Engineering News 7360

Going with the flow

Developments in laminar-flow technology may dramatically reduce drag on next-generation aircraft

Seal Beach, CA--"To the aerodynamicist, the achievement of laminar flow represents perhaps the last remaining major challenge in a science considered by many to be mature and therefore lacking in opportunities for advancement." --Heinz Gerhardt, design engineer, Northrop Grumman Advanced Technology and Development Center.

Laminar flow sometimes appears like an aeronautic magic trick--you can't see it happen, but the result is quite remarkable. In the boundary layer, that thin region of air next to the surface of the aircraft, smooth, laminar flow produces roughly 1/10 the skin drag of its nemesis, turbulent flow. Yet nature abhors order, and a turbulent boundary layer blankets most of the exterior of today's high-speed airplanes. Some recent advancements, however, point to possible ways of generating more laminar flow, a critical factor in the design of many next-generation aircrafts.

One such aircraft is the High Speed Civil Transport (HSCT), a supersonic airliner concept being studied by NASA and all the major aerospace manufacturers. Since skin friction creates about half the total drag on such an aircraft, the difference between laminar flow and turbulent flow represents a 7-10% reduction in direct operating costs.

"That's a huge gain," says Marta Bohn-Meyer, program manager for NASA's F-16 XL supersonic laminar-flow research program. "It's the kind of thing that would allow the industry to offer a seat on the High Speed Civil Transport for almost the same price as a subsonic transport today."

Yet only the leading 2-3% of a highly swept supersonic airfoil is blessed with natural laminar flow. "Anything beyond that requires some creative manipulation," says Bohn-Meyer. Such manipulation appears on the left wing of the F-16XL research aircraft. There, engineers have mounted a close-fitting panel, or "glove," that consists of a titanium panel laser-drilled with more than 10 million tiny holes--about 2,500-3,000 per inch. Below the panel lies ducting connected to a vacuum system powered by a converted Allied Signal cabin pressurization pump mounted in what used to be the fighter's ammunition bay. The pump operates off of engine bleed air, and by sucking the turbulent boundary layer through the tiny holes, it can maintain extended runs of laminar flow over the F-16XL's wing.

Such a system poses several practical challenges, many still to be resolved. For instance, how do you prevent smashed bugs from "tripping" the boundary layer into becoming turbulent? How do you keep dirt from plugging the holes? And what if the suction system fails?

"Bug strikes are a disaster to laminar flow, but it might be that if you go fast enough, the bugs will just burn off," Bohn-Meyer hypothesizes. To prevent pore-clogs on the F-16XL, the holes are tapered--smaller on the outside than inside--so that dirt tends to suck through. And should the system fail, the wing can be designed to continue providing lift, though less efficiently.

Cool and noisy solutions. Engineers at Northrop Grumman looked at another method of achieving laminar flow in their Laminar Supersonic Transport (LST) design study. It has an unusual reverse-delta wing planform that uses thermal laminar-flow control (TLFC) instead of suction. This contrasts with the severely swept wings of proposed HSCT designs that, though they have other advantages, inherently generate lots of span-wise airflow that prevents laminar flow from occurring naturally.

The TLFC system would work by cooling the rear half of the wing surface, a method that has been shown to maintain laminar flow in wind tunnels and CFD analysis. Heinz Gerhardt, engineer on the program, claims that the temperature reduction required would be modest. And, in combination with the natural laminar flow over the front of the reverse-delta wing, the TLFC would be simpler, lighter, and more robust than a suction system, he says.

Others think it's not so simple. "The predictability of thermal systems is much worse," says Mark Maughmer, associate professor of aerospace engineering at Penn State University. "It would depend much more on the environment. I turn on a suction system, I know I'll get laminar flow."

Maughmer describes yet a third method of maintaining attached laminar flow: sound. It seems that the transition from laminar to turbulent flow is accompanied by an increase in particular instability frequencies. These frequencies vary with the wing planform and other factors, but fall within a predictable range. "Bombard the boundary layer with a strong-enough acoustic signal that cancels these frequencies," says Maughmer, "and you can prevent the transition."

Peter Masak, designer of the Scimitar sailplane (Design News, 6/26/96, p. 30), is working on such a system, though the details are still proprietary. Its role, however, isn't to address lami-nar-to-turbulent transition. Sailplanes live in the world of low-speed aerodynamics, where large regions of laminar flow are the norm. The challenge is trying to design an airfoil that attains high lift, yet prevents the laminar flow from detaching from the wing. "It's always good to have laminar flow," says Masak, "but you'd rather have turbulent attached flow than laminar separation."

His boundary layer control system prevents laminar separation, allowing, theoretically, a smaller, more highly cambered airfoil with a greater lift coefficient. It would also generate 15% less drag--a staggering improvement in the mature sailplane world.

On airliners, laminar-flow control will probably debut in incremental steps over many years, as did composites. Boeing has studied 757s outfitted with suction devices, and Airbus is flying an A320 with a suction-based vertical stabilizer. "We've pretty much done everything there is to do to cut the drag of commercial transports with conventional aerodynamic means," says Maughmer. "The next thing with the most promise is laminar flow."

--Mark A. Gottschalk, Western Technical Editor

Compaq launches workstation division

Houston, TX--Compaq Computer is gunning for the engineering desktop.

The company--last year's number-one seller of personal computers--just formed a division to build workstations for the technical marketplace. One of the first areas to be targeted by the Windows NT machines: computer-aided design.

Compaq announced several technical alliances for its new workstation venture, including a graphics partnership with Intergraph Computer Systems. The new workstations will offer "extremely attractive price/performance," Compaq officials say, and should be on the market by the end of this year.

This marks the latest effort aimed at winning over power-using engineers from UNIX systems. Intergraph and Digital Equipment Corp., among others, already offer popular lines of NT workstations targeted for CAD and analysis.

MEMS cuts guidance-system cost, size

Anaheim, CA--Rockwell says its latest MEMS-based inertial measurement unit for "smart" weapons systems dramatically boosts guidance accuracy while cutting back on size and manufacturing costs. The technology may also find its way into civilian systems that might prevent accidents such as the crash of Commerce Secretary Ron Brown's plane in Croatia, company officials say.

The Digital Quartz Inertial Measurement Unit (DQI) uses microelectromechanical sensors (MEMS) batch processed from a quartz substrate, a method something like that used to create semiconductor devices from silicon. This means the DQI sensors are only 1/10 the size of sensors that use other technologies, such as fiber-optic gyros, while also being less expensive to manufacture, according to Program Manager Charles Dutch.

MEMS technology has been used in such systems before, Dutch says, but proprietary Rockwell processes substantially improved accuracy. While previous generations of the technology used sensors with hundreds of deg/hr bias stability, the current unit boasts short-term bias stability of 1 deg/hr.

The DQI recently passed 14 major qualification tests for the AGM-130 Guided Munition Weapons System program, and has been picked up for several other military programs, including an unmanned aerial vehicle.

In addition, airlines are looking at the technology for possible use in autonomous precision approach landing systems, which would help planes land in poor visibility at airports with less-than-ideal ground-based guidance systems. "A number of commercial airlines are very interested," Dutch said.

Analysis ensures optimal balance

Darmstadt, Germany--Testing in a wind tunnel lets design engineers simulate the conditions an airplane endures during flight. Engineers use a measuring device called a balance to gauge how the plane will react to drag, lift, and side forces while the wind tunnel operates.

At the Technical University of Darmstadt, Germany, Junnai Zhai builds precise balances based on designs analyzed and optimized with Algor software from Algor, Inc., Pittsburgh, PA.

Zhai recently used Algor Superdraw and Supersurf software to design a balance and generate a finite element model (FEM). He then used several Algor analysis packages to determine how the balance would respond to various forces of the wind tunnel.

For example, linear stress analysis software helped determine that the balance was not stiff enough to withstand the stresses of the testing. To solve the problem, Zhai increased stiffness by 12%.

During testing, temperature change is a big influence on the balance, the airplane prototype, and the wind tunnel. The temperature drops dramatically when the wind tunnel is operating.

Zhai created a model of the entire system in Superdraw and used Algor's heat transfer analysis software to test the system over a temperature range of 104F. First results showed that the wind tunnel, airplane, sting, and balance would withstand the temperature changes, but that the balance would not. This prompted him to use a different metal alloy to build the balance.

"The eventual design of an airplane depends on the balance design," says Zhai. "Algor software gave me the most accurate results to optimize my balance design."

CFD optimizes complex inlet design

Albuquerque, NM--To reduce costly wind-tunnel testing on the inlet of a missile engine, engineers at Sandia National Laboratories wanted to refine the design with computational fluid dynamics (CFD) before progressing to physical testing.

Engineers needed to develop a design that would be undetectable by radar, yet still provide good performance characteristics such as a uniform air flow with no separation. The air coming out of the inlet and into the turbine had to be uniform as well.

Sandia's intent was to study different design configurations on the computer instead of using the wind tunnel. To create the complex inlet design, engineers chose RAMPANT CFD software from Fluent, Inc., Lebanon, NH. RAMPANT uses an unstructured grid to provide automatic mesh generation. Once the surface is defined, the grid is generated without user intervention.

The first step in solving the problem was to import an I-DEAS-generated surface mesh into RAMPANT. There, engineers created an initial concept model with approximately 200,000 cells in about four hours. Sandia engineers estimate that the same model could be created in as little as 15 to 20 minutes with the current version of the software.

Initial analysis showed a region of air separation. More importantly, the pressure did not recover well at the outlet and was not uniform. Sandia engineers went back to I-DEAS and adjusted the geometry in about one half hour. The result of the second analysis: a more uniform distribution of pressure at the outlet.

Using CFD showed engineers that the original design could be improved. Without the analysis, it would have taken a week of testing the original design in the wind tunnel to arrive at this conclusion, Sandia engineers say. Overall, CFD saved Sandia five days of wind tunnel testing and at least $100,000.

FEA resolves maintenance issues

Seattle, WA--Just like an automobile, airplanes suffer damage from dents and corrosion over years of use. But how serious is a small dent to the safety of a huge 747?

To answer such questions, engineers in Boeing's 747 Post Production group use finite element analysis tools. Structure engineer Patrick Safarian says FEA helps Boeing provide sound, quantitative responses to its customers' questions about maintenance and repair.

In one instance, Boeing customers were concerned about corrosion in the metal tracks that hold the seats in place. The companies wanted to know if sanding off the corrosion was sufficient to solve the problem. Boeing responded with a simple "No," requiring instead that the tracks be replaced.

But the customers weren't satisfied. The issue: Could a specific amount of material be removed without reducing the integrity of the tracks, before resorting to costly and difficult replacement?

Using ANSYS FEA software from ANSYS, Inc., Houston, PA, Safarian created the geometry of the seat track cross-section from design specifications. He used symmetry boundary conditions, then performed structural analyses to study how the track responded to extreme loads with various amounts of material removed.

Within a few days, Safarian had an answer. The FEA results verified the original recommendation. This time, however, Boeing was able to quantify exactly how much material could be removed before replacement was clearly mandatory.

"When we can quantify our recommendations, the customer is much happier," says Safarian. "That is a major benefit of using FEA."

Hybrid maker OKs second-source silicon in two weeks

Orlando, FL--The Low Altitude Navigation and Targeting Infrared system for Night (LANTIRN) is designed to let fighter pilots fly beneath enemy radar in total darkness, at speeds exceeding 500 mph, and destroy several targets on the first pass. Lockheed Martin, which developed and builds LANTIRN, had a fast-turnaround requirement for prototype units of a multichip-module (MCM) amplifier circuit.

The primary vendor of 5-channel silicon chips was running out of stock, so Lockheed-Martin asked Micro Networks Corp. (MNC), Worcester, MA, to assemble 25 prototype amplifiers to evaluate a new source for the discontinued silicon chip. Because MNC had previously supplied components for the program, it was already tooled up to do electrical testing and characterization of the amplifiers.

"We had to quickly make up an evaluation lot for Lockheed Martin, and do all the testing and characterization so they could then try the units in their end application," says Mark Sullivan, MNC hybrid development design engineering manager.

The hybrid amplifier used in the LANTIRN program is a custom device. It provides 15 channels of audio-range amplification and is based on a custom IC that incorporates five channels per IC. (A hybrid is a collection of several silicon chips put into one package.) In addition to the custom IC, the multichip device contains passive components such as resistors and capacitors. The interconnect pattern was incorporated within the 0.7x0.7-inch package itself--no internal substrate was needed.

Under Micro Networks' MN2-20 program, the company supplies the design rules needed for integration into the customer's CAD program. With these rules, the customer develops the MCM interconnect pattern using its own CAD software. The customer then sends the CAD circuit showing die and wire placement with components for 20 prototypes to Micro Networks, which fabricates the substrate, mounts and wires the circuit elements, and inspects and seals the units.

"Since these prototypes are built on our ISO 9001 line, they can serve as pro-duction verification units, enabling customers to proceed directly to manufacturing," notes Lee R. Allain, Micro Networks president.

Tiny jets deliver big performance

Seal Beach, CA--For all those engineers who've dreamed of owning their own jet, the wait may be over. Within the past few months, three companies have announced or introduced tiny turbojet engines, ranging in thrust from 7.5 to 100 lbs with price tags of thousands of dollars instead of millions. Of course, none will power that Gulfstream V you've been drooling over, but it's a start.

Technical Directions Inc. (TDI), based in Ortonville, MI, targeted the TDI-J7 for unmanned-vehicle applications such as decoys and drones. The 7-inch-diameter turbojet weighs 16 lbs and generates 100 lbs of thrust at 96,000 rpm. "We've identified the need for an inexpensive, expendable engine, and nobody out there makes one," says Vern Brooks, TSI's president.

"The duty cycle is minutes, not hours," he explains. "This gave us some design flexibility in the engine."

Such flexibility appears in the engineering of the rotating components, which were adapted from a diesel-truck turbocharger. One wheel of the turbocharger acts as the compressor, the other as the turbine. Brooks designed the combustor to fit in the limited 3-inch space between the two stages, a significant constraint.

To further keep down cost and weight, he patented a system whereby the fuel is used for both cooling and lubrication. Fuel passes through the aft bearing to the front bearing and then to a fuel slinger, a rotating cup on the drive shaft that throws the fuel into the combustion chamber via centrifugal force. "This accomplishes two things," he says. "We cool and lubricate the bearings and pre-warm the fuel in one step."

TSI's engine contrasts with the SR-30 of Turbine Technologies (TTL) of Chetek, WI. It's engineered for long life, with the turbine wheel and vane guide ring being vacuum formed at TTL's facility from super alloys such as MAR-M-247 and Inconel 718. The company designed the engine several years ago for drones or UAVs, but it flew for the first time just a few months ago.

The SR-30 produces 40-lbs thrust from a 10-lb package measuring 10 inches long by 6.75 inches in diameter. TTL recently began offering the engine as part of a Mini-Lab^TM, a self-contained, turn-key propulsion laboratory for universities and technical training programs. All that's needed is 110V electric power and 100-psi shop air for starting. To date, both Embry-Riddle Aeronautical University and the University of Alabama at Tuscaloosa have purchased one for teaching. Engine life is expected to be more than 2,000 hours.

Tackling the truly tiny-turbine market is Golden West Models' FD 3/67 LS, an 8.7-inch-long, 4.3-inch-diameter jewel that weighs 26.5 oz and develops 7.5 lbs of thrust at 100,000 rpm. It builds on Kurt Schreckling's model-aircraft turbojet design, later put into production by Schneider-Sanchez of Austria.

"For one, we re-engineered the turbine wheel," says Bob Wilcox, GWM's cofounder. "It was flat plates welded to the hub (in the Schneider-Sanchez motor). Now it's punched blades with curvature." Engineers also upgraded many of the materials to Inconel, titanium, and stainless steel. Interestingly, Schreckling's home-built plans called for a wooden hub reinforced with composite and a housing made from a gas-bottle container.

The greatest addition may be the patented electronic control. It regulates fuel flow depending on inputs from an exhaust-gas temperature sensor, engine tachometer and throttle setting. A second patent covers the fuel vaporization system that carries fuel through the combustion chamber inside 3/16-inch-diameter Inconel tubing. Fuel consists of a blend of kerosene and Coleman gas. Engine cost: $2995.

--Mark A. Gottschalk, Western Technical Editor

Nintendo dazzles with 64-bit graphics

Redmond, WA--For $250 you can buy a video game whose graphics rival those of high-end workstations.

"Nintendo 64 puts Silicon Graphics' visual computing technology into the hands of everyday consumers," says Silicon Graphics President and COO Tom Jermoluk. The company plans to provide MIPS RISC-processor technology to consumer companies worldwide, enabling a new generation of affordable, media-rich 64-bit electronic products, he says.

"This game offers the graphics capability of a $100,000 workstation five years ago," says J. Scott Carr, principal engineer for Silicon Graphics' Project Reality. "Ten years ago, this technology would have cost you $14 million and it would have been classified."

In fact, Nintendo developed the game on an emulator running on an SGI workstation. The unit has 4 Mbytes DRAM, plus a memory expansion slot.

At the heart of the Nintendo 64, which Nintendo claims is "the world's first true 64-bit home video game system," is Silicon Graphics' "Reality Immersion Technology." It consists of a custom version of the MIPS 64-bit R4300i RISC microprocessor manufactured by NEC Corp., a MIPS "Reality Co-Processor," and an embedded software layer--all designed to power a new class of interactive consumer applications. (MIPS Technologies, Mountain View, CA, is a subsidiary of Silicon Graphics.)

Speeding along at 93.75 MHz, the R4300i controls the system and game environment. It runs the MIPS 3 instruction set, has a 16-kbyte instruction cache and 8-kbyte data cache, and features 0.35-micron process technology. The chip is rated at SPECint92 and 45 SPECfp92 (measures of integer and floating-point performance).

The 62.5-MHz copro-cessor, which MIPS designed for Nintendo, implements Silicon Graphics' digital media and visualization capabilities and performs audio synthesis. Specifically, it offers:

- High frame rate for complex images, which enables interactive movement through virtual 3-D space via a combination of high fill and transform rates;

- Dynamic media mixing, which allocates the resources necessary for real-time processing of stereo audio and 2- and 3-D graphics;

- Real-time antialiasing to remove the jagged edges from objects;

- Advanced texture mapping techniques that generate high-quality textures and retain the natural texture of every object in the scene, independent of how close the player is to an object;

- Real-time depth buffering to remove hidden surfaces during the real-time rendering of a scene;

- Alpha blending to show transparent objects, such as water and dust clouds; and

- Automatic load management to enable objects to move smoothly and realistically by automatically tuning graphics processing.

Also new is the analog joy stick with 14 digital inputs. A slight push on the stick makes Mario walk; a full push gets him to run. Buttons let you change camera perspective to see what's hidden behind a corner--or to find Mario in his 3-D world.

I'm new to video games--and Mario--but was blown away by the naturalness with which Mario makes his way through various obstacles. Especially impressive were the smooth and rich graphics, and the realistic way water is presented. When Mario swims, you can tell by the water's transparency and reflectiveness that he's underwater, and bubbles slowly rise from his breathing.

--Julie Anne Schofield, Associate Editor

NASA ventures on aerospike for VentureStar

Palmdale, CA--The X-33 "flying wedge" configuration NASA selected as a subscale reusable launch vehicle (RLV) technology demonstrator dusts off a number of concepts shelved for decades. While the VentureStar under development at Lockheed-Martin's Skunk Works appears suitably futuristic, one can find its roots back in the 1960s.

The genealogy of VentureStar aerodynamics can be traced back to the M2-F1 lifting body designed in 1962 at NASA's Ames Research Center. Martin's rocket-powered X-24B, the last in the series, was built in 1973. According to Skunk Works spokesman Ron Lindeke, the X-33 is a natural extension of the lifting body program.

VentureStar is being fitted with a metallic thermal-protection system (TPS) manufactured by Rohr Inc., St. Louis, that eventually would shield a full-scale RLV during re-entry. Brad Moore, project engineer at Rohr, says the mechanically alloyed titanium and Inconel 617 shells have honeycombed cores between two sheets and will provide VentureStar with a TPS for less weight than the ceramic tiles used on the Space Shuttle. Rohr has used this technology for jet-engine nacelles and nozzles: X-33/VentureStar is the first time metallic TPS will be used for spacecraft.

Getting down again is only half the challenge. Lockheed-Martin unearthed an experimental engine technology it hopes will propel VentureStar into orbit. Designers are hitching their wagon to an aerospike rocket engine developed by Rockwell International's Rocketdyne Division, Canoga Park, CA, that has never flown before.

Rocketdyne's aerospike engine is much like a conventional gas generator cycle rocket engine, except it has no bell-shaped nozzle. Exhaust gas flow is directed inward from seven thrusters along opposite walls of a truncated spike nozzle. Exhaust also presses in on the base of the nozzle, producing additional thrust. An aero-spike nozzle creates a plume that expands as outside pressure falls, resulting in efficient performance at all altitudes.

The X-33 will have two such engines, while the operation-al VentureStar RLV would have seven. Rocketdyne is incorporating a number of advanced features into its aerospikes, such as electromech-anical valve actuators instead of hydraulic ones and laser ignition instead of spark plugs. The RLV aerospike nozzle will be made from a carbon-silicon composite to provide weight savings.

Steve Cook, NASA's deputy program manager for the X-33, said the turbo-machinery in aerospikes is simpler than in closed gas cycle engines, such as those used on the Space Shuttle. "Most of the design challenges are integration-related," Cook says.

Mike Hampson, Rocketdyne's phase one program manager, says the aerospike is by design a perfect fit for the X-33. Unlike most rockets, which are designed to accommodate several possible engine configurations, VentureStar is designed specifically for the aerospike. "We have a design team at the Skunk Works and electronic links between Palmdale and Canoga Park to coordinate our efforts," Hampson says.

Rocketdyne has performed ground-fire tests of various aerospike engines since the early 1970s. But in picking the X-33, NASA is gambling on the technology flight unseen. A prototype aerospike was to have flown aboard an SR-71 testbed aircraft last April, but those tests were postponed until this Fall due to technical problems with an engine valve.

Nevertheless, NASA's Cook is confident that aerospike will perform as advertised, citing computational-fluid-dynamics analysis and cold-flow wind tunnel test results.

--Michael Puttre, Associate Editor