This article highlights that computational fluid dynamics (CFD) software has become a widely used tool in engineering, biomedical, and environmental research and development in the past few years. CFD might be coupled with multiphysical applications, so users can solve for more than one phenomenon at a time. CFD methods can predict the areas of a component that will see the most damage from the sand, The CFD-based erosion model let the component manufacturer predict the areas of highest erosion and take steps to either reinforce those areas with a suitable material or else redesign the component. CFD will be used even more widely across disciplines and different types of technologies than it is now, but the computational method is already invaluable to many engineers working with varied, unrelated applications.
Because of its ability to simulate fluid flow digitally, computational fluid dynamics software has become a widely used tool in engineering, biomedical, and environmental research and development in the past few years. But Milovan Peric, who until last year was a professor of fluid dynamics at Harburg-Hamburg University in Germany, sees more to Come.
Peric, a coauthor of the book Computational Methods Jor Fluid Mechanics (Springer-Verlag, 1999), predicts that CFD will be come even more widely used in these fields because universities are teaching CFD techniques more than ever. In addition, the technology is getting easier to use, which makes it more accessible to nonexpert analyst. The ease of use is in keeping with a trend for many computer-aided engineering technologies, such as multi physics or finite element analys is programs, which vendors now design for use by engineers not specifically trained in analysis. In many cases, engineers no longer need special training to apply CFD techniques to their work, Peric said.
Like FEA, CFD applications divide a model into a grid of small parts and calculate how it will react to certain conditions. Often, vendor packages hide the mesh when simulating fluid or airflow for a more user-friendly way to animate fluid dynamics. Peric says that technology vendors can now automate grid generation more than they could in the past, meaning that simulations run faster and are more faithful representations of the model's geometric details than was previously possible. He expects the automation to be stepped up even further in the future.
"Furthermore, CFD tools will become part of a global computer-aided engineering environment, interacting with other design tools to make optimization tasks faster, easier, and more reliable," Peric said.
That means CFD might be coupled with multiphysical applications, so users can solve for more than one phenomenon at a time. For instance, they could look at a simulation that shows fluid-structure interaction phase change, and electromagnetic effects with multipie phases, Peric said.
Advances in computer power are already making these complicated simulations possible and Peric doesn't expect growth in computer memory and storage capacity to taper off any time soon. In fact, as he sees data storage and memory capacity increase, Peric predicts that virtual reality CFD simulations will become commonplace.
With a virtual reality application, for example, engineers might stand before a three-dimensional representation of fluid or air flowing through a model. They could then put their hands in the stream and immediately see how that affects air or fluid flow. Such technologies are now being worked on at university research centers.
A Boiler from the Inside Out
Until that day comes, engineers still have to rely on standard digitized CFD simulations. But that's certainly good enough for today's users, including Francois McKenty, director of numerical simulation services at Brais Malouin and Associates, an engineering consulting firm in Montreal. His company recently helped improve boiler performance and eliminate some classic problems with a tangentially fired boiler for Cerrey S.A. de C. V., an industrial boiler manufacturer in Monterrey, Mexico. Cerrey asked Brais Malouin to help find ways to increase boiler performance.
Tangentially fired boilers were originally developed in the 1920s to burn pulverized coal within a confined area by means of an intense fireball in the middle of he furnace. Firing the burners tangentially to a central target circle creates the fireball, McKenty explained. These boilers are used by industries that need a large amount of steam for their heating or manufacturing processes, he said.
Through the years, the boilers have been adapted to burn heavy oil and natural gas, although problems with flame stability, flame impingement, uneven wall-heat transfer, and pollutant formation have caused difficulties.
"Tangentially fired boilers are subject to the same operational difficulties as other types of boilers," McKenty said. "Some factors that limit efficiency and can damage a boiler are unevenly distributed heat flux to the water side, which results in circulation problems-pollutant formation, incomplete combustion, flame impingement, and soot deposits."
Adjusting the boiler on site after its completion can help overcome operational problems, but this often results in decreased boiler efficiency and increased pollutant emission levels, McKenty said. Boilers can never be adjusted precisely for one important reason.
"It's difficult to know exactly what's happening inside the furnace," McKenty said. "The boiler constitutes the proverbial black box. We know what goes in and what comes out, but what happens in between is difficult to measure precisely. Engineers and operators can often make only educated guesses about what's going wrong in problem areas. Attempts to correct unfavorable behavior sometimes have limited success because of the lack of precise information about the cause of the problem."
In other words, no one can get inside the fireball-filled boiler to see exactly what's going on in there.
That's where modeling boiler operation with CFD comes in.
"CFD now gives us a tool to open up the black box and take a peek at what's happening inside," McKenty said.
"It not only lets us witness the physical phenomena that are creating problems or limiting performance, but it also lets us economically test possible solutions."
McKenty's company uses CFD technology called Star-CD, from CD Adapco Group in London, to view the combustion and heat transfer processes within the boiler. Engineers couple combustion simulations with BMA's in-house water-circulation-simulation technology to model the entire boiler as it would operate in the field. They then can see how design changes would affect the performance and efficiency of the overall design, including water circulation.
For a recent project, BMA simulated and tested nine different boiler-firing configurations to find the best one.
"We wanted to find a firing configuration conducive to creating a stable, well-centered flame envelope for the fireball, while enhancing heat transfer and avoiding hot spots on the walls. Of course, all this was to be done while keeping pollutant levels to a minimum," McKenty said.
Variables measured within the nine simulations included gas and metal temperatures, heat fluxes, and combustion product concentrations. The tangentially fired boiler optimi zation study is ongoing, although Cerrey has put into effect a number of BMA's findings on burner firing configurations.
"Our job as specialists in combustion aerodynamics is to devise and test new configurations for tangential firing," McKenty said. "We used CFD to evaluate, within six months, ideas that would have required millions of dollars and several years to investigate in the field. In a nutshell, the results were an elimination of flame impingement on the walls. We also reduced peak metal temperatures by 40°F and enhanced furnace heat transfer by 15 percent."
Designing Aerodynamic Autos
Brais Malouin and Associates used CFD to model combustion. Chalmers University of Technology in Gothenburg, Sweden, uses CFD technology to examine the effects of aerodynamics on Volvo automobiles. Using a combination of CFD and visualization software, the university is close to finding a way to predict airflow around cars to help design a truly aerodynamic automobile.
Sinisa Krajnovic, a CFD researcher at Chalmers, said that aerodynamics affects cars in several ways. For example, it plays a role in driving stability. Auto designers naturally want the ride to be as smooth as possible. They need to study airflow because a smooth flow around the auto body reduces wind noise. In addition, dirt and water will accumulate more readily on a vehicle with an unsteady airflow.
Volvo asked the university to come up with a good method of digitally modeling aerodynamic conditions, Krajnovic said.
The largest aerodynamic force on a car is drag. Drag is the more difficult aerodynamic force to study, though it's the largest, because it's caused by wakes, which can 't be predicted with the equation that's used to simulate other aerodynamic forces, Krajnovic said.
Chalmers scientists studied wakes using Large Eddy Simulation, a CFD method. They limited their initial LES studies to airflow around a three-dimensional cube, since the simulation of a real-life car required a great amount of computer memory and time. For this study, they used CFD software from ICEM CFD Engineering of Berkeley, Calif. They used separate software, EnSight Gold, from Computational Engineering International of Apex, N.C., to depict the airflow so they could conceptualize it better. Krajnovic called the visualizations high-powered movies that let him and his team study flow in detail.
After their simulation with the cube, the researchers stepped up to running simulations using a simplified model of a bus. Those tests were successful, Krajnovic said, and the research team is now ready to use the process on a simplified passenger car model.
Sand Through a Pipeline
Malcolm Wallace, meanwhile, polished off his doctorate by looking at sand. Wallace is a development engineer at Computational Dynamics, which is part of the CD Adapco Group. For his Ph.D. project, he used the Star-CD software to model a component developed by Wood Group Pressure Control Ltd., a Houston manufacturer of wellhead and flow-control equipment for oil and gas pipelines.
One of the problems with these pipeline components, Wallace said, is that the sand particles sucked in along with the oil or gas scrape the components and erode them, shortening their lives.
CFD methods can predict the areas of a component that will see the most damage from the sand, Wallace said. To find erosion profiles on the Wood component, he first modeled its fluid flow.
Once Wallace had determined where the sand hit the structure, he used impact data to determine the force and then applied a separate erosion equation to find particle impact velocity and angle of erosion rate-in other words, how the component eroded as a result of the way the sand ran through it.
"By tracking thousands of particles through the system and recording the cumulative erosion rates, it's possible to build up a picture sh owing the areas of erosive wear in a component," Wallace said. "The method will generally highlight the regions of a component most susceptible to solid particle erosion, which can allow more intelligent design."
The CFD-based erosion model let the component manufacturer predict the areas of highest erosion and take steps to either reinforce those areas with a suitable material or else redesign the component.
Peric looks forward to the day when CFD will be used even more widely across disciplines and different types of technologies than it is now, but the computational method is already invaluable to many engineers working with varied, unrelated applications.