Investigation of Concrete Constitutive Models for Ultra-High Performance Fiber-Reinforced Concrete under Low-Velocity Impact

摘要

Among various alternatives developed for conventional reinforced concrete, ultra-high performance fiberreinforced concrete (UHPFRC) is proven to provide superior mechanical properties, making it appropriate for impact-resistant structures. While several experimental tests have been conducted to determine the static and dynamic response of UHPFRC, the simulation of this important category of concrete materials, especially under impact loads, is still in need of extensive effort. To address this issue, the current study investigates how two widely-used concrete constitutive models, i.e., Continuous Surface Cap Model (CSCM) and Karagozian and Case Concrete (KCC), can be reliably employed for modeling UHPFRC subjected to low-velocity impact. Utilizing the available experimental test data, a rigorous calibration process has been developed in the current study for the two constitutive models, capturing the main strength parameters, damage evolution parameters, and strain rate effects. This process begins with single-element simulations performed under various stress paths to generate the information necessary for post-peak softening and confinement factors. The investigations with single-element simulations consist of four element sizes to study the mesh size sensitivity of both constitutive models. The study is then extended to examine the capabilities of the calibrated models in simulating the response of full-scale structural elements made with UHPFRC. For this purpose, the drop hammer tests are replicated on both plain and reinforced UHPFRC, and the performance of the two constitutive models is evaluated in comparison to the experimental tests. Specifically, this evaluation examines peak impact forces and displacements, considering various hourglass coefficients and drop hammer heights. Furthermore, a metamodel-based sensitivity analysis is conducted to quantify the effects of uncertainty inherent in the input parameters on the predicted impact response measures, in terms of force, displacement, and duration.