4.3 Discussion on application of newly proposed model in predicting erosion from tests conducted in literature
To further verify the applicability of the proposed model, the available data from literature work is collected and performed the redetermination of erosion characteristics, as shown in Fig. 10 and Fig. 11. The studies of Indraratna et al. (2008) and Xie et al. (2018) recorded the temporal variation of the concerned parameters in the HET. The current model shows considerable consistency and adaptation in predicting erosion variation. The instantaneous recording included erosion rate, hydraulic shear stress, water pressure, eroded depth, and radius. Indraratna et al. (2008) plotted logarithmic coordinates to find the linear relationship between erosion rate/hydraulic shear stress and logarithm of time. Since the equations of concerned parameters are found as power functions in this study, it is reasonable to obtain the linearity in the logarithmetic scale. Although the experimental results of Xie et al. (2018) indicated that the hole enlargement was not perfectly symmetrical and uniform during continuous erosion due to the friction loss at the entrance of the hole (Říha and Jandora 2015), the erosion process appears to be generally predicted by the new model. It might be attributed that the energy loss at the entrance of the hole is minor as compared to the energy dissipation of sediment detachment and transport. Nguyen and Indraratna (2020) also suggested that a large percentage of the input energy in the fluid flow was dissipated by soil erosion. Further, the energy was transferred to the kinetic component of soil particles was negligible. Ouyang and Takahashi (2016), Fattahi et al. (2017), Jiang and Soga (2019), and Zhang et al. (2020) also performed HETs and explored internal erosion process on various soils. Cumulative eroded soil loss (from equation [36]), shows considerable performance of fitting. However, it should be noted that prediction ability cannot be judged completely since, the soil properties and experimental setups in the above studies were different. The erosion trend of these studies can be generally captured. However, the cumulative eroded soil loss at the final phase is slightly overestimated by proposed equations, as shown in Fig. 10 (d) and Fig. 11 (a-d). This might be resulted from the effects of sediment clogging in long-scale soil pipes (Wilson and Fox 2013). Soil particles were accumulated and entrapped in the latter part of soil pipes, when the residual fluid energy after dissipating in the sediment detachment could not support the entire process of sediment transport. Therefore, the estimated soil loss is marginally larger than the measured sediment and the water pressure is subtly higher than the prediction.
Seepage tests have been used to investigate the influence of internal erosion (Tomlinson and Vaid 2000). New model might be useful to estimate the settlement and strain of the soil body subjected to internal erosion by following the mass conservation law. The dynamic measurement of soil settlement and strain in the previous studies (Tomlinson and Vaid 2000; Marot et al. 2010; Ouyang and Takahashi 2016; Sato and Kuwano 2018; Indiketiya et al. 2019) was approximately observed to obey the equation [36]. This is because the deformation of soil body is attributed to the soil loss. If the subtle change of soil strength properties is ignored, the prediction of the soil loss would roughly explain the settlement and strain of the soil (Indraratna et al. 2009). Hence, the theoretical model might be significant to explain the mechanism and influence of erosion process in the HET, soil erosion, soil deformation, and even the instability of earth structures.