching construction of underground structure.

Comparison with Numerical Results
In order to compare the influence of the scale model test and the prototype model on the construction of the underground engineering, eight different finite element models of the prototype excavation tests are carried out with the finite element software MIDAS/GTS based on the geometric similarity principle of the first similarity theorem, as shown in Figure 21.And the test models I, III, V, and VII and the test models II, IV, VI, and VIII are different in the size of contiguous piles, not shown in the finite element models.
e dimensions of the model were selected so that the effect of boundary conditions on the results of the numerical analysis was negligible.Except for the ground surface, boundary conditions were totally fixed; only the top surface boundary was free.Eight-node isoperimetric hexahedral elements were used to represent the soil mass.Four-node shell elements with reduced integration points were used to model the plates of the surrounding buildings and the underground structure.Two-node line elements were used to simulate the beams, columns, foundations, and support structures.An elastoplastic constitutive model is needed for the elastoplastic stress and deformation analysis of the soil.
e numbers of nodes and elements of prototype models are about 31620∼34340 and 29195∼32369, respectively.e material parameters of the soil are same as each scale model test, the concrete floor of adjacent buildings is C25 (the elastic modulus is 2.8 × 10 4 MPa), and the underground  In this paper, the Drucker-Prager elastoplastic constitutive model was used for natural soil in the numerical analysis.Surrounding buildings, underground structure, and support structures were assumed to have a linear-elastic behavior.Key construction conditions of the scale model test are simulated, such as the initial state, excavation of cover soil, construction of support structure, excavation of soil, construction of underground structure, and back ll of cover soil.
Comparison of horizontal displacement of existing buildings between the experimental and numerical simulation results is shown in Figure 22, where the solid lines and dotted lines represent the results of the scale model tests and numerical calculations respectively.e simulated values of the horizontal displacements of the adjacent buildings during the excavation construction are larger than the results of the scale model tests, since the nite element models are the original size for each scale test models, and the horizontal displacement curves of nite element models obviously change as the soil excavation.Furthermore, the numerical results are about two to ve times of the experimental results, in uenced by the factors such as the discreteness of sand soil and the sti ness of frame structure model, and the trend of horizontal displacement is consistent.Also, it can be seen that the maximum horizontal displacement of the left frame structure is less than 5 mm, and the maximum horizontal displacement on the bottom is less than 2.5 mm. e horizontal displacement of the right frame structure changes obviously due to the small structural load relatively, and the maximum horizontal displacement on the top is about 12.2 mm, the maximum value on the bottom is about 4.9 mm.
As shown in Figure 23, it can be seen that the vertical displacement curves of scale tests and numerical results change in a similar way except the frame structure nearest to the excavation on the right side in tests III and IV, and the di erence between the scale tests and the results of nite element models is small.However, the maximum vertical displacement of the numerical results in test III and test IV is 8.3 mm and 9.3 mm, respectively because the load of the frame structure on the right side is smaller and closer to the excavation relatively.
From the above comparative analysis, we can see that the changes of the data curve measured by the eight parallel scale model tests can reflect qualitatively the influence of underground excavation construction on the adjacent buildings and the soil surrounding the excavation with different influence parameters, such as the disturbance of the soil by construction of the support piles, the settlement of the existing structure close to the excavation, and other phenomena.
Although the indoor scale model tests satisfy the geometric similarity based on the first similarity theorem, the experimental results have a certain degree of dispersion and do not fulfill the strict proportional relationship with the numerical calculation results due to the poor cohesion of the sand soil in the model box and the stiffness of the frame model itself.e influence zone induced by the underground excavation is mainly concentrated at the bottom and both sides of the excavation.e finite element results of the prototype excavation construction show that the soil in the 7.5 m area directly under the excavation is disturbed to varying degrees, and the maximum vertical displacement value of soil reaches to 11 mm.e disturbance of the soil on both sides of the underground excavation is mainly concentrated within 1.5 m from the sidewall of the underground structure due to supporting effect of contiguous piles.Furthermore, the influence of excavation construction decreases with the increase of the relative distance to the excavation, and the change trend is consistent with the results of the indoor scale tests.

Conclusion
is paper presents the impacts of underground excavation on adjacent existing buildings and surrounding soil through eight parallel scale model experiments and numerical simulation when considering different key influence parameters, which are the size of support piles, the load and layer of adjacent existing buildings, and the relative horizontal distance from the excavation, respectively.e main findings from this work are summarized as follows.
e experimental data show that the horizontal and vertical displacement variations of the existing structures are small, and the fluctuations are within a relatively small subrange.Besides, the displacement variation of the adjacent building models related to the stiffness of the structure from the scale model.e obtained soil pressure from experiments reflected the pressure changes of surrounding soil in approaching construction of underground structure.
e results show  Advances in Civil Engineering that the soil pressure has different changes at each parallel scale model test with different key influence parameters.e disturbance of the surrounding soil related to the size of support piles, and the soil pressure under the excavation is affected by the unloading of the excavated soil and the loading of the underground structure; also the peak value of the fluctuation is greater than the soil under the adjacent building foundation.
Moreover, not only the measured data but also the prototype finite element models can qualitatively reflect the influence of underground excavation construction on the adjacent buildings and the soil surrounding the excavation with these key influence parameters.
is work can be extended in future to explore the displacement and the soil pressure when consider other influence parameters such as the depth and width of the excavation and the soil parameters.We believe that apart from looking for the relationship with influence parameters, future research should look for the disturbance of soil as underground structure construction.

Figure 7 Figure 1
Figure 1: e concrete test box.

Figure 10 :Figure 11 :
Figure 10: Existing building models with load blocks: (a) support structure and excavation; (b) the underground structure; (c) backfill of cover soil.

Figure 12 :
Figure 12: Comparison of horizontal displacement of existing buildings: (a) the top of the left building; (b) the bottom of the left building; (c) the top of the right building; (d) the bottom of the right building.

Figure 13 :Figure 14 :
Figure 13: Comparison of vertical displacement of existing buildings: (a) left building farthest from excavation; (b) left building nearest from excavation; (c) right building farthest from excavation; (d) right building nearest from excavation.

Figure 15 :
Figure 15: Relative vertical displacement of adjacent buildings: (a) relative vertical displacement of the left building; (b) relative vertical displacement of the right building.

Figure 22 :
Figure 22: Comparison of horizontal displacement of existing buildings between experimental and numerical results: (a) the top of left building; (b) the bottom of left building; (c) the top of right building; (d) the bottom of right building.

Table 2 :
e steps of dial indicators data acquisition.

Table 3 :
e steps of soil pressure sensors data acquisition.