Embedded Files
Current Projects
High performance data-base crystal plasticity approach using Machine Learning
The rapid advancement of artificial intelligence and machine learning (AI/ML) techniques has brought about significant progress in their application to materials science. Particularly, in the realm of crystal plasticity finite element models, which are instrumental in predicting plastic responses, the complex and nonlinear nature of materials and geometries demands substantial computational resources. To address this challenge, a data-driven approach to crystal plasticity has been introduced, leveraging machine learning methods. This approach involves the creation of a database that captures constitutive variables and slip system details influencing equilibrium variables, particularly in plastic deformation, across various crystal orientations. Subsequently, this database is utilized in AI algorithms, resulting in a substantial acceleration of computational processes and a remarkable reduction in resource requirements.
Multi-scale Crystal Plasticity Finite Element Framework for Superalloys
In this project's initial phase, a hierarchical crystal plasticity constitutive model is developed to simulate both single and polycrystalline microstructures of Nickel-based superalloys. The model encompasses three distinct scales crucial for modeling polycrystalline Ni-based superalloys:
(i) the sub-grain scale capturing the γ-γ′ microstructure, characterized by the size and spacing of γ′ precipitates;
(ii) the grain-scale representing the size of individual crystals; and
(iii) the scale of polycrystalline representative volume elements.
At the sub-grain scale, a crystal plasticity model incorporating dislocation density is enhanced with mechanisms for anti-phase boundary (APB) shearing of precipitates. This sub-grain model is homogenized to derive parametric functions for morphological variables in evolution laws at the next level. An activation energy-based crystal plasticity finite element method (AE-CP FEM) model is formulated for the grain-scale, considering characteristic parameters of the sub-grain scale γ-γ′ morphology. An advantageous feature of this AE-CP model is its high efficiency, making it suitable for incorporation into polycrystalline crystal plasticity FE simulations while maintaining the accuracy of detailed sub-grain level representative volume element (SG-RVE) models. SG-RVE models are generated to account for variable morphologies, such as volume fraction, precipitate shape, and channel widths. Finally, at the next scale, a polycrystalline microstructure of Ni-based superalloys is simulated using an augmented AE-CP FE model.
The developed constitutive models are designed to capture the thermo-mechanical behavior of Nickel-based superalloys across a broad temperature range from 300 K to 1300 K. These models incorporate orientation dependencies, introducing asymmetry in tension and compression behaviors. This means that the response to tension and compression loads is not symmetrical for nearly all orientations.
Recently, the developed model has been employed in the study of Cobalt-based superalloys as potential alternatives to Nickel-based superalloys. The continuous advancements in nickel-based superalloys have brought their compositions close to their melting temperatures, necessitating the exploration of new high-temperature alloys. The discovery of γ-γ′ microstructure compositions similar to those in nickel-based superalloys has spurred significant research efforts to establish a new class of high-temperature, high-strength materials utilizing cobalt in the creation of cobalt-based superalloys. Experimental data indicates that the addition of certain elements to the composition can yield significant effects on the mechanical properties of Cobalt-based superalloys. Therefore, this study investigates the impacts of elements such as Tantalum, Titanium, and Chromium, addressing both the locking mechanism for the cross-slip of screw dislocations and the glide and climb mechanism to simulate diffusions at higher temperatures.
Multi-Scale Framework to simulate Superalloys
Implementing Large Deformation Crystal Plasticity Framework into Object Oriented Finite Element Software (OOF)
Presently, I collaborate with the OOF team at the National Institute of Standards and Technology (NIST) within the Materials Science Division, specifically in the Thermodynamics and Kinetics group. The focus of our work involves the implementation of a 3D Crystal Plasticity framework into OOF. OOF, a Finite Element software, is tailored for simulating microstructural-based problems. Its functionality includes the analysis of macroscopic properties derived from real or simulated microstructures. OOF processes images, assigns material properties to features within them, and conducts virtual experiments to ascertain the macroscopic properties of the microstructure. Subsequently, the Finite Element tool is employed to simulate the microstructure under various mechanical loadings.
Past Research
Simulation and Optimization of hot and cold isostatic pressing of metal powder
1. Simulation of Cold Powder Compaction Process: Simulating the cold powder compaction process with a focus on understanding the effects of container friction by utilizing finite element analysis (FEA) techniques to model the compaction behavior, taking into account container friction effects.
2. Development of Non-Associated Contact Frictional Model: Developing a frictional model specific to the non-associated nature of contacts in powder compaction by formulating a frictional model that accurately represents the non-associated contact conditions prevalent in powder compaction processes.
3. Topology Optimization of Cold Powder Compaction Process: Optimizing the topology of the cold powder compaction process to achieve components with improved surface finish and dimensional tolerance by implementing optimization algorithms to find the optimal arrangement of powder particles during compaction.
4. Coupled Thermo-Mechanical Analysis for Hot Powder Compaction: Developing a coupled thermo-mechanical analysis to simulate hot powder compaction processes by integrating thermal and mechanical aspects in the simulation model to accurately represent the conditions during hot compaction.
5. Topology Optimization of Hot Powder Compaction Processes: Optimizing the topology of the hot powder compaction process for enhanced efficiency and improved component properties by applying topology optimization techniques to find the ideal distribution of heat and pressure during hot compaction.
The approach yields in
Reduced production time and costs through optimized compaction processes.
Components with superior surface finish and dimensional accuracy.
Advancement in simulation techniques for powder metallurgy processes.
A conical-shaped charge liner: geometry, initial FE mesh, the half stage of compaction and the final stage of compaction
Double-surface cap plasticity model
Page updated
Google Sites
Report abuse