Mechanical Engineering Machine Learning

Rasika Deshpande

Greater Chicago Area 779-***-**** **********@*****.*** My LinkedIn My GitHub EDUCATION

Northern Illinois University Dekalb, IL

Master of Science in Mechanical Engineering Graduated: May 2024

● Finite Element Methods, Advanced Mechanical Engineering Analysis, Design of Thermal Systems, Advanced Thermodynamics, Convection Heat Transfer, Data Analytics for Engineers, CAD University of Pune Nashik, India

Bachelor of Science in Mechanical Engineering Graduated: May 2018

● Fluid Dynamics, Thermodynamics, Engineering, Material Science, Heat Transfer, Turbo Machines, Manufacturing Processes, FEA, Refrigeration and Air Conditioning PROJECT EXPERIENCE

Northern Illinois University Dekalb, IL

Thermal Systems Engineer - PINNs Application in Heat Transfer May 2023 - May 2024

● Developed Physics-Informed Neural Networks (PINNs) using Python and TensorFlow/PyTorch to model heat transfer in complex 1D and 2D geometries, focusing on applications in cooling systems. This included effectively integrating nonlinear boundary conditions to enhance model accuracy.

● Optimized hyperparameters—such as learning rate, neural architecture, activation functions, L2 regularization, and loss function—resulting in 100% accuracy for 1D and 2D models within the time domain, and around 70% accuracy beyond it, demonstrating enhanced precision and robustness in heat transfer predictions for cooling applications.

● Validated the PINNs model against COMSOL 5.6 PDE solver, identifying and addressing discrepancies in temperature, heat flux, and spatial patterns, thereby improving accuracy in heat transfer predictions. Northern Illinois University Dekalb, IL

Machine Learning Engineer - Academic Success Dataset April 2024 - May 2024

● Implemented machine learning models in Python with Pandas, sci-kit-learn, and TensorFlow/Keras to predict student dropout rates from a dataset of 4,424 records, effectively managing data processing, feature selection, and hyperparameter tuning to aid educational strategy and retention planning

● Performed data exploration, preparation, and modeling using logistic regression, SVM, neural networks, and random forest to compare model accuracy on a dataset, addressing missing values, feature scaling, selection, and encoding, and evaluated performance using accuracy, feature importance, confusion matrix, and McNemar test

● Optimized a Random Forest model to 77.09% accuracy by tuning hyperparameters and using Out-of-Bag evaluation, identified key factors like tuition fees and scholarship status affecting student success, and recommended focusing on financial aid programs to enhance retention

Northern Illinois University Dekalb, IL

Thermal Systems Engineer-Bio-Inspired PCHE( Printed Circuit Heat Exchanger) Design April 2023 - May 2023

• Engineered a bio-inspired PCHE, modeled after tree trunk structures, to enhance heat transfer efficiency while minimizing the system footprint, a key requirement for compact cooling systems.

• Performed detailed analyses of pressure drop, heat transfer surface area, and pumping power, selecting a Gould’s pump optimized for precise flow rate and pressure conditions within the cooling system.

• The PCHE design offered superior thermal performance due to its high surface area-to-volume ratio, enabling efficient cooling with minimal energy consumption.

• Validated the cooling system through comprehensive CAD modeling, confirming enhanced functionality, structural integrity, and operational efficiency within constrained geometries. University of Pune Nashik, India

Mechanical Design Engineer - Vehicle Suspension Operated Air Cooling System May 2017 - May 2018

● Designed and tested a 1:10 scaled air-cooling system (HVAC) prototype using vehicle suspension energy, converting kinetic motion into compressed air through double-acting cylinders and a compact heat exchanger, achieving cooling between 20 C to 30 C and air pressure of 1-2 bar, with simulations and performance data guiding optimization for energy efficiency and vehicle comfort

● Evaluated a suspension-powered HVAC system's efficiency by measuring compressed air output, calculating cooling loads, and comparing performance to conventional systems; results showed a 5-7% increase in vehicle mileage and reduced emissions, demonstrating sustainability potential based on prototype analysis SKILLS

FEA, Thermodynamics, Heat Transfer, Fluid Dynamics, Thermal System Design, Mechanical Design, AutoCAD, COMSOL, NX, SolidWorks, Manufacturing Processes, TensorFlow/PyTorch, Machine Learning, Cross-functional, Simulation and Modeling, Assembly, Stress Analysis, Product Design & Development, Mechanical System Optimization, Prototype Development &Testing, Root Cause Analysis, GD&T, Six Sigma, Detail-oriented.

Contact this candidate