Design News is part of the Informa Markets Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.


Articles from 1995 In August

Engineering Megatrends

Engineering Megatrends

What a difference a generation makes!

Twenty years ago, you shaped your designs on a drafting board, probably using a calculator or slide rule. You had to build prototype after prototype to verify your designs. You balked at sharing data with others inside or outside the company. Standard materials were the norm, and microprocessors were still a novelty. Few engineers cared about designing for export, what with America's huge captive market.

It's a completely different world today. First, the engineering toolbox brims over with computer and software aids to make you more productive. You can get virtually any material you want, made to order. And the emergence of microcontrollers has helped lower part counts and costs, while increasing the reliability and functionality of all kinds of products.

And those are just the tangible changes. The way you work today is different too. Globalization demands that you design with foreign customers in mind. Pressure to get to market fast is fierce, and one of the drivers of design for manufacturing. Environmentalism and the quest for quality have led to designs incorporating recyclability and ease of maintenance. Demands for efficiency and cost control have brought about outsourcing and technical alliances.

The changes are mind-boggling, says Victor Poirier, president of Thermocardio Systems and a former Design News Engineer of the Year. "We have to consider things we never thought about 25 years ago."

Adds Norman Augustine, president of Lockheed Martin Corp.: There has been "an explosion of technical knowledge engineers must deal with." And, he says, "they must increasingly bear the burden of selling the projects they wish to undertake."

This issue of Design News identifies 10 megatrends that are dramatically changing engineering. Let us know how they are affecting your job.


Tasks that used to take days now take hours

To fully grasp the staggering advances in computer performance, it helps to make comparisons. One oft-cited example: If auto price/performance had made similar leaps in the past few decades, a Rolls-Royce today would cost $2.75 and run 3 million miles per gallon.

What has all this affordable computer power done for users?

Physical prototypes are being replaced by simulations. Paper drawings are giving way to 3-D solids models. And, high-performance workstations now allow engineers to perform sophisticated analyses on their desktops--and easily share information with colleagues.

Computers were a key part of Boeing's 777 design. The 777 passenger jet is the world's first airplane to be 100% digitally designed and preassembled on computer. "Designers could see parts as solid images and then simulate the assembly of those parts on the screen--easily correcting misalignments and other fit or interference problems," the company explains. Thanks to a powerful computer network, 238 different design/build teams could work concurrently, so problems could be found early.

The results: Engineering changes, errors, and rework dropped to less than half of other programs; while parts and systems fit together better than expected. In fact, the first 777 was just 0.023 inches away from perfect alignment--about the width of a playing card, the company says, while most airplines line up to about half an inch. "Digital preassembly helped us significantly improve our engineering and manufacturing processes, and overall quality," says Charlie Kyle, Boeing chief project engineer, Airplane Integration.

In its constant push to cut time to market, General Motors has also used high-powered computer technology to "redesign the entire design process," according to Larry Howell, director of body systems and research at GM's R&D Center, Warren, MI. GM is using its computer systems and networks to help with concurrent engineering: allowing engineers from different departments, and even outside suppliers, to share information--and fix problems--early on.

The kindest cut. GM has cut its auto-design time from 5-6 years at the beginning of the decade, down to "four, perhaps three," he says; and a two-year cycle may be possible.

Using faster computer simulation tools has helped. Now, for example, automakers can simulate a car crash in considerable detail, to test the safety of a proposed design, with a few hours of supercomputing time. "Five years ago, we couldn't even do that," he notes.

At this year's North American International Auto Show, GM revealed another significant new tool: a powerful proprietary computer program that allows engineers to easily simulate the complex surface of a complete auto body, not just one part. "In a recent benchmark, a commercial CAD supplier took 39 hours to do what SurfSeg could accomplish in 90 seconds," says Paul Besl, staff research scientist at GM's Research & Development center.

Besides cutting time, this helps improve quality, Howell says, by testing for fit and finish very early in the process. SurfSeg was used on several 1995 GM models, including the award-winning Chevy Blazer.

How much faster are computers today? Processor power has been increasing about 1.7-fold per year; and the accumulating advances offer staggering improvements. Jim Glass, an engineer at Rockwell International's Rocketdyne Division, recalls an early assignment to develop some "engine-balance" software. "I can remember when it used to run overnight to generate one case on a CDC mainframe," he says. "Now, it runs in minutes or an hour (maximum) on a workstation."

That means engineers can run hundreds of iterations to explore their designs. "In the old days, we didn't have this luxury," he notes. In addition, Glass says, all this analysis helps an engineer "begin to internalize a 'feel' for rocket-engine systems; that is, the codes teach the humans how the engines behave. One's intuition becomes sharper; we come up with better design concepts earlier in the cycle."

Design implications:

  • Share design data with colleagues internally and at supplier locations.

  • Evaluate more design options than ever before.

  • Sharpen knowledge of product behavior.

-Sharon Machlis, Senior Editor


Standards, quality, customer preferences the key

In today's world, companies scramble to sell their goods in international markets. Likewise, engineers are designing and redesigning to meet foreign standards and tastes.

All too often, companies find that if you don't design for export, you can't sell at all. The 12 nations of the European Union, for example, increasingly ban a growing variety of products that don't conform to their formal standards.

Of broader impact is the worldwide popularity of global standards--especially the ISO 9000 series for quality management. Many engineers now must search for component suppliers that are "ISO-certified." Comments IBM's Lawrence L. Wills, who chairs the American National Standards Institute: "International standards are becoming national standards."

What's the trickiest part of designing for export? Fulfilling the special needs of customers in other lands, manufacturers say. It often means working closely with foreign engineers.

Kodak's case. Eastman Kodak Co., Rochester, NY, is among many giant firms that have embraced design-for-export as company policy. "We aim to assure that our designs meet worldwide criteria," Michael Tennity, manager of Kodak's Design Resource Center, told Design News. "We are finding that the Japanese and European markets have higher expectations than what you would find if you were designing for a U.S. market."

To illustrate, Tennity points to how Kodak units responded to management's decision in late 1993 to widen markets for its one-time-use cameras. The firm sent teams of researchers to Japan, France, the United Kingdom, and parts of the U.S. They interviewed consumers, retailers, and photofinishers to find out what they wanted in next-generation, single-use cameras--including looks, performance, and price.

Steve Chapman, Kodak's lead designer on the program, describes the results of the surveys abroad: "They said our 'Fun' cameras of that time were like big, bulky bricks. They had sharp corners, and didn't fit well in pocket or hand. Some said the camera felt like a lower-end product, and they were kind of embarrassed to be seen with it."

Size was the main concern among the Japanese, Chapman continues. They wanted a thinner, lighter camera that would grip well, without fingers looping over the flash. And they wanted a flash that took less time to recharge.

"They basically didn't want a cardboard box," Chapman recalls. "Foreigners seem more aware of how things are put together. We in the U.S. are more price driven. The American consumer will give up looks or feel or ergonomics for lower prices."

A major concern of Germans: the recyclability of the cameras. They wanted a camera designed so a larger portion of its material could be reused after photofinishers returned the emptied cameras to Kodak.

Multiple choice. Kodak engineers came up with 30 different camera concepts. Armed with block models of the versions, survey teams went back to their interviewees around the globe. They repeated the process at various stages as designers narrowed the candidates to four.

In testing and verifying the models, Chapman's design team worked closely with foreign counterparts. "I got to know them very well over the phone and face-to-face," Chapman relates.

Kodak wanted to make sure the final version of the new camera conformed with all standards in prospective world markets. Among proposals rejected for being nonstandard: use of a smaller film cartridge. Instead, Kodak chose to install 35-mm, 400-speed film. That allows photofinishers in different countries to process more conveniently. As an added benefit to photofinishers, engineers designed a cover that opens easily.

Once they chose the film cartridge, the designers shaped the camera to contour around the film. They smoothed corners to make the device smaller and easier to hold and use. And they surrounded the lens with a raised plastic ellipse to protect the lens and prevent the user's hand from blocking the picture.

Also, the engineers installed a flash unit that automatically recharges for the next shot. It remains in a ready state throughout the picture-taking event. Kodak replaced the cardboard body with a sleek, black cover made of lightweight polystyrene. It designed virtually every major component, including circuit board and polystrene label, to be reused. Recyclers can grind up the other parts.

From conception to shipping, the new product--the Fun Saver--took Kodak just over a year.

Design implications:

  • Consider tastes of foreign customers.

  • Choose suppliers that comply with international standards.

-Walter S.Wingo, Washington Editor


Using DFM, General Motors reduces part counts in bumpers and doors

When General Motors engineers began designing the doors for their 1995 J-car, few thought they could improve on the existing configuration. After all, GM's tried-and-true door designs had changed little over the past decade.

After applying Design For Manufacture techniques, however, they learned otherwise. Engineers at GM's Lansing automotive division reduced the number of parts in the door by 50%. More important, they eliminated the need for several stamping dies, replacing them with lower cost rolled steel tubing. In all, they cut the unit cost for the 1995 Chevrolet Cavalier and the Pontiac Sunfire by 13%.

"Going in, I was convinced that they couldn't improve the doors," notes Joe Joseph, manager of General Motors Knowledge Center and a former door designer for Fisher Body. "But by applying creative techniques, they succeeded."

The J-car's success was by no means an isolated one. During the past five model years, GM engineers have applied Design For Manufacture (DFM) to scores of automotive systems, wringing substantial costs from many vehicles. In the 1992 Cadillac Seville, for example, they found new ways to attach the bumper fascias, thus eliminating 61 fasteners and retainers from the front bumper fascia, saving $4 per car on labor and $51 per car on materials. Similarly, engineers for the 1992 Caprice saved $17 and 3.3 kg per car by moving the ABS module from the trunk to a location beneath the instrument panel.

Such manufacturability wasn't always the norm for GM, or any other major American manufacturer, for that matter. In decades past, design engineers and manufacturing engineers often failed to communicate, resulting in the well-known "throw it over the wall" mentality. As a result, many products were maddeningly difficult to build. Many others went through numerous, time-consuming design changes before they reached manufacturing.

The early 1980s, however, changed the approach of many heavy manufacturers, particularly in the automotive field. At that time, droves of American carbuyers were opting for foreign designs. "We knew we had to improve the productivity of engineering, design, and manufacturing efforts," Joseph recalls.

Emphasis on assembly. Many companies achieved that by shifting the emphasis to the assembly level, rather than the component level. "If you look at the collection of items that make up your product, and then force yourself to ask critical questions, you find that you can actually reduce the number of individual parts," notes Peter Dewhurst, a partner in Boothroyd Dewhurst Inc., a Wakefield, Rhode Island-based consultant in DFMA (Design For Manufacture and Assembly). "And if you do that during the early stages of the design, it can result in very large savings."

Dewhurst says the key to DFMA success is the formation of teams that view the product holistically. "The important thing is to gather all the people who will be responsible for the product at the beginning, so that the reliability, service, and maintenance issues are considered at the same time as design performance," Dewhurst says.

The results can be startling, as a recent Boothroyd Dewhurst study showed. The study documented the following average successes across several industries as reported in DFMA forums, conferences, newsletters, and internal papers: 56% reductions in parts counts; 62% reductions in assembly times; 45% reductions in assembly costs; and 72% reductions in the number of fasteners.

GM's multi-disciplinary teams typically include materials engineers, product designers, product engineers, and manufacturing engineers, as well as participants from purchasing, financial, and supplier companies.

Ultimately, DFM efforts could spell the difference between success and failure for many products. As the trend gains momentum, most large companies now have DFM programs in place, giving them an edge over competitors that don't. Boothroyd Dewhurst, for example, has counseled such giants as GM, Ford, AT&T, 3M, Lockheed, McDonnell Douglas, Hewlett-Packard, and scores of others.

"The industry leaders are reaching a point where they are all doing it," Dewhurst says. "If they don't have DFMA in some form, their product development efforts simply aren't competitive."

Design implications:

  • Cut costs by cutting unnecessary parts.

  • Form interdisciplinary teams to consider all aspects of product design.

  • Make even complex designs easy to manufacture.

-Charles J. Murray, Senior Regional Editor


Design engineers take on new roles

It's a straightforward concept. If you manage a big OEM, push your suppliers to do more design work, and relieve your company of the need to perform tasks outside its core competencies. Outsourcing has become commonplace in these days of belt-tightening and corporate re-engineering. It has, of course, created new opportunities for suppliers. Engineering consulting groups have also seen increases in billings, and changes in their relationships with clients, attributable to outsourcing.

The Brian J. Lewis Company, Castle Rock, CO, is one such firm. After a study of published data on consulting firms, principal Brian Lewis says that in 1974 only three U.S. engineering consulting firms had billings greater than $50 million ($150 million in today's dollars). In 1994, he says, more than 50 consulting firms exceeded $150 million in billings. The most recent data from the U.S. Bureau of Labor Statistics confirm the trend toward bigness:

  • Between 1987 and 1992 the number of engineering services firms with annual receipts exceeding $10 million grew from 515 to 886.

  • Total employment in the field, at firms operated for the entire year, grew from 544,686 to 644,110.

  • The total number of companies in the field operating for a full year rose from 31,472 to 35,893, or 14%.

Even some software companies are taking on outsourced work. For example, Algor, Inc., Pittsburgh, PA, has established a service called Speed Mesh(TM) to undertake the time-consuming creation of eight-node hexahedral finite element meshes from manufacturers' CAD solid models. Algor uses its Houdini software to do the work. This activity is a case, says Algor President Michael L. Bussler, where manufacturers are outsourcing to gain speed as well as offload work.

Driving to outsource. The automotive industry has probably committed itself to outsourcing more completely than other industries. And Milwaukee-based A.O. Smith, the self-declared largest supplier in North America of suspension and structural modules is poised to capitalize on the trend. (A module consists of a system of integrated components and assemblies that can be installed as a unit.)

In fact, Charles Chapman, senior vice president of sales and marketing at A.O. Smith, declares that his company intends to lead the way in outsourcing. "OEMs want the major suppliers to do more of the design work for them," says Chapman. But meeting that demand calls for changes in suppliers' approaches to OEMs.

In 1992, A.O. Smith opened a facility dedicated to supplying front- and rear-suspension modules for Chrysler's popular midsize LH car line. Some 33 vendors supply as many as 70 components for the modules, and A.O. Smith coordinates their efforts. The program is the largest Chrysler module program managed by an outside supplier.

"Yet for years we just did the assembly portion of module production," Chapman says. "With Chrysler, we're well into Phase Two, the phase where we administer all the efforts of the Tier 2 suppliers. And now we're just poking into Phase 3, which we call systems integration engineering."

At A.O. Smith, plans exist for a Tierless Engineering Co-Location (TEC) Center, where engineers from suppliers and OEMs will sit shoulder to shoulder. A facilitator from A.O. Smith will help the engineers, all employees of other companies, integrate components to optimize the design of each OEM's module. "By looking at the module as a total system instead of a bunch of stand-alone components, we can take cost and weight out," says Chapman. "And we're beginning to get the OEMs interested."

What does outsourcing mean to design engineers? Some engineers will find themselves functioning as coordinators. Others will receive design data from colleagues located at a customer's site. Operations not considered part of the company's core areas of competence--like FEA mesh construction--will be farmed out. But, no matter how much of the job gets outsourced, someone will still have to meet the project's deadline and handle recalls. And more likely than not that someone will be the design engineer.

Design implications:

  • Work for or with large consulting firms.

  • Consider working on-site at a customer's facility.

-Brian J. Hogan, Managing Editor


First step is engineering the design process

The pursuit of quality has become a sprint since the 1980s, when changing consumer demands and Japanese competition in the auto industry drove it into the American consciousness. Of course, closer inspection shows that many successful firms had incorporated quality aims in their business philosophy long before it was fashionable to do so.

The advent of the Malcolm Baldrige National Quality Award gave us shining examples of quality in action; yet the methods behind achieving high quality are still elusive. Now, standards such as QS9000--ISO 9000 standards with additional automotive-industry requirements--are expanding the impact of quality on competition and design.

Cutting warranty costs. "The automotive industry is pushing for absolutely transparent product," observes Jonathan Slass, general manager at Rotor Clip, Somerset, NJ. "They want to do anything they can to reduce warranties--even if it means paying a bit more for the product."

Rotor Clip, a producer of spring clips, is no stranger to automotive quality awards. Their accolades include top awards from GM, Chrysler, and Ford, a Saturn Supplier Recognition Award, and quality awards from Allied Signal, Borg-Warner Automotive, and Warner Electric.

Rotor Clip uses Pareto tools to identify the sources of quality problems and prioritize them as part of the corrective process. The technique allows engineers to chart, and therefore visualize, the specific causes of defects--then determine the best way to eliminate them.

Another veteran of engineering's quality crusade is Cherry Electric, Waukegan, IL. Cherry supplies electronic and electrical products and assemblies to automakers, and has been decorated with such awards as General Motors' "Worldwide Supplier of the Year," Ford's "Q1," and a World Class Manufacturing Award from the Advanced Manufacturing Systems Exposition.

How has Cherry used quality to distinguish its products in the intensely competitive automotive industry? "Quality has to be considered right up front, during the conceptual stages of the design process--typically before there are even drawings," says Ken Kunin, manager of electro-mechanical engineering at Cherry's Automotive Group. A good design idea is not enough, he adds, noting that a well-refined design concept takes the manufacturing process and repeatability into account.

Emphasis on process. As they gear up for QS9000 certification, Cherry engineers are paying close attention to the processes that surround design.

"The engineer establishes a concept for a product, and in general terms, he really is also establishing the manufacturing plans for that product," Kunin says. "Design can drive process, just as process can drive design, and one of the keys to high quality is to let the process drive the design; Understand what processes are going to best be repeatable."

For example, when designing ganged high-current switch assemblies, Cherry engineers examined methods for interconnecting the individual switches to a group carrier. They weighed hand- and wave-soldering, mechanical twist-locks, and toy-tab features based on historical parts-per-million reject rates.

Engineers at Cherry also use formal evaluation techniques such as FMEAs (failure mode effects analyses) and software-based, worst-case design analyses.

Cherry recently combined quality-improvement techniques in the early stages of development of a product supplied to GM. Within twelve weeks, engineers brought the defective-parts-per-million count down from 70,000 to 124.

With results like those, today's automotive quality standards may well be tomorrow's general manufacturing standards. In the next months, Cherry and other auto suppliers will participate in audits for QS9000. "It's a logical extension of ISO 9000," says Kunin. "And I'm pretty confident we'll receive it the first time through."

Design implications:

  • Focus on core technology and improve it incrementally.

  • Integrate production, manufacturing, automation, and testing considerations into the design process.

  • Test at every step.

  • Set guidelines that allow design freedom

  • Review the process every six months.

-Andrea L. Baker, Associate Editor


Microcontrollers give machines brains-and more

Microcontrollers are everywhere. They're in your telephone, answering machine, VCR, TV, power drill, stereo, thermostat, security system, utility meters, camera, microwave, remote controls, pager, exercise equipment, coffee maker, and even your electric toothbrush. Your car alone has dozens of microcontrollers--you could call it a computer on wheels.

Motorola--the worldwide leader in the microcontroller market with a 19% market share--recently sold its one billionth 8-bit microcontroller. And there's no end in sight.

In fact, for every microprocessor sold in a desktop computer, four processors or controllers go into embedded systems, says Tom Franz, general manager of the Intel Embedded Processor Operation.

One reason for the proliferation of microcontrollers is the electronics explosion, says Motorola's Gary Daniels, senior vice president and general manager of the Microcontroller Technologies Group. "We're all using more electronic devices in the home, office, and cars and there's a huge market waiting in the developing countries."

In addition to the increase in the electronics market, many typically mechanical or electromechanical systems are starting to use microcontrollers, says Mark Throndson, strategic marketing manager for embedded controllers at National Semiconductor, Santa Clara, CA. "A microcontroller makes products easier to use by making the user interface friendlier and simpler," he notes.

"Microcontroller applications really started to take off about 10 years ago when semiconductor makers started designing chips for applications other than computers," says Dataquest Principal Analyst Tom Starnes. Microcontroller companies are now making chips for specialized applications to fit the needs of OEMs.

Advantages abound. Many designers are surprised to find that adding a microcontroller to a design can actually lower cost and complexity and speed time to market. Other advantages include: increased reliability and lower parts count. One microcontroller can replace many discrete logic devices, timers, gates, and electromechanical switches and relays. Microcontrollers also offer programmability: You can make changes to software instead of ripping out logic circuits. And there are specialized chips for specific applications. The chip designer has already done some of your work for you.

Recent examples of traditionally mechanical designs that have benefited from microcontrollers include:

  • A scooter controller from Curtis PMC, Dublin, CA. The company was pleasantly surprised that as current levels and complexity increased, microcontroller designs became less expensive. In the high end of their product line, the company went from 350 parts to 250 parts. Using Motorola's 68HC11 decreased the cost of the product, gave the controller more features, and let the company meet new European safety regulations.

  • A bulldozer design that John Deere engineers are developing uses a microcontroller to implement a fuzzy-logic control system. This technology uses an experienced operator's knowledge as well as sensor inputs from the equipment to help the bulldozer do a better job--even under the control of a mediocre operator.

  • Automated meter reading. For 100 years, home power meters have been mechanical devices. Several companies have added microcontrollers, which can measure power electronically, drive the digital display, store the data, and transfer the data to the utility via modem, over the power line, or using radio signals.

After embedding a 4-bit microcontroller in its QuadPacer Sonicare electric toothbrush, Optiva Corp., Bellevue, WA, is evaluating an 8-bit National Semiconductor chip for the next version.

A previous version used discrete ICs, and engineers were looking for a way to reduce the design's cost. "I couldn't believe a microcontroller would let me do that--and add lots of features." says Electrical Engineer Ryan McMahon. He designed the circuitry and wrote the specs for the controller's software. A programming consultant wrote the actual code.

"The cool thing about this new design," adds McMahon, "is that it let us reduce the total number of parts--passives and ICs--from 32 to 15. We had five discrete logic ICs and a timer, and we replaced them with the 4-bit micro." He also notes that reducing parts count means the company isn't as subject to parts going out of allocation, which had caused problems in the past.

"Microcontrollers give you what you need to control almost anything," explains Daniels. "They contain the brains, or CPU; some scratch-pad memory, or RAM; program instructions in ROM; timers; and input and output ports to communicate with other components. The beauty of it is that for a company that builds lots of one thing, they can use one controller board and change the software."

Design implications:

  • Lower parts counts and increase reliability with microcontrollers

  • Use microcontrollers to make products easier to use.

-Julie Anne Schofield, Associate Editor


Software gives engineers a leg up in race to market

Since the first caveman decided to capitalize on his best idea for a new club, businesses have operated on the principle that the first to get to market owns the market--at least for awhile. The theory still holds. With increased competition from all corners of the globe, and the nearly universal consumer fascination with having the latest, most innovative products, cutting time to market is now a critical element of competitive advantage.

"Marketing and finance drive the trend," says Chris Stergiou, president of engineering firm Global Design and Procurement, North Andover, MA. And they are driving hard. Indeed, 64% of the respondents to the most recent Design News/Simmons Market Research survey of the design engineering universe said shortening design cycles was one of the biggest engineering challenges they face.

How they meet that challenge can change their companies' future. Speed in new product development ranks highest among eight organizational skills in its effect on both market share and growth of market share, according to a recent survey of nearly 600 U.S., European, and Japanese manufacturing companies by the Boston Consulting Group and Cambridge, MA-based Product Development, Inc. (PDC). Quality ranked second.

The strategies companies employ to win the race to market include better understanding of customer needs, simultaneous engineering, and early prototyping, says PDC's Shelia Mellow. Adds engineer Michael Black, of Seattle-based Stratos Product Development, "software is one of the tools that make shorter development cycles possible by helping us do faster iterations." Examples abound:

  • With the aid of SDRC's I-DEAS CAD software, Stratos helped Fluke Corp. take two new multimeter designs from concept to molds in nine months rather than the normal year to year-and-a-half for the process.

  • Engineers at Artco, Inc., Thief River Falls, MN, used EDS' Unigraphics to cut development time for its watercraft hulls and decks from eight weeks to four weeks. Chief Engineer Charlie McCarty expects to eventually get the time down to four days.

  • And, Navistar International recently used Algor's Houdini software to cut the time for creating the finite element mesh of a crank shaft. Creating the mesh is among the most time-consuming tasks in FEA, which itself is a necessary-but-time-eating development step. "This was our first use of Houdini," says engineer Anton Calash, "and we still created the 13,000-element mesh in less than three days rather than the normal one to two weeks."

Blending technologies. Most often, time-to-market success depends on a multi-faceted, cross-departmental effort. A classic example is Waring Products, New Haven, CT. The company combined corporate re-engineering with extensive use of software to cut design time 40% for its blenders and kitchen equipment.

Such improvements are critical in the fiercely competitive consumer-goods industry, says Robin Ruck, Waring's director of engineering. "You need to be there first to capture the market before competitive prices eliminate profits." Waring achieved that with its new Pro-Line drink mixer, taking the product from design to launch in just three months. That rapid turn-around required Waring to tear down walls separating departments, employ concurrent-engineering principles, and experiment with its CAD/CAM software.

Waring evolved the design by simultaneously creating hand-crafted urethane foam and wooden models, and capturing the geometry in Pro/ENGINEER mechanical design software from Parametric Technology Corp. Systems integrator Rand Technologies laser-scanned product geometry from the industrial design model into Pro/ENGINEER. They held important areas of the stand to one-sixteenth of an inch to ensure that all design subtleties were recorded.

Ruck says the scanning saved time by avoiding multiple iterations between machining and solid modeling. Engineers used Pro/MANUFACTURING to machine for casting patterns. At the same time, they developed stereolithography models from the database as a master pattern in preparation for urethane casting of the head. 3-D Systems, Valencia, CA, provided the stereolithography machine.

Ruck says that because industrial designers worked on the patent of the stand at the same time design engineers created the Pro/ENGINEER solid models of the entire product, everything was ready for limited as well as full production in the same time frames.

Using the software to quickly evaluate several design alternatives helped engineers avoid downstream design problems that could have slowed the product-development process, Ruck says. "With our mold designs, for example, we evaluated more options in the early design stages. As a result, we caught flaws that would have gone to the tool-making stage." Additionally, they avoided shrinkage flaws, shearing problems, and cold flow lines.

"Previously, we would spend three or four months just getting the tooling right," Ruck notes. "We've dramatically cut that time, and we know our tool will be right the first time."

Design implications:

  • Cut your design cycle--but don't cut quality.

  • Use CAD and FEA to automate time-consuming engineering design tasks.

  • Perform different design tasks simultaneously to save time and catch problems early.

-Paul E. Teague, Executive Editor


Choices snowball as suppliers fight for market share and end-use needs become more specialized

Special orders don't upset us. That line, made famous by a hamburger chain, applies increasingly to the materials available to design engineers.

The fact is, there have been no major new inventions in metals, ceramics, rubber, or plastics in recent years. On the other hand, there has been an almost unbelievable proliferation of types and grades, all offering some incremental advantage in either the final-use application or processing.

The result? Opportunity--or chaos for the unsuspecting.

"There are 21,000 off-the-shelf plastic materials in the U.S.," reveals Mike Kmetz, president of IDES, Inc., a Laramie, WY, database firm. "That's up from 11,000-12,000 just five years ago."

The trend toward tailoring also prevails in other materials groups. When Advanced Refractory Technologies, Buffalo, NY, went to market recently with a series of advanced ceramics, emphasis was placed on their tailorability. The A500 series not only offers outstanding thermal conductivity and corrosion resistance, but the materials can be specifically sized, treated to be water-resistant, and spray dried. "As we talked with the customers, it was a clear case of one size does not fit all," comments Mary Spohn, director of marketing.

Few steel-industry companies escaped the well-publicized round of R&D budget cutting in the 1980s. Still, ferrous specialists are targeting growth through customization. One such leader is Carpenter Technology Corp., Reading, PA. New efforts are underway at CarTech to develop special alloys for all-fuels compatibility in cars and/or for specialized medical requirements, such as easily machinable, very-tough hip implants.

Innovative new product. CCM-plus, a cobalt chrome molybdenum alloy made in bar form through powder metallurgy, solved a specific customer's request, says Ron Gower, manager of stainless and high temperature alloy R&D at CarTech.

Several materials experts reveal that the most intense efforts in custom formulation are under way in the medical market. Identifying the correct material for a specific medical application is a ten-step process, explains Tom Huitema, senior engineer at Ethicon Endo-Surgery, Cincinnati, a Johnoson & Johnson company. He recently completed a materials change for a next-generation stapler used for intestinal surgery. The device features a flexible shaft, articulating stapler head, and trigger-operated handle.

In the original design, a pair of laser-cut, stainless-steel support plates acted as structural members and provided support and guidance to other components. These plates were fastened and separated by six double-ended shoulder rivets. The riveted assembly was strong, but 20% heavier than the design objective. In addition, the steel plates were costly, assembly was difficult, and the design did not support disassembly for salvage.

Before beginning the process of selecting from alloy steel, aluminum, or hundreds of possible plastics, the engineering team at Johnson & Johnson determined critical performance.

Then, the engineers reviewed opportunities to improve the design through use of alternative materials. A switch to injection molding would allow integration of spacers and bosses into a plate and the incorporation of several assembly aids.

Engineers then analyzed the initial computer concept model, generated rapid prototypes with stereolithography, and cast urethane prototypes for limited structural testing. Example: Making certain that bosses would not be crushed or split when barbed pins were inserted during manufacturing.

Ethicon contacted a plastics materials supplier, molder, and aluminum die caster to help identify specific materials candidates. Three manufacturing-related issues emerged with the die caster. Secondary machining would be required of the die-cast parts to achieve close tolerances on holes, for example.

Initial top choices, based on very high stiffness requirements, included die-cast aluminum and carbon-filled nylon. Preliminary testing on the urethane prototypes indicated that other filled plastics could do the job at a lower cost. Engineers reviewed two other plastics: 30% glass-filled liquid crystal polymer (LCP) and 40% long-glass-filled engineering thermoplastic urethane (ETU). The 40% ETU fell short on stiffness, so Ethicon looked at an even more exotic grade, 60% long-glass-fiber ETU, which offers flexural strength close to carbon-filled nylon.

Ethicon developed a matrix to compare all of the final materials candidates against each other and the minimum and desired performance requirements. Five materials, including four filled high-end engineering plastics, were compared. Two were eliminated: aluminum, due to weight, cost, poor tribological properties, and lower flexural strength than the plastics; and nylon, due to cost and past experience with shrinkage and warpage.

Ethicon picked the 60% ETU material (Isoplast from Dow Plastics) because of its toughness, stiffness, strength, dimensional stability, and cost. One caveat Huitema offers design engineers: Published properties for plastics and other materials are based on tests conducted using standard test samples. "The mechanical properties of molded parts may vary significantly from the published values," says Huitema. "The material that looks the best on paper may not be the one which performs the best in the part," adds Karen L. Winkler, the Dow medical specialist who worked on the project.

Still, use of large materials databases provides a good starting point when it comes to making a rough cut on materials selection. Most major resin suppliers offer electronic databases of their products, sometimes as part of an industry collection of materials.

It boils down to this. Pushing the envelope on materials' capabilities through a custom formulation or newly developed grade can be a time-consuming affair, but it could give your company a competitive advantage.

Design implications:

  • Specify detailed performance requirements.

  • Don't make assumptions based on data sheets.

  • Accelerate use of rapid prototyping to make certain parts are dimensionally accurate.

-Doug Smock, Contributing Editor


Tomorrow's products to feature ease of maintenance and recyclability

It may be a bumper-sticker truism, but even the highest-quality designs eventually break down, grow obsolete, or get mandated out of existence. Tomorrow's successful engineers must keep entropy in mind: Not only will their products be efficient and manufacturable, they also must be easy to dispose of and to service.

As the range between survivors in various quality surveys narrows, the consumer's choice hinges on how simple a product is to keep up, rather than how often it breaks down.

Whether or not a product's recycling benefits the environment remains a moot point. For designers, recyclability will become an imperative for economic reasons. Consumer environmentalism and government mandates may soon dictate component recycling. But more significantly, improved serviceability and OEM downsizing will lead to smaller service departments, resulting in more widespread equipment leasing.

Future is now. Examples of increased emphasis on recycling and serviceability abound. According to the American Automobile Manufacturers Association, 76% of the average car is already recycled. New SAE standards for plastics marking will raise that figure in the near future.

The proliferation of engines good for 100,000 miles between tune ups underscores the new emphasis on serviceability in the automobile industry. Better transmission fluids and coolants with 10-year life spans will debut in 1996. "We're trying to please the customer by making cars as maintenance-free as possible," says Edward Zeller, Cadillac's chief engineer.

Further, 1996 introduces the second phase of government-mandated On-Board Diagnostics (OBD II). The law requires computerized monitoring of emission-system performance that warns operators when service is required. General Motors, for one, has broadened the scope of OBD II diagnostic capabilities to alert operators of impending mechanical difficulties. Technicians will interrogate the system to pinpoint problems and speed repairs.

Switching fixes. Recyclability and serviceability issues merged in the recent upgrading of power supplies used in digital central-office phone switching equipment used by GTE. As these units aged, their failures created an increasing percentage of equipment down time.

How to handle the problem--recycle or redesign? To answer the question, the switching-system OEM, AG Communications, Inc., Loraine, OH, asked GTE's Electronics Repair Services, Ontario, CA, for help. GTE ERS, a separate business unit of the telecommunications giant, provides depot-level maintenance expertise to a variety of electronics manufacturers. As a result, its personnel has gained significant expertise in end-of-life design issues.

ERS examined a prototype redesign of the power supply, performing a failure analysis, testability analysis, and critique for ease of maintenance. Doug Suda, manager of technical support at ERS, says the analyses resulted in several design changes. Among them: use of de-rated electronic components to improve mean time before failure.

ERS also recommended that the power supply's heat sink include holes to allow access to all nodes beneath it for in-circuit testing and simplified troubleshooting. "I'd estimate we'll save a half-hour or more in diagnosing the boards in the field, since the device won't have to be removed for access," explains Suda. Overall, he estimates that the in-service life of the power supply went from seven years to between 12 and 15 years.

Half the power supplies will be recycled. ERS replaces five components on each board that are the most vulnerable to electrical and thermal stress, including all electrolytic capacitors older than four years. The better-than-new power supplies reduce the company's procurement costs and cut waste at the same time. It's a lesson with an easy to understand moral.

Design implications:

  • Make service life and ease of repair differentiators.

  • Communicate with service personnel.

  • Design products so they can easily be recycled.

-Terrence Lynch, Northeast Technical Editor


To compete, collaborate

At companies throughout the world, the walls are tumbling down. In the '80's, they fell between engineering departments within companies; now they're falling between the companies themselves.

This new cooperation is often called partnering or a technical alliance. At its purest, an alliance is an agreement between two or more firms to pool resources--engineering, marketing, manufacturing, or otherwise--to create a new product that each could not produce easily or at all alone.

What's driving this trend? The fall of trade barriers is one cattle prod. "Global competition is the source of all of this," says Dr. Jordan Lewis, industry analyst and author of The Connected Corporation: How Leading Companies Win Through Customer-Supplier Alliances. "It's not us against them anymore, it's everybody against everybody else."

Another prod: the uncompromising demands of customers for the latest technology delivered in the shortest time. "One way to do this is to share the expense of product development with partners," says Phil Pompa, director of marketing at Motorola's RISC (reduced instruction set computing) processor division in Austin, TX.

To that end, Motorola entered into the AIM alliance in 1991, teaming with engineers from IBM and Apple Computer. The goal: to develop next-generation microprocessors and create a new standard for future computers. Given the seeming dominance of Intel-based PCs at the time, it was quite an ambitious undertaking.

As with any good partnership, each company carried particular strengths into the marriage. Apple brought the Macintosh operating system and its base of PCs. IBM contributed its advanced semiconductor development and fabrication, RISC microprocessor architecture, and array of design tools geared towards rapid IC development. And Motorola delivered its bus and I/O expertise, experience with high-volume production, and knowledge of compact IC design.

"It's increasingly difficult to find all forms of expertise in one firm or design group," says Pompa. "Other partners may have already developed technology, and it doesn't make sense to reinvent it."

In just 12 months--far less than typical in the industry--the engineers at AIM completed the design and fabrication of the PowerPC 601. Containing 2.8 million transistors and measuring 0.4-in square, the chip was said to be both smaller and more powerful than those from Intel. In addition, it would run several operating systems, including Mac OS, SunSoft's Solaris, IBM's AIX, and Microsoft's Windows NT.

Don't call home. To facilitate cooperation and communication, AIM partners built the 80,000-sq-ft Somerset Design Center in Austin, TX. More than 300 engineers from the three companies converged on the site, sporting ID badges not from IBM or Motorola, but from the consortium. "You checked your ego at the door," says Pompa. "Had we set up in an Apple, or Motorola, or IBM building, things would have been very different," Pompa says.

Master documents spelled out security and legal issues as well as means to resolving conflicts. Not once did AIM engineers have to refer to them, a tribute to the close cooperation fostered by co-locating designers at Somerset.

Could the PowerPC have been developed without AIM? Possibly. But Pompa doubts any one partner would have taken the risk. "The project would have cost more, taken longer, and would not have had the market influence of all three companies," he surmises.

Not every technical alliance can afford to build a separate facility like Somerset. Nor does every one require it.

At Enterprise Integration Technology (EIT) in Palo Alto, CA, principal scientist Jay Glicksman concerns himself more with how engineers will collaborate than why. An advocate of the "virtual corporation," he's working on creating computer tools that will allow engineers at remote locations to communicate and share engineering information over the Internet. "It's a way to overcome geographic limitations and team with people that you would not or could not otherwise," he explains.

Several academic examples point to the future of such virtual collaborations. SHARE, an ARPA-sponsored project, funds EIT and Stanford University's Center for Design Research (CDR) to come up with a methodology and environment for collaborative product development. SHARE researchers envision a world in which teams of engineers from multiple organizations work together over networks.

To demonstrate and develop such capabilities, ARPA also funded the MADEFAST project in which researchers from Stanford and the University of Utah modified a missile seeker head into a target tracker for civilian applications such as parking-lot surveillance. Collaboration occurred entirely over the Internet using a variety of computer programs that allowed the sharing of voice, video, CAD, and data right at an engineer's workstation.

The point of all technical alliances--virtual or co-located--is survival. And to survive, designers must think beyond their organization and hone their relationship skills. "Engineers will not make it in the future without those two attributes," says Lewis. "Having good technical skills is no longer adequate. Those who don't cooperate won't be in business."

Design implications:

  • Look beyond your own organization for technical capabilities.

  • Your competitor in one project can be your ally in another.

  • Focus on the goal of the alliance. It will further your corporate goals.

-Mark A. Gottschalk, Western Technical Editor

Metal-hydride system heats catalytic converters

Metal-hydride system heats catalytic converters

Ringwood, NJ--Automobile catalytic converters are heated electrically upon cold start to help bring the catalyst to "light-off" temperature (300 to 350C). Because the catalyst converts unburned hydrocarbons into non-toxic water and carbon dioxide only after it reaches light-off temperature, heating the catalyst reduces pollution during the first few minutes of vehicle operation.

Using electricity to heat the catalyst loads a vehicle's battery and charging system. Also, electrical heating is too slow to meet proposed low-emission vehicle (LEV) and ultra-low-emission vehicle (ULEV) standards, which are 70% and 80% cleaner than 1995 standards.

  • Heat pumps

  • Air conditioners

This new metal-hydride cold start heater (CSH) generates three to four times more heat than electrical heaters, enabling it to bring catalysts to light- off temperature in seconds. Further, the device does not load a car's electrical system. The reversible exothermic/endothermic chemical-reaction system reportedly lowers automobile hydrocarbon pollution as much as 80%, surpassing LEV standards and meeting ULEV standards.

Ergenics Inc. is currently marketing its patent-pending CSH technology to automobile builders. The company plans to design versions of the nominally 5-lb, 12-inch-long CSH system for small displacement engines, as well as for six-cylinder and V-8 engines. Analysis indicates that the technology will cost about $100 to $150 per car, some $150 to $200 less per car than current electrical systems, according to Ergenics.

The metal-hydride CSH works by placing an exothermic reaction in the path of the exhaust gas, ahead of the catalytic converter. Each heater consists of two beds connected by a solenoid valve from Kip Valves, Farmington, CT. One bed, a low-temperature alloy "source" bed, is contained within an aluminum bottle from Parker-Hannifin's Cliff Division. It operates at about atmospheric pressure. The second, a proprietary high-temperature alloy "heater" bed, is contained within stainless steel coils.

When the solenoid valve opens, hydrogen escapes from the source bed's metal hydride (MmNi4.3Al0.7). The heater bed's low-pressure alloy then absorbs the hydrogen. This exothermic reaction raises the heater bed's temperature dramatically, and brings its coils to 400C in less than five seconds.

About 20 seconds after engine startup, the solenoid valve closes. As the engine warms up, its exhaust reaches temperatures ranging from 500 to 650C. That stream of hot gas liberates hydrogen from the heater bed's metal hydride. After separating from the hydride, the hydrogen returns to the source bed through a check valve, and awaits the next startup cycle. Ergenics says the reversible, self-contained CSH consumes just 6W--to operate the solenoid valve. The company estimates a 25,000-start or 100,000-mile lifetime (the latter specification is a federal requirement and applies to all emission-control systems).

Although the CSH is elegantly simple, "we've had to overcome a lot of problems," says Mark Golben, the CSH's principal designer. "We had to develop an alloy impervious to high temperatures, and find a way to rapidly transfer heat from the heater bed's metal hydride to the exhaust gas." Because Ergenics has a patent pending on their hydride production process, Golben won't say how the firm made the high-temperature alloy. But he claims Ergenics' proprietary metal alloy resists the tendency of hydrides to break down into a stable alloy, which would destroy the material's reversible-process potential.

Normally, smaller tubing restricts flow in heat-transfer systems like air conditioner coils. But hydrogen flows far more easily than Freon(TM). So, contrary to conventional thinking, "I realized that there was no penalty associated with going to a smaller size tube," says Golben. By using 24 ft of 1/8-inch-diameter stainless-steel tubing for the heater-bed coil, he made the coil's surface area/volume ratio about four times greater than that of coils that use .50-inch-diameter tubing. Naturally this increase in surface area quadruples the coil heat-transfer rate.

In tests, the CSH delivered 50 kW to a test vehicle's exhaust gas, far surpassing the 3 to 12 kW produced by conventional electrical heating systems. According to Ergenics, testing of the CSH by a respected national labratory conformed to Federal Test Procedures. It demonstrated that using the CSH resulted in non-methane hydrocarbon emissions of between 0.019 and 0.026 gm/mile--which corresponds to ULEV performance.

Because the CSH reaction involves only one gram of hydrogen, and because the hydrogen exists within the system at sub-atmospheric pressure, there is no danger of a hydrogen release. Inevitably, some of the system's hydrogen will migrate through the containment.To offset this unavoidable loss of working fluid, and also to guarantee that proper operation of the device will be ensured throughout its lifetime, Ergenics uses twice the quantity of hydrogen needed by the process.

Additional details...Contact Philip A. Burghart, Ergenics Inc., 247 Margaret King Ave., Ringwood, NJ 07456, (201) 962-4325.

System integrates data for vehicle navigation

System integrates data for vehicle navigation

Middletown, RI--Combining an electronic compass and a GPS receiver with intelligent displays allows engineers to produce a navigation system capable of either dead reckoning or GPS navigation. In addition, the system can display the orientation of an armored vehicle's hull and turret simultaneously.

Called TACNAV, the system's compass is accurate to two degrees. It uses a sensor antenna that contains a toroidal-ring-core fluxgate magnetometer to detect magnetic north. Supported by a heavy helical spring, the antenna is located in a position where the magnetic field will remain as stable as possible. Helmholtz cage compensation helps the toroidal core see a uniform field.

Robert W. B. Kits van Heyningen, vice president of engineering at KVH Industries, Inc. explains that the compass was originally developed for use on racing yachts. Later, KVH adapted it for use on large naval vessels such as aircraft carriers. In the late 1980s, Picatinny Arsenal contacted KVH and asked the company to build a compass capable of functioning on a main battle tank. Success in that project created interest in a more capable system for armored military vehicles, and KVH responded with TACNAV.

Basically, the design challenge required that engineers take data from several different sensors and combine those data streams to produce navigation and directional information. Software provided the key to interfacing the sensors and generating useful outputs. TACNAV, the end product of this effort, consists of five subsystems: a sensor antenna; a driver's control box; a turret angle interface box (called TANGO); a commander's display; and a driver's display. The system can interface with the vehicle's odometer, turret angle encoder, laser rangefinder and GPS to provide interactive navigation data and target location data.

Given that KVH developed TACNAV for the military, ruggedness and reliability became important design issues. "Our entire design philosophy is based on a modular approach," says Ron Paradis, a component and military products engineer at KVH. Each subsystem in TACNAV contains an Intel 8051 eight-bit microprocessor. Internal communications employ messages output in ASCII. Kits van Heyningen explains that its low cost and the ready availability of tools and compilers drove the selection of the 8051.

Each of the intelligent subsystems carries out internal tests to determine its state of health. Placing intelligence in each subsystem makes the system highly modular and easy to service. Rated at 10,000 hours MTBF, the system's mean time to repair is 30 minutes.

Displays for the driver and vehicle commander employ heated and backlighted four-plane LCDs designed for n20-degree readability. Heating the displays enables them to operate in temperatures as low as 10C. Military usage required placing the displays in rugged metal housings, and also made it necessary to do tests with night vision goggles to ensure display visibility.

In an armored vehicle equipped with TACNAV, data from an absolute turret angle encoder and a GPS receiver are received by the TANGO. That information goes directly to the vehicle commander's display, along with compass and odometer data. Laser rangefinder information can also be presented to the commander. TANGO output goes to the driver's display via a pair of processor boards, along with odometer data and directional information derived from the fluxgate magnetometer's output. The processors calculate compass headings, turret position, and other parameters, and deliver that information to other subsystems.

According to Kits van Heyningen and Paradis, interfacing with the GPS system was one of the most difficult parts of the design project. They found it necessary to use a great deal of microprocessor overhead to interrogate the GPS.

  • Emergency vehicles

  • Aiming satellite dishes

  • Intelligent vehicle highway systems

By capturing the output of a laser rangefinder, TACNAV enables an armored vehicle's crew to navigate to a new location without losing the target's location. So they can illuminate a target, then move to a different spot and fire on it, without using the rangefinder a second time. This capability can reduce the danger of being shot at by hostile folks who observe the laser.

Engineers at KVH believe TACNAV is a military system with considerable potential in civil markets. Consider just one potential area of use: Because TACNAV collects data from several sources, concentrates it, and then communicates it in a digital form, it can be transmitted to a central location for analysis. This characteristic might make many types of mass transportation systems more user-friendly. A bus or rail vehicle could be interrogated, and information on its location presented to the public. Thus, although your bus might be late, you would have a fair idea of when you could expect its arrival.

Additional details...Chris Burnett, V.P., Business Development, KVH Industries, Inc., 110 Enterprise Center, Middletown, RI 02842, (401) 847-3347.

Gage design improves liquid-level monitoring

Gage design improves liquid-level monitoring

Mequon, WI--Few things annoy a driver more than a flickering display light on an oil gage. Flickering can mean that the vehicle's oil level has dipped too low, or it might simply indicate a faulty gage. To settle the matter, drivers must manually check the oil level, which defeats the purpose of having a display light.

By using a new design, however, engineers from Kelch Assemblies say they eliminate the primary cause of flickering oil lights. Employed on snowmobiles and various types of agricultural and construction equipment, the new gage is not affected by the splashing and movement of oil or other liquid. As a result, it provides a steady, reliable signal to the dashboard display.

Key to the new gage's performance is a design that provides a built-in hydraulic damping effect. Unlike conventional liquid level gages, the new model allows liquid to enter only through a single port at the bottom of the unit. That represents a departure from the designs used in competing liquid level sensors, which allow liquid to splash over multiple openings. Because the new unit, known as the E-Gauge, only allows liquid ingress through a single hole, changes in terrain do not affect it. As a result, it doesn't flicker on and off every time the vehicle traverses a rolling hill.

The E-Gauge consists of a 1-inch-diameter plastic tube containing a float, a magnet, and electronics. During operation, liquid enters or exits through the bottom port, lifting or dropping the float. A thin plastic stem atop the float holds the magnet. When the liquid level drops, the magnet moves down. When it approaches a reed switch, the magnet closes a circuit, turning on a warning light or buzzer. When fuel or oil are added to the tank being monitored, the float moves back up, breaking the circuit.

Reliability represents another inherent advantage of the new design, say Kelch engineers. Separated from the active area by a plastic wall, the E-Gauge's reed switch and electronics are potted in epoxy. "All of the electronics are isolated from the liquid by the nature of this design," notes Michael J. Holz, applications engineer for Kelch Assemblies.

  • Golf carts

  • Lawn tractors

  • Watercraft

Kelch engineers accomplish that isolation by placing the magnet high in the float stem. In contrast, conventional designs typically incorporate the magnet in the float and, consequently, engineers must also position the unit's sensing electronics there. As a result, conventional designs expose the electronics and conductors to the liquid in the tank.

By employing high-pressure injection-molding techniques, Kelch engineers also endowed the unit with a smoother exterior finish, resulting in less leakage of fluid. In the past, rougher, "orange peel" surfaces sometimes permitted fluid to seep around the tank seal, they say.

For users, however, the E-Gauge's most important feature remains its ability to provide a steady and reliable monitoring signal. "By restricting the way fluid flows into the tube, we've isolated it from the sloshing that normally occurs in a tank," Holz says. "So you don't have the constant on-off flickering of the oil or fuel light."

Additional details...Contact Doreen Lettau, Kelch Assemblies, 11035 No. Industrial Dr., Mequon, WI 53092, (414) 238-6080.

CADKEY 7 for Windows

CADKEY 7 for Windows

In the latest version of CADKEY, now on Windows, I think the company has done a nice job of blending all of its tools into an industry standard user interface. The installation is much more streamlined and simpler than with the DOS version. However, it does give the user less control over configuring locations for directories; and customized mouse button mapping is not offered in the Windows version.

User interface. CADKEY has kept the main components and techniques of their existing user interface intact, but placed them into the Windows world. The history of last commands, conversation line, settings area, and status area are all present.

New for CADKEY users is a menu bar and icon bar, where pull-down menus for many settings and view controls can be found. Overall, the new U.I. takes some getting used to for existing CADKEY users. But all of the things CADKEY users are accustomed to are included in one form or another.

Sacrifices. While CADKEY has migrated their traditional DOS product to Windows in a very easy-to-use fashion, they have had to give up some user interface techniques in the process. Mouse buttons no longer perform the main controls of accept, backup, and escape of commands. These controls are now provided as buttons on the top of the menu. There is also no ability to map other functions with the mouse buttons. However, the Windows version does allow for an Accelerator Key that maps almost any command or function to keyboard keys for quick access to their operation.

3-D functions. CADKEY offers a full set of 3-D wireframe and surface construction tools. It supports multiple viewports, as well as a Layout function for extracting 2-D drawings from the 3-D model. A dynamic part rotation command that was buried in the menu structure in the DOS version is now conveniently accessible with its own icon. Shaded image and hidden line removal are provided in the Picture It utility, which is now one of the menu picks.


CADKEY 7 for Windows

The program needs an Intel 386, 486, or Pentium, and a math co-processor. It runs under Windows 3.1 or higher, Windows NT 3.5 or greater, or DOS 5.0 or higher. 8M bytes of RAM and 30M bytes of free disk space are required.

List Price: $795

CADKEY Inc., 4 Griffin Rd. N., Windsor, CT 06095; ph. (203) 298-8888; fax (203) 298-6590.

Since this is a Windows edition of a previous version, I did not see any new functions. In fact, several extra modules that were provided with the DOS version are not in this one. Fastsurf, which provided advanced surfacing tools, is not present. An Advanced Drafting Module that gave geometric tolerencing abilities and a complete set of pre-drawn symbols for mechanical and electrical design is missing. And QuickSnap, which offered quick visual feedback for snapping to locations and aligning entities when creating geometry, has also been left out.

Sharing data. Support for reading DWG or DXF files is included. Exporting to DWG and DXF formats, along with IGES read/write with level mapping, is also provided. Stereolithography output, conversion of files to CADKEY's Advanced Modeler, and mass property analysis are available through the Picture It module.

I imported several complex solid models from AutoCAD to CADKEY. The models came across intact with converting times of only a few minutes each. The models are converted to 3-D wireframe geometry in CADKEY. Although no longer solids, I was still glad that I could reuse the geometry in CADKEY.

Though CADKEY 7 for Windows has left out several extra items available to the DOS purchaser, it has successfully carried the basic CADKEY functions over to the Windows world. Existing CADKEY users will find the product is very easy to use, since it retains most of the quick-action tools and functions of the DOS interface. The only difference is the use of pull-down menus and buttons verses a text hiearchical menu.

A similar product:

AutoCAD 13 for Windows - Autodesk Inc., 111 McInnis Pkwy., San Rafael, CA 94903; ph.: (800) 964-6432; fax: (415) 507-5100.

Rustbelt becomes the 'Powerbelt'

Rustbelt becomes the 'Powerbelt'

Albert Moore became president of AMT--formerly the National Machine Tool Builders' Association--in 1989. Before assuming his present position with the organization, and while still employed by Gleason Corporation, Moore was active on AMT's Board of Directors. He joined the board in 1982, and worked in a number of different positions, finally serving a term as chairman from 1987 to 1988. Moore's industrial career began with Gleason Corporation, Rochester, NY, in 1959. He rose through the ranks at Gleason, and served in a series of executive positions, finally becoming executive vice president--operations for the corporation. Moore received an undergraduate degree in business administration from Clarkson College, Potsdam, NY, and an M.B.A from the University of Rochester, Rochester, NY.

Manufacturing is critical to the future of the U.S., says Al Moore. And sensible behavior by government and industry can keep manufacturing a prominent part of our industrial landscape.

Design News: What trends do you see emerging in the design of modern manufacturing equipment?

Moore: More computer-control-based machinery is in the offing for the future. It will incorporate a wide array of sensors for process control, and also for maintenance diagnostics. These fast-operating computer controls, and the incorporation of the sensors, will put the final touches on machine design in the future. They will make the Up time on equipment extremely high. Also, we're probably going to see more composites used in machinery design. We as an industry will probably not have enough volume to develop these composites ourselves. It depends on whether this material will be designed by large-scale users, and therefore generate cost reductions that will allow us to employ them.

Q: What future do you see for the "lights-out" factory?

A: It's probably more an interesting philosophical idea than a practical concept. Most of those who have tested that idea, and it's been tested now for 15 or more years, have probably come to the conclusion that there are some drawbacks. Now there are things that can be done in a lights-out mode. The real question is: Will the approach be more widespread in the future, both here and abroad, than currently is the case? While it's a fascinating subject, I'm not so sure that you're going to see more application of the concept than you see today.

Q: How important is software to U.S. manufacturing technology today?

A: Today it's critical, tomorrow it'll be more critical. As we computerize more of the shop floor, software is going to be tremendously important. As you work to drive out inefficiencies in the factory, you're going to have to knock out the inefficiencies already present in the current software. But software is front and center.

Q: How do you respond to the argument that American companies are foolish to invest in countries that will soon give birth to low-cost competitors?

A: You can't ignore the global market. A lot of people who try to sell in foreign markets, for instance aircraft manufacturers, find that they cannot sell unless they're willing to put some manufacturing capability there. They're forced to invest in those countries whether they want to or not. Playing in the international arena, albeit not easy, clearly is the way to go. You find out what your potential competition is doing, and the needs of the different marketplaces. And you develop products that can generally handle the requirements of the whole world.

Q: What future do you see for "smokestack" companies in the U.S.?

A: There's always going to be a manufacturing base here in the United States. And maybe we ought to stop using terms like "rustbelt" and "smokestack" to refer to manufacturing companies. We ought to be using a term like "powerbase." We somehow have to get everybody to realize that there's value in being a manufacturer, as opposed to being a non-manufacturing country. If we get rid of a lot of the manufacturing in this country we will more rapidly become the victim of what the rest of the world wants us to do or not do. We've forgotten what made us strong, and getting away from the formula has created a big problem for us.

Q: How can the government help U.S. manufacturing survive and prosper?

A: Other than get smaller and stay out of our way, which are always good ways to help, a classic example in our industry is product liability. We've spent 20 years trying to come up with some reasonable product liability reform legislation. It's just now happening. We hope that by the end of this congress we will have successful change. But why should something that's so obvious take so long? Probably our biggest issue is tax reform. All of a sudden there's this realization that the tax system in the United States is broken. We've got to scrap it. Starting this fall sometime, you'll see a lot of discussion on this issue. And everybody in this country needs to stay tuned.

Engineering News

Engineering News

Machine vision eyes new applications

Newton, MA--Once reserved for the programming elite, machine vision is maturing into a flexible, powerful automation tool for electronic, aerospace, medical, and automotive OEMs--just to name a few. Now, with RISC architectures, faster co-processors, and more user-friendly software, machine vision is expanding into new applications.

The technology's earliest use was inspecting products in advanced manufacturing applications such as semiconductor fabrication and electronics assembly. Although such applications remain a core of the industry, performance improvements are broadening machine vision's scope.

For example, at Cognex Corp., Natick, MA, engineers recently designed a PC-based machine-vision system that combines a user-friendly "point-and-click" Windows(TM) development interface with grey-scale processing software.

  • Machine vision is increasingly being used for robot guidance and assembly verification

  • Machine vision will become more integrated with production processes

Checkpoint software digitizes and analyzes images in 256 shades of grey and works with custom ASICs to accommodate manufacturing-process changes such as lighting changes and variations in part appearance and position, say Cognex engineers. The top-of-the-line 800 model uses a Motorola 68040 CPU running at 40 MHz for fast image processing.

The Checkpoint system lets manufacturing engineers create and implement advanced applications without extensive programming expertise or in-depth knowledge of machine vision, says Bill Silver, vice president of Cognex R&D. Such applications aren't limited to inspection. Cognex customers now use machine vision to calibrate automotive speedometers, verify camera assemblies, align integrated circuits, and count pills in membrane packaging. In many cases, these tasks had to be performed by hand in the past: a tedious and error-prone job.

At Aquity Imaging, Inc., Nashua, NH, engineers are applying neural-network technology to machine vision. Their PC-based Mentorvision(TM) system uses neural networks to generate a filter that discriminates between aesthetically acceptable and unacceptable products. This "intelligent vision" brings machine vision into applications previously reserved for human judgment.

For instance, at Colgate Palmolive, Paris, engineers are using Mentorvision to inspect multiple labels on household cleaners--as well as adapt to changes in bottle size, label language, and position. The vision system's neural networks "learn" a product by viewing good and bad products on the line and receiving instructions as to which is which from the operator. After processing 10-20 good samples this way, operators can refine the inspection by accepting or rejecting any non-conforming bottle.

Previously, Palmolive didn't use machine vision for inspection because frequent changes in label language and bottle size would have required time-intensive re-programming and de-bugging. Mentorvision's "show and go" approach lets Palmolive change the line to add promotional labels or bottle sizes in less than half an hour, say Acuity engineers.

Machine vision is also coming to laboratory automation. To better handle hazardous fluids such as blood samples, Acuity engineers are designing a robot that uses machine vision for guidance and to identify clinical specimens. The "AutoPrep" removes samples from a rack and either sorts them or places them in a centrifuge. The research is sponsored by the National Institutes of Health, and is typical of new arenas for machine vision.

John Agapakis, vice president of R&D at Acuity, explains, "Machine vision is always integrated with manufacturing, but we see machine vision being more and more integrated with the production process, and therefore more beneficial." The AutoPrep and similar systems protect workers from infection or repetitive-stress injuries, and can keep them out of dangerous environments such as hazardous-waste disposal operations.

Higher yields. At Ismeca U.S.A., Inc., Carlsbad, CA, engineers have incorporated machine vision in an electrical test system to boost through-put rates for surface-mount devices. The G116 combines electric testing and machine-vision inspection. Using a rotary vacuum turret to move the devices, it inspects in 2D or 3D to monitor lead condition, coplanarity, and packaging. The system handles as many as 6,000 parts per hour and reduces the risk of component damage, say Ismeca engineers. "In years past, the rate might have been 1,500 or 2,500 parts per hour," explains Ismeca marketing mananger Tom Clerici. The system also reduces cost by combining several operations. Previously, SMD makers relied on slower human operators to transfer products from test to inspection fixtures, or inspected every second or third device to save time.

Some machine-vision manufacturers, such as Imaging Technology, Inc., Bedford, MA; Vision Modules, Inc., Campbell, CA; and DI/MAC Technologies, Orlando, FL; supply modular machine-vision components that allow the user to select individual image-analysis components and upgrade in increments. Active Imaging, Incline Village, NV, specializes in IP67-rated stainless-steel cameras for machine vision in harsh industrial installations.

With enhanced performance, inspection at assembly-line speeds, and flexibility, machine-vision technology is gaining acceptance in mainstream manufacturing applications, as well as creating new ones. For machine-vision designers, the challenge will be to keep pace with fast automated processes, and use inexpensive processors to keep prices down.

--Andrea Baker, Associate Editor

IC vendors offer Fast-Ethernet designs

Houston, TX--In a move to speed development of high-speed networking schemes for desktop computers, four leading chip makers--Texas Instruments, National Semiconductor, AT&T Microelectronics, and Broadcom Corp.--are offering free reference designs to help engineers develop systems for the three 100-Mbps LAN standards.

The designs show how to link TI's TNETE100 ThunderLAN(TM) 100-Mbps Ethernet controller chip to physical-layer devices from the other three companies. These devices include: National Semiconductor's DP83223/DP83840 TWISTER(TM) chip for 100Base-TX networks, AT&T's ATT2X01 transceiver chip for its Regatta(TM) 100VG-AnyLAN family, and Broadcom's BCM5000 10/100BASE-T4 Fast-PHY transceiver to 100Base-T4 networks.

"The day of homogeneous networks is gone," says Joe Valente, TI's Networking Business Unit manager. "All four companies want to ensure that OEMs have the technology to deliver a variety of 100-Mbps Ethernet products to the market quickly."

Simulation boosts helicopter design

Leovil, England--Engineers at Westland Helicopters Ltd. are using computer analysis of flight-test data to improve designs, develop performance specs, and produce flight-manual information. In addition, the company's Aerodynamics Performance Group (APG) creates helicopter flight simulations for test pilots, allowing them to develop effective emergency flying maneuvers--before trying them out on riskier, expensive test craft.

"It is truly an outstanding tool helping pilots generate effective flight techniques," says Grant Matthews, APG prinicpal engineer.

Westland's simulation system presents a high-resolution reproduction of a cockpit instrumentation panel, as well as control box and message window, and a real-world outside view while "flying."

"Using the display, the pilots can input parameters to test new maneuvers and see how the simulation reacts," Matthews explains. In one case, pilots simulated take-offs and landings on an off-shore oil rig if an engine failed.

After actual test flights, information captured by the aircraft's on-board tape system is converted into thousands of small data files and stored in a mainframe computer. Westland engineers then transfer selected files to their workstation network, and have the option of placing up to 12 separate traces on screen.

"Once the data is analyzed, we produce a data set that tells us the power required to fly a particular aircraft model at any given weight, air speed, and ambient condition," Matthews says.

A new network of powerful workstations has cut the time required to display graphics simulation data from 30 minutes to 30 seconds, the company says. The company is using a network of Sun Microsystems SPARCserver 2, SPARCstation IPX, SPARCstation IPs, SPARCstation ELCs, and SPARCstation 10s, as well as Evans & Sutherland graphics workstation and Digital VAX computer.

For the future, the company is working on an enhanced simulation program with more 3-D graphics. "We are developing a pilot model that will actually fly the Coupled Rotor-Fuselage Program through various maneuvers to determine how the rotor interacts with the fuselage," Matthews says.

IBM serves up affordable CAD

Sommers, NY--IBM is wooing engineers on a budget with a slew of new workstations ranging from $6,000 to $11,000. Among them: six models built around the PowerPC processor, a chip jointly developed with Apple Computer and Motorola.

IBM RISC/6000 PowerPC 604
Workstation Performance
43P-133 42T
Base Price $7,620 $10,945
Memory 16-192 Mbytes 16-192 Mbytes
Disk 540M-5G 1.1-4.4G
Power PC 133 MHz 120 MHz
L2 Cache 512k optional 0.5M
SPECint92 176.4 118.2
SPECfp92 156.5 116.5

"I think they're absolutely dynamite," says Tom Copeland, director of workstation research at International Data Corp., Framingham, MA. "They've really redefined performance at the low end of the workstation market."

While IBM has not been known for low-cost workstations before, Peter Lowber at Datapro, Lexington, MA, says Big Blue has changed its strategy. "Suddenly, they're the price leader for comparable performance at the low end," he notes. "They have very aggressive pricing." A key reason: PowerPC, which allows the company to take advantage of high volumes to lower costs.

The Model 42T, at $10,945, features 16M memory, 1.1G, and a 120-MHz processor for $10,945. Company officials rate the machine at 118.2 SPECint92 and 116.5 SPECfp92, measures of integer and floating-point performance. The workstation is designed with 3-D graphics. The lower end 43P with 133-MHz CPU, at $7,620, does not yet feature 3-D graphics acceleration; but boasts performance of 176.4 SPECint92 and 156.5 SPECfp92.

IBM also introduced a new server, the Model C20, offering 32M memory and 1G disk storage for $19,200.

Powder metal lowers pump-component costs

Dexter, MI--A powder metallurgy (P/M) process designed for a variety of alloys, mild steels, and tool steels is helping design engineers use P/M in new pump applications.

The F2(TM) powder metal process, developed at Krupp Engineering, gives engineers the option to design with low-alloy and car- bon steels in strucural components. "The process expands P/M beyond the traditional territory of tool steels," explains President Philip Krupp.

For example, at Vickers Aerospace Marine Defense Group, Jackson, MI, engineers are using the process to lower the cost of fuel-injector pumps for aircraft. By switching from wire EDM to the F2 process for fabricating vane pump rotors, engineers lowered the cost of the preform. Parts produced with the P/M process require only finish grinding of the slots and opposing faces, and the internal spline no longer requires finishing, say Vickers engineers.

Likewise, using the P/M process improved the design of clips used to attach the motor and pump housing of a centrifugal pump developed by Marley Pump Co., Lenexa, KS. Stamped clips proved not strong enough, and engineers ruled out cast or machined clips as too costly and time-consuming. The F2 process reduced overall design time, say Marley engineers. The clips have a hardness of Rc 15 and exceed load-test requirements.

At the Sorenson Research Division of Abbott Laboratories, Salt Lake City, UT, engineers replaced machined aluminum components with P/M parts made of 316L stainless steel in a nutrition pump. The switch improved part quality, says Chief Process Engineer John Kirk. "We got a part that had a better appearance and greater durability," he adds.

The difference between the F2 process and convention- al P/M occurs during the second half of manufacturing, say Krupp engineers. The process uses higher, variable temperatures and pressures during sintering to boost finished densities by as much as 20%, they say, for final densities of 98 to 100%.

Thermoelectric units cool cellular base stations

Dallas, TX--Cooling an enclosure not only prevents components from frying, but can also let you use commercial-grade components rather than military grade. The difference? Commercial-grade electronics operate at up to 70C; military electronics operate at up to 125C, but cost twice as much.

Motorola's Cellular Infrastructure Group is developing an extensive wireless personal communications system. The system will use bay stations as relay points to transmit signals from cellular phones and other devices. The stations will be housed in 2.5x2x1-ft enclosures located on building exteriors and telephone poles.

When Motorola wanted to have an enclosure cooled to 50C to use commercial electronics, the company turned to Marlow Industries, which developed a design that uses thermoelectric coolers in the enclosure doors.

A standard compressor system wouldn't work in the enclosure because of its size and the inability of the working fluid to recondense. Thermoelectric coolers are much smaller and inherently more reliable because they have no moving parts. They are essentially heat pumps made of bismuth telluride semiconductor material. Direct current moves heat from one side of the module to the other.

Inside the enclosure door, three groups of six thermoelectric coolers wired in series draw 9.4A at 156V. Each group is sandwiched between two heat exchangers. A squirrel cage fan moves air throughout the enclosure, stabilizing the temperature at 50C in ambient conditions.

"We had to specially design the coolers to fit this application," says Marlow Project Manager Lance Criscuolo. "The design offers ease of manufacturability for Motorola. For units that require active cooling, the thermoelectric door is attached. Units that will be located in cold climates will have a standard door. The units can be manufactured on one assembly line."

Cavity-alignment jig helps illuminate subatomic particles

Port Washington, NY--Since July, physicists at CEBAF, the $600 million superconducting continuous-electron-beam accelerator facility in Newport News, VA, have been studying the behavior of gluons and quarks within a nucleus. The research harnesses the nearly speed- of-light beam's particle nature.

CEBAF's 4-billion-volt electron beam gains its energy from five "laps" around the accelerator's 7/8th-mile-long "oval:" two electron "drag strips" containing 160 superconducting oscillating- electrical-field niobium cavities connected by two 180 degrees magnetic-force "turns." Each trip down a drag strip boosts the beam by 400 million volts to add 800 million volts per lap. Because the niobium cavities superconduct when they are cooled to -456 degrees F with liquid helium, they do not heat up when the accelerator is operating. This allows CEBAF to run continuously, which is unique among particle accelerators.

For CEBAF to accelerate its beam properly, the axes of each individual bowl-to-bowl shaped niobium cavity must be aligned to within 10 microns during assembly. To accomplish this, Danny Machie, a senior engineering associate at CEBAF, designed a cavity-pair assembly and alignment fixture that uses chrome-plated rails and pillow blocks from Thomson Industries, Inc.

C-clamps mounted on four of the fixture's Super Ball Bushing bearing pillow blocks hold the cavities in place, and the 1.5-in 60 Case Tubular Lite LinearRace Ways allow the cavities to be brought smoothly together along a single axis. By adjusting three plastic screws on each clamp and measuring from reference points on the rails, each fixture's eight cavities can be aligned in just three iterations. After alignment, the cavities can be transfered to either a test bench or the accelerator itself, where mounting flanges maintain the alignment and the fixture is removed.

Omron supports Design News Foundation again

Schaumburg, IL--Omron Electronics, Inc. has announced that it will once again support the annual Design News Engineering Awards program with a $10,000 gift to the Engineering Education Foundation. The gift marks Omron's fourth year of participation in the program benefiting engineering students at the university level.

"Omron is proud to again sponsor this powerful program to lift the awareness of the engineering arts and support the pursuits of dedicated students. By supporting the growth of young engineers, we know we are contributing to the overall growth and prosperity of the country and the world," comments Frank Newburn, executive vice president of Omron.

Omron employs more than 1,300 engineers worldwide who work in areas including material science, mechanical design, bioengineering, electronic design, and optical system design. "Our life-blood and our value to our customers depends on there being a continuous stream of talented young engineers," says Newburn.

Omron ranks as one of the world's leading suppliers of factory automation and control components including sensors, process controllers, vision systems, relays, PLCs, and optical switches.

In addition to the EEF donation, Omron continues to support the engineering arts in other ways. For example, the Omron Foundation, Omron's philanthropic arm, offers continued support to engineering students via ongoing scholarship programs with five universities in Illinois.

Omron also continues to direct its local product engineering staff toward the design and development of software and hardware products for the North and Central American markets. Within the last two years alone, Omron has nearly doubled its product development and application engineering staff.

The results of these investments include improved support services for Omron's engineering customers and new product solutions such as the new E3JU general purpose photoelectric sensor and new Analog Input module for their large rack PLCs.

'Smart' sketchpad captures drawings electronically

Cestas, France--An "intelligent" sketchpad with patented pressure-sensitive electronic pen enables designers to develop ideas from doodles to concepts--and then transfer them to a 3-D CAD system.

Using Lectra Systemes' Graphic Instinct, the designer sketches freehand directly onto a high-resolution (1,280 x 1,024) screen. The electronic pen can simulate a felt-tip marker, pencil, pen, or brush in a wide variety of colors.

As the sketch proceeds, a unit built into the top of the screen detects infrared and ultrasound signals emitted from the pen. This ensures that details of the drawing are captured electronically. So stored sketches can be subsequently modified without redrawing.

A system of triangulation determines the pen's position in space, as well as information concerning its rotation. This system is based on detecting ultrasonic signals transmitted from a number of sources within the pen. The receivers situate along the top of the screen.

By building a resistance sensor into the pen, Lectra Systemes differentiates between light and heavy pen strokes. The resistance sensor digitizes the pressure applied and sends the signals to the computer via ultrasound.

Once a design is finalized, it can be exported to a CAD workstation for further detailed design work. Intended for designers with no computer skills, the "intelligent" sketch pad is based entirely on visual icons.

Following demonstrations of the first prototype in Milan in March, the pad was installed at a number of test sites. First commercial products should be ready for delivery this summer, at prices ranging from 250,000 to 300,000 French francs (about $50,000 to $60,000 U.S.).

--Anna Kochan, European Editor, France

Sporty Sunfire won't empty your wallet

Watertown, MA--You've likely seen Pontiac's new Sunfire everywhere this summer: streaking down the Great Wall of China, doing donuts around the Leaning Tower of Pisa--and touring the Boston area with me behind the wheel. The $13,764 SE Coupe I drove for a week had the usual standard features, plus some you'd only expect to find in a more expensive car.

For example, in the Sunfire you'll be pleasantly surprised by rear-seat heating and ventilation ducts, battery-rundown protection, theater dimming of interior lights, a rear folding seat, front- and rear-seat cupholders, and an oil-level sensor. One feature Pontiac is especially proud of is the 4.9-l glove compartment, which can hold 12 cans of the beverage of your choice.

Controlled by a 5-speed manual transmission, the 133-cu-in, 2.2-l OHV four-cylinder engine delivers 120 hp at 5,200 rpm and 130 lb-ft of torque at 4,000 rpm. Pontiac estimates mileage at 24 mpg city and 35 mpg highway. A 2.3-l DOHC in-line four-cylinder engine is also available, as are three- and four-speed automatic transmissions.

Safety features include dual front-seat air bags, 4-wheel antilock brakes, side-guard door beams, and a slotted-frame rail design for increased occupant protection during frontal collisions. The front suspension has a stabilizer bar; the rear suspension features a coil-over-shock design. Having the shock and spring "in line" isolates passengers from bumps and bounces, resulting in a solid and controlled ride.

Although Senior Editor Sharon Machlis prefers the ride and interior of the Dodge/Plymouth Neon, I prefer the aesthetics and sportiness of the Sunfire. The suspension was good and stiff, shifting was easy, and I had a ball zooming around in this hot little number. I enjoyed the experience even more knowing that I could actually afford to buy the car.

--Julie Anne Schofield, Associate Editor

Electric roadster delivers low-cost, fun transportation

Palm Bay, FL--In June, Renaissance Cars Inc. (RCI) began shipping its first electric roadster, the Tropica. Weighing 2,160 lbs, the sporty-looking two-seater easily reaches 55- to 60-mph cruising speeds.

Two advanced-technology 24.5-hp DC motors provide the motive force, driving one rear wheel each via cogged belts. Twelve six-volt lead-acid batteries offer a range of 60 to 80 miles, and the company says that optional advanced batteries extend that figure to more than 100 miles.

Most important--as any Detroit engineer knows--is cost. "The goal was to demonstrate that it's feasible to build, from the ground up, an electric vehicle for under $20,000, battery pack and all," says Jack Guy, manager of electric-transportation commercialization at the Electric Power Research Institute (Palo Alto, CA). Production of the Tropicas beat that figure considerably, and they will grace RCI's initial 21 showrooms at an suggested retail price of $17,500.

--Mark A. Gottschalk, Western Technical Editor

Application Digest

Application Digest

For flexible automation choose linear modules

Steven J. Annen, AdeptModules Business Manager, Adept Technology Inc.

When designing automated postioning equipment, engineers have access to a variety of flexible, off-the-shelf robotics and linear modules. In contrast to dedicated positioning systems, flexible solutions can be re-programmed and re-deployed, increasing a manufacturer's return on investment. Several questions help determine which type of automation engineers need:

  • How many axes of motion are required?

  • Is programmable positioning necessary?

  • How big is the work envelope?

  • What are the payload requirements?

  • What is the budget?

Linear modules prove to be most cost-effective for applications with one to three axes, a large work envelope and high payloads. Payload affects the type of linear module needed. Most come in three styles: a high payload, X-axis; a lighter duty, Y-axis; and a small Z-axis. AdeptModules combine to provide 15 two- and three-axis configurations with payload ranges from 15 to 40 kg.

The drive mechanism is also critical. Belt drive modules offer speed but limited configurability, repeatablity and lower payload capacity. Ball-screw drives are slower but more precise, with higher capacities and greater configurability. To ensure long life, select a system with at least twice the capacity of the actual payload.

To speak with an Adept applications engineer, call: (800) 226-6385.

Leverage workstation CAD on the PC

Brad Weinert, Engineering Product Manager, AGE Logic

The increased demand for workstation-based CAE applications can easily tax any organization's high-end computing resources. As engineers vie for time on a limited number of workstations, frustrations and bottlenecks result. Many companies, consequently, are implementing a PC X-server solution that leverages workstation CAD across their PC network.

With the advent of more powerful PCs and video cards, PC-based X servers are a viable alternative to X terminals and UNIX workstations. A few simple guidelines will assist in determining the hardware requirements for your PC and which PC X server is optimal for your application:

The faster the processor speed of the PC the better. A minimum configuration of a 486DX/2 66 with 16Mb of RAM is recommended for CAD applications. To attain acceptable graphics performance, select a high performance video card (at least 32 bit) with 2Mb of RAM and support for 256 colors. The video card is important because the drawing speed of your X server depends on it.

To choose a PC X Server, contact your CAD/CAM/EDA software vendor to see what they recommend. You should only consider a native 32 bit X server (vs. 16-bit) and be sure to try it with the applications you're running before purchasing anything. Remember, benchmarks are only good at determining how fast the benchmark runs, not your application. You should always run your applications in a production environment before making a decision.

To speak with an AGE Logic applications engineer, call: (619) 755-1000.

Post-processing software fine-tunes defroster

Post-processing software fine-tunes defroster

Research Triangle Park, NC--Not long ago, the way to improve a system such as an automotive defroster was to build a prototype, test it, build another prototype, test it.

Chrysler engineers now use computational fluid dynamics (CFD) software to speed up the design process. In one minivan project, they sought to improve flow distribution, reduce pressure drop, and quiet a defroster duct.

To generate the defroster duct geometry and finite element mesh, the engineering team used PATRAN from PDA Engineering. FIDAP code from Fluid Dynamics International provided the analysis. With 31,000 elements in the model, the solution ran on a single CPU of Chrysler's CRAY Y-MP 8I system.

To display and animate the data set from the 3-D analysis, the engineers selected EnSight from Computational Engineering International (CEI). EnSight is a post-processing system that displays and animates the results of sophisticated finite element modeling techniques. EnSight ran in a distributed mode with the graphics on a Silicon Graphics Personal IRIS and the compute-intensive portion running on the CRAY.

In a related project, Chrysler engineers studied the de-icing action of the defroster in the LH series sedans. Realistic simulation required a model of the passenger compartment.

  • Vehicle aerodynamics

  • Engine flow patterns

  • Combustion research

Starting with CAD surface data generated by CATIA, the geometry was imported to ICEM-CFD to generate the multi-domain mesh. The model contained 110,000 cells and 50 domains. Then, CFD-ACE code from CFD Research Corp. solved the steady-state flow field and the transient ice-clearing process.

Post-processing the CFD-ACE code results with EnSight produced the animation of ice-melting patterns. "The wealth of flow field information enables us to make better design decisions earlier in the design cycle," says John Tripp, Chrysler fluid dynamics engineer.

Additional details, program...Contact Tom Palmer, Computational Engineering International, Box 14306, Research Triangle Park, NC 27709, (919) 481-4301.

Additional details, application...Contact John Tripp, Chrysler Corp., 800 Chrysler Drive East, Auburn Hills, MI 48326.

Slippery coating replaces pneumatic force

Slippery coating replaces pneumatic force

Huntington, IN--The voice coil is critical to an audio speaker's performance; coil winding efficiency is critical to production. At Pyle Industries, home of "Made in the U.S.A." sound systems for cars, a problem with sticky arbors slowed voice coil production. Solution? A "synergistic" coating from General Magnaplate, Linden, NJ.

Pyle's coil-manufacturing operation works like this: A polyimide film, measuring 3 or 5 mils thick, wraps around a 6061-T6 aluminum arbor, which ranges in size from .50 to 3 inches in diameter. Insulated magnet wire, coated with a proprietary adhesive system, is then wound wet onto the arbor.

After alignment and application of reinforcing tape, the coils bake for 40 minutes at about 400F. Lead wires are chemically stripped and solder tinned, coils are removed from the arbors, and the arbors recycle back to the beginning of the wet-winding process.

Problem: Following baking, Pyle's workers discovered that the coils adhered to the arbors. "At first we could do the separationby hand," explains Production Manager Brian Miller, "but eventually we found it necessary to set up a pneumatic cylinder with a tooling arrangement that allowed us to step on a pedal and physically drive the arbor out of the coil."

Repeated recycling and the buildup of adhesive residues, however, required more and more separation force. The procedure resulted in damaged arbors. Other attempts to solve the problem also proved unsuccessful.

  • Conveyor system components

  • Food processing rollers

  • Sealing dies

Solution: Miller found the specialized surface enhancement coating applied by General Magnaplate. Named Lectrofluor(R) 615, the elastomeric treatment process involves applying proprietary blends of polymers to the cleaned surface of the aluminum arbors. Polymer selection depends on several factors: End-use application of the part, its base metal, the kind of hostile environment it will encounter, and the permitted coating buildup.

The "synergistic" coating results in a surface superior in performance to the base metal and to the individual components of the surface enhancement.

"With the Lectrofluor 615-treated arbors," says Miller, "the coils just slide off. By eliminating the pneumatic ramming process, Pyle Industries expects to reduce arbor damage by 75%, and at least double the number of coils they can separate from arbors in a given time.

Additional details...Contact General Magnaplate Corp., 1331 Route 1, Linden, NJ 07036, (908) 862-6200.