After focusing on the study of first-order hyperbolic turbulence modelling for moment closures during my doctoral research, I am currently working on aerodynamic and aeroacoustic optimization for helicopter rotor blade design. The acoustic noise generated by helicopter units severely limits its service including altitude

and time. To minimize its operating noise, we first focus on generating accurate numerical approximation for the fluid and acoustic field. This will be done by simulating the acoustic source using the Amiet model based on the Reynolds-averaged Navier-Stokes (RANS) solutions and validating it against results from Large Eddy Simulation (LES). To reach the final goal, an adjoint-based optimization will be performed to optimize blade geometry for noise minimization purposes.

Calista recently graduated from the University of Toronto with a BASc in Engineering Science, Majoring in Aerospace Engineering. She is now pursuing an MSc degree in Mechanical Engineering at McGill University. The focus of her research is on the development of reduced order models (ROMs), and their applications in the fields of design and aerodynamic shape optimization (ASO). Her work will build on previous research from this lab relating to projection-based reduced order models, specifically a method which uses Proper Orthogonal Decomposition coupled with a Galerkin projection. These models are useful in cases where the solution of a high-dimensional problem can, with high accuracy, be modeled on a much lower-dimensional subspace. This is desirable in areas like ASO, as these problems typically involve a high number of design parameters and have full-order solutions which are computationally expensive to compute. 

Ph.D. Student in Mechanical Engineering conducting research in high-order methods for the large eddy simulation (LES) of turbulent flows; specifically on entropy stable flux reconstruction (ESFR) schemes for implicit LES. Previously conducted research in modelling nonequilibrium hypersonic flows as part of my Masters in Aeronautics and Astronautics at Purdue University. Aside from research, my passions are Formula 1, skateboarding, and snowboarding. 

Computational fluid dynamics enthusiast of every stage, from numerical analysis and algorithms to the engineering applications. When solving unsteady, nonlinear partial differential equations, there is a constant tension between computational cost and stability. An attractive, feasible, and high-order approach, with the purpose of increasing the maximum time-step is known as flux reconstruction (FR). Although the scheme significantly decreases the computational cost, it introduces stability issues manifested through dense-norms. Using entropy conserving two-point fluxes and/or split-forms, we derived and developed a nonlinearly stable FR high-order framework to ensure stability in both affine and curvilinear domains. This framework allows us to solve unsteady turbulent problems efficiently and accurately.

Calvin is an enthusiast of aerospace engineering and computer science. He obtained his bachelor’s in mechanical engineering at McGill University and is currently pursuing a Master of Science Degree at McGill University in the field of Computational Fluid Dynamics. His current work focuses on Mesh Generation applications in Higher-Order Methods for Three-Dimensional Problems.

I am M.Sc. student in Mechanical engineering at McGill University. After graduating from Osaka Prefecture University (In Japan) with an Aerospace Engineering degree, I joined the Computational Aerodynamics group under the supervision of Prof. Nadarajah. Through previous research internships and undergraduate research, CFD captured my interest and it led me to pursue my passion in this field. I am currently working on the development of provably stable schemes for multi-species gas flows.

In the pursuit of efficiency and emissions, people have been actively researching techniques to enforce a laminar boundary layer over aerodynamic surfaces to reduce the skin-friction drag. However, over the past decade, very few have been able to implement it in a real world design; the Boeing 787 has laminar flow over its nacelles, while the 737MAX has NLF wingtips and Airbus with laminar flow wings in their BLADE project. This is because, laminar boundary layers are incredibly unstable and require a careful profile design and specific flow conditions to maintain stability and delay transition to turbulent boundary layers. For commercial air-crafts two modes of transition are of great interest- Natural Transition and Crossflow transition. Traditionally, these modes have been modelled with a semi-physical model based on linear stability that is non-local and difficult to parallelize, making it tricky to implement efficiently in most CFD codes. We use a fully-local, versatile model instead, in a parallel unstructured flow solver, to predict Natural or Crossflow induced transitional onset in real-world flows. To be able to efficiently obtain NLF designs of aerodynamics surfaces, we aim to couple the model to a discrete adjoint gradient-based optimization framework. However, in complex unstructured grids, such a coupling can result in a highly ill-conditioned system that can prove difficult to solve using traditional iterative methods, even with Newton-Krylov methods. We therefore are investigating various iterative techniques to obtain a robust and efficient way to solve the optimization problem.

I am currently working on a project titled, “Matrix-free preconditioners for Newton-Krylov methods for Large Eddy Simulations” supervised by Prof. Nadarajah as a doctoral student. Preconditioners play a vital role in the convergence of Newton-Krylov methods. For large-scale problems, it is very difficult to get convergence using Netwon-Krylov methods without properly preconditioning the Jacobian matrix. Therefore, preconditioning the system is a necessity. The popular preconditioners currently in use need explicit formation of Jacobian matrices which is costly when it comes to storage.  This is where the matrix-free preconditioning comes in handy. My current research goal is to obtain matrix-free preconditioners and to apply them in a matrix-free fashion (i.e., relying on matrix-vector products only) in an asynchronous manner. My love for applied mathematics drove me towards this project. As a long-term goal, I wish to be a professor someday. In my free time, I am an avid consumer of stories of any kind (books, movies, TV shows, anime etc.)  and I like to write my thoughts as poems sometimes. 

Coming from the University of Alberta, with an Undergraduate in Mechanical Engineering, I am now in a M.Sc. program for Mechanical engineering at McGill. For my research, I am working on methods to reduce the number of variables needed to solve systems that are slightly different to those that are already solved. These systems often take much longer the more variables there are. Therefore, there is a lot of interest in reducing the number of variables and to make the computations much faster. This can have practical applications such as the changing of the shape of a car for aesthetics and how it affects the drag force on the vehicle. Other than school, I enjoy cross-country skiing, downhill skiing, and running.

Carolyn is currently pursuing a M.Sc. with the Computational Aerodynamics group. Her research interests are in higher-order entropy-stable temporal discretization. By ensuring that the time-stepping method maintains entropy stability, the group will be able to perform fully discrete entropy stable aerodynamics simulations, allowing for higher accuracy and better stability than previously possible. Her research has applications in fields such as turbulence, supersonic aerodynamics, and turbomachinery. Carolyn recently graduated from the University of Calgary with a B.Sc. in Mechanical Engineering specializing in Energy and Environmental Engineering. Outside of her academic interests, Carolyn enjoys spending her free time hiking and backpacking in the Canadian outdoors. 

Dominic is a Master’s student within the Computational Aerodynamics Group. After a short time in the manufacturing industry, he decided to follow his passion and pursue a master’s degree in the aerospace field. His current work is on the development and application of high-order methods for Large Eddy Simulation, with a particular focus on the industrialization of the methods for industry. Originally from New Brunswick, Dominic enjoys cooking and gastronomy, reading and listening to music in his free time.

A recent graduate of the BASc program in Honours Mechatronics Engineering at the University of Waterloo, Shruthi is now pursuing a MSc degree at McGill University with the Computational Aerodynamics Group. Her current research interests are in the field of computational aeroacoustics with a focus on limiter functions.

Adaptive mesh adaptation techniques for higher order methods is a promising field to improve efficiency and robustness of numerical algorithms. As Industrial CFD applications typically focus on evaluating a quantity of interest (lift or drag), I work on goal-oriented mesh adaptation. In particular, my work focusses on developing techniques to generate isotropic/anisotropic hp meshes for the discontinuous galerkin high order framework optimized to predict a quantity of interest.

I am an undergraduate honors mechanical engineering student looking at the application of Reduced Order Models (ROMs) to the aerodynamic shape optimization of aircraft wings. High-fidelity solutions to the flow over an aircraft are extremely computationally expensive and thus there is a need for ROMs to provide accurate and computationally cheap solutions. Currently I am working to use a Proper Orthogonal Decomposition with Galerkin Projection ROM as the input to the optimization scheme. The end goal of my work is to be able to perform shape optimization of an aircraft wing for varying flight conditions using the ROM model.