Hello, I'm Pragya Goyal

OpenFOAM-Based Comparative Study of Drag and Lift Behaviour in Bluff and Streamlined Flows

Unpublished

Fluid flow around solid bodies is a common phenomenon in everyday life, and understanding the resulting drag and lift forces is key to the performance and efficiency of many engineering systems. This study investigates incompressible, steady, laminar flow over three geometries—a circular cylinder, a square cylinder, and a thin flat plate—simulated using OpenFOAM, focusing on how velocity and viscosity influence drag and lift, and how this differs between bluff and streamlined bodies. In bluff bodies, pressure forces dominate, so drag is primarily due to pressure differences with minimal contribution from viscosity, while lift arises from pressure asymmetry between the upper and lower surfaces and is therefore more sensitive to viscosity. In streamlined bodies, pressure differences are minimal, so drag is caused mainly by viscous shear and responds directly to viscosity, while lift is comparatively unaffected.

Machine Learning Classification of Quantum Particle Energy States

Unpublished

This study tests whether classical classifiers can identify the first four stationary states of a particle in a one-dimensional infinite well from simulated detector observations. Analytical probability densities generate 1,200 stochastic 32 × 32 images (80 detections each, peak-normalized). k-NN, Decision Tree, Random Forest, and standardized RBF-SVM are trained once on nominal data and tested on held-out observations under reduced resolution, Gaussian blur, and additional noise. Nominal accuracy is 100.0%, 95.8%, 99.2%, and 100.0%, respectively. At 25% resolution and high blur, tree-based accuracy falls to about 46–48%, while SVM and k-NN remain higher; their gap at 25% resolution is within 20-seed variability, and rankings depend on preprocessing. For resolution and blur, retraining on degraded images restores 96–100% accuracy, indicating train/test mismatch rather than lost class information. k-NN and Random Forest remain robust to the tested noise. This is a controlled synthetic robustness benchmark, not a new classification method.

Parkinson’s Resting Tremors: A NetLogo-Based Neural Simulation of Dopamine and DBS

Published in IJASR, October 2025

This paper visualises the neural activity during resting tremors and the effects of different treatments using a custom-built agent-based simulation in NetLogo to model a simplified neural resting oscillator.

Among other symptoms of Parkinson’s Disease, this study focuses on Parkinsonian Resting Tremors and investigates two treatments namely Deep Brain Stimulator and an Automatic Dopamine Dosage. Among other compounds effecting tremors, this study looks at dopamine levels as a primary factor for the frequency of tremors. It analyses the effects of treatments when various parameters are changed to observe the efficiency of both the treatments.

The simulation demonstrates that Deep Brain Stimulation (DBS) is significantly more effective at suppressing Parkinsonian tremors when applied to the thalamic region compared to the cortical region. It also shows that increasing disease severity demands higher dopamine dosage frequency and quantity, with diminishing effectiveness at later stages.

Bio-Enzymatic Decomposition of Crop Residue: A Scalable Solution to Stubble Burning in India

Published in IJSR, October 2025

Stubble burning in Northern India remains a serious environmental and public health concern. This study investigates the potential of bio-enzyme formulations as a scalable, cost-effective alternative for crop residue decomposition. By combining expert interviews, field visits, and controlled experiments using three different bio-enzyme compositions, the study finds that higher concentrations of jaggery and yeast substantially enhance microbial activity and accelerate stubble breakdown. The findings underscore the viability of integrating bio-enzyme use with agricultural machinery and supportive policies, thereby offering a feasible path to reducing reliance on stubble burning.