Investigation of the effect of cyclic loading on the soil arching using damped spring-based trapdoor model

Numerical multi‐span segmented trapdoor test with viscoelastic boundary for soil arching within pile‐supported embankment under cyclic loading

“…The compromise of two competing effects makes the SAR similar in models of R‐STM and R‐TTM after unloading. In addition, soil arch is more unstable in cases of low‐spring stiffness and can be more easily disturbed under cyclic loading, 7 where most particles are directly driven under the cyclic loading to displace vertically and the effect of lateral boundaries on the soil arch is relatively insignificant. Therefore, the SAR is also observed similarly in F‐STM and R‐STM and all three models behave similarly in terms of SAR .…”

“…2,5 The dynamic wave reflection in both vertical and horizontal directions would cause great interference if the traditional trapdoor model was used to investigate soil arching under cyclic loading of the pile-supported embankment. 4,6,7 In the vertical direction, the numerical trapdoor model by the discrete element method (DEM) can be improved by including underlying soft soil particles to reflect the energy assumption of soft soil under cyclic loading 8,9 ; but the soft soil represented by numerous particles requires cumbersome parameter calibration and great calculation capacity. 10 Thus, the underlying soft soil was replaced by springs, and the model was improved to a spring-based trapdoor to investigate the soil arching without cyclic loading.…”

“…10 Thus, the underlying soft soil was replaced by springs, and the model was improved to a spring-based trapdoor to investigate the soil arching without cyclic loading. 10,11 Subsequently, the spring-based trapdoor model was further used to investigate the influence of cyclic loading on soil arching, 6,7 where the rebound of the spring-based trapdoors generates a negative soil arching to worsen the soil arch effect at post-loading, but the elastic spring was still unrealistic. In the horizontal direction, the lateral rigid wall was remotely placed to reduce the interference from the trapdoor, 6,7 where the distance from the rigid wall to the edge of the trapdoor was increased to about 2∼3 times the width of the trapdoor.…”

“…10,11 Subsequently, the spring-based trapdoor model was further used to investigate the influence of cyclic loading on soil arching, 6,7 where the rebound of the spring-based trapdoors generates a negative soil arching to worsen the soil arch effect at post-loading, but the elastic spring was still unrealistic. In the horizontal direction, the lateral rigid wall was remotely placed to reduce the interference from the trapdoor, 6,7 where the distance from the rigid wall to the edge of the trapdoor was increased to about 2∼3 times the width of the trapdoor. This treatment method, however, may exaggerate the width of stationary support and thus violate the design of the pile-supported embankment in terms of the ratio of pile width over pile net spacing.…”

“…12,13 Besides, a wider calculation domain would significantly increase the number of soil particles, thus exponentially increasing calculation work and seriously reducing computation efficiency. In addition, the localized load would result in a trough near the loading plate 2,7 and the resultant settlement would be unrealistic because of the absence of pavement structure. In summary, reasonable settings for the boundaries are the keys to ensuring the precision and efficiency of observations.…”