A plane creates a lot of noise as it lands. NASA scientist Mehdi Khorrami studies which components are causing all that racket.

In November, Khorrami led a presentation at SC2017  , an international conference showcasing how high-performance computing supports everyday applications – like how to keep things quieter at the airport.

This week, the Langley Research Center scientist spoke with Tech Briefs about how he’s using supercomputers and simulation software to determine the sources of airframe and landing gear noise.

Mehdi R. Khorrami

Tech Briefs: Why is it so critical to address aircraft noise?

Khorrami: Airframe noise is quite dominant when aircraft land. As the landing gear and the high-lift devices are deployed, the engines are at reduced thrust. The high-lift devices come out at the leading edge and the trailing edge of the wing, and allow the aircraft to slow down. So, as the airflow goes over the body of the aircraft, these components make a lot of noise.

Tech Briefs: What role has simulation played in aircraft noise detection?

Khorrami: For many years, we’ve been trying to apply computational simulation to these components. The aircraft undercarriage, which is one of the most prominent airframe noise sources, does not lend itself very well to computational simulation due to its sheer geometric complexity. In recent years, the aerospace community has made a lot of progress in terms of numerical simulation and modeling.

In order for civil aviation to grow, we need a better understanding of the noise that aircraft generate, so that we can reduce or minimize it. In the future, our studies will extend to small unmanned aircraft systems (UAS), personal air vehicles, you name it.

Tech Briefs: Why is modeling so valuable?

Khorrami: With simulations, we can do a lot of the guesswork before we get to flight testing. For a particular component, the hope is that we can first understand the noise sources, the noise generation mechanism, and which components of the landing gear are making the most noise. Once you know that and simulate that, then you develop your noise reduction concept, which you can evaluate in virtual space.

Once you have a certain level of confidence that the noise reduction concepts or technology will performance well, with regard to the acoustics, then you can do the flight testing. Increasingly, we are trying to use modeling and simulation in order to be more productive, as well as reduce the risk of some of the hardware and the technologies we develop.

Tech Briefs: How is the simulation and modeling created?

Simulated sound waves emanating from a full-scale Gulfstream aircraft in landing configuration with nose landing gear retracted.

Khorrami: The prerequisite for an accurate prediction of airframe noise is the availability of a high-fidelity geometry definition for the component of interest where the smallest features are represented. The volume surrounding the aircraft body is subdivided into billions of cells of varying sizes, with the smallest cells adjacent to the aircraft solid surfaces. The mathematical equations that govern the flow of air over the aircraft body are solved in their discrete form within each volumetric cell. Starting with an initial condition of uniform flow, the governing equations are advanced in time until a state of fully developed, unsteady flow is reached. Beyond this state, the flow field is sampled in time to obtain the flow quantities of interest, such as pressure fluctuations on the landing gear surface.

Tech Briefs: On a basic level, what technology enables the kind of simulation that wasn’t possible previously?

Khorrami: Simulations such as these are possible only with the simultaneous availability of 1) supercomputers with hundreds of thousands of processors; 2) computational algorithms able to efficiently distribute billions of calculations over thousands of processors; 3) highly parallelizable equations that statistically describe the behavior of fluid media, such as the lattice Boltzmann equations; and 4) agile surface preparation and volume discretization software that permit the rapid creation of meshes around extremely complex configurations. The computational fluid dynamics (CFD) suite used for the nose landing gear simulation possesses the last two requirements.

Tech Briefs: What are your biggest simulation challenges?

Khorrami: The challenges are the extreme geometric complexities of the components that we are working on. In order to do a good evaluation of some of the noise reduction concepts that we are modeling with our simulations, you have to evaluate the concept on a system-level basis; you have to take into account the effects of the wing, the body, and how they change or alter the flow fields around the main landing gear or any other aircraft component. You also have to make sure that these changes in the flow field do not adversely impact your noise reduction concept performance. A system-level modeling and simulation means that you have to include a lot of the aircraft components. Until a few years ago, that was challenging and not a possibility for us to do.

Tech Briefs: How does simulation address system complexity?

Simulated vorticity field generated by flow past a full-scale Gulfstream aircraft in landing configuration.

Khorrami: Our simulation has to consider larger structures, such as the wheels, but also very tiny structures on that landing gear. The aircraft noise that gets generated is very broadband, meaning it goes from very low frequency all the way to 7, 8, or 10 kilohertz. Modeling the noise requires a lot of computational resources to not only resolve the large flow structures but also very tiny ones; the tiny ones, in fact, generate the noise in the mid- to high-frequency range. Our work here is cutting-edge and pushes the envelope, in terms of system complexity, by including a lot of the other aircraft components that may affect the flow field for the component of interest.

Tech Briefs: What are you working on currently?

Khorrami: I hope that, within a year or year and a half, we will model the full aircraft in landing configuration, the airframe part of it: the nose gear, the two main landing gears, and all of the high-lift devices deployed exactly as the aircraft lands at the airport. And that, among our community, is the “Holy Grail.”

This is an effort that Boeing and NASA are doing collaboratively, just to see how far we can make inroads in applying computational simulation to the most geometrically complex type of aircraft being flown today.

Tech Briefs: Do you have any advice for fellow aerospace engineers?

Khorrami: With this type of work, you really have to have tenacity. I tell a lot of the younger generation: If we stood in front of a technical group of people seven years ago and I said, “In about 7 or 8 years, we’d be attempting a full-scale aircraft simulation of a 777-class transport,” they would have laughed at us.

You have to keep looking for the right tools. You have to keep looking for small gains here and there. Never lose focus and just have the tenacity to say, “Ok, they say it’s not doable, but let’s look at it in another light.” Little by little, you just push that boundary, and lo and behold, one day you say “Wow, we are doing the stuff we dreamed of a few years ago.”

What are your biggest aerospace simulation challenges? Share your thoughts below.

Related Content: