by Douglas Clark

At the Air Force Research Laboratory's Computational Sciences Center in Ohio, research aerospace engineer José Camberos and his colleagues are diligently working to coax accurate calculations of electromagnetic equations from Cobra, a computer program based on the code Cobalt 60, which was originally written for solving fluid dynamics equations. The effort represents a significant step towards to the Computational Science Center’s ultimate goal: computer-simulated aerospace experimentation that is both reliable and comprehensive.

Simulations that will guide real-world experiments
Although the Center is developing the capability for simulated experiments, its goal is by no means to replace conventional experimentation. The aim is, in fact, to develop dependable methods that will ultimately help guide real-world experimentation. By using preliminary virtual experiments to pinpoint the most promising directions for research and to weed out dead ends, researchers will be able to winnow down the list of experiments that are actually conducted in the lab or field, thus saving considerable time and money.

Currently a number of physical forces can be digitally simulated, but one of the great challenges in designing or testing complex aerospace equipment is to develop methods that can simulate all relevant factors at once. It’s possible, for example, to model how an aircraft’s wing will respond to air flowing over it, and also possible to model the structural response, but separate programs are currently used for each task. And therein lies one of the great difficulties, according to the researchers: perfectly good data but from separate and disparate sources.

“We have computer codes that can solve problems within different disciplines,” points out Dr. Camberos, “but in order to integrate those results, there’s a lot of complicated back and forth data exchange that has to take place first.”

Integrating results is labor intensive because the various programs that model different physical forces were all developed independently and thus use grids in different ways and often have different approaches to handling data. Because of these differences, each program also has its own margin of error. When putting data from different programs together into a comprehensive view, it’s incredibly difficult to compensate for the different type and degree of error associated with each program. For this reason, even relatively successful attempts to combine data from different sources often leave researchers with results that are less reliable than desired.

Development of a single code
To reduce error and the amount of time involved in the process, Dr. Camberos and his colleagues are trying to find a way to look at any given experiment as a whole, instead of breaking it down by various disciplines, and then having to piece it back together. This will allow engineers to test and develop equipment in a more holistic manner, instead of looking at individual parts and physical forces one at a time. With this approach, the landing gear of an aircraft, for example, could be understood as a part of the entire aircraft, and in relation to all the forces acting upon it, not just as a separate piece of equipment operating on its own. This will bring not only savings of time and money but will ultimately produce superior designs.

So why hasn’t this approach been used before? One of the great motivating factors behind the project is feasibility. While the goal was nearly impossible to accomplish in the past, today’s faster computer processors and the power of parallel computing now put this goal within reach. And so the pursuit of better computer modeling is under way at the Air Force Research Laboratory.  This step is just one in the process to bring multiple disciplines together in a unified code.

Dr. Camberos and his colleagues have adapted their CFD code by reworking it to solve Maxwell’s equations, which are a set of partial differential equations—much like those used in fluid dynamics—that describe the behavior of electric and magnetic fields, and their interactions with matter.

The researchers solved the equations in the time-domain, meaning that the models they created were devised to describe how electromagnetic waves travel in terms of time, as opposed to understanding wave behavior in terms of frequency. If this method were applied to visible light—a form of electromagnetic energy—coming from a bulb as it was switched on, it would describe in steps of time how the light moves from the switched-on bulb to an onlooker’s eye.

Visualizations aid research
To test the accuracy of the newly devised code, the researchers simulated experiments that had already been conducted in the field or were simple enough to be conducted repeatedly in the lab. This gave them two sets of data to compare and the means to measure error in the simulated experiments.

One of the experiments they simulated was how an experimental aircraft shape would react to radar. A second measured electromagnetic interactions with a simple length of a metal pipe. When the calculations were complete, data were fed into EnSight, a software program from Computational Engineering International (CEI), to allow the researchers to view the results.

The visualizations, which use bands of color to represent the different periods of the electromagnetic wave, make it possible to study how different parts, such as the nose or fins of the missile, interact with a radar wave. 

The visualizations have helped the researchers in fine-tuning their work, such as in developing an appropriately fine grid for more precise simulations. Clues such as blurred edges or corners of visualized experiments signaled in a few instances that the grids the researchers were using did not supply precise enough data.

While adapting the CFD code was a difficult task, producing animated images with EnSight fortunately proved to be simple. “The process that EnSight describes in its user’s manual for making movies is very straightforward and intuitive,” explained Dr. Camberos. “An undergraduate student working with us on the project made our movies, and it was very easy for him to learn how to create them.”

The EnSight visualizations also allowed the researchers to look at cross-sections of the objects they were studying.  “We used the plane-cut feature not just to visualize the surface but to show the vertical plane, so we could see the scattering on the surface as well as through the object. Being able to cut away parts and map a variable onto that was very convenient,” he noted.

Thanks to diligent work and the assistance of EnSight visualizations, the researchers have successfully obtained reliable data from their newly adapted CFD code. While they continue to refine the results, work won’t be completed once they finish their project in the area of electromagnetism. They’ll no doubt soon be looking to tackle another area of the design process, as they work toward assembling a unified code that will bring new efficiency to all aerospace design and testing at the Air Force Research Laboratory.


More Information

Riding the Waves
How does a gamma ray compare to a radio wave? Read up on the basics of electromagnetics at http://en.wikipedia.org/wiki/Electromagnetism, or for the more technically inclined reader, brush up on Maxwell’s equations http://hyperphysics.phy-astr.gsu.edu/hbase/waves/emwv.html.

Learn more about James Clerk Maxwell, the 19th-centry Scottish mathematician and physicist behind the equations that explain electromagnetic behavior. http://en.wikipedia.org/wiki/James_Clerk_Maxwell.

View a mapping of the complete electromagnetic spectrum http://hyperphysics.phy-astr.gsu.edu/hbase/ems1.html#c1.

Beneath the Radar
Electromagnetism plays a role in aircraft design in a number of ways, among them, how an aircraft appears on radar. While most aircraft are meant to be clearly visible on radar, stealth craft are designed to be virtually invisible. Understand the basics of how these aircraft fool radar. http://en.wikipedia.org/wiki/Stealth_aircraft.

Learn more about how radar works here http://encarta.msn.com/encyclopedia_761569568/Radar.html#461530763.

Complex Calculations
Other researchers are harnessing the power of parallel computing to solve highly complex problems. Read about how researchers at North Carolina State University are using EnSight visualizations to unravel the mysteries of supernovas. www.ensight.com/images/stories/application_stories/ncsu.pdf.   

Ever wonder what happens to aging nuclear warheads? Maybe not, but someone has to—and it’s a real computational challenge. Learn how EnSight and parallel computing are helping scientists keep an eye on nuclear weapons at the molecular level.  www.ensight.com/images/stories/application_stories/beowulf.pdf.