Sarven Yardim

Master's Thesis: Physics Informed Neural Network for Thermal Simulation of Laser Powder-Bed Fusion Process - Implementation of a novel Machine Learning algorithm based on Physics Informed Neural Network (PINN) in order to solve the 3D Heat Equation for thermal simulation of laser powder-bed fusion additive manufacturing process. - Comparison of the outcome of PINN with the results of Finite Element Analysis to assess the effectiveness and reliability of the developed PINN-based solution.

- Analyzing the mixing of two liquids in various blood pumps using OpenFOAM - Conducting the relevant changes in the existing OpenFOAM Multiphase and Lagrangian solvers in order to perform necessary simulations that offer an appropriate understanding of the physical processes in the cases investigated. - Creating the fundamental post-processing tools to acquire the required physical fields.

- Studied different failure modes in fluid machines and inspected existing data to help develop preventative maintenance schedules in order to standardize the work process. - Participated in the preparation of maintenance and inspection plans to mitigate the risk of failure in fluid machines in the assigned area of fuel oil transformation units.

- Researched different mechanisms in the context of aircraft rotation (yaw, pitch and roll). - Prepared the conceptual design of the mechanical flight control system of the new Very Light Aircraft (VLA) which consisted of parts such as the stick, pedal, pulleys and cables used to move control surfaces (ailerons, rudder and elevator) smoothly. - Contributed to the 3D design of the control system of the VLA project in accordance with the requirements of EASA CS-VLA using CATIA.

Semester Project: Implementation of Data-driven Regression methods for the Rosenbluth-Fokker-Planck Equation