Application of near infra-red Spectroscopy in Engineering Soil Classification

Abstract

Soil classification is very important in engineering applications since it categorizes soil into groups of similar characteristics. There are various classification systems that have been developed, but the most commonly used in engineering soil are the AASHTO and USCS. These classification schemes are based on the particle size, liquid limit and plasticity because these index properties of soil are fairly stable. At the moment the laboratory methods for obtaining these soil properties are costly and time consuming hence making the engineering classification an involving process. This study developed rapid protocol for expediting engineering classification of soil through the use of near-infrared (NIR) spectroscopy. NIR spectroscopy in determining chemical, organic and physical composition of soil has been proven to be useful; which is rapid, economical and reduces drudgery. This research focused on the application ofthe NIR within the range of 350nm and 2500nm and was found with over 67% correlation with soil - properties used in engineering classificationif'h.¢ correlation encouraged the application of near ..• infrared spectral reflectance in identifying engineering classes of soil. More than 70% overall correct identification of AASHTO engineering classes of soil was registered. This predictive accuracy demonstrated that near infrared spectral reflectance can be useful in classifying soil into different engineering classes. Together with the time and cost gains that researchers have

recently shown on spectral reflectance over convectional soil testing, the results from this study reflectance. Furthermore, the spectral reflectance was able to detect subtle differences between neighbouring engineering classes which could have been the result of hard boundaries imposed by the existing engineering classification schemes. Consequently, reconstruction of classes in engineering classification was suggested by use of near infrared spectral reflectance.