Test Studies on Geogrid–Soil Interface Behavior under Static and Dynamic Loads

Abstract

Pullout tests on geogrids have been regarded as the most direct way to investigate geogrid–soil interaction. In the pullout tests on geogrids, either static or dynamic load is commonly used for applying the vertical loads. In order to investigate the influence of static and dynamic load on pullout test results, pullout tests are carried out to analyze the mechanical response of geogrids and soils under static and dynamic load from the large-scale pullout tester. The results show that frequency and amplitude have significant effects on the pullout test results under the dynamic load. The interface cohesion and friction coefficients under dynamic loads are smaller than those under static loads. The reinforcement effect of geogrids is reduced by dynamic load. Therefore, the strength of geogrids should be reduced when quasi-static analysis is used for reinforced structures. Knockdown factor is recommended for the corresponding reduction. The investigation results of this study may provide scientific references for regulating the design method of reinforced structures.

1. Introduction

Geogrid, an important geosynthetic material, is widely used in reinforced earth retaining walls, bridge approaches and embankment side batch-retaining structures. The interface parameters between geogrid and filler are the key indicators that characterize the reinforcement effect of geogrid, which determines the safety and stability of reinforced soil engineering. Therefore, the characteristics of the geogrid–soil interface are important factors [1] affecting the reinforcement mechanism of geogrid. There are still some deficiencies in the design method of reinforced soil structure, and the understanding of the interface mechanism of reinforced soil is not comprehensive enough. Domestic and foreign scholars have carried out direct shear tests [2,3,4,5], biaxial compression tests [6] and triaxial compression tests [7,8]. The pullout test is one of the important ways to study and reveal the characteristics of the reinforced soil interface, and is adopted by many researchers [9,10].

In the pullout tests, the type of geogrid, mesh shape, longitudinal and transverse rib spacing, and filler have significant influence on the pullout test results. Abdelmawla [11] studied the interaction between high density polyethylene (HDPE) and polyethylene terephthalate (PET) geogrids and silica sand and calcareous sand. Zheng [12] analyzed the influence of geogrids with different mesh shapes on the mechanism of the reinforcement–soil interface. Cao [13] and Zheng [14] carried out indoor pullout tests of three-dimensional geogrids to study the deformation of geogrids and the characteristics of the reinforcement–soil interface during pullout. Cardile [15] analyzed the influence of geogrid transverse ribs on the pullout force in the pullout test of geogrid and sand. Through pull-out test and DEM analysis, Wang [16] found that geogrid ribs bear 90 % and 70 % of pull-out force respectively. Pant [17] studied the influence of transverse ribs and longitudinal ribs on the pullout characteristics of PET geogrid, and the pullout force increased with the increase of transverse ribs number and normal stress. Through a series of pull-out tests of geogrid, Jin [18] and Zuo [19] deeply analyzed the mechanical behavior of geogrid–soil interface considering the effect of ribs spacing and boundary conditions.

Due to the limitations of test equipment, test conditions and other factors, the interface characteristics of reinforced soil are mostly studied under normal static load, and the interface characteristics and stress transfer mechanisms of reinforced soil under normal dynamic load are relatively few. Koshy [20] analyzed the influence of stress level and frequency of normal dynamic load on the pullout characteristics of geogrid. Ferreira [21,22] studied the pullout force and strain of geogrid under monotonic and cyclic pullout loads. Razzazan and Cardile [23,24,25] analyzed the influence of normal stress, load cycles, load frequency and amplitude on the pullout test.

The existing research result of the interaction mechanism of the reinforced–soil interface was mostly obtained under normal static load conditions. However, with the applicability of reinforced-earth retaining structures in high-speed railways, clearly put forward in the Code for Design of High-Speed Railway (TB10621-2014) and Code for Design of Intercity Railway (TB10623-2014), reinforced-earth technology is used more and more in railway engineering. It is necessary to study the mechanism of the reinforced soil interface under normal cyclic load from the perspective of safety or economic rationality. In view of this, this paper conducts an in-depth and systematic study on the interface effect of geogrid reinforcement and the geogrid-reinforcement effect under static and dynamic loads.

2. Materials and Methods

2.1. Pullout Equipment

In this paper, the large-scale pullout equipment was self-made by Shijiazhuang Tiedao University. The pullout equipment was mainly composed of four parts: pullout box, normal loading system, horizontal loading system, data acquisition system. The pullout box was made of 10 mm thick steel plate and was 600 mm in length, 400 mm in width, and 500 mm in height in internal dimensions. Moreover, to increase the rigidity of the pullout box, rectangular steel bars were added around the box to prevent the side wall of the test box from deforming during test loading. The reaction frame was arranged above the box to fix the normal loading equipment and apply normal pressure to the soil in the box. In the middle of the front and rear end of the test, a slot was reserved to extend the geogrid outside the test box, in order to connect a clamp on the front with a displacement gauge on the back. The normal loading was imposed by the inverted hydraulic jack and transferred to test soil through the loading board.

Normal stress included static and cyclic dynamic loads. The horizontal loading system was composed of a tension sensor and front and rear displacement sensors. The maximum normal stress and horizontal pullout force could reach 800 kPa and 100 kN, respectively. Pullout rate ranged from 0 to 30.0 mm/min under static and dynamic loads, and horizontal pullout displacement could reach 150 mm. The pullout test data could be collected by a data acquisition system connected to the computer, to achieve the real-time monitoring and recording during the pullout test. A schematic diagram of large-scale pullout equipment and a detailed drawing of the pullout box are shown in Figure 1.

2.2. Tested Material

The uniaxial geogrid, composed of high-density polyethylene (HDPE), was used in the current research, as shown in Figure 2. The spacing of geogrid transverse ribs s was 25 cm, and the thickness of transverse ribs h was 4 mm. Effective length L and width S of geogrid pullout sample in the test box were 60 cm and 30 cm, respectively, as shown in Figure 2. The mechanical indexes of the geogrid were showed in

Table 1. Mechanical property indexes of geogrid.

Specification	Tensile Strength/(kN/m)	Tensile Strength of 2% Strain/(kN/m)	Tensile Strength of 5% Strain/(kN/m)	Peak Strain/ %
TGDG 90	98.38	33.25	60.54	11.28

.

In the pullout test, coarse sand was employed as infill material. According to ASTM D6913/D6913M-17, the curve of the grain size distribution of coarse sand was obtained using sieve analysis, as shown in Figure 3. The direct shear experiment was used to determine the cohesion and friction angle, and a standard Proctor test was used in the compaction experiment to determine the maximum dry density, and the indexes are presented in

Table 2. The physical and mechanical indexes of soil.

Soil Particle Proportion	Minimum Dry Density/cm3	Maximum Dry Density/cm3	Void Ratio	Uniformity Coefficient/Cu	Curvature Coefficient/Cc	Cohesion/kPa	Internal Friction Angle/°
2.65	1.42	1.83	0.74	5.31	1.42	1	26.5

.

2.3. Experimental Program

The geogrid–soil interface behavior was studied by a pullout test of geogrid under static and dynamic loads. The waveforms used for normal cyclic load in this test were sine waves, and the specific test scheme is shown in

Table 3. Experimental Program under static and dynamic load.

Loading Type	Normal Stress/kPa	Frequency/Hz	Amplitude/kPa	Pullout Velocity/mm·min−1
Static Load	80, 100, 120	—	—	2.0
Dynamic Load		2	20	2.0
	60~80	4
		6
		2	40	2.0
	60~100	4
		6
		2	60	2.0
	60~120	4
		6

Note: In Table 3, under dynamic load, the initial normal stress is 60 kPa, the amplitude is 20 kPa and the corresponding lower and upper limits are 60 kPa and 80 kPa, respectively.

. The expression of normal stress was σ = c under normal static load and σ = σ₀ + Asin2πft under dynamic load, where σ is normal stress, c is normal stress constant of static load, σ₀ is the valley value of sinusoidal cyclic load, A is the amplitude, f is the frequency, t is loading time.

The height of infill sand in the test box was 50 cm and the upper and lower parts of the geogrid were 25 cm, respectively. Coarse sand was compacted in four layers, and compaction degree was controlled to 90%. According to the Test Code for Geosynthetics in Highway Engineering (JTG E50-2006), the geogrid was pulled out at a constant rate of 2 mm/min until one of the following conditions was achieved: (1) Tensile failure of geogrid. (2) The maximum displacement of the clamp up to 100 mm. The inner surface of test box was smeared with lubricant, and it was ensured that the geogrid had a certain distance from the side wall of the box.

3. Results and Discussion

This section describes and discusses the test data and conclusions obtained from laboratory pullout tests of HDPE geogrid-reinforced coarse sand under static or dynamic load. Firstly, the effect of different static loads on the mechanical characteristics of the geogrid–soil interface, including three kinds of normal stress, were investigated. Meanwhile, with the rise in normal stress, the geogrid–soil interface parameters were also studied. Then, the curve characteristics of dynamic load and the variation laws of parameters of geogrid–soil interface under different frequencies and amplitudes were analyzed. Finally, the effects of static and cyclic dynamic load on the reinforcement effect of the geogrid were compared and analyzed.

3.1. Pullout Test Results under Static Load

3.1.1. Effect of Normal Stress on the Mechanical Response of the Geogrid–Soil Interface

Figure 4 describes the curve of pullout displacement vs. pullout resistance obtained from the pullout test of HDPE uniaxial geogrid-reinforced coarse sand under various normal stresses.

The curves of pullout displacement–resistance increased dramatically in the beginning stage of pullout test, as shown in Figure 4. In the middle stage of pullout test, when pullout displacements were 3 mm~20 mm, the pullout resistance curve increased but the growth rate slowed down gradually, and the effect of pullout resistance became more and more prominent. In the later stage of the pullout test, after pullout displacement reached 20 mm, the curve gradually began to level off, the pullout resistances were close to peak, and as a whole, the relationship curves show strain hardening. When the normal stress increased from 80 kPa to 120 kPa, the maximum pullout resistance increased gradually while the corresponding maximum pullout displacement decreased gradually. The reason was that under the lower normal stress, the embedding effect of geogrid on soil particles was weaker, leading to the smaller maximum pullout resistance and the larger pullout displacement.

The curves between pullout resistance and normal stress of geogrid are shown in Figure 5. It can be seen that the pullout resistance of geogrid increased gradually when the same pullout displacement was reached. The reason was that sand particles tended to be dense with the increase of normal stress, resulting in the gradual increase of the reinforcement effect of geogrid and the greater reinforcement–soil interface effect. Meanwhile, pullout resistance also increased with the increase of pullout displacement, which showed that the effect of the geogrid–soil interface was a gradual development process with the pullout test.

In the design of the reinforced soil structure, geogrids were subjected to different normal stress levels in the reinforced-earth retaining wall of different heights. Therefore, the maximum pullout resistance was also different, meaning that the strength parameters of the reinforcement–soil interface were not at a constant value. Therefore, it is suggested that the influence of retaining wall height should be considered when determining the strength parameters of the geogrid–soil interface in the code for design of reinforced soil structures.

3.1.2. The Parameter of the Geogrid–Soil Interface under Static Load

(1) Shear strength of geogrid–soil interface

The shear stress and normal stress of the geogrid–soil interface was linearly fitted according to the Mohr–Coulomb strength theory, as shown in Figure 6. The shear strength parameters of geogrid–soil interface were c_gs = 15.27 kPa and φ_gs = 5.95°. Relative to the coarse sand, the cohesion was increased but the internal friction angle decreased in geogrid–soil composite.

(2) Friction coefficient of geogrid–soil interface

The friction coefficient of the geogrid–soil interface was not only related to internal friction angle but was also affected by normal stress, which was a comprehensive strength parameter for evaluating the reinforcement effect. The friction coefficient curve of the geogrid–soil interface under different normal stress levels is presented in Figure 7.

It can be seen that the friction coefficient of the geogrid–soil interface decreased with the increase of normal stress, which is due to the shear dilatancy of dense sand under low normal stress. Soil particles gradually moved or rotated with the pullout of geogrid, and shear dilatancy of the sand was obvious under the lower normal stress. With the increase of normal stress, shear dilatancy is weakened as the movement of the sand particles was limited, and the interface friction coefficient was gradually reduced. The friction coefficient of the reinforcement–soil interface was not a fixed value but changed with the change of normal stress. Therefore, the reinforcement effect should not be evaluated only by the interface friction coefficient, but also by normal stress. The research results provided a reference for the interface friction coefficient in the design of a reinforced soil retaining wall. For example, the French Standard Guide (NF P 94–270) suggested that the different interface friction coefficients when the retaining wall height was greater than 6 m were because the influence of sand dilatancy must be considered, as the normal stress increased with the increase of wall height.

3.2. Pullout Test Results under Cyclic Dynamic Load

3.2.1. Effect of Amplitude on the Characteristics of the Geogrid-Soil Interface

The plots of the pullout displacement vs. pullout resistance are presented in Figure 8. Where the frequency of cyclic dynamic load was 2 Hz, 4 Hz and 6 Hz, the corresponding amplitudes were 20 kPa, 40 kPa and 60 kPa, respectively. The curves of pullout displacement-resistance increased dramatically in the beginning stage of the pullout test, and then the increasing trend of pullout resistance gradually slowed until the change trend was basically flat. The curve change trend under cyclic dynamic load was generally consistent with that of static load. Under the normal cyclic load, the curves showed a significant wave cyclic change trend, and the distance from wave peak to wave valley gradually increased with the increase of amplitude; that is, the wave cyclic vibration trend was more obvious. Therefore, the amplitude of cyclic dynamic load had a significant effect on the characteristics of geogrid–soil interface.

The curves of the maximum pullout resistance of the peak and valley with the amplitude are shown in Figure 9. When the corresponding frequencies were 2 Hz, 4 Hz and 6 Hz, the maximum peak values were 5.71 kN, 6.54 kN and 7.06 kN at amplitude 20 kPa, and 6.48 kN, 7.34 kN and 8.31 kN, and at amplitude 60 kPa, they increased by 13.48%, 12.23% and 17.7%, respectively. The maximum values of both peak and valley increased with the increase of the amplitude, in that cyclic load amplitude had an important impact on the characteristics of the geogrid–soil interface.

3.2.2. Effect of Frequency on the Characteristics of the Geogrid–Soil Interface

The curve of the pullout displacement vs. pullout resistance is presented in Figure 10. The amplitudes of cyclic dynamic load were 20 kPa, 40 kPa and 60 kPa, and the corresponding frequencies were 2 Hz, 4 Hz and 6 Hz, respectively. The range of wave oscillation with frequency at 2 Hz was significantly greater than that with frequency at 6 Hz. If the amplitudes were fixed, the smaller the frequency, the greater the impact of cyclic dynamic load on the characteristics of the reinforcement–soil interface, and vice versa. In conclusion, the influence of cyclic dynamic load on the interface characteristics increased as frequency got smaller.

The curves of the maximum pullout resistance of the peak and valley with the frequency are shown in Figure 11. When the corresponding amplitudes were 20 kPa, 40 kPa, and 60 kPa, the maximum valley values were 5.52 kN, 6.45 kN, and 6.92 kN at frequency 2 Hz, and 5.96 kN, 7.16 kN, and 7.97 kN at frequency 6 Hz, increasing by 7.98%, 11.01%, and 15.17%, respectively. The maximum value of both peak and valley increased with the increase of the frequency, in that cyclic load frequency had an important impact on the characteristics of the geogrid–soil interface.

3.2.3. The Parameter of the Geogrid–Soil Interface under Cyclic Dynamic Load

(1) Shear strength of geogrid–soil interface

The shear strength curve of the geogrid–soil interface under cyclic dynamic load is shown in Figure 12. The peak and valley of the maximum shear stress had a good linear relationship with normal stress, which approximately conformed to the Mohr–Coulomb strength criterion. The peak and valley of shear stress increased with the increase of frequency, and the difference between the peak and valley of shear stress decreased with the increase of frequency.

According to the Mohr–Coulomb strength theory, the shear strength parameters of the geogrid–soil interface under cyclic loading are shown in

Table 4. Interface strength parameters of geogrid-reinforced interface.

Load Type	Frequency/Hz	Cohesion/kPa	Friction Angle/°
Static load	—	15.27	5.95
Cyclic dynamic load (peak value)	2	11.14	3.82
	4	13.613	3.54
	6	10.891	6.19
Cyclic dynamic load (valley value)	2	11.938	2.88
	4	14.607	2.59
	6	11.094	5.94

. The interface cohesion and interface friction angle corresponding to the peak or valley of cyclic dynamic load were less than that under static load. It can be concluded that compared with static load, cyclic dynamic load weakened the reinforcement effect of geogrid.

(2) Friction coefficient of geogrid-soil interface

The curve of friction coefficient vs. normal stress under cyclic dynamic load is shown in Figure 13. The interface friction coefficient of both peak and valley decreased with the increase of amplitude under the same frequency. When frequency was 2 Hz, amplitude increased from 20 kPa to 60 kPa, interfacial friction coefficient corresponding to peak decreased from 0.221 to 0.175, with the reduced rate of 20.81%, and the friction coefficient corresponding to the valley decreased from 0.229 to 0.182, with the reduced rate of 20.52%. For frequency 6 Hz, amplitude increased from 20 kPa to 60 kPa, interfacial friction coefficient corresponding to peak decreased from 0.223 to 0.206, with the reduced rate of 7.6%, and friction coefficient corresponding to the valley decreases from 0.246 to 0.209, with the reduced rate of 15.04%. Therefore, the smaller the frequency, the more obvious the influence of cyclic dynamic load on the parameters of the geogrid–soil interface. In addition, the friction coefficient under cyclic dynamic load was less than that under static load, which showed that cyclic dynamic load weakened the reinforcement effect.

According to the variation of strength parameters and friction coefficient under cyclic dynamic load, it can be concluded that the strength of geogrid should be reduced when quasi-static analysis is used to reinforced soil structure design, and the reduction coefficient was 0.85.

The curve of interfacial friction coefficient with amplitude under cyclic loading is shown in Figure 14. The interfacial friction coefficient decreases with the increase of amplitude. When the amplitude was 20 kPa, the friction coefficient corresponding to the peak or valley value was greater than that when the amplitude was 60 kPa. The variation law of the interfacial friction coefficient conforms to the progressive pullout mechanism of the geogrid–soil interface.

4. Conclusions

The pullout test of HDPE uniaxial geogrid under static and dynamic loads was carried out by self-made, large-scale pullout equipment. The relationship between the pullout resistance and displacement of the geogrid was described macroscopically. The variation law between pullout resistance vs. pullout displacement and interaction mechanism of the geogrid–soil interface were studied in depth. The following conclusions were obtained.

(1) Under static load, the maximum pullout resistance of geogrid increases with the increase of normal stress. The normal stress significantly affected the characteristics of the geogrid–soil interface. The influence of the height of the reinforced soil wall on the strength of geogrid should be considered.

(2) Under cyclic dynamic load, the pullout resistance and displacement curve was basically consistent with the overall trend under static load, but it showed obvious wave-cyclic characteristics. Cyclic dynamic load significantly weakened the reinforcement effect of geogrid.

(3) Under cyclic dynamic load, the interfacial cohesion and interfacial friction angle corresponding to peak or valley values were less than those under static load. The smaller the frequency, the more obvious the influence of cyclic dynamic load on the interface parameters. The interfacial friction coefficient decreased with the increase of amplitude. When quasi-static analysis is used in reinforced soil structure design, the interface design parameters should be reduced and the reduction coefficient can be about 0.85.