Engineering News 8005

Torino scale has impact By Rick DeMeis, Senior Editor

Cambridge, MA-No, it's not a device for measuring old Ford muscle cars. The Torino Scale is a risk-assessment measure, similar to the Richter Scale, that assigns values to celestial objects moving near the Earth (see Design News, 9/21/98, p. 35). Developed by MIT professor Richard Binzel, it is officially the Torino Impact Hazard Scale, named for the Italian city in which the International Astronomical Union adopted it this past June.

Here's how the scale combines size, speed, and probability of collision with Earth:

Events having no likely consequences

0. Collision probability is zero (well below random) or for small objects unlikely to reach Earth's surface

Events meriting careful monitoring

1. Collision unlikely, same as for random object of same size striking Earth

Events meriting concern

2. Somewhat close, but not unusual, encounter; collision very unlikely

3. Close encounter with 1% or greater chance of collision causing localized destruction

4. Close encounter with 1% or greater chance of collision causing regional devastation

Threatening events

5. Close encounter with significant threat of causing regional devastation

6. Close encounter with significant threat of causing global catastrophe

7. Close encounter with extremely significant threat of causing global catastrophe

Certain Collisions

8. Collision capable of localized destruction (between once per 50 to 1,000 years)

9. Collision capable of regional devastation (between once per 1,000 to 100,000 years)

10. Collision capable of global climatic catastrophe (once per 100,000 years or longer)

To date, no asteroid has ever had a value greater than one. Several objects with initial values of one were later reclassified to zero after additional orbit measurements.

Piezoceramic makes vacuum motion a snap

By Jean Gonzalez, Western Regional Editor

Yokneam, Israel-Most conventional electric motors can't operate in a vacuum. Problems with winding surface areas, bearings, brushes, and heat dissipation make it difficult. Not to mention electromagnetic or electrostatic interference. Up until now, the answer has been to mount the motor outside the vacuum chamber, but this can be expensive and cumbersome.

Nanomotion offers a novel solution a motor which uses piezoceramic standing waves for vacuum applications typical in robotic, semiconductor, disk drive, and biomedical equipment. The motor is compact, weighing in at 91 grams and measuring 26 mm high, and has no moving parts. Suitable for both linear and rotary motion, two motors are offered: the SP-8V which operates down to 10^-7 torr, and the SP-8UHV, rated to 10^-10 torr for ultra high vacuum compatibility. Both offer a wide velocity range of 1 mu/sec to 250 mm/sec at 5 nanometer resolution. The motors can interface with any off-the-shelf servo PID controller.

How it works. Under special excitation drive and ceramic geometry, it is possible to excite a transverse bending vibration mode in close proximity to the longitudinal mode frequency. The simultaneous excitation of these two modes creates a small elliptical trajectory of the ceramic edge. By coupling the ceramic plate to a precision stage, a non-symmetrical driving force is exerted on the stage, causing it to move. The periodic nature of the driving force at frequencies much higher than the mechanical resonance of the stage allows continuous smooth motion for unlimited travel, with high resolution and positioning accuracy. Travel can be linear or rotary, depending on the coupling mechanism. Typical applications are microscopy and light industrial machines.

Advantages of Piezoceramic Versus Electric Motors

Expert offers Cool Hand to refuse cans

By Rick DeMeis, Senior Editor

Saint-Nicolas, Quebec-What has four wheels, flies, and is operated by one person? Labrie Equipment's latest side-loading, compactor-type refuse truck, the Expert 2000. One option on the truck is a robotic arm, the Cool Hand, which lifts and dumps containers of many shapes.

The driver controls the Cool Hand from either the left or right seat, or a right-side standing station. The arm is hydraulically powered, as is the refuse compactor system. However, while ram-type compaction and load dumping is best done with a fixed displacement, vane-type pump (for constant hydraulic flow and lower cost), the arm functions require a variable displacement, piston-type pump (for proportional flow). The latter takes joystick movements and translates them into proportional "arm" and "hand" extensions and retractions. Such conflicting requirements are usually met by a compound system of a stacked vane unit and a piston-type pump having their drive shafts linked via a coupling. But this is not an ideal solution because of compound pumps' larger size, noise, plumbing and installation difficulty, maintenance costs, and lower reliability than the individual pumps.

Enter Denison Hydraulics (Marysville, OH), which in the early '90s began developing a pump that would solve such difficulties. The result is its Hybrid line of pumps, which combines a piston and vane-type pump mounted on a single shaft, in a common housing, having a single hydraulic inlet for simplified piping. Challenges Denison designers faced in combining two types of pump technology in a single package, according to Bill Johanson, product and marketing manager, included feeding the pumps from the same inlet without adverse interactions from combining variable and fixed volume flows. "The contour of the transition section between pump inlets and the flow paths to both pumps were designed to minimize pressure drops," he notes.

In the Expert truck application, the pump mounts in front of the engine, coupled directly to the crankshaft-driven power take-off. Other installation benefits are the device's compact size and the fact that pressure outlet ports can be mounted in any of eight different positions.

Air comes to light

By Kenneth O. Gibbs, Staff Editor

Albuquerque, NM-Triggering light paths in the sky may have all kinds of aviation safety uses. The key to these may be an atmospheric light guide or light string, an invisible, low-energy beam of laser light. Ewan M. Wright, in collaboration with light string discoverer Jean-Claude Diels of the University of New Mexico, and Jerry V. Maloney of the University of Arizona (UA) Center for Mathematical Sciences, is currently advancing the technology in UA's Optical Sciences Center.

Discovered by Diels and a group of University of Michigan researchers in 1995, light strings only as wide as a human hair have already been used by German scientists to produce white light sources higher than six miles in the sky. Wright believes the German's accomplishment is proof that light strings may be valuable in commercial applications, such as aerospace navigation.

The phenomenon occurs when low-energy, electrically charged channels of air are created that allow short laser pulses to travel within them. "The key is using high power but extremely short laser pulses," says Wright, "If we use longer pulses the laser doesn't propagate as a string; it just produces an electrical discharge that breaks down instantly and terminates the propagation. We call this optical breakdown." In optical breakdown, the atoms of air are ripped apart by pulses from high-powered lasers, but if too short, the pulses can't pull enough electrons from surrounding air molecules to produce light strings. The laser pulses that create the light strings are only a trillionth of a second long.

James Murray of Lite Cycles (Tucson, AZ), which develops LADAR (Laser Detection and Ranging) systems, which are like radar but use lasers, expects to benefit from the research being done. Clouds and fog scatter light, a problem that Murray believes light strings may be able to solve by penetrating them.

"There will still be light scattering, but we think we can bring it down by orders of magnitude so the laser, for all intents and purposes, can get through," says Wright. If so, light strings may one day be used to help pilots avoid turbulence in flight paths and windshear during take-off, by creating visible paths of light that expose these threats, saving money, time, and lives.

Students show skills in NASA/FAA contest

Oshkosh, WI-Embry-Riddle Aeronautical University (Daytona Beach, FL) students captured the first place award in the National General Aviation Design Competition. Their design of an affordable jet aims to attract customers interested in moving up from turboprops without increasing pilot training and skills.

Embry-Riddle was one of four university engineering teams who received aircraft-design awards in the joint NASA and FAA-sponsored design contest.

Student designers submit concepts for a fixed wing, single engine and single pilot, 2- to 6-passenger aircraft. The yearly contest aims to stimulate engineering schools to participate in NASA's AGATE program, a major national effort to rebuild the U.S. general aviation sector (see Design News 9/6/99, p. 88). Teams addressed design challenges for integrated cockpits, propulsion, noise and emission, integrated design and manufacturing, aerodynamics, and operating infrastructure.

Powder metal industry enjoys record growth

Vancouver, British Columbia-"The powder metallurgy industry has fulfilled the promises of the past." Those were the encouraging words of Donald G. White, executive director of the Metal Powder Industries Federation (MPIF), as he addressed more than 1,200 attendees at the 1999 International Conference on Powder Metallurgy & Particulate Materials.

White was referring to the latest figures that showed 1998 metal powder shipments in North America hit 511,185 short tons, which not only represented a 5% increase over 1997, but set a new industry record. Breaking this down, White noted that North American iron powder shipments last year increased 5.4% over the previous year to 410,555 short tons. The powder metallurgy (P/M) parts share of this figure grew to 376,769 short tons, which amounts to 91.7% of the total iron powder market. White also reported that iron powder shipments for the first four months of 1999 enjoyed a 4.7% increase compared with the same period in 1998.

What does all this add up to in dollars? White estimated that the North American P/M parts business exceeded $2 billion last year. He attributes this to an increase in P/M parts usage in autos and light trucks, strong growth in new home building, and a vibrant national economy. For instance, White reports that the typical family vehicle now contains more than 35 lbs of P/M parts. He cites General Motors' transverse-mounted electronic transmission that contains 26 lbs of P/M parts as an example.

One area that is experiencing an exceptionally large industry growth at present is powder-injection-molded parts. White estimates current sales of these components at $100 million annually, and growing at a rate of about 20% a year. Major markets include medical, automotive, business machines, and electronic components.

Other sectors of the P/M market enjoying favorable growth levels: rapid prototyping, spray forming, metal-matrix composites, metal foams, nanoscale powder, and stainless steel metal fibers.

Heavy equipment relies on rugged pressure sensors

Plainville, CT-High pressures, surges, and harsh operating environments are the daily enemies of hydraulic systems since these factors can cause component failures from pressure overload. Now sensor manufacturers are under increased pressure themselves to meet OEM needs for rugged, durable components at reduced prices.

Responding to the needs of hydraulic systems designers, Gems Sensors (Plainville, CT) has produced a pressure transducer sensor that meets the challenges of daily use in agricultural and construction equipment, and off-highway and railway vehicles. Gems' psibar^(R) pressure transducer offers long-term stability in adverse ambient conditions at a price below $100.

"OEMs said they would be willing to sacrifice some performance for reduced price," says Carl Schoenherr, Gems Sensors product manager. Key "must have" requirements are all stainless steel wetted parts, no O-rings, no liquid filling, high fatigue life, low thermal errors, good overpressure characteristics, and freedom from drift. "We developed a more relaxed sensor with 0.5% accuracy versus 0.25% for applications where repeatability and drift are critical but accuracy is not a dominant factor," notes Schoenherr.

Efficient assembly. "We did not invent anything new, instead, we combined many things from our 30-year knowledge of developing sensors," Schoenherr says. Traditional technologies sputtered thin film and chemical vapor deposition (CVD)were used, but Gems' engineers rethought their way of calibrating, assembling, and manufacturing.

A key psibar transducer component is the strain gauge sensor, manufactured with CVD technology. The result is a polysilicon gauge atomically fused to a stainless steel beam. Each sensor is produced on a 50-mm diameter wafer, with 100 sensors per wafer and 12 wafers per batch. The common beam sensor has only two versions, covering the 1-400 bar pressure range. "This arrangement gives the benefit of low cost, large quantity production," Schoenherr adds.

The sensor beam is laid across a nickel-brazed, stainless steel diaphragm. Stainless steel is used for the sensor beam, diaphragm, and pressure port with all welded construction. "This composition allows the transducer's body and diaphragm to expand and contract at the same time enabling excellent thermal properties," Schoenherr says. The ends of the sensor beam lie on the edge of the diaphragm capsule and support. A spot weld is made at the center of each end and at both sides of the beam. Then a stitch weld is applied to ensure the beam is held flat and does not twist or tilt during the laser-welding process. "The thicker diaphragm provides higher overpressure capabilities," Schoenherr explains. "It allows ten times the burst rating and four times the pressure rating of conventional units."

Simple setup. During assembly, Gems Sensors takes the capsule/sensor assemblies and fits them with an Application Specific Integrated Circuit (ASIC) compensation board. ASIC technology allows the compensation data to be programmed into the transducer. "This enables simple and fast set-up," Schoenherr says. The core unit is then assembled with an RFI board and serialized, and the outer sleeve is welded.

"In the field the psibar transducer offers low long-term drift, is accurate to within 0.5% and has an operating life of more than 100 million cycles," Schoenherr says. "This high level of stability is important to OEMs."

Schoenherr adds that the psibar transducer will not vary by more than 2% over the temperature range and will maintain a reasonable thermal performance. It is compatible with all common fluid media and its stainless steel parts eliminate the need for O-rings or oil screens. "Thus, it offers freedom from possible leak paths and oil media contamination," Schoenherr concludes.

ANSYS joins foundation sponsors

Canonsburg, PA-ANSYS Inc., a global provider of advanced computer-aided engineering (CAE) software is joining the Design News Engineering Education Award Foundation team with a $10,000 grant that will benefit engineering students at the university level.

"ANSYS is honored to participate in this commendable and worthy foundation and awards program," says President James E. CashmanIII."Technical and economic developments are changing the nature of manufacturing and creating a demand for engineers who understand advanced computational techniques. Thousands of engineers will be needed to meet this demand, and our institutions of higher education must be capable of preparing them. The goal of our educational initiative is to ensure this happens."

The company was founded in 1970 by Dr. John A. Swanson to develop, support, and market the ANSYS^(R) program, a finite-element analysis code used in the CAE environment. Today, the company's software products help customers meet engineering challenges as well as compete in a global marketplace.

In June, the company announced its release of DesignSpace^(R) v5, which adds assembly analysis functionality. The software uses both automatic contact surface detection and surface-to-surface contact element creation to provide easy setup of an assembly analysis. These capabilities allow users to specify one part in the assembly to be non-critical, while dialing in the adjacent part to be critical for higher accuracy.

The ANSYS Inc. Educational Program aids more than 2,000 colleges, universities, and educational institutions worldwide in teaching the fundamentals of finite element analysis. The company will move into the next millennium increasing the awareness of the educational Program, which stresses the importance of collaborative engineering by providing affordable, computer-aided and advanced simulation technology.

Deep Space 1 is a success

Pasadena, CA-"This is a dramatic finale to an amazingly successful mission," said Marc Rayman, chief mission engineer and deputy mission manager, after NASA's Deep Space 1 experimental spacecraft (See Design News, 9/6/99, pg. 92) successfully flew above the surface of asteroid 9969 Braille using a new sophisticated autopilot system. This was the closest flyby of an asteroid ever attempted, at less than 15 km (10 miles). Ten minutes after the flyby, the spacecraft turned back to face the asteroid. The turn, indicated by a marked Doppler shift, was a clear early indicator of a successful encounter.

"With AutoNav's successful piloting of the spacecraft, we've completed the testing and validation of the 12 new technologies onboard and acquired important science data, including photos," says Rayman from NASA's Jet Propulsion Laboratory (JPL).

Launched Oct. 24, 1998, Deep Space 1 is the first mission under NASA's New Millennium Program, which tests new technologies for future space and Earth-observing missions. The technologies that have been tested on Deep Space 1 will help make future science spacecraft smaller, less expensive, and more autonomous so that they rely less on tracking and intervention by ground controllers.

Industrial PC fits on a chip

By David J. Bak, International Editor

Wetzler, Germany-What is claimed to be the world's smallest IPC, the world's smallest Web Server, and the world's smallest IEC 6 1131 controller opens up a world of possibilities for industrial automation technology. Specifically, integration into mass-produced articles, into sensors and actuators, and into the smallest of control systems.

Beck IPC GmbH has put the basics of an industrial PC on a chip. Introduced in prototype form at the Hannover Fair last Spring, the IPC@CHIP offers the essential components of conventional industrial PCs. The first version features:

A 20-MHz AMD 186 CPU

Flash memory is split into two drives: drive A and drive B. The former is read only and contains the operating system and application kernel. For example, PLC, DOS, or Web Server. Drive B holds the application program such as HTML files. There is communication software to download into drive B and communication software to change the operating system.

"Initially, the customer will receive an IPC@CHIP, which is only a very basic functional device a TCP/IP stack plus CD ROM," explains Bernhard Plagemann of Beck IPC customer relations. "With the TCP/IP stack and CD ROM, however, the customer can then download the desired operating system via Ethernet."

Once the OS is downloaded, the IPC@CHIP can function as a PLC, a traditional PC using DOS, or a Web Server. "The two TTL serial communication ports can be converted to RS232, RS485, CAN, Interbus, Profibus, or anything you like," adds Plagemann.

The first IPC@CHIPs will be a 3.3 V version with I/O pins compatible to 5.0 V. Out of the 32 pins on the chip, 20 will be multiuse and defined by software. These can be used as traditional digital I/Os or they can be multiplexed for more I/Os. The system permits a synchronous or asynchronous serial interface depending on how the pins are defined.

"The customers decide how to program the system," Plagemann notes. "If they want a PC, they can use C or C++; if they want to use the IPC@CHIP as a Web Server, they can program with HTML or JAVA.

First applications for the IPC@CHIP, predicts Plagemann, will be within an intra-, not Internet, setting. In fact, Beck's parent company FESTO GmbH, is looking to use the chip-size IPC in their service units. The chips can be everywhere, allowing the system operator to look at each valve or valve terminal via Internet technology, which is much less expensive than traditional visualization systems.

Samples of the IPC@CHIP are presently available with production planned for the end of 1999. Unit price for individual deliveries is approximately $50.

Dynamic simulation lowers landing gear vibration

Troy, OH-Chatter, squeal, shimmy, and other vibrations in aircraft landing gear systems are not only disconcerting, but also may affect the stability of the plane during take-off and landing. So engineers designing these aircraft systems must track down and correct vibration sources among numerous components and subassemblies.

Although mathematical equations representing various parts of landing systems are well established, solving the problems manually can be slow and laborious. Simplifications made to reduce problem size may introduce inaccuracies, such that a design modification to correct a problem in one area can cause unforeseen vibration in other parts of the structure. In many cases, vibration problems may not be uncovered until physical prototypes are built and tested, adding considerable time and expense to the product development cycle.

At the Aircraft Wheel and Brake Technology Division of BFGoodrich Aerospace, engineers create computer models of landing systems using ADAMS mechanical system simulation software from Mechanical Dynamics Inc. (Ann Arbor, MI) to study and correct these complex vibration problems.

Designers build models by combining representations of individual parts such as tires and wheels with connecting joints. Engineers apply input driving loads to represent runway and aircraft forces. The software then solves equations of motion for the system and computes displacements, velocities, accelerations, and reaction forces for all parts. Results are plotted in graphs or displayed as animated models to show the dynamic behavior of the entire landing gear or selected subassemblies.

Dave Moser, manager of engineering technology at the division, says that using computer simulation and analysis is key to the company goal of reducing the design cycle time by 50% and lowering customer product cost by 35%. "Meeting these goals is critical to our business," says Moser. "In the past, aircraft manufacturers allowed four to five years from proposal to product delivery. The OEMs typically have cut that time in half, so we don't have the luxury of iteratively refining the design experimentally with physical testing of prototype hardware. Instead, we rely on simulation to optimize the product faster and less expensively in the early stages of development."

Plane makers formerly subcontracted the various components and subsystems and integrated the landing system themselves, says Moser, but are now relegating the responsibility for the whole system to individual first-tier suppliers. "In that respect, we're moving to position BFGoodrich Aerospace as an integrated landing-system supplier," he says. "Crucial to this strategy is ability to perform dynamic simulation on the entire landing system to evaluate overall performance, especially the many vibration modes possible in these complex systems."

According to Harsh Vinayak, analysis group leader at the Aircraft Wheel and Brake Technology Division, landing systems can have 16 different dynamometer test configurations for a four-brake gear truck such as the Boeing 747, or 68 configurations for those with a six brake gear truck like the 777.

"Physically testing each condition would take three to four weeks, plus time to build the hardware. So previously we tested only four of the configurations we suspected would be troublesome," says Vinayak. "Now we analyze all of them with dynamic simulation, with each run taking only a few hours. We can do a more thorough evaluation of the landing system design in much less time, plus we eliminate the risk of trouble surfacing later in a configuration we didn't test."

The analysis group has used ADAMS since 1995 to study a variety of vibration phenomenon, including low-frequency gear walk, wheel chatter, and brake squeal as well as higher frequency disk squeal. "Dynamic simulation is a particularly valuable tool in understanding the effect of energy-sharing and force-coupling between the different vibration modes," explains Vinayak.

He says the parametric capabilities of ADAMS are useful in performing multiple analyses on the same landing gear, allowing engineers to run "what-if" simulations by merely changing tabular values in a matrix of conditions. Also, dynamic models can be created quickly for different aircraft because the basic braking system configuration is already put together in the computer.

Optimization capabilities in ADAMS enable engineers to refine configurations quickly around a selected set of critical variables and design constraints. In one project to determine the best fit for a landing gear retraction, Vinayak says that the optimization routine ran through more permutations in eight hours than engineers would have possibly investigated in months. "Most importantly, the simulation showed us that one variable we had previously regarded as insignificant was actually a critical parameter in determining the overall stability of the system," he adds.

Wire mills make precision shapes, cut costs

North Haven, CT-Ever wonder how manufacturers come up with long, slender, but intricately detailed parts like eyeglass frames, ski edges, headbands, or slinkies? Many of them make these parts by cutting and forming sheet. But that might not be the least expensive way.

Just like hydroforming and silicon-based electrolytic surface coating (see Design News, 4/19/99, p. 63), having parts produced on a wire mill is a little-known way that engineers can use to keep their costs lower than competitors. You can create the design once in the dies and reproduce it to required tolerances on literally miles of wire. When the end product consists of stainless steel or nickel alloy, the savings could multiply, there is no scrap.

You do have to buy dies, says Robert Maley, general manager of Ulbrich Shaped Wire, whose firm produces the wire-formed components. Instead of the round hole for round wire, dies are made to the desired cross section. But more than offsetting that cost are savings in machining, forming, coil changeovers, and whatever tooling might be needed for an alternative process, Maley adds.

"This method always reproduces the exact shape of the die, and, therefore, has excellent dimensional control of the cross section," Maley explains. "Tolerance of up to plus or minus 0.001 inch is achievable."

There is a limitation, Maley warns. "Our company and most other wire mills can't produce (parts of) more than one inch, but this is sufficient for a number of applications," he stresses.

Maley cites an example of a long, slender part that many people can relate to the metal edges of skis and snowboards. The edges have an L-shaped cross section, typically 0.080 inch on one side and 0.120 inch on the other. He notes that a wire mill can create a highly accurate L shape with almost perfectly smooth corners and edges on both sides all in one pass and in lengths that can make up an 800-lb coil.

Blades of all kinds also make good candidates for the wire mill. Circular seals and mechanical rings may have a design detail around their circumference an inset or a raised portion that can represent significant costs if machined into the final shape.

Even parts that have no resemblance to a wire can be produced using this process fasteners, for instance. Hexagon cross sections can make small alloy nuts, half rounds for cotter-pin stock, and full rounds for cold-headed alloy bolts and screws.

"The growing availability of special-alloy wire from the producing mills is a key factor in designing parts for wire-mill production," says Maley. "Series 300 and 400 stainless steel, nickel, and cobalt alloys, titanium and titanium alloys, and even some proprietary alloys are now made in a range of wire diameters. This makes it even easier for engineers who design narrow-width or very slim parts with cross sectional design details to take advantage of the benefits of wire-mill production."

Transition to 3D made easy

Ann Arbor, MI-Making the transition to 3D modeling as easy as possible for engineers is Peter Smith's goal. Smith, the original founder of CADKEY Inc., and Walter Silva, a writer of CAD books and CADKEY user, started Distance Engineering to make 3D engineering software accessible.

To help engineers at Ford transition from 2D wireframe to 3D solid modeling, Smith and Silva developed a course. The program, independent of any CAD system, begins by teaching how to think through a project in 3D.

"We start with, 'How do you begin creating a model?' " says Smith. "We say: 'Start in 2D.' Engineers try to model like an inexperienced person might begin building a house. They hold up a beam in one hand and try to hammer and nail boards together with the other. It can't be done." Instead, build a framework. "Design the outline in 2D in free space, and later add the walls, surfaces, and shapes. This takes the mystery out of the process."

Smith uses a 3-tiered wedding cake as an example.

One can look at that and see that it has 1, 2, or 3 features. If a designer sees it as one feature, ask, "What tools do I have to make it?" Start in 2D. First define an axis. Second, draw the outer contour, moving perpendicular to the axis, vertical for a couple of inches, perpendicular, and vertical again until you meet the top point. Then rotate the profile around the axis.

If someone sees the cake as two features, maybe he will find wheel shapes in the CAD library and join these with a Boolean joint.

A third person will see the same image as three features. She may draw the cake as three extruded circles joined together.

No approach is wrong. One just has to look at an object and ask, "How many features are in it?" Think about what you are making in the simplest, most fundamental manner and find what keystrokes or icons will make those features happen. "No matter what the CAD program, you typically start in 2D," says Smith.

So why transition to 3D in the first place? "2D is nothing more than a bunch of dead lines," he continues. "One may think 2D is easier because he or she is used to it. But with 2D you have to rely on your brain to make the connections between lines and parts. This opens up the opportunity for mistakes." With 3D, nothing is left to the imagination. Everything is visible hidden lines, balance points, shadows, rendering. One can modify and see relationships quickly, easily. Thus, fewer mistakes, and greater productivity, he continues.

For potential users, Smith and Silva developed a quarterly publication on CD-ROM called CADInsight. Here animated parrots and genies walk people through sheet metal, molding, helixes, and other applications using CADKEY software. For $495, users can get an educational copy of CADKEY, a one-year subscription to CADInsight, and a copy of the course Distance Engineering developed for Ford. If interested, visit www.distance-eng.com.

A thunderbolt in blue

By Jude DeMeis, Contributing Editor

Brookline, MA-In the 20 years since Saab introduced its first turbocharged car, the company has embraced the technology to the point where every Saab now sold is a turbo. Recently Saab turned up the "turbo boost" by introducing its most potent car ever, the 9³ Viggen. The name, shared with the Swedish fighter plane, derives from the thunderbolts sparked by the hammer of Thor. The earthbound Viggen is a high-performance, limited-edition (only 400 will be available in the U.S. this year) version of Saab's compact 9³ three-door coupe. Five-door and convertible versions will follow starting this October.

Initially, all Viggens will be painted a metallic medium blue, which highlights a deep front spoiler and aggressively lowered stance for improved aerodynamics, namely lower drag and rear lifting forces. Inside, a unique pair of side-bolstered and cleverly-textured leather seats grips the driver and front passenger during cornering.

Saab developed the suspension dynamics in conjunction with Tom Walkinshaw Racing (TWR, Oxfordshire, England), whose engineers tailored the Viggen for speed with stiffer suspension springs, and shocks tuned to match. In addition, the designers revised the anti-roll bars and steering to make the Viggen turn better and enhance steering feel. For example, stiffer steering rack mounting bushings provide a more direct feel of the road. A lighter front swaybar, plus 25% stiffer rear and 5% stiffer front springs together flatten body roll in turns and reduce understeer. The team also added oversized front brakes within a set of special 17-inch aluminum wheels.

Report card. How well did they do? Immediately underway from our Museum of Transportation starting point, I discover that the turbocharged engine is the soul of the Viggen. Full-throttle acceleration in third gear is tremendous. After a sprint up to speed, it's obvious why this car has a fighter-aircraft namesake thrust from the wide maximum torque band from 2,500 to 4,000 rpm. The highly refined powerplant is strong, yet smooth and responsive. Saab's combination of 32-bit Trionic engine management, direct cylinder ignition, new turbocharger, and electronic throttle control produce very little turbo-boost lag, excellent throttle response, low-end torque for pulling in any gear, and smooth, even acceleration, for a top-notch combination among the best I have ever driven. Because the engine is small, starting off from a complete standstill is dull unless you slip the clutch slightly. At highway speed, however, few passenger cars can muster the kind of blistering acceleration that the Viggen achieves.

Turning on some twisty back roads, the Viggen is well composed and easy to drive at a fast clip. The TWR suspension is taut and controlled but never jarring. While not luxury car plush, the ride is comfortable enough for long-distance driving. With 60% of the weight on its front wheels, the Saab has less than ideal 50/50 balance. Although the car is more responsive and balanced than a standard 9³, like any nose-heavy car, the front wheels still push (loose grip) through sharp corners if I use too much steering angle.

The large brakes are reassuringly strong and do not fade under hard use. I do notice that the pedal must be depressed slightly before any braking effect occurs. Eventually this quirk becomes transparent as I grow accustomed to the slight dead spot.

Low-profile Dunlop tires afford adequate grip for fast constant-speed cornering. Unfortunately the powerful engine quickly overwhelms the rubber if too much power is applied in a turn or on sandy pavement. The car has no traction control system but fights wheel spin by limiting torque in the first two gears.

While special-edition performance coupes like the BMW M3 and the Acura Integra Type R have undeniably better handling and traction, the Viggen is far more practical with its cavernous cargo space. Furthermore, the Viggen gives up nothing to the competition in terms of straight-line acceleration because of its turbo engine. And how's this for a bonus all buyers are automatically enrolled in the "Viggen Flight Academy," a complimentary two-day driving school at the Road Atlanta race track complex. With that said, expect this rare and fun Swedish hatchback to sell quickly.

When not working as a software engineer, Jude DeMeis is a performance-car technician and auto cross driver.

Polymer membrane protects glucose sensor

By Dennis Normile, Japan

Tokyo, Japan-Home-based monitoring of urine glucose levels, a potentially valuable procedure for those suffering from diabetes and hypoglycemia, has not been practical because of the lack of a reusable, inexpensive, and accurate test device. NEC Corp. claims to have surmounted technical problems with a thin-film single chip semiconductor glucose sensor that is protected from contaminants by a new perfluorocarbon polymer membrane.

The sensor is made up of platinum electrodes formed on a quartz substrate and coated with a number of membranes, including glucose oxidase enzyme. When placed in a urine sample, the glucose in the urine is converted to hydrogen peroxide by the glucose oxidase enzyme (GOE). The hydrogen peroxide then decomposes on the platinum electrode at an applied voltage of 700mV, leading to an electric current. The current produced in this reaction is proportional to the glucose concentration in the urine.

The principle of the sensor reaction has been well known. But previous sensors were quickly degraded by the buildup of contaminants that affected sensitivity. Narushi Ito, who heads the NEC development team, says a new perfluorocarbon polymer developed by NEC has proven to have the right porosity to allow the glucose to react with the enzyme accurately measuring glucose concentrations from 0 to 3,000 milligrams per deciliter. Yet the membrane effectively blocks contaminants so the sensor has a life of more than 100 tests or roughly 2 months. The device is suitable for mass production, with the sensor elements formed using conventional semiconductor processes and the membranes laid on top by spin-coating.

NEC expects to manufacture the sensor chip at its own semiconductor fabrication plants and to source most of the electronics components from within the company. Production of the perfluorocarbon polymer will be commissioned to one or more chemical makers which NEC declines to identify. NEC expects to put the sensor on the market in Japan at a price below $250 within 1999, pending approval by the Ministry of Health and Welfare. It will then consider taking the product to other markets.