CUDA Spotlight: Monica Syal




Monica Syal

GPU-Accelerated Brownout Simulations

This week's Spotlight is on Monica Syal, an aerospace engineer at Sunnyvale, Calif.-based Advanced Rotorcraft Technology (ART).

ART is leading designer and developer of state-of-the-art flight simulation facilities for fixed-wing and rotary-wing aircraft.

This interview is part of the CUDA Spotlight Series.


Q & A with Monical Syal

NVIDIA: Monica, what is your role at Advanced Rotorcraft Technology (ART)?
Monica: My responsibilities include the development of a real-time pilot-in-the-loop rotorcraft brownout simulation and mitigation methodology for flight simulator applications.

This is a very interesting and challenging problem, and we are working on it in collaboration with the University of Maryland (UMD) at College Park. This project is being funded by the Air Force Office of Scientific Research (AFOSR) under a Multidisciplinary University Research Initiative (MURI) grant.

NVIDIA: Who are the end-users of your research?
Monica: The end users will be the U.S. Air Force, Army, and Navy, defense companies, rotorcraft manufacturers/operators and flight simulator companies.

NVIDIA: What is brownout and what is its significance?
Monica: Brownout is a fluid mechanics phenomenon that involves the development of a large and dense dust cloud that forms around a rotorcraft, such as a helicopter, when it is landing or taking off from surfaces covered with loose dust or sand, such as when operating in a desert-like environment.

These dust clouds develop because of the rotor downwash flow, which impinges upon the ground. The fluid dynamic effects produced there can mobilize and uplift a very large quantity of dust particles. The underlying physics is basically a dual-phase fluids problem, one fluid phase being the air and the other being the dust.

Helicopter landing in brownout conditions. Source: OADS

As a consequence of the formation of these dense dust clouds, the pilot may lose visibility of the takeoff or landing zones, and may also experience spurious sensory cues and, in some cases, spatial disorientation or even vertigo.

The onset of brownout conditions has led to many accidents with military helicopters when they are operating in desert environments. The uplifted dust can also cause abrasion of the rotor blades and to the compressor blades in the engine, resulting in costly maintenance issues. Therefore, the brownout problem is one of great significance in rotorcraft flight operations, and it is critical to explore ways to mitigate this very serious problem.

NVIDIA: What is your approach to tackling this problem?
Monica: At ART, we are developing a real-time, physics-based brownout simulation capability that can be used to understand the underlying physics of rotorcraft brownout and to explore means of mitigating its effects.

The approach being used is to integrate the flight simulation methodology (FLIGHTLAB), which was developed at ART, with the comprehensive rotorcraft brownout simulation model that was developed at UMD. This is an extremely challenging project with ambitious goals because eventually we want to achieve high-fidelity brownout dust cloud simulations in real-time.

Simulation showing development

Simulation showing development of brownout dust cloud during an approach maneuver

This simulation capability will be used to understand the effects of different rotor design parameters (e.g., rotorcraft), soil parameters (e.g., particle size and mineralogy), and flight conditions (e.g., landing and takeoff maneuvers) on the development of the dust clouds. Different approaches will be explored to mitigate brownout conditions, e.g., by means of flight path management, or by modifying certain aspects of the rotor design, or even the flight control system.

The brownout flight simulation capability that we are developing can also be used to train pilots to recognize the onset of brownout conditions, and to develop piloting strategies that could be used to mitigate the effects of brownout and thus minimize the impact on flight operations.

NVIDIA: Why is it important to accelerate the simulations?
Monica: In this research project we want to develop a high-fidelity methodology that can represent and simulate the fundamental physics of the brownout problem. To this end, a realistic dust cloud simulation requires tracking billions of individual particles.

The fact that we want this methodology to be coupled to a flight simulation code (FLIGHTLAB) and used in a piloted simulator, requires the code to run in real-time, which is a great computational challenge. Therefore, it is important to accelerate the simulations as much as possible to achieve the needed fidelity.
 
NVIDIA: Why did you decide to use GPUs?
Monica: The simulation of the individual particle motions in the dust clouds is equivalent to an N-body problem, where the number of bodies (or particles) is very large, in this case of the order of 1014. We are using several techniques to expedite such simulations, some of these being smart algorithms (e.g., fast multipole methods), particle clustering algorithms, and high-performance parallel computing techniques.

An obvious way to achieve the needed computational accelerations by using parallel computing is to conduct the simulations using as many computing resources as possible. The number of cores used in a CPU is relatively few and the CPUs are optimized for serial processing.

On the other hand, a GPU consists of hundreds of cores, which can be used to parallelize the computationally intensive parts of the simulations. Therefore, we decided to use high-end Tesla GPUs to conduct these simulations. This has provided us with about two orders of magnitude speedup in the computational time compared to the serial execution of the code.

NVIDIA: What role does CUDA play in your work?
Monica: CUDA C is easy to learn and allows fast computations on the GPUs. Our code was originally written in FORTRAN, and only the computationally expensive parts were computed on the GPUs by using CUDA. Our ultimate goal in this research project is to achieve real-time speedups and, to this end, we will now be implementing most of the code on multiple GPUs by using OpenMP, MPI and CUDA.

NVIDIA: How did you become interested in this area?
Monica: I became interested in the rotorcraft brownout problem during my studies in the Aerospace Engineering Department of the University of Maryland (UMD) College Park. My advisor, Dr. J. Gordon Leishman, hired me to work on this interesting project for my doctoral research, in which I had to develop a simulation of the dust cloud generated by a helicopter rotor when it was operating near the ground. I found the problem fascinating and decided to make it my Ph.D. research topic.

Regarding GPU programming, I was first introduced to it in a course offered by Dr. Ramani Duraiswami from the Computer Science department at UMD, whose team specializes in this area. We collaborated with his team, and one of his graduate students, Qi Hu, helped us to implement our methodology on GPUs. Our current goal at ART is to achieve such brownout simulations in real-time, and so we are integrating a multi-GPU system to achieve that goal.
 
NVIDIA: What is the potential impact of your work in the rotorcraft industry?
Monica: The present work will provide the rotorcraft industry with a physics-based high-fidelity tool to simulate and understand brownout, and possibly lead to the development of various means of mitigating this problem.

This research will benefit the rotorcraft industry in several ways. Incorporating these simulations into flight simulators will be extremely helpful in training pilots to recognize the onset of brownout conditions and to learn the correct piloting tactics, techniques, and procedures to avoid the undesirable consequences of brownout.

Eventually, this research may also help improve the handling qualities of rotorcraft by incorporating certain design changes into the flight control system, such as helping to reduce the workload imposed on the pilot while landing or taking off in conditions where brownout may be unavoidable.

Overall, this research will help enhance flight safety and reduce the large number of brownout related accidents that occur in both military and civil rotorcraft flight operations.


Bio for Monica Syal

Monica Syal is an Aerospace Engineer at Advanced Rotorcraft Technology (ART), and is working on the development of a real-time rotorcraft brownout simulation facility for flight simulator applications. Monica completed her Ph.D. and M.S. from the University of Maryland College Park in 2012 and 2008, respectively.

Her research interests include helicopter aerodynamics, computational fluid dynamics, optimization techniques, and high performance parallel computing using MPI and CUDA. Monica is a recipient of the Amelia Earhart Fellowship (2009, 2010), Vertical Flight Foundation Scholarship (2008), and Dr. Kalpana Chawla Award (2004).

Relevant Links
http://www.flightlab.com/
www.aero.umd.edu
http://www.aero.umd.edu/news/news_story.php?id=6504
http://www.aero.umd.edu/news/news_story.php?id=3781
http://www.zontadistrictthree.org/AE_Poster_11x17in.pdf

Contact
syal[at]flightlab[dot]com
syal.mona[at]gmail[dot]com