(4)

The energy absorption capacity (EAC) of earthen materials significantly influences the safety of civil projects. Furthermore, the development of machine learning techniques, including Artificial Neural Network (ANN) and Multiple Linear Regression (MLR) models, entails financial and non-financial benefits by reducing the need for performing expensive, exhausting and time-consuming laboratory tests. This study investigates the EAC of sandy soil reinforced by three different forms of processed lignocellulosic fiber pulps. The studied influence parameters included fiber type, curing time, effective confining pressure, and fiber content. Artificial neural network (ANN) models were developed to assess the EAC of the reinforced specimens and evaluate the impact of studied parameters. The analysis of each fiber type was carried out using Multiple Linear Regression (MLR) methods. The specimens, subjected to a 7-day curing period and reinforced with 2% of lignocellulosic fibers of 1.5 mm in length, exhibited the greatest EAC values. Sensitivity analysis identified effective confining pressure as the most influential factor on the EAC of the reinforced specimens. This study demonstrates the advantageous impact of processed lignocellulosic fibers, which are environmentally harmless substances, in enhancing the EAC of sandy soil and its ductility response. As a result, this decreases the likelihood of unexpected and catastrophic failures. This research also demonstrates the high capability of ANN-based models in predicting EAC at various influence parameters.

来源平台：INTERNATIONAL JOURNAL OF GEOSYNTHETICS AND GROUND ENGINEERING