Research | Neeraj Balachandar | Research Portfolio

VarFIExI - Variable Fidelity Aeroelastic Solver

Thu, 11 Sep 2025 00:00:00 +0000

Problem

VarFIExI is a variable-fidelity aeroelastic framework that couples vortex particle flow models with structural solvers. The aim is to retain key aeroelastic physics while keeping the computational cost low enough for design and control studies.

Approach

rVPM flow solver coupled to FEM via preCICE
Implicit and partitioned explicit coupling strategies
Validation on benchmark aeroelastic cases

Nonlinear Model Predictive Control via Sequential Convex Programming for Drone-to-Drone Docking

Thu, 11 Sep 2025 00:00:00 +0000

Problem

This project develops sequential convex programming formulations for drone-to-drone docking. The focus is on safe, feasible trajectories that respect disturbance bounds and actuator limits while preserving computational tractability.

Approach

Convexified constraints for safety and docking geometry
Robust thrust profiles under bounded disturbances
Numerical validation on docking scenarios

Quasi-static Elbow Flexion Study

Sat, 01 Jun 2024 00:00:00 +0000

Problem

Designed an exoskeleton system and validated muscle activity models under simulated microgravity conditions using biomechanical analysis.

Approach

Built a quasi-static biomechanical model for elbow flexion with exoskeleton interaction.
Compared simulated muscle activation trends against controlled experimental observations.

Neural Operator Surrogates for Fast Multipole Vortex Simulations

Thu, 11 Sep 2025 00:00:00 +0000

Problem

This project builds neural-operator-based surrogates for fast multipole vortex simulations. The emphasis is on conservation-preserving learning and super-resolution reconstruction of velocity fields.

Approach

Neural operators for N-body vortex interactions
Conservation-aware training objectives
Super-resolution velocity reconstruction

Analytical Model for Soft Robotic Membranes

Thu, 11 Sep 2025 00:00:00 +0000

Problem

This project develops analytical models for tendon-actuated soft membranes inspired by jellyfish propulsion. The goal is to capture dominant deformation and propulsion mechanisms with compact, interpretable models.

Approach

Cosserat rod modeling of membrane deformation
Hydrodynamic force estimation via Lighthill’s theory
Validation using immersed boundary simulations

Chaotic Wake Dynamics of Flapping Wings

Thu, 11 Sep 2025 00:00:00 +0000

Problem

This project studies chaotic wake evolution in flapping-wing flows. The goal is to extract low-dimensional descriptors that capture transitions to chaos and inform control-oriented modeling.

Approach

Graph-based representation of vortex interactions
Information-theoretic metrics for chaos detection
Comparative analysis across operating regimes

Drone Swarm Rescue Simulation

Thu, 11 Sep 2025 00:00:00 +0000

Problem

This project focuses on multi-agent rescue and transport in simulated environments. I explored navigation, coordination, and clustering methods for robust swarm behavior.

Approach

Artificial potential fields for navigation
RRT-based transport strategies
Clustering comparisons (DBSCAN, K-means)

Sparus AUV Modelling and Control

Sat, 01 Jun 2024 00:00:00 +0000

Problem

Developed a hydrodynamic model for the Sparus AUV and implemented a Simulink-based stabilization controller for underactuated dynamics.

Approach

Derived a reduced hydrodynamic model suitable for control-oriented simulation.
Implemented and tuned a Simulink stabilization loop for underactuated operating conditions.

Neuro-Fuzzy PID Control

Sat, 01 Jun 2024 00:00:00 +0000

Problem

Developed a neuro-fuzzy-based adaptive PID controller for a 2R manipulator with varying trajectories, emphasizing stability and performance across operating conditions.

Approach

Designed a neuro-fuzzy adaptive tuning strategy to update PID gains under changing trajectories.
Evaluated tracking stability and control smoothness across multiple excitation profiles.

Parallelised SVD (C, OpenMP)

Sat, 01 Jun 2024 00:00:00 +0000

Problem

Implemented and optimized a parallel SVD routine in C with OpenMP, focusing on scaling behavior and performance profiling.

Approach

Parallelized core matrix decomposition steps using OpenMP and optimized workload partitioning.
Benchmarked strong and weak scaling to identify bottlenecks and quantify speedup.