Engineering News

PCs take on new roles in automobile manufacturing

Industrial-grade PCs are challenging the rank of traditional industrial control equipment

Michael Baab, Editor at Large

Bremen, Germany--Engineers here at Johann A. Krause Maschinenfabrik GmbH are putting the finishing touches on an unusual engine assembly line they will soon ship to Ford's Dagenham plant near London. Mechanically, the line assembles four-cylinder diesel engines, but electrically it will not operate like anything else this OEM machine builder has shipped before.

Instead of the traditional gear-programmable controllers, switch cabinets, and thick bundles of I/O wiring, one Interbus line runs from station to station. At one end, the grey plastic cable disappears inside an industrial personal computer.

"Most automobile engine manufacturers, like Ford, want to be able to react to new market requirements in a quick, flexible, and cost-saving manner," says Burkhard Werner, automobile project manager of Phoenix Contact, (Blomberg, Germany). Phoenix supplied approximately 500 Interbus modules and the "soft-logic" PC software that provides real-time control of the assembly line. "Ford intends to automate the production of engines with open and manufacturer-independent control systems," says Werner, "and has decided in favor of the industrial PC because of its substantial cost advantages compared to conventional PLCs."

It was in the 1980s that OEM machine builders first brought the PC to the factory floor--initially as PLC programmers, and later for data acquisition, operator interface, and diagnostic applications. Since then, the PC has become one of the most important peripherals for industrial control systems.

At Audi's plant in Ingoldstadt, for example, PCs are used in a two-stage aggregate assembly system where transmissions and auxiliary systems are assembled during the first stage, and the telescopic strut component group is put together during the second. The individual steps are carried out by assembly line workers, with transport between the two stages automated and controlled by PLCs and Interbus-connected component groups. Workers monitor the progress of the operation with personal computer workstations.

At Dagenham, the PCs will take on a much larger role. They will control the diesel- engine assembly line in place of traditional PLCs. Where, all of a sudden, did OEMs get the idea they could replace PLCs with PCs? Five technology enablers of the mid-1990s have contributed to this trend:

Thinking ahead, Ford engineers say they want future generations of engines to be manufactured in a uniform manner at production sites around the world. The PC's worldwide availability and acceptance, which is much greater than that of manufacturer-specific PLCs, weighs heavily in their favor.

Another factor is the increasing acceptance of I/O buses such as Interbus, which will be used by Ford on its first PC project in Europe. "Interbus means that the I/O cards do not have to be integrated into the PC," says Werner. "They go out into the machine, reducing the requirement of switch cabinets."

In the U.S., similar thinking has achieved major proportions at GM Powertrain, where in half a dozen large Detroit plants, thousands of PCs are deployed as machine controllers. Division managers have, in fact, opted to exclude PLCs on all new projects in favor of PCs. One of Powertrain's newest projects is at the Romulus plant, where 350 PCs machine and assemble V8 engines.

Powertrain's need for PCs, explains GM's Jack Lienesch, include lower cost and global manufacturing, but go far beyond. GM's manufacturing strategy will shift production to aluminum engine blocks, a softer material that requires closer tolerances and a higher degree of data-intensive control. "It's like machining egg shells," says Lienesch. "We have to be very careful how we do it." Modern PC technology is particularly well suited for such data-intensive applications.

Another reason, says Lienesch, is the need to incorporate factory-floor controllers into GM's "math-based manufacturing" agenda, which in theory has a network of computers, from the design table to the production floor, sharing the same data files. The proprietary PLC has, until now, been a major obstacle to this program.

Powertrain's soft-logic implementation has one distinctive advantage: the flowchart control language developed with inventor Ron Lavallee over the past 10 years. With this software, called FloPro, complex machine control algorithms are represented as a series of parallel flowcharts, which are compiled and executed directly from the PC. More than a programming tool, FloPro's diagnostic and operator interface capabilities allow it to penetrate deeply into the life cycle of the machine. Lavallee's product is now owned by Nematron Corp. in Ann Arbor, MI, and several other vendors offer flowchart programs.

Standard PLC software. In Europe, the desire for standardization has driven the development of PLC software that conforms to the IEC-1131-3 programming standard. Many software vendors have created Windows-based IEC-1131-3 editors for the purpose of downloading executable programs into PLCs. But today's powerful PCs execute the programs just as fast and as reliably as most PLCs, and easily connect to both proprietary and "open" I/O systems. It wasn't long before the Europeans had their own version of soft logic: Why not just forget the PLC altogether, they reasoned, and use a PC for everything: programming tool, controller, operator interface, alarm monitor, and machine diagnostics? Phoenix Contact's software package PC WorX performs all of these functions simultaneously.

Futurists see yet another advantage of the PCs-turned-controllers: With a modest investment in hardware and software, each PC can become a site on the World Wide Web. OEMs would be able to point their web browsers at machines thousands of miles away, inquire of their status, diagnose problems, and advise their manufacturing clients about preventive maintenance. By e-mail, no less.

What this means to you

Camera details rubber meeting road

Rick DeMeis, Associate Editor

Akron, OH--The most important real estate to any driver is the four 6-inch- wide by 8- to 12-inch-long ovals formed by tires against the road. Particularly important in precipitation, the friction forces generated in these patches determine whether a vehicle weighing a few thousand pounds moving at high speed is under control.

Understanding the behavior of individual tire-tread lugs under dynamic conditions, such as acceleration, braking, and cornering, gives tire designers insight to optimize vehicle handling and responsiveness. Other considerations are ride quality, traction, wear, and low noise. The effectiveness of lug patterns on rain tires to channel away water, preventing hydro-planing, or on snow tires to grip, are also important. "The contact patch is highly dynamic," says John Turner an engineer with the Advanced Tire Technology Division of Bridgestone/Firestone. "Tread lugs require intimate contact with the ground." Improper contact de-grades traction and increases wear.

Bridgestone/Firestone is using new high-speed, high-light-sensitivity FlashCam CCD cameras from The Cooke Corporation (Tonawanda, NY) to better understand tread and footprint phenomena. As a tire is driven over a viewing window, the camera captures digital images of rolling tire footprints during vehicle operation. "We can determine how each lug is placed against the ground, and overall footprint shape and how it changes during maneuvering," notes Turner.

The FlashCam provides a combination of speed and light sensitivity not available before. A Sony 8-bit, 640- x 480-pixel array CCD scientific-grade sensor is used for imaging, with pixels registering 255 levels of gray. Electronic shutter speeds as fast as 1 musec are possible, and Bridgestone/Firestone uses multiple FlashCams to capture images separated by as few as 10 musec. A frame grabber retains the images until they are downloaded into a PC, where the tire-maker's image-processing algorithms extract shape and light-intensity information.

Standard CCD cameras cannot provide the extremely short shutter speeds and still have sensitivity to image the footprint under reasonable lighting conditions. "Earlier CCD cameras had light sensitivities of ISO 125 to 300," says Cooke Vice President of Sales Gerald Lilly. "Now they're into the thousands." But the FlashCam has a "binning" feature that clusters four pixels, treating their output as one, for four times the sensitivity, he adds, with sufficient resolution. "Also, the short shutter speed can be utilized to produce multiple exposures on a single video readout frame."

With older digital video technology, tires had to be rolled slowly with a special machine because of the standard 30-Hz frame rate. Such methods were unable to duplicate high-speed dynamics of a moving vehicle. As for using high-speed films, with "push processing" they have an ISO just over 1,000 and have been used for stop-action pictures of hydroplaning. However, tests are limited by the amount of film, camera loading, and processing--which slows tests and analysis.

Digital cameras allow nearly instant data manipulation and analysis via PC. Engineers can also repeat or modify tests with minimal delays.

Polymer cuts 'big bang' research costs

Brookhaven, NY--The Relativistic Heavy Ion Collider under construction at the Brookhaven National Laboratory will use a superconducting cable called a cold crossing bus (CCB) to supply up to 27,000 amperes of current. The current will surge into large superconducting magnets that produce in-tense magnetic fields. They will bring crashing together ions for the study of the origin of "the other universe," among other things.

Robert Meserve, vice president for engineering at New England Electric Wire (Lisbon, NH), working with Brookhaven scientists and engineers, designed the cable. Robert Lambiase of the lab will apply the CCB design to the collider's electrical design.

The CCB measures just one inch (25 mm) in diameter and costs $50 per foot. That's about $200 less expensive per foot than copper cables of equivalent capacity, according to Arthur Greene,who serves as engineering manager for New England Electric Wire.

The cable requires no trays, because it resides in piping that supplies helium at5 atmospheres pressure to cool the accelerator's superconducting magnets. At the same time, it chills the CCB to the cryogenic temperature needed for superconductivity.

Greene estimates that the use of the superconducting CCB also will save the lab about $600,000 in power costs. By reducing resistive losses, power use will drop to 6,000 MW a year.

What makes the cost reduction possible? "The toughness of Tefzel(R) ETFE fluoropolymer resin (supplied by DuPont Fluoroproducts, Wilmington, DE) at cryogenic, ambient, and elevated temperatures is crucial to system reliability," says Greene. The temperature of the helium in which the cable is installed drops as low as -269C (4.2K), and the cables can be subjected to temperatures up to 100C in the event of an emergency shutdown, according to Greene.

The CCB consists of nine or 10 individually insulated cables enclosed in an outer sheath of wire braid and two DuPont films, 0.001 inch (25 mum) of Kapton(R) polyimide film, and 0.002 inch (51 mum) of Mylar(R) polyester film. Seven or eight of the cables contain niobium-titanium alloy conductors that become superconducting at cryogenic temperature. The other two, used for instrumentation leads, are designed around the use of conventional copper conductors.

The insulation's toughness and flexibility are crucial during installation. The pipe in which the CCB is placed measures four inches (102 mm) in diameter. It has bends with radii as much as nine inches (229 mm).

The superconducting cables carry currents of 150, 450, and 6,300 amperes. The 6,300-ampere cables use Kapton in a 0.001 inch thickness for primary insulation, and contain an outer jacket of Tefzel. The other cables use Tefzel as the only insulator, with insulation thickness measuring 0.012 or 0.015 inch.

CAD, materials spur racing wheelchairs

Rick DeMeis, Associate Editor

Georgetown, CT--Joe Montgomery had a frustrating problem. The wheelchairs he had begun purchasing for his then four-year-old son Michael, who has cerebral palsy, were breaking frequently. This was in addition to their high cost and weight.

Fortunately, Montgomery realized he also had the means to do something about it. Being the founder and president of Cannondale, he figured the company's experience in producing light, welded aluminum frames for bicycles could be applied to improve wheelchair--and specifically sport wheelchair--technology. The fruits of the company's six-year long application of its bike-fabrication methods, materials, and components are seen in the 1998 availability of its new S.S.T. (Seated-Sports Technology) line for various sports, racing, and "casual" daily use.

What does it take to make a world-class wheelchair? First off, Cannondale prides itself on its flexible manufacturing technology. Once a journalist visiting the Bedford, PA, manufacturing plant complained he could never find a bike that fit. Cannondale took his measurements, and a couple hours later when he was ready to leave, had a made-to-order frame on hand. For the wheelchairs, frames are also built to the custom dimensions of each owner, according to Cannondale designer, and wheelchair user, Alan Ludovici. Similarly, the dimensions are input to Pro/ENGINEER(R) CAD software (Parametric Technology, Waltham, MA), which then guides laser cutting of the large-diameter, thin-wall aluminum frame tubes. In addition, a series of tabs and slots are cut on the tubes so that they "snap-fit together" when placed in a modular alignment fixture. "The frame pretty much jigs itself," notes Ludovici.

Achieving Cannondale's light but strong frame comes next, according to Ludovici, by welding and heat treatment--annealing and a chemical-solution aging--for stress relief and tensile strength. "Final frame alignment is before the final treatment, and some critical areas, such as the rear axle and head tube, are then machined afterward. We can use lighter weight tubes because of the heat treat and stress relief."

Also important material-wise are the racing chair's carbon composite rear disk wheels. Light and strong, they produce less drag than spoked wheels which "fan" the air. Such a disk is not used at the front of the three-wheeled racers--it would act like a sail ahead of the center of gravity, turning the chair in brisk crosswinds.

A knob-adjusted spring compensator puts a slight "steer" into the front wheel to allow for racing along crowned roads. Thus, the athlete can put full power from both arms into the wheels, rather than hold back on one side to account for the crown. For oval-track racing, an "Integrated Track Stop" can be preset in practice to steer the front wheel at the appropriate distances along the track, again permitting full power from both arms.

Cannondale also notes it has access to better bicycle-derived brakes, bearings, and carbon-fiber wheel rims than those found in wheelchairs.

What about using composite frames for the chairs, as is done on some bicycles? Ludovici says, "Composites would be harder to integrate. Bicycles are essentially two-dimensional frames, while chairs are more complex three-dimensional objects. At this point, alignment and straightening make it more difficult to build with composites."

In conclusion, Ludovici notes, "I've worked for other chair manufacturers and nobody is making anything as stiff and light. They out perform anything I've ever ridden."

Hybrid bat delivers more power to the ball

Chicopee, MA--Softball bat technology has come a long way since those wooden sandlot sluggers. In the 1970s, lumber bats pretty much gave way to aluminum. The high-yield strength of aluminum allows for a thinner barrel, giving the bat more flex in the contact area and helping transfer additional energy to the ball.

In order to increase the flex of the barrel and reduce weight, designers typically minimize the wall thickness. But it's a Catch-22: The thinner the aluminum (softball bat walls are typically 1/10 of an inch or less), the more likely the possibility of permanent deformation--a problem that may prematurely sideline a bat. The extrusion and subforming operations also result in a handle that is thicker and heavier than optimal from a kinematics perspective.

Not satisfied with the status quo, engineers at Spaulding Sports Worldwide began thinking about how to design a better bat. "We wanted to be able to exploit the advantages of aluminum, while at the same time overcome some of its deficiencies," says Brian Feeney, director of product development at Spaulding.

Engineers adapted their bat design from another sport where swing speed and power are critical: golf. To help players drive the ball further down the fairway, designers of golf clubs want to eliminate weight in the shaft and concentrate most of the mass in the head.

The design of the new bat, called the Fusion, features a body made of graphite/epoxy with a thin (0.065-inch-thick) C-405 aluminum sheath fused to the barrel area. The design, explains Feeney, produces a bat that has a better weight distribution, is stiffer in bending, less prone to permanent deformation (even with thinner aluminum), and more flexible in the barrel than a standard single-wall aluminum bat.

During trials, players swung the bat faster and hit the ball further. How? By using composites in the handle, says Feeney, the bat is lighter in weight, yet, because the weight distribution changes the bat's moment of inertia (to match that of an all-aluminum bat), it still feels like an aluminum bat when swinging. More energy delivered to the ball results in more home-run hits.

To fabricate the new bat, designers came up with a unique bladder molding process. Sheets of glass fabric, uni-graphite tape, and graphite fabric (oriented in a specific direction for strength) in an epoxy matrix are laid up. The aluminum sheathe is then positioned over the composite, which is placed into a mold and cured.

The development of this new manufacturing process has far-reaching implications for Spalding's design team. "Historically, our material options were limited because of the need to form the relatively complex shape of a bat," says Feeney. "With our new fabrication technique, I think we can expect to see the use of some pretty radical materials in the future."

For now, the Fusion, which is sold under the Dudley name for about $250, is helping everyone hit more home runs. But even the best bat technology in the world, admits Feeney, won't help those players who have trouble hitting the ball.

Polyketone leaps regulatory hurdles

Ann Arbor, MI--The U.S. Environmental Protection Agency (EPA) has mandated reduced automotive hydrocarbon emissions at the pump. This directive presented a tough challenge for the automotive industry. But a tougher thermoplastic and a patented design for onboard refueling vapor recovery valves (ORVRs) met the test.

The EPA has ruled that 40% of all 1998-model vehicles must be equipped with ORVRs. By the year 2000, every new car must have an ORVR. By 2003, light trucks and sport utility vehicles have to have the valves installed.

The ORVRs mount directly on the automobile's fuel tank. Fuel travels from the pump nozzle through the car's filler neck to the tank, through the ORVR, and into a charcoal canister. The hydrocarbons are burned off by the vehicle's engine. Without the valves, five to six grams of hydrocarbons per gallon of gasoline are released to the atmosphere during refueling.

The new valve design, patented by GT Products, drops that amount to less than the EPA's 0.2 grams per gallon. The component, made from CARILON aliphatic polyketones, supplied by Shell Chemical (Houston), lowers the amount to 0.06 to 0.08 grams per gallon. The device also acts as an automatic shut-off for the fueling nozzle, once the fuel in the tank reaches 95%.

"The ORVR component forms an integral part of the automobile fuel system," says Ken Zander, GT Products sales manager. "The part needs to be made from a material that resists degradation in a harsh fuel environment."

"We originally specified polyester for our ORVR design," recalls Bob Benjay, GT Products chief engineer. "But polyester, although commonly used in fuel systems, is being phased out as new, more robust fuel-resistant materials become available. Then we tried acetal--it was too brittle--and a host of other materials."

"Of all the materials we tested, CARILON gave us all the properties we needed, notably strength and fuel resistance, at the right value," adds Paul Wrona, GT Products project engineer. Before making the selection, however, GT Products put the material through vigorous drop, shock, durability, and fuel-soak testing.

Shell is consulting with automotive customers on added applications for the material--fuel tanks, lines, and pumps; gears for actuator motors; and under-the-hood components. Other end-uses for the polymer include gears for business machines and industrial molded parts.

Kayak foot brace moves to head of class

Easley, SC--In 1988, Keepers Products Inc. introduced a foot brace for kayakers that has become the standard of excellence. The brace consists of a rail screwed into the bottom of the boat, and a foot pad that slides up and down the rail to fit the kayaker's leg. Today, most kayak makers feature the brace as standard equipment.

Not content to rest on its laurels, Keepers now markets a new, improved version of the brace called the Max. Critical to the Max's enhanced performance are glass-reinforced, nylon-based, injection-molding resins.

"I'm a great believer in nylon resins and their properties," says Janet Master, president of Keepers. "The two nylon-based compounds we use have helped us deliver an outstanding product with great performance, strength, and durability at a lower cost than would have been possible with competitive products. Our new brace also is lighter, a great benefit for kayak enthusiasts who always try to shave as much weight as possible off their boats."

The glass-reinforced nylon materials that Master refers to, Capron(R) and Nypel(R), are supplied by AlliedSignal Plastics (Morristown, NJ).

The Max took shape when Masters brought a challenge to Injection Technology Corp. (ITECH) President Carl Morris. To move the existing foot-brace technology to a new level, Masters wanted an improved locking system so the foot pad would remain in place on the rail. She also required a system with greater overall strength that included a dimpled surface on the pad to reduce slippage, and a product that would solve the shrinkage problem associated with the old design.

The glass-reinforced nylons helped Morris deliver on all fronts. For instance, as the greatest stresses are transmitted to the foot pedal, it had to be the strongest component. Nypel 6033G, a heat-stabilized, injection-molding compound, has a 33% glass content for added strength. The rail and spring are made of Capron 8331G Blend BK 102, a 14% glass-reinforced, injection-molding compound based on recycled nylon 6.

Nypel combines high-strength, stiffness, and heat-deflection characteristics, while extending the retention of these properties at high temperatures, and a lower cost than competitive materials, according to Morris. Moreover, since it is made from recycled feedstocks, Masters adds, "we stamp the 'recycled' logo on both the rail and foot pad." Masters sees that as an added product benefit and a sales driver.

"Field testing was key to developing this product," Morris notes. "You can beat and jump on the kayak all you want, but nothing can duplicate raging whitewater or ocean current. The Max withstands these conditions."

Finite-element analysis for high-speed steel

Saxony, Germany--In order for high-speed steel parts to see more widespread use, the behavior of components made from them has to be better understood. To this end, a partnership of business and academia in Germany is seeking to characterize the interaction of internal notches and external defects under static loading by modeling the microstructure using the finite element method (FEM). The partnership includes Freiberg University of Mining and Technology, the University of Stuttgart, and Siemens AG of Munich.

According to an abstract describing the partnership's FEM approach, the fracture behavior of high-speed steel tools, which considerably influences their performance, is determined by external loading, chemical composition, and structure of the steel. A characteristic feature of the structure of high-speed steels are carbides produced during solidification and forming.

The amount and distribution of these carbides influence the fracture behavior, acting as internal notches as a result of their elastic behavior which differs from the matrix. Fracture of high-speed steels is induced by cracks that form at internal or external notches. A theoretical comprehension of the fracture process is important for the materials optimization with regard to the fracture behavior.

To this end, a variety of analysis techniques can be used to simulate crack initiation and growth in high-speed steels, including modeling of a brittle fracture with nonlinear spring elements. The partnership predicted instable crack propagation, calculated stress intensity with the experimental critical stress intensity of the model, and reported a good agreement of calculated and experimental failure load obtained using the multi-phase elements combined with the element elimination technique.

External loading considerably influences the performance of tools made from high-speed steel. By studying the fracture behavior of such steel using FEM, the partnership aims to improve the design and performance of this useful material.

Sometimes the book isn't right

Newton, MA--The Head Work problem appearing in DN 10/20/97, p. 176 turned out to be a real stumper. Thanks to William, Curtis, Robert, and the many other readers who e-mailed, wrote, and called in to tell us that the correct answer was not among the given choices--which appear with the problem in its original form at left. According to the Selected Fundamentals of Engineering Examination, copyright 1986, the answer should be E (17.5 hp). That is incorrect.

A number of readers came up with the correct solution, which is 40.42 hp. Design News reader Florian Wisinski will receive a $50 gift certificate for being the first to submit the right answer. Not surprisingly, Wisinski is Chief Engineer for Ampco Pumps.

Here's our solution: The simplest way to do this problem is to realize that the change in pressure ({DELTA}P) is essentially a measure of the force being applied to the water going through the pump. Once you convert that force into the right units, all you have to do is multiply the "unit {DELTA}P" by the rate it is being applied (i.e., the water flowing through), as follows: