About
Welcome to my homepage! I am Nishan Parvez, a recent graduate with PhD in Mechanical Engineering from Rensselaer Polytechnic Institute, P.I: Dr. Catalin Picu. My interests lie in the broad domain of computational solid mechanics, with extensive research experience in computational modeling and analysis of large deformation and damage in soft materials (e.g., soft tissue, paper, hydrogels, etc.) using FEA and Machine Learning techiniques. The up-to-date list of my published works is available at google scholar.
Please feel free to browse around or contact me for questions related to my work.
Recent Research Projects & Research Highlights
Constitutive modeling of fibrous materials using neural network: Soft fibrous materials exhibit novel non-linearities due to their microstructural features. Conventional hyperelastic models are often inadequate as constitutive models for these materials. This project introduced a neural network-based surrogate computational model that accurately predicts the mechanical response of such materials. Read more at: https://doi.org/10.1007/s00366-024-02095-8
Technical Summary:
- Obtains data from a highly accurate but computationally expensive multiscale simulation package (MuMFiM)
- Trains an input convex neural network based hyperelastic model.
- To improve the prediction accuracy, incorporate Sobolev space training that considers higher derivates in the loss function.
- The trained neural network is used as a material model in traditional FEA procedures.
The surrogate material model allows very efficient and accurate FEA simulation of soft material with few orders of magnitude cost reduction compared to MuMFiM.
Mechanism of damage and fracture in network materials with pre-existing cracks: Soft materials exhibit novel failure behavior compared to traditional engineering materials due to factors such as non-linearity and microstructural features. This leads to phenomena such as notch insensitivity, transition in the microscale failure mechanism, crack tip blunting, etc. This project explores the mechanical origin of these behaviors and provides guidelines for the optimation of material strength and toughness. Read more at: https://doi.org/10.1016/j.ijsolstr.2025.113445
Upcoming articles include a broader discussion on the nature of crack propagation and deflection in these materials.
Technical Summary:
- Fibrous material composed of fiber and crosslink exhibit a transition from fiber to crosslink failure as a function of material parameters.
- A single material parameter is identified that accurately predicts the microscale failure mechanism.
- Cracks in these materials exhibit notch-insensitive behavior. This is linked to the non-linear deformation of fibers.
- A transition from brittle to ductile failure is observed as a function of material property.
- In a ductile failure regime, the material fails due to reduction of strength instead of crack propagtaion. This is facilitated by the formation of damage clusters in the unbroken ligament of the materials that effectively kill the stress concentration.
Peer-Reviewed Publications
For a list of my publicly available works on large deformation, damage, plasticity, and constitutive modeling of material, please visit my google scholar profile
Last Updated: Jan, 2024 | Powered by Ankit Sultana’s Researcher theme