Proceedings of the 6th World Congress on Civil, Structural, and Environmental Engineering (CSEE'21) Lisbon, Portugal Virtual Conference – June 21 – 23, 2021 Paper No. ICGRE 116 DOI: 10.11159/icgre21.lx.116 Material Modelling of Pare rock mass by Finite Element Back Calculation Technique Pawan Kumar Singh 1 , Diganta Goswami 2 , Ranjeet Bahadur Singh 3 , Pragati Goswami 4 , Sylvia Kashyap 5 12 34 5 Assam Engineering College Department of Civil Engineering, Assam Engineering College, Guwahati 781013, Guwahati, India singhpawank123@gmail.com, digantagoswami.ce@aec.ac.in, rbsinghntpc@gmail.com, pragatigoswami900@gmail.com, sylvia.kashyap@gmail.com Abstract-River Valley Projects possess great significance for the development of developing countries. They are used for various purposes like irrigation, hydroelectricity generation, water supply, flood control etc. In this regard the determination of the properties of the rock mass of the site becomes essential, which can be done with the help of in-situ tests as well as laboratory tests. But, conducting a greater number of in-situ tests in the field is not feasible, forcing one to use the tools like Finite element codes like Plaxis 3D AE for determining the properties of the rock mass. In this research work, an attempt has been made to derive the properties of the rock mass by back calculation technique using the results of In-situ Plate load test. An appropriate material modelling is very important in any Finite Element analysis to arrive at solutions close to the exact values, necessitating the selection of appropriate constitutive model. In this context, the suitability of the four constitutive models viz. Jointed Rock mass model, Hoek-Brown model, Mohr-coulomb model and Hardening Soil model, appropriate to Pare Rock mass is examined. Through this study, a systematic approach has been adopted for the simulation of rock mass properties of the Pare Hydroelectric Project site with the help of numerical modelling. Keywords: Pare Hydroelectric Project, Plate load test, Finite Element Method, Back Calculation Technique, Root Mean Square Error 1. Introduction The demand for power in India is rapidly increasing due to development requirements and so also the construction of more hydroelectric power projects. Construction of the various components of a river valley project tend to create instability in the rock mass, which is often water charged too, primarily due to alteration of in-situ stress conditions. For a safe design and execution of these elements, the behaviour of the geo-material involved, namely the rock mass, needs to be understood thoroughly and predicted well. As most of the hydro-electric power projects of India are in the Himalayan region and many more are likely to come up in this region, the behaviour of the Himalayan rock mass, with respect to construction of their various components such as dam foundation, powerhouse cavern, Headrace tunnels, surge shaft etc., needs to be understood and predicted with higher confidence level. In-situ tests like Plate load test, Tri-axial shear test, Direct shear test etc. represent the properties in a more sophisticated way than the other methods. However, multiple numbers of tests requires more resources which becomes an economic restrain, as the resources available are limited [1]. Further, the zone of influence in case of plate load test is limited only to a minimum depth from the top surface of rock formation. Therefore detailed evaluation of design models of the site cannot be made from this limited number of test results, encouraging the geotechnical engineers to go for numerical modelling or forces them to develop empirical correlations [2]. Back analyses are also often needed to be carried out to ascertain appropriate rock mass parameters like deformation modulus, sub-grade modulus, poisson’s ratio etc. The deflection profile is often taken into account in the determination of the in-situ elastic modulus while performing back calculation analysis [3].Singh &Geol (2011) [4] states that repeated cycles of back analysis and forward analysis eliminates many uncertainties in understanding the rockmass behaviour. The results obtained from back calculation techniques may differ depending on the assumptions, methods, and material models considered in the study [5]. So, a combination of forward calculation and backward analysis, from the field instrumented data capturing the actual field behaviour, during construction needs to be incorporated, to arrive at some practically correct geo-material properties. Inspite of the fact that the soil mass is highly heterogeneous, the notion of a single soil model to accurately predict the soil behaviour seems to be inappropriate [6].