Mesoscale modeling of recycled aggregate concrete under uniaxial compression and tension using discrete element method

This paper presents a computational study to underline how the uniaxial behaviors of recycled aggregate concrete (RAC) are affected by various factors. Particularly, RAC is modeled using a discrete element method (DEM) wherein each phase of RAC (i.e., new and old aggregates, mortars, and in-between interfaces) is represented explicitly. By doing so, the mesoscopic cracking process of RAC can be accurately traced. The development, calibration and validation of the modeling method are presented. Then a parametric study is conducted, considering the following major factors: the attributes of new mortar and old mortar, the old mortar content, the severity of damages produced in manufacture of recycled aggregate (RA), and the bond property of interfacial transition zones (ITZs). It is shown that the relative strength of old mortar to new mortar plays a vitally important role in defining the failure mechanism of RAC. Specifically, if old mortar is stronger than new mortar, cracking initiates at their interfaces. By contrast, for relatively weak old mortar incipient cracks arise instead from inside of RA particles, localizing mostly at old aggregate-old mortar ITZs. Yet in some extreme cases with very weak old mortar, early damages are trapped in old mortar alone. Those different origins of crack formation affect the macro-level behavior of RAC considerably. Moreover, the relative amount of old mortar has a significant effect-a higher content of such constituent reduces both strength and modulus of RAC, but concurrently delays the onset of mesoscale cracking. Results also show that a weak ITZ between RA and new mortar confers a pronounced reduction in RAC's mechanical properties; conversely, the properties' enhancement from improving those ITZs is capped though. This research illustrates how the developed DEM-based approach can be conducive to a better understanding of RAC's mechanical response to uniaxial loadings. (C) 2020 Elsevier Ltd. All rights reserved.