Projects
Selected research projects from my Master’s and PhD work, focused on granular flow, DEM simulations, and pattern formation.
Impact of friction and grain shape on the morphology of sheared granular media (PhD)
How particle shape and friction determine surface morphology in granular shear flow
- Spheres → flat surface; elongated particles → pronounced surface depression
- Competition between alignment-driven compaction and friction-driven dilatancy
- Links microstructure (alignment) to macroscopic surface evolution
Alignment-induced depression and shear thinning in granular matter of nonspherical particles (PhD)
How particle alignment simultaneously controls compaction and rheology
- Alignment inside shear band drives surface depression
- Transition from dilatancy (spheres) to alignment-induced compaction
- Reveals shear-thinning behavior via inertial-number scaling
Shear-induced pressure anisotropy in granular materials of nonspherical particles (PhD)
How particle shape generates localized pressure through stress anisotropy
- Elongated particles develop a pressure peak inside the shear band
- Caused by alignment → higher packing density → normal stress anisotropy
- Spheres remain nearly isotropic under identical conditions
Dilatancy-Induced Heap Formation in Dense Cohesive Granular Media (PhD)
How cohesion transforms dilatancy into surface heap formation
- Dry systems → flat surface; cohesive systems → heap formation
- Controlled by balance of capillary forces, gravity, and shear rate
- Identified Bond number as governing parameter
Pattern selection of three components Gray-Scott model (2019)
How diffusion and nonlinear interactions select spatial patterns
- Derived conditions for Turing instability
- Predicted unstable wavelength bands via dispersion analysis
- Showed how nonlinear effects determine stripes vs hexagonal patterns
Applied & Engineering Projects
Teaching a Robot to Build Patterns
A robot that picks, moves, and places particles to form structured designs
- Programmed robotic arm for precise positioning
- Built a pick-and-place loop (detect → pick → move → place)
- Used vacuum suction for reliable gripping
- Generated patterns (e.g., Yin–Yang) from discrete particles
- Works with both spherical and irregular shapes
Sensor Fusion for Accurate Motion Tracking
Fusing noisy GPS and accelerometer data using a Kalman filter
- Implemented state-space model and Kalman filter
- Fused accelerometer and GPS measurements
- Reduced position error by ~20%
- Modular Python architecture
Battery Thermal Digital Twin
Estimating internal temperature from noisy measurements
- Simulated battery thermal dynamics (heating + cooling)
- Added realistic sensor noise
- Built a digital twin for state estimation
- Compared threshold and exponential degradation models