The capabilities of the Plaxis Shotcrete material model for designing laterally loaded reinforced concrete structures in the subsurface
More Info
expand_more
Abstract
A hot topic in the Netherlands is the safety reassessment of dikes. Currently several (river)dikes are being reinforced to withstand a given normative high water level. The environment of those dikes is often characterised by small floodplains and urban areas right next to the dike, wherefore standard geo-technical measures cannot be applied. Installing a diaphragm wall in the crest of the dike is in such a case a measure that can guarantee the safety, by preventing the development of a shear plane when the dike is subjected to a normative high water level. In engineering practice modelling of those structures is simplified by using a linear elastic plate element with a reduced stiffness, to account for concrete cracking upon loading, or a linear elastic-perfectly plastic volume element in the geotechnical finite element code Plaxis. In this master thesis attention will be paid to the possibility of modelling and designing such (geotechnical) reinforced concrete structures in Plaxis more accurately with the use of a new concrete material model. Recently the Shotcrete material model for volume elements is released by Plaxis. The Shotcrete material model accounts for non-linear and time dependent behaviour, i.e. more realistic behaviour, of concrete. The material model is however only been used for modelling unreinforced shotcrete in a NATM application. The capabilities of using the Shotcrete material model to design laterally loaded reinforced concrete structures in the subsurface is investigated. At first the theoretical applicability of the Shotcrete material model to design reinforced concrete structures in Plaxis 2D is being researched. Verification against guidelines and model codes for describing concrete behaviour showed that the material model has the prospect to model non-linear concrete material behaviour, including cracking, realistically. By means of analysing a statically determined reinforced concrete bending beam model it is found that using a by elastoplastic plate element(s) reinforced Shotcrete volume element has additional value over the use of a linear elastic and a linear elastic-perfectly plastic material model when modelling a reinforced concrete structure. Nevertheless it is observed that the results given by the Shotcrete model are slightly mesh sensitive. The use of a by elastoplastic plate element(s) reinforced Shotcrete volume element is subsequently verified by comparing the results to commonly for design used M-kappa and M-N-kappa hand calculations. It is demonstrated that in case the reinforced concrete structure is subjected to bending, as well as a normal force, the model gives accurate and realistic results. Deviations can be explained by differences in calculation method, hand calculation versus numerical model. When the structure is subjected to a normal force attention has to be paid to pre-stressing of the reinforcement and the resulting stresses in the concrete. An elastoplastic plate reinforced Shotcrete volume element is used to model a diaphragm wall next to an excavation. Evaluation of an excavation next to a diaphragm wall, such that structural bending failure is the normative failure mode, showed that using an elastoplastic plate reinforced Shotcrete volume element gives improved results with respect to commonly used modelling methods. The other used modelling methods are a linear elastic, respectively equivalent to the reinforced diaphragm wall M-kappa, plate element and a linear elastic volume element. A lower bending moment is obtained in the structure at the same deformation, which is partly the effect of higher (soil-structure) interface shear stresses. During the analysis it is found that obtaining the correct bending moment in the structure is not straightforward. It is found that only by taking an internal force equilibrium at the neutral line in the normative cross-section of the diaphragm wall the correct bending moment can be obtained. Applying the model in the Kinderdijk-Schoonhovenseveer (KIS) dike reinforcement project demonstrates the advantage of applying a more realistic, nevertheless more complex, way of modelling reinforced concrete structures in Plaxis 2D. The obtained bending moment in the structure increases with 25 % in SLS and decreases with 10 % in ULS with respect to modelling the diaphragm wall by means of a linear elastic plate element with a reduced stiffness, to account for cracking. To conclude, a stepwise approach is proposed to model a laterally loaded reinforced concrete structure in the subsurface realistically, by means of implementing a by elastoplastic plate elements reinforced Shotcrete volume element.