Artificial neural networks approach for swell pressure versus soil suction behaviour

Abstract

In this study, the swell pressure versus soil suction behaviour was investigated using artificial neural networks (ANNs). To achieve this, the results of the total suction measurements using thermocouple psychrometer technique and constant-volume swell tests in oedometers performed on statically compacted specimens of Bentonite–Kaolinite clay mixtures with varying soil properties were used. Two different ANN models have been developed to predict the total suction and swell pressure. The ANNs results were compared with the experimental values and found close to the experimental results. Moreover, several performance indices such as correlation coefficient, variance account for (VAF), and root mean square error (RMSE) were calculated to check the prediction capacity of the ANN models developed. Both ANN models have shown a high prediction performance based on the performance indices. Therefore, it can be concluded that the initial soil suction is the most relevant state of suction that characterizes the potential swell pressures.

The use of neural networks for CPT-based liquefaction screening

“…5, and the root mean square error RMSE, represented by Eq. 6, were also computed to assess the performance of the developed models (Grima and Babuska 1999;Finol et al 2001;Gokceoglu 2002;Erzin 2007;Erzin and Yukselen 2009;Erzin et al 2008Erzin et al , 2010 …”

“…This modeling capability, as well as the ability to learn from experience, have given ANNs superiority over most traditional methods since there is no need for making assumptions about what the underlying rules that govern the problem in hand could be (Shahain et al 2008). Since the early 1990s, ANNs have been effectively applied to almost every problem in geotechnical engineering (Shahain et al 2008), including constitutive modeling (Najjar and Ali 1999;Penumadu and Zhao 1999); geo-material properties (Ozer et al 2008;Park and Kim 2010); bearing capacity of pile (Das and Basudhar 2006;Park and Cho 2010); slope stability (Zhao 2008;Cho 2009;Erzin and Cetin 2012, 2014, shallow foundations Erzin andGul 2012, 2013), and tunnels and underground openings (Shi 2000;Yoo and Kim 2007). The ANN approach was also found to be suitable in the field of liquefaction potential assessment by various researchers such as Goh (1994Goh ( , 1996Goh ( , 2002, , , Wang and Rahman (1999), Barai and Agarwal (2002), Baziar and Nilipour (2003), Neaupane and Achet (2004), Baziar and Ghorbani (2005), Das andBasudhar (2006), Young-Su andByung-Tak (2006), Hanna et al (2007a, b), Rao and Satyam (2007), Ramakrishnan et al (2008), Farrokhzad et al (2010), Pathak andDalvi (2011), Moradi et al (2011), Kumar et al (2012), Venkatesh et al (2013).…”

Artificial neural network models for predicting soil thermal resistivity

“…(2), and the root mean square error (RMSE), represented by Eq. (3), were also computed to check the performance of the developed models [46][47][48][49][50]. …”

The use of neural networks for the prediction of cone penetration resistance of silty sands

“…The piezocone penetration test (CPTu) measures additional parameter that is the pore water pressure. These measurements can be effectively used for the following applications: (1) to classify soil identification, (2) to directly estimate pile capacity from the CPTu, (3) to evaluate soil properties through an appropriate correlation, especially the undrained shear strength, (4) to determine bearing capacity and settlement of the shallow foundations, (5) to control compaction in ground improvement, (6) to design wick or sand drains, and (7) to evaluate the soil liquefaction [2]. Therefore, the CPTu can be used for a wide range of geotechnical engineering applications [1].…”