Current Projects:
Continuum models:
Strain localization and size effect in metals and nano-metallic laminates.
Creep, fatigue and creep-fatigue interaction in Nickel superalloys.
Data-driven modelling of constitutive equations in metallic and hyperelastic materials.
Time-averaging of molecular dynamics to study mechanical response at slow time scales.
Past projects:
Structural materials used inside nuclear reactors undergo extreme irradiation at very high temperature which causes severe degradation of their mechanical strength. Irradion damage at the microscale results in the formation of defects such as vacancy clusters, interstitial loops, voids and dislocation loops. At the large-scale, this manifests in the form of phenomena such as embrittlement, swelling and hardening which contributes to significant loss of mechanical strength. In this project, we propose a multiscale coupling framework of microscopic defect kinetics with macroscopic crystal deformation. The effect of neutron irradiation dose and temperature on the thermomechanical properties of tungsten is investigated. The model is validated using widely available data of uniaxial tension and nano-indentation tests of irradiated materials.
Relevant Paper: Qianran Yu, Sabyasachi Chatterjee, Kennethe Roche, Giacomo Po, Jaime Marian, Coupling crystal plasticity and stochastic cluster dynamics models of irradiation damage in tungsten, Modelling and Simulation in Materials Science and Engineering, (2021), 29:055021 (doi: 10.1088/1361-651X/ac01ba)
Load-depth curve
Indentation test simulation
Plasticity is inherently a complex and multiscale process. What happens at the small scale significantly governs the large-scale behaviour. Crystal plasticity simulations using finite element method has paved the way for major breakthrough in unravelling the behaviour of metals at high stress and temperature as well as cyclic loading. However, inspite of the success, the phenomenological assumptions used in these simulations act as their prime limitations and render them not entirely predictive. The objective of this work is to bridge the gap between microscopic dislocation dynamics simulations with large-scale crystal plasticity models. The biggest challenge is the separation in length and time scales that these models operate in. The work presented in Chatterjee et. al. (2020) defines one such approach in which averages obtained from dislocation dynamics simulations are used to replace the phenomenological assumptions in crystal plasticity. Other possible approaches may include using data driven methods.
Stress-strain curve
Stress
Constitutive material models are approximate in nature and are therefore limited in their predictive capabilities. A data-driven surrogate constitutive model offers two advantages. Firstly, they are trained using data from microscopic models, which are more accurate. Secondly, they reduce the computational cost of the microscopic models and therefore are less restricted in their length scale of application. The developed surrogate model must also be physically reasonable and satisty the basic governing equations and the laws of thermodynamics. We are aiming to develop such data-driven models for hyperelastic and elastoplastic materials.
Nickel-based superalloys are high temperature materials with improved strength and creep resistance, typically used in gas turbines, jet engines and other applications. At the micro-scale, plastic deformation occurs due to the motion of line defects called dislocations. The improved strength of superalloys is due to the presence of precipitates which act as obstacle to the motion of dislocations. In this project, the correlation between various dislocation bypass mechanisms and their associated bypass stress with precipitate size, volume fraction and lattice misfit are investigated using a computational modeling approach called 3D discrete dislocation dynamics. A generalized gamma surface is constructed which is used to obtain the stacking fault forces acting on dislocations.
Relevant Paper: Sabyasachi Chatterjee, Yang Li, Giacomo Po, A discrete dislocation dynamics study of precipitate bypass mechanisms in nickel-based superalloys, International Journal of plasticity, (2021), 145:103062 (doi: 10.1016/j.ijplas.2021.103062)
A long-standing limitation in the use of molecular dynamics (MD) simulation is that it can only be applied directly to processes that take place on very short timescales. Many important processes in chemistry, physics and materials science take place on time scales that cannot be reached by molecular dynamics, which is limited to nanoseconds (or a few microseconds for very small systems). Often, these processes can be characterized by infrequent activated events. The dynamics on these timescales is typically characterized by infrequent-event transitions, from state to state, usually involving an energy barrier.
In this project, we plan to apply a recently developed time-averaging framework to accelerate Molecular Dynamics simulations. This can help us understand many real-world practical problems, such as thin-film deposition and crystal-growth from microscopic fundamentals.
*Image adapted from Tan et. al. (2014)
