An Eco-Sustainable Stabilization of Clayey Road Subgrades by Lignin Treatment: An Overview and a Comparative Experimental Investigation

Subgrade conditions significantly affect functionality of the road pavement during its service life. Among the different stabilization techniques for upgrading poorly performing in-situ soil subgrades, an economically attractive example involves the use of waste materials, such as lignin. A deep bibliographic analysis of previous studies is carried out in the first section of this paper. The literature review suggests that use of lignin as a stabilizing agent of road subgrade soils is not completely consolidated. In addition, this study reports an investigation on the strength and performance characteristics of a lignin-treated clayey soil. Several experimental tests were carried out on both the untreated and lignin-treated soils in order to shed some light on different aspects with limited knowledge available, such as the behaviour of the stabilised soil in specific conditions (e.g., the presence of water). Finally, the test results are discussed and compared with those obtained when the same soil is treated with lime, which is more widely used. The most relevant finding is the poor ability of lignin to upgrade the bearing capacity of the soil in wet conditions compared to lime; on the contrary, the presence of lignin helped in controlling the swelling potential of this type of soil.

Comparative Experimental Investigation of Deep Convolutional Neural Networks for Latent Fingerprint Pattern Classification

Fingerprint pattern recognition is of great importance in forensic examinations and in helping diagnose some diseases. The automatic realization of fingerprint recognition processes can take time due to the feature extraction process in classical machine learning or deep learning methods. In this study, the effective use of deep convolutional neural networks (DCNN) in fingerprint pattern recognition and classification, in which feature extraction takes place automatically, was examined experimentally and comparatively. Five DCNN models have been designed and implemented with a transfer learning approach. Four of these five models are Alexnet, Googlenet, Resnet-18, and Squeezenet pre-trained DCNN models. The fifth model is the DCNN model designed from the ground up. It was concluded that the designed DCNN models can be used effectively in fingerprint recognition and classification, and that fast results can be obtained and generalized with advanced DCNN models.

Comparative Experimental Investigation on Optimal Parametric Array Types

As a sound transmitting device that relies on the nonlinearity of a medium, a parametric array (PA) can generate high-directivity low-frequency signals using a small aperture transducer and high-frequency signals. Despite their relatively low source level, the PA is frequently used to measure the acoustic properties of materials in low-frequency regions owing to their high directivity in confined acoustic water tanks. Therefore, methods for improving the source level of secondary signals are of interest. Currently, there are two driving methods for PA: the dual-frequency PA and the broadband PA with amplitude modulation. In this study, we share the results of an elaborate and comparative experimental investigation of these two driving methods. Comparisons are made and discussed in terms of the intensity of the generated secondary signal and its characteristics in the frequency domain. Based on these factors, we confirmed that the broadband PA was more suitable as the sound source of the low-frequency characteristic measurement system of acoustic materials.