CFD Scientist with AI-Driven Turbulence Closure Expertise

*

Adrin Issai Arasu

Guangdong-Technion Israel Institute of Technology,

Shantou, China.

Phone: +86-134******** +91-958*******

E-Mail: ***********@*****.***

EDUCATION

Program University GPA Year

PhD in Aerospace

Engineering

Indian Institute of Technology, Madras 9/10 July 2019 – Dec 2024 ME in Aerospace

Engineering

Iowa State University, Ames, Iowa, USA. 3.31/4 January 2015 – May 2017

BE in Aeronautical

Engineering

Anna University, Chennai 8.5/10 August 2009 – May

2013

RESEARCH AREAS

o Computational Fluid Dynamics

o Turbulent flow control and Aerodynamics

o High-Fidelity simulations

o Gas turbine engines

o Deep Learning and AI

o High Performance Computing and code development

RESEARCH INTERESTS

o Development of High-Order GPU-Accelerated Navier-Stokes Solvers: Focusing on advancing numerical algorithms for high-fidelity simulations (LES/DNS) of complex flows, with applications to practical aerospace and energy systems. My goal is to bridge the gap between cutting-edge computational science and industrial-scale engineering challenges.

o Active and Passive Flow Control for Propulsion and Energy Systems: Investigating novel control strategies, including vortex generator jets and plasma actuators, to manage flow separation, reduce losses, and enhance stability in gas turbine engines and other high-speed intakes. o Sustainable Aviation and Renewable Energy Aerodynamics: Applying advanced computational methods to address challenges in next-generation sustainable technologies, including wind turbine wake dynamics, performance optimization, and the aerodynamics of low-emission aircraft configurations. o Data-Driven Modeling and Scientific Machine Learning: Developing and applying deep learning techniques, such as Physics Informed Neural Networks (PINNs) and convolutional models, for turbulence closure, flow field reconstruction, and other complex physical systems. 2

RESEARCH PAPERS AND PUBLICATIONS

Journal publications:

• Adrin Issai Arasu, Sumit Sarvankar, Vadlamani N R - “Distortion Control in Intakes Subject to Strong Crosswinds Using Steady Vortex Generator Jets” – AIAA. DOI: https://doi.org/10.2514/1.J064548

• Adrin Issai Arasu, Vadlamani N R - “Subsonic Intake Under Crosswinds: Flow Control Using Pulsed Vortex Generator Jets” – Physics of Fluids, Volume 36, Issue 6, 065115 (2024). DOI: https://doi.org/10.1063/5.0209518.

• Sumit Sarvankar, Drik Sarkar, Adrin Issai Arasu, Chetankumar Sureshbhai Mistry, Vadlamani N R -

“Characteristics of Laminar Separation Bubble with Varying Leading-Edge Shapes and Deflections of the Trailing-Edge Flap” – ASME Journal of Engineering for Gas Turbines and Power, Oct 2024. DOI: https://doi.org/10.1115/1.4065408

• Vijayananth, Adrin Issai Arasu, Dinesh Kumar, Kanmaniraja, Senthil Kumar - “Numerical and Experimental Investigation on the Aerodynamic Performance of roller airfoil” at International Journal of Engineering and Technology, Oct 2018. DOI: 10.14419/ijet.v7i4.10.21302. Journals (Under preparation):

• Adrin Issai Arasu, Vadlamani N R - “Influence of Vortex Jet Configuration on the Control of Crosswind-Induced Intake Distortion” – RAeS Special Issue (Abstract Accepted)

• Adrin Issai Arasu, Vadlamani N R - “Aerodynamics of Intakes under Crosswinds – Effects of Reynolds number and Mach number” – Theoretical and Computational Fluid Dynamics Conference papers:

• Abraham Benjamin Britto, Vikrant Gupta, Adrin Issai Arasu - “State estimation in turbulent channel flow using ensemble Kalman filter” - APS Division of Fluid Dynamics Annual Meeting 2025, Nov. 23– 25, 2025, Texas, USA

• Shivam Saini, Adrin Issai Arasu, Yuan Fang, Maximillian Reissman, Richard D Sandberg, Nagabhushana Rao Vadlamani – “Uncertainty Quantification of GEP-based Turbulence Closure Models for Intake Subjected to Extreme Crosswinds” - Cambridge Unsteady Flow Symposium, March 2026 (Full Manuscript submitted)

• Adrin Issai Arasu, Sumit Sarvankar, Vadlamani N R - “Mitigating Distortion in Subsonic Intakes Using Vortex Generator Jets” – Cambridge Unsteady Flow Symposium, 4-5 March 2024

• Adrin Issai Arasu, Manish Modani, Vadlamani N R - “Application of Machine Learning Techniques in Temperature Forecast” – 21st IEEE International Conference on Machine Learning and Applications

(ICMLA), Nassau, Bahamas

• Sumit Sarvankar, Adrin Issai Arasu, Vadlamani N R - “Crosswind aerodynamic analysis using Quasi 3D ducts” – Turbomachinery Tech-nical Conference and Exposition, Proceedings of ASME Turbo Expo 2022

• Sumit Sarvankar, Adrin Issai Arasu, Vadlamani N R - “Effect of crosswind flows on Intake Aerodynamics” at Proceedings of the 8

th

International and 47

th

National Conference on Fluid Mechanics

and Fluid Power(FMFP-2020), IIT Guwahati, India.

• Adrin Issai Arasu, Dinesh Kumar, Kanmaniraja, Gowsik N - “Aerodynamic Performance Analysis of a Dynamic Upper Surface NACA 4412 Airfoil” at International Conference on Engineering Materials 3

and Processes (ICEMAP-2013) in Chennai, India.

• Dharun Lingam, Adrin Issai Arasu, Sanal Kumar - “Conceptual design of Short Takeoff Supersonic Aircraft with cold flow nozzles” at 3rd International Conference on Trends in Mechanical and Industrial Engineering (ICTMIE’2013) held at Kuala Lumpur, Malaysia.

• Adrin Issai Arasu, Dinesh Kumar - “Conceptual design of solar-powered Unmanned Aerial Vehicle” at National Conference on Engineering Systems, July 2011, Sathyamangalam, India. WORKSHOPS AND TECHNICAL EVENTS

• NPTEL Masterclass workshop on “Machine Learning for Fluid Mechanics”, conducted by Dr. Ricardo Vinuesa, KTH Royal Institute of Technology, Stockholm, Apr 30 – May 06.

• Ananth, S.M., Ganesh N., Adrin Issai Arasu, and Vadlamani, N. R., (2020) “Optimization of High– Fidelity High Order Solvers on Multiple GPUs”, in NSM Workshop on High–Performance Computing in Computational Fluid Dynamics (HPCCFD), Dec 01 – 03.

• Ananth, S.M., Ganesh N., Adrin A., and Vadlamani, N. R., (2020) “Optimization of COMP-SQUARE on Multi-GPUs”, CDAC GPU Hackathon, Aug 31 – Sept 07.

• OpenACC Bootcamp conducted by Center for Development of Advanced Computing (C-DAC) in association with NVIDIA and OpenACC – June 16 and June 17, 2020.

• Course on Interactive Usage of Finite Element Analysis Software for Mechanical and Aerospace design organized by National Agency for Finite Element Methods and Standards (NAFEMS), Bangalore, India - May 2011 to July 2011.

RESEARCH PROJECTS

• Atmospheric Boundary Layer Flows and Wind Turbine Simulations. (Postdoctoral research) Mentor: Dr. Vikrant Gupta (April 2025 to present)

The in-house solver COMP-SQUARE is enhanced to perform Atmospheric Boundary Layer simulations. Actuator Line model and Actuator Disc models with and without rotation were added to the solver to perform wind turbine simulations to study the wake dynamics. These improvements allow for more accurate representation of wind turbine operation within realistic atmospheric conditions, enabling detailed analysis of wake effects and their impact on turbine performance and array layout.

• High-Fidelity simulation of crosswind flows over intake and flow control. (Doctoral research) Guide: Dr. Vadlamani Nagabhushana Rao (July 2019 to Dec 2024) In this project, an existing numerical framework has been enhanced to simulate the dynamics of the crosswind flows in gas turbine intakes. The effects of strong crosswinds on the intake and its influence on the fan are investigated using high-fidelity simulations. Due to crosswind flowing over the intake, the flow separates at the intake lip and causes flow distortion at the fan-face. The flow distortion is mitigated using active and passive control strategies such as plasma actuators and Vortex Generator Jets.

• Application of Machine Learning techniques in temperature forecast. (Internship with NVIDIA) Guide: Dr. Manish Modani, Dr. Vadlamani Nagabhushana Rao (Jan 2020 to Dec 2022) Temperature prediction is critical for many industrial and everyday applications. Numerical Weather Prediction (NWP) models using high-performance computing is the most sought technique to forecast 4

weather, including temperature. However, NWP is complex in nature and computationally expensive. In this project, the temperature is forecast using data-driven Machine Learning techniques, which are not computationally intensive and are further accelerated using GPUs. Two deep learning models: A stacked Long Short-Term Memory (LSTM) and Random Forest Regressor (RFR), are developed and validated using the standard ERA5 data (at 850hPa, above the atmospheric boundary layer). In addition, the models are tested against the ground-level observations (inside the atmospheric boundary layer) for twenty different locations in India. The performance of univariate and multivariate models is also analyzed for the real-time dataset.

• Tensor-based Neural Networks for turbulence closure. (Collaborative work with Nihal S Manvi) Guide: Dr. Vadlamani Nagabhushana Rao (Jan 2020 to Dec 2021) RANS models are far less computationally expensive than Direct Numerical Methods (DNS), but they fail to provide accurate predictions for several canonical flow fields. Attempts to create more sophisticated RANS models (non-linear eddy viscosity models) have failed to give consistent results. There has been an increasing interest in applying machine learning techniques for Reynolds stress closure. Specialised neural networks - Tensor Basis Neural Networks (TBNNs) - have shown promising results by predicting the entire Reynolds stress anisotropy tensor while preserving Galilean invariance. In this project, TBNNs were used to predict the Reynolds stress anisotropy tensor on two different canonical flow fields - flow through a square duct and flow over a flat plate.

• Numerical Investigation on the performance of roller airfoil Guide: Dr. Senthil Kumar (May 2012 to March 2013)

In this project, commercial software ANSYS Fluent is used to investigate the flow features of NACA 2412 airfoil at different angles of attack. In order to mitigate the flow separation at high incidences and delay stall, a dynamic active control device is utilized.

• Conceptual design of short take-off supersonic aircraft with cold flow nozzles Guide: Dr. Sanal Kumar (May 2011 to Dec 2011)

A supersonic rectangular CD nozzle with multi-purpose frame is designed in such a way that during the early taxiing time the internal flow choking will be established without forming any shockwave in the divergent region for facilitating the aircraft for a smooth and short takeoff. Parametric analytical studies have been carried out for both internal and external flows using suitable turbulence models. The results from parametric study indicate that cold flow choked nozzles with conventional jet engines will complement the short take-off of supersonic aircraft lucratively without having any additional propulsion systems.

DEVELOPMENT OF IMPORTANT CODES AND KEY CONTRIBUTIONS

• Enhancement of COMP-SQUARE solver to perform wind turbine simulations: Actuator Line method

(ALM) and Actuator Disc Methods (ADM) with and without rotation were added to the solver.

• Implementation of the Characteristic Interface Boundary Condition (CIBC) in the in-house solver COMP- SQUARE – Communication between blocks of the computational domain is established using the non- reflective boundary condition – CIBC.

• Standard k-omega model – The standard RANS model is added to the in-house solver, and the enhancement is validated for various canonical cases. 5

• TBNN-enhanced RANS model – The standard RANS model is enhanced using the Tensor Based Neural Networks and tested on canonical cases such as turbulent square duct and flat plate.

• MPI and OpenACC parallelization of the in-house solver.

• Implementation of Convective Outflow boundary condition to the solver – The stability of the solver is improved and reflections from the outflow is mitigated using the non-reflective boundary condition.

• Implementation of the flow control (Vortex Generator Jets (VGJ)) boundary condition – Steady and pulsed VGJ boundary condition is added to the in-house solver.

• Development of Machine Learning tools using stacked LSTM and Random Forest to predict temperature forecast.

• GPU and CPU parallelization of WeatherBench, an open-source tool for weather prediction – WeatherBench uses Convolutional Neural Network and Linear regression models to forecast weather. PROFESSIONAL EXPERIENCE

Guangdong-Technion Israel Institute of Technology – Postdoc Fellowship (April 2025 – present)

Development of numerical methods for wind turbine research.

Flow field reconstruction using Deep Learning Techniques. NVIDIA – Research intern (Feb 2020 – Dec 2022)

CPU and GPU parallelization of open-source weather forecasting software WeatherBench.

Developed two ML models to predict real-time temperature data at 20 locations in UP, India. NPTEL – Teaching Assistant (Jan 2022 – April 2022 & Jan 2023 – April 2023)

Assist instructor with lectures, assignment preparations and exams. International Students and Scholars Office, Iowa State University – Instructor (April 2015 – May 2016)

Lecture, lead a group of students through pre-assigned material. AWARDS AND RECOGNITIONS

• Recognized with Research Excellence Award during Research Scholars Day for impactful contributions to research endeavors.

EXTRACURRICULAR ACTIVITIES

• Peer-reviewed 8 manuscripts for AIAA Aviation 2023 and 3 manuscripts for ASME GT-Turboexpo 2023 conference and 4 manuscripts for Physics of Fluids 2024.

• Member of ISU PrISUm Solar Car Team – Design solar cars and participate in Solar Challenge competitions.

Contact this candidate