Cadillac's databus: Solution in search of a problem
Washington D.C.--Ask an engineer to ex-pound on the advantages of databuses, and you're likely to hear a dissertation about wiring reductions and space savings. Databuses, after all, have repeatedly proven that they can save time and money by shrinking huge wiring bundles to single strands.
But when Cadillac introduced its 1998 Seville a few weeks ago, it showed that databuses offer more than mere wiring reductions. They offer a foundation for innovation. "When you've got the databus and computing power and sensors in place, it changes everything," notes Scott W. Badenoch, manager of vehicle performance for General Motors' Delphi Chassis Systems. "With the design power that it gives us, we're finding that the 'car guys' can take over again."
By allowing the "car guys" to take over, Cadillac has endowed its new Seville with a multitude of technical features. Its highly publicized anti-skid feature, known as StabiliTrak, is joined by road- sensing suspension, speed- sensitive steering, traction control, antilock brakes, adaptive seating, side air bags and a series of electronic-based concepts almost too numerous to list. Better yet, all of those features are neatly tied together by the Seville's Class II databus. "Seville is adding content while other luxury manufacturers are 'de-contenting' their cars," says Jack K. Horvath, vehicle chief engineer for GM's prestige cars. "And the best part is that we're keeping the customers' costs stable while we do it."
Solution in search of a problem. Engineers say that the databus is the enabler for Cadillac's "up-content" scheme. In essence, they say, it's a solution in search of a problem. Because it allows existing sensors and microprocessors to talk to one another, it gives engineers an opportunity to dream up and implement performance-enhancing features without the addition of hardware, in many cases.
The Seville powertrain, for example, offers a feature never before seen in an automatic transmission. Known as Performance Shift Algorithm, it "knows" when the transmission should switch to a more aggressive shift pattern. It recognizes when a driver steps deeply into the accelerator pedal, or brakes aggressively, or corners hard. It then responds with more aggressive transmission performance--most often by downshifting at the right time.
Incredibly, this feature employs no additional hardware. Yaw sensors and lateral accelerometers, both of which play a key role in the new shifting scheme, were already in place on StabiliTrak. Similarly, Seville's electronic transmission already used speed and throttle sensors.By placing the signals from those sensors on the bus and making them available to every other system in the car, Performance Shift Algorithm was born.
Within GM, the success of Performance Shift Algorithm serves as an example of the fast development schedule for software-related features. Invented by GM engineer Ed Lansinger in May, 1995, the Performance Shift Algorithm was incorporated into the Seville in 26 months--an extraordinarily quick turnaround for any automotive feature. The reason for its fast turnaround, Lansinger says, is that it required no new hardware. As a result, there were no packaging issues and development cost was minimal.
Electronic enabler. The idea for the databus first surfaced nine years ago, when GM engineers noticed that electronically based systems--from ABS to traction control--were quickly multiplying throughout the company's vehicle lines. Because each of the systems had their own microprocessors and sensors, engineers realized they would soon face a wiring nightmare. With a databus, they reasoned, they could reduce their hardware needs and cut costs.
Still, no one foresaw the design dividends that would accrue from their databus decision. As the number of systems increased, however, and more hardware was installed, new possibilities eventually emerged. "Now engineers can add algorithms as new ideas come up," Badenoch says. "As a result, Cadillac is exploding with new capabilities."
Among those capabilities are enhancements to systems ranging from steering to cell phones. If, for example, hard braking causes a pitch or roll of the vehicle, StabiliTrak's sensors recognize it, then share their data with the suspension system. As a result, the road sensing suspension can firm up one corner of the vehicle to compensate for the pitch or roll. Similarly, StabiliTrak's yaw sensors can share data with speed-sensitive steering. Then, when the car yaws, the steering system can change the turning effort required by the driver.
| Seville’s Class II databus allows for data sharing between 16 modules (18, if optional OnStar and Navigation Modules are included). To put things into perspective, if the databus hadn’t been available, each door would have needed 60 wires, rather than nine. Overall, 65% of the wiring was eliminated.
Data-sharing even works with General Motors' OnStar system. Linking the GPS-telephone with the databus enables OnStar agents in another part of the country to access the doorlocks. Hence, when an owner in Albuquerque calls the OnStar center to say he's been locked out, an agent in Michigan can remotely unlock the doors. In the near future, GM engineers also say that the OnStar-databus link will enable remote computers to help diagnose car troubles.
Wiring reductions, too. As expected, GM engineers also reaped the traditional advantages of databus architecture--that is, they dramatically reduced wiring. They estimate that they cut wiring splices by 92%, reduced the number of wiring terminals by 25%, and decreased the average wire diameter by 15%. Overall, they say, they eliminated an astounding 65% of Seville's wiring mass.
Better yet, the databus enabled them to add features that couldn't otherwise have been installed. Seville's doors, for example, now employ only nine wires. Those nine wires carry conventional electrical current and data. They light the bulbs in the door-mounted switches, and carry information for the seats' memory functions, windows, and mirrors.
But if the databus hadn't been available, each door would have needed 60 wires, rather than nine. "You simply cannot build a car by running 60 wires into the door," Horvath says. "It's not achievable in the assembly plant because it increases the opportunity for manufacturing error. Without the databus, there would be no way to incorporate all these features in the door."
In that sense, Cadillac's Class II databus is not much different than the ones used in countless automated factories and giant aircraft. There, databuses have made their mark through wiring reduction. But the ability to tie features together through data sharing should make databuses more attractive to engineers in other industries. "Bus networks that have been previously used in factory settings are finding their ways into other areas," notes Steve Glaze, program director for Micro Switch Division of Honeywell, the Freeport, IL-based maker of the Smart Distributed System. CAN-based databuses, Glaze says, have recently been used in off-road vehicles, dock loaders, and train brakes.
In the future, GM engineers say that the databus could be an enabler for countless algorithms that will link seemingly unrelated parts of the car. One day, they say, they can imagine linking a vehicle's rain-sensing wipers to its brakes or traction control, in order to tell those systems that it's raining outside. They predict that such connections will grow rapidly in the next five years. "Now that we have the databus in place, we have to ask ourselves: 'What problems can we solve today that we couldn't solve before?'" Badenoch says. "A lot of the solutions haven't even been dreamed of yet."
What this means to you
Databases can be applied to vehicles, both off-road and on-road
Using databuses, vehicle designers can add features without adding hardware
Photoelectric sensor shines in harsh environments
Eden Prairie, MN--Extreme cold, heat, moisture, fog, dirt, bright sunlight and salt spray: Apply all of those to a conventional photoelectric sensor, and you're likely to get poor performance.
But a new photoelectric sensor from Sick Optic-Electronic, Inc. successfully faces all of those conditions in a new "touchless" automatic car wash. The sensor, part of a family of devices known as the W.2000 series, moves a spray nozzle inside the touchless car wash. In that environment, it must not only stand up to dirt and moisture, but must also "see" through shrouds of fog. And it must distinguish between its own light and bright sunlight.
The W.2000 sensor family accomplishes all that through a number of design innovations. Primary among those is a new optical design that employs vertical orientation of the sender and receiver lenses. By orienting the lenses vertically, rather than in conventional horizontal fashion, Sick Optic engineers say the new sensor offers greater range and consistency. "Since the lens orientation is parallel to the plane of the sensed object, you get a more consistent switch point," notes David Lagerstrom, division manager for industrial sensors at Sick Optic.
In the car wash application, one of the most difficult tasks for the sensor is to pick out incoming light during the winter, when cold outside air creates a foggy shroud inside. To deal with such situations, the sensor employs a special Application Specific Integrated Circuit (ASIC) designed by Sick Optic's engineers. The analog section of the ASIC incorporates a special pre-amplifier that analyzes incoming light and provides a better switch point. The result, Lagerstrom says, is a higher "excess gain"--a value commonly used to determine a sensor's capabilities in difficult environments. "With the longer range and the higher excess gain, we have the ability to punch through the fog in the air," he says.
The digital portion of the W.2000's ASIC also provides the sensor with the ability to deal with another condition of the car wash: bright sunlight. Special digital circuitry gives the sensor greater immunity to high frequency light and defeats potential sensor confusion.
To enhance the ability to deal with dirt and moisture, Sick Optic engineers ultrasonically welded the optical window to the sensor body and provided a special silicone rubber gasket for the top of the enclosure.
With its longer range, high excess gain, and enhanced sealing, engineers say the sensor is ideal for harsh environments.
Record-setting electric cars go farther and faster
by Mark A. Gottschalk, Western Technical Editor
Seal Beach, CA--The recent achievements of two electric cars might help transform the image of electric vehicles as inherently slow and short-ranged into one of speed and endurance. One car is taking a shot at the electric-vehicle Land Speed Record (LSR)--currently 183 mph--while the other eclipsed the world distance record of 325 miles by traveling 1,043 miles on a single charge.
World Record Performance Associates of Santa Ana, CA, commissioned the design of the 2,450-lb LSR vehicle called White Lightning. Its chassis and carbon-fiber bodywork were developed by the Arivett Brothers (San Bernardino, CA). With a Cd of 0.13, White Lightning is expected to top 300 mph and had reached 237 mph as of press time.
Amazingly, the drive train consists of a pair of stock, commercially available electric-car motors and controllers supplied by AC Propulsion (San Dimas, CA). Called the AC-150, each contains an AC induction motor rated at 200 hp from 6,000-12,000 rpm and 165 ft-lb of torque from 0-5,000 rpm.
They were developed by AC Propulsion's founder, Alan Cocconi, who also designed the controller for the original GM Impact that evolved into today's Saturn EV-1. "Induction motors are well suited for electric vehicles," he says, "because you want high power, yet for freeway cruise you want high efficiency at low power and high rpm."
For the record runs the motors will draw 1,000 amps at 300 volts for three minutes. This power is supplied by a battery system consisting of 6,120 small NiCd batteries, weighing 55-gm each, which are normally sold for use in radio-controlled cars. Masterstroke, a development company in Montebello, CA, assembled the batteries into groups of 306 cells all soldered together inside 20 fiberglass tubes. "They were the only rechargeable batteries we could find that could be discharged fully in 90 seconds," Cocconi says.
In addition to the White Lightning, the AC-150 drive train has been installed in everything from buses in Africa to Honda Civics. Cocconi's latest invention, a 2,400-lb roadster called the Tzero, uses a 220-hp version to bolt from 0-60 mph in 4.9 seconds.
White Lightning won't be chasing the record alone. Donald Wales, grandson of land-speed legend Sir Malcolm Campbell, recently unveiled another LSRelectric car, Bluebird.
Keeps going and going… At the other end of the spectrum is B.A.T. International's (Burbank, CA) AT-1. Fitted with Zoxy Blitz® zinc-air batteries from Kummerow Corporation (Burbank, CA), the vehicle ran for 1,043 miles at 20-25 mph.
The batteries consist of 180 zinc-air cells weighing 4-lb 1-oz and outputting 1.1 volts each. On the record run they delivered 76 kW-hr of power with 10% remaining, compared to the Saturn EV-1's total capacity of 16.2 kW-hr from a lead-acid battery pack which weighs 50% more than that in the AT-1. A unique aspect of Kummerow's batteries is that they are intended to be physically swapped from the vehicle in minutes instead of recharged onboard for hours.
Intended for emerging-market countries, the AT-1 is a five-passenger utility vehicle that can be configured in a variety of ways. Weighing 1,400-lbs, it can carry a 1,300-lb load. Depending on motor, it has a top speed of 40-70 mph.
While zinc-air batteries offer about four times the energy density of lead acid, their weakness is specific power. Hans Kummerow, the company's president, notes that this could be addressed by combining zinc-air batteries with other batteries or an ultra-capacitor. "As vehicles based on crude oil lose their selling position," he says, "I expect zinc-air fuel cells to capture a substantial part of the market."
The supply-chain goes online
Dearborn, MI--Ford Motor Co. is reaching out beyond the confines of its internal engineering organization to make suppliers an integral part of the design and manufacturing process. The idea is to get key component-makers on-board Ford's global computer integration initiative, dubbed C3P.
According to Richard Riff, manager of Ford's C3P Project Office, the objective is to unify CAD, CAM, CAE, and product information management (PIM) so as to make them effectively a single application from the user's point of view. "There is no reason to differentiate design, analysis, manufacturing, and information management," Riff says. "All of these activities are necessary to create a product."
Structural Dynamics Research Corp. (SDRC), Milford, OH, supplied the software components of the C3P architecture. The company's I-DEAS Master Series CAD/CAM software provides design and manufacturing functions. Ford is also using the FEA solver integrated into I-DEAS for engineering analysis.
Perhaps the most important element of C3P is SDRC's Metaphase PIM system. Riff describes Metaphase as the "bus" through which engineering and product information is passed throughout Ford and exchanged with suppliers. "This is the enabling technology that gets our suppliers involved," Riff says. "We are asking all our full-service suppliers to get on-board with C3P."
Supplier involvement is seen as essential because many suppliers have test, analysis, and manufacturing responsibilities for the components they design. The engineering data produced from these activities can, through C3P, be assessed by other suppliers and engineers at Ford in the context of the assembly.
To date, more than 3,000 Ford employees worldwide have received C3P training. More than 800 companies supplying Ford have purchased I-DEAS Masters Series and Metaphase Series 2.3 software to support the C3P effort.
WE ARE HERE!!!Stress analysis assists cancer detection
Houston--Researchers at the University of Texas Medical School (UTMS), the University of Kansas Medical Center (UKMC), and Ecole Polytechnique in Montreal are developing a new type of medical imaging called elastography. The process uses ultrasound and other technologies to create an image that describes the elastic properties of certain tissues.
Applied to breast cancer, an elastogram shows how a lesion looks when pressure is applied. Malignant tumors are 10 to 100 times stiffer than other breast tissue, and thus move in a different manor than benign tissue.
To create an elastogram, two ultrasound images of breast tissue are taken: one of the tissue in its regular state, one in a compressed state. These images are compared by signal processing, which determines how the tissue moved and then converts these signals into an elastogram.
"For this method to be effective, we need to test a variety of scenarios," says Dr. Michael Insana, researcher at UKMC. "Algor software (Algor Inc., Pittsburgh, PA) enables us to create samples that represent the gamut of possible tissue arrangements."
For each hypothetical placement of tissues, Insana uses Algor to create a computer model of the tissue in its normal state. He then compresses the tissue 1%. By applying mathematical formulas that represent movement in the real world, Algor's linear stress analysis software tests this compressed model to see how the tissue moves.
"The resulting analysis image indicates if a particular tissue arrangement is difficult to detect," says Insana. "If that's the case, we perform a real life test using gelatinous materials that imitate the various lesions and breast tissues."
Setting up a test takes about one week to create the gelatinous form to the team's exact specifications, explains Insana. "Considering it only takes Algor software about a half hour to run an analysis, it's pretty clear how much time and money we're saving by using it."
"We also know we can depend on Algor since we've found it to be very accurate," says Insana. "Our real-life tests of gelatinous materials have been in close agreement with the Algor analyses."
WE ARE HERE!!!Bypass surgery explores new dimensions
Linköping, Sweden--This fall the University Hospital of Linköping will perform coronary bypass surgery on a human for the first time in medical history. Obviously, a procedure of this intensity requires a great deal of preparation. To this effect, the medical team is collaborating with Proslavia Clarus AB, Gothenburg, Sweden, on the development of Virtual Cardiothoracic Endosurgery (VCE), an endoscopic surgical simulation tool.
The set-up for the simulator mimics the real-world scenario of this medical procedure. In one hand users hold a camera, which serves as a force feedback system. When surgeons move the camera, it sends signals to the computer to update the image. Additionally, they will "feel" the different objects and body structures as the camera collides with them. The other hand operates the tools necessary for practicing the procedure.
Although most of the procedure is conducted by looking at a 2-D screen, advanced processes such as sewing require stereo vision. A head-mounted display (HMD), supplied by a unit of Kaiser Electro Optics, meets this requirement by providing 3-D vision.
The VCE application runs on a Silicon Graphics Onyx workstation. "Simulating medical procedures is a complex task. We think the power of Onyx offers the best technology," says Frederick Gustafson, executive VP at Proslavia Clarus AB.
The software for VCE utilizes Proslavia Clarus AE's development toolkit, Oxygen. This toolkit consists of three basic parts: the Oxygen Base; Oxygen Technology plug-ins for specific technologies such as CAD data import, virtual-reality peripherals, Newtonian physics, and real-time collision detection; and ready-to-use Oxygen applications including Assembly, Showroom, Machine, Man Machine Interface (MMI), and Medical.
"Since all of Clarus AE's medical applications are based on our Oxygen software package, these simulators will soon run on platforms other than SGI," says Gustafson.
"For the first time in history it will be possible to perform minimally invasive cardiothoracic surgery in a simulator. Surgeons will have the ability to enhance their skills at any time in a fully realistic environment, without a real patient," says Ulf Hermansson, M.D. at the University Hospital of Linköping.
The next step is to train new surgeons on the different techniques. These skills include the entire procedure, as well as just sewing, cutting, or the handling of different instruments. Any procedure within the heart or lung area can be practiced. When the system is finished, it will be simple to add or change the procedures. In addition, follow up with medical procedures is simple since sessions are recordable.
"Because endoscopic procedures are extremely demanding, training tools are urgently needed," says Stefan Hallin, president of Proslavia Clarus. "Proslavia Clarus understands the importance of making endoscopic surgical procedures available to medical communities globally, rather than only a few experts in the field."
Developers will preview the VCE simulation system at the 11th annual meeting of the European Association for Cardiothoracic Surgery, scheduled for this fall in Copenhagen. Release is set for the spring of 1998.
Sensors help safeguard against potential hazards
Hingham, MA--Telescada Inc. recently developed a remote terminal unit (RTU) that enables utility companies to monitor the regulation and flow of gas from the main pipeline to a house or commercial building. The goal was to minimize operating costs while improving system performance with accurate real-time data.
In the past, monitoring natural gas pressure was an inefficient and labor-intensive process. Mechanical paper chart recorders installed at regulator vault locations recorded inlet and outlet pressures of the gas. (A change in pressure can demonstrate a potential problem with the regulator station.) Every seven days technicians retrieved the recorded results and replaced the paper chart. Engineering departments later reviewed the charts for discrepancies. However, due to the high number of vaults on a given system that need to be reviewed, engineers typically studied results at a much later date and often identified problems after the fact.
"Too much time and labor was spent retrieving the information rather than reviewing results. A remote system solution was a necessary change for the better," notes Dennis Mantia, sales manager at Telescada.
When designing its automated system, Telescada needed to determine the most accurate way to obtain pressure readings. The company wanted pressure sensors that provided minimum power consumption and excitation voltage, 0 to 5V dc full-scale output, ease of calibration, repeatability, high accuracy, and stability. Model 209 pressure sensors from Setra Systems Inc. (Boxborough, MA) met those criteria and came in a variety of off-the-shelf ranges Telescada required. The compact pressure transducers also maintain their accuracy in the face of mechanical shock and vibration, thermal shock, RFI, corrosive media, and other extremes common in industrial environments.
"One strong feature of the 209 Series is that the output device is electrically identical, regardless of the selected range. This enabled us to manufacture a single analog input module for all required pressure ranges. In addition, Setra provided a low-pressure sensor that maintained temperature stability and accuracy over range for very small pressure readings--as low as 0-1 psi (0-6.89 kPa). And Setra was willing to work with us to ensure the results we needed," says Mantia.
The remote monitoring unit provides an early warning for system failures and potential danger. A user dials into the RTU and downloads the recorded information on demand, a process which takes only minutes per week rather than hours per day. Another benefit: Immediate action can be taken if fluctuations in pressure or gas flow are detected because the remote unit automatically calls in alarm conditions as they occur.
Each unit has between one and four sensors for monitoring gas pressure. One sensor monitors inlet pressure with varied pressure ranges from 0-100 to 0-1,000 psi. Another monitors outlet pressures, which can have readings as low as 0.25 psi or 1.72 kPa and up. "The Model 209 plays a crucial role in this system," says Mantia. "We recognize the important function of our monitoring device and want to ensure its proper performance. Setra's 209 guarantees us we are providing the best system possible."
Bearing is forgiving in rugged applications
Livermore Falls, ME--When the floor began to vibrate severely at Wausau Paper's Maine-based papermill, maintenance workers knew it couldn't be an earthquake in this area of low seismic-risk. Ultimately identifying a 10 year-old drying cylinder as the source, maintenance supervisor Larry Castonguay suspected that the guilty party might be a settling of the floor in the turn-of-the-century papermill. "Even a slight angular misalignment will change the load distribution on the cylinder's non-locating bearing," says Castonguay.
Depending on the bearing arrangement, a minor axial displacement could cause similar stresses. But the bearing in question was not axially restricted, in order to accommodate a 0.39 inch expansion of the shaft during high temperature operation.
Opening the cylindrical roller bearing on the front (non-drive) side of the drying can, maintenance workers were stunned to see that the lock nut was broken and that the inner raceway had worked itself nearly halfway out of the bearing assembly.
Although today's bearings are designed to have extremely long service lives, they rarely fail due to material fatigue--as the case at Wausau Paper illustrates. In addition to culprits like moisture, vibration, contamination, and lubrication distress, the force imbalance resulting from angular misalignment or axial displacement can send a bearing into premature retirement. A bearing on the verge of failure can wreak havoc on the shaft as well.
The buildup of stresses on a bearing is particularly problematic in applications involving high operating temperatures, heavy loads, settling floors, and long shafts--just like the drying cylinder at Wausau Paper. Until recently, designers have been able to address only part of the problem.
Standard cylindrical and needle bearings can accommodate axial displacement (up to approximately 8% of the bearing width), but cannot tolerate much misalignment without special enhancements. And while spherical roller bearings can manage misalignment and displacement, overcoming the latter also requires special engineering of the housing and adjacent components.
Wausau Paper's solution, however, was a new type of self-aligning bearing (SKF USA Inc.) designed to accommodate both. This is the first U.S. application of the bearing, called CARBTM (for Compact Aligning Roller Bearing), which was introduced in Europe two years ago. The design itself is based on a systematic way of equilibrating the force imbalance caused by a misalignment or displacement. CARBTM utilizes a long, barrel-shaped roller with a radius of curvature slightly less than that of the outer ring. When a misalignment or displacement occurs, the rollers move freely to the location at which the forces are once again in balance. An inner raceway serves to maintain correct spacing between rollers. This degree of freedom sets the bearing apart from any other designs.
Maintenance workers at Wausau are satisfied with the performance of the new bearing. In fact, they are closely monitoring the vibration levels on all 35 drying cans, planning to install CARBTM bearings as soon as the existing bearings show signs of wear. Reception has also been good elsewhere; several motor manufacturers and a maker of generators are currently conducting field tests. Future tests are scheduled for continuous casters, industrial fans, and planetary gearboxes.
Now if bearing makers could only solve another nagging problem: Although CARBTM handles high loads and accommodates misalignment and displacement, it is not designed to carry an axial load--precluding its use as a locating bearing.
Plastic pulleys reduce inertia for servo system
by Charles J. Murray, Senior Regional Editor
Fairport Harbor, OH--In most power transmission applications, component cost is critical. It's even more important in servomotor applications, where the inherent cost of servo technology can become a large percentage of an overall machine.
For engineers from Mexico Plastics, Mexico, MO, servomotor cost even represented a make-or-break situation in a bag-cutting application. The company, which makes plastic products, needed the precision of a servomotor to accurately cut plastic bags. Problem was, use of a large servomotor was not deemed cost effective. "We needed a servomotor," notes Jay Johansen, plant engineer for Mexico Plastics, "but if we used one that was too large, it could have doubled the overall cost of the project."
To obtain the necessary precision, and still keep costs down, the company's engineers minimized the size of the servomotor by reducing the inertia of the system. The solution: lightweight pulleys.
The system, which feeds plastic sheet and then cuts it into bags, employs a 2.5-inch-diameter pulley on the servomotor shaft and a 7-inch pulley on the feed roller. The pulleys are connected by a timing belt. A 7-inch steel pulley for the application weighed approximately 10 lbs, Johansen says. But by switching to a plastic-and-steel pulley, the weight dropped to about a pound.
The pulleys, designed and manufactured by Torque Transmission, a division of Rampe Manufacturing, Fairport, OH, employ an outer plastic rim, and a steel hub, web, and locking collar. The locking collar behaves much like a machine tool collet, locking the hub down on the shaft so it doesn't slip during the fast start-stop action of the bag-cutting application.
The locking collar is critical for such applications, says Torque Transmission engineer Doug Gryczan, because a keyway-type system would be prone to widening under start-stop loading. "When you clamp down with this locking collar, you are grabbing the shaft around all 360 degrees, as opposed to just a few selected locations," Gryczan says.
By employing the plastic pulleys, Mexico Plastics could use a mid-sized servomotor to keep costs in check. The servomotor, in turn, provided the company with better cutting precision than the previous clutch-brake system. "The pulleys are an important component of the low inertia design of the entire system," Johansen adds. "Overall, we have increased productivity on each machine by about 35%. We also significantly reduced our downtime and maintenance costs."
TI puts networking silicon to the test
by Julie Anne Schofield, Senior Editor
Dallas--Texas Instruments' Network Technology Center (NTC) is home to millions of dollars of networking equipment, more than 10 miles of cable, and 4,000 to 5,000 connect-ion points. Its purpose: to develop and test TI network silicon chips and to test OEM networking equipment in a real-world environment. The result: reduced cycle time for both chips and equipment and a much smaller chance of failures in the field.
The key word at the NTC is interoperability--testing to make sure that TI's silicon works across the board, in every possible networking configuration.
"Customers have an implicit expectation that equipment will work the first time no matter what their system configuration," says NTC Director David Brenner.
There are literally millions of possible equipment combinations. "The more equipment we can expose our designs to, the better," says Brenner. In fact the lab is connected to TI's internal data network for testing in a "real world" environment.
The NTC tests new chip designs by using hardware emulation and its network systems test capability. This lets silicon designers test a functional model of a chip before committing to silicon. Modeling new designs this way allows TI to reduce development costs and quickly bring new products and technologies to market.
After silicon prototypes of a device have been manufactured, the NTC often serves as the alpha test site. TI extensively tests new devices in a variety of network environments--multiple vendors' 10Base-T, switched Ethernet, 100Base-T, 100VG-Any LAN, token ring, and ATM equipment--to verify standards conformance, function, and interoperability. "We want to catch flaws before the chips leave TI," says Brenner.
Customers who use TI silicon can use the data from the NTC's extensive characterization and interoperability test database to support testing requirements for their end equipment. This, in turn, decreases the time needed to take their products to market.
The third major task of the NTC is testing reference designs that application teams create for networking chips. Such designs let new entrants who don't have the experience to design networking equipment from scratch get to market fast.
Take the tour. Walking through the NTC is a dream come true for networking aficionados--or a living nightmare for IS managers. Modeled on a star topology, the network has equipment from all major vendors, as well as some offbeat or defunct ones. "If companies are using it, we're testing it," Brenner says. Once the lab was established, he says, companies were more willing to donate new equipment. Ratio of what is donated to what the NTC buys is approaching 50:50.
Hardware includes PC clients, servers, and RISC machines from Compaq, Hewlett-Packard, ACER, Dell, IBM, Micron, Zeos, AST, NEC, Digital Equipment Corp., Gateway, and Apple. The "server farm" has close to 100 different machines. Servers plug into a video/keyboard switching unit, so eight or so units can share one monitor and keyboard, thus cutting down on cables. Server software varies from Netware, varieties of Windows, UNIX, OS/2, and AppleTalk for Macs.
Connecting the hardware is Type 1 shielded twisted-pair copper cable, voice- (Category 3) and data-grade (Category 5) unshielded twisted-pair copper cable, and single- and multi-mode fiber-optic cable.
For proper testing, cables can't be rolled up, so one weekend workers ripped up the floor of one room and laid cables underneath. The maximum copper cable length for a valid configuration is 100m, but TI's lab also has 120 and 200m cables lying under the floor. Not everyone follows the rules, and the lab has to take that into account.
"People expect equipment to work in a network--even one that doesn't meet standards or has rats running around the wiring closets--the first time," stresses Brenner. "If it doesn't, they'll buy something else."
Self-lubricating bearing eliminates need for maintenance
by Charles J. Murray, Senior Regional Editor
Downers Grove, IL--In some applications, there's no easy way to grease a bearing. On large excavators, for example, maintenance personnel must often climb ladders to reach bearings. Or they may install complex lubrication systems, complete with pumps and tubing lines, to deliver a steady stream of lubricants.
Now there's a better way. Engineers at Rexnord Corp. have developed a self-lubricating, composite spherical bearing. The bearing, which employs a filament-wound ball within a fractured steel outer race, is said to be ideal for use on hydraulic cylinder devices. As a result, it could see use on construction equipment and agricultural vehicles. The bearing could also serve on conveyors, packaging machinery, scissors lifts, and a host of other applications.
Rexnord engineers say that the composite ball can be retrofitted into existing applications using a special tool. "Size for size, it's the same as a steel bearing," notes Bernard Harris, manager, engineering and marketing for composite bearings at Rexnord. "So people who use a lubricated metal bearing can now have a self-lubricating bearing without major changes."
Key to the design is the spherical, filament-wound Duralon ball. During use, the ball's convex outer surface mates with the steel race's concave inner surface, enbaling it to rotate or move axially within the inner race. This, along with its self-lubricating quality, enables the new bearing to handle installation misalignment and oscillatory loads.
Up to now, the vast majority of such bearings have employed steel-on-steel construction. The reason: Composites couldn't handle large loads. Rexnord, however, solved that problem by helically winding the ball with glass fibers along its long axis. Fibers are wound over the resin-saturated Duralon fabric, then machined off in an operation that gives the ball its spherical shape.
| Static tests of the composite bearing showed an ultimate strength of 108,000 pounds.
That process also endows the ball with surprising strength. Tests done by engineers at Rexnord Technical Services, Milwaukee, WI, showed that the bearing has a compressive ultimate load capacity of 108,000 pounds. In dynamic loading, it features a capacity of more than 30,000 psi. In contrast, steel-on-steel spherical bearings typically begin losing lubricant at about 8,000 psi, Harris says. Beyond that, metal-to-metal wear and galling can occur, ultimately followed by seizing of the bearing.
To employ the new bearing, users must make one special accommodation, Harris says. Mating shaft surfaces must be made from ground and polished bar stock or tubing. Rexnord engineers recommend chrome plating, electroless nickel coatings, or nitrocarbrized steel.
By eliminating the need for special lubricating pumps and plumbing lines, the self-lubricating bearing also disposes of a potential failure mode. "In construction, it's not unusual for a plumbing line to be damaged," Harris notes. "And when that happens, the user has bigger problems because then the bearing fails, too."
Although the bearing's initial cost is slightly more than steel-on-steel designs, it compensates for that cost across its useful life. Says Harris: "The goal of this design is to eliminate the need for maintenance."
Astute engineering sticks new Volvo to unimproved roads
by Rick DeMeis, Associate Editor
Girdwood, AK--Short of another gas crisis, the plague of vision-blocking so-called sport-utility vehicles (SUVs) continues. Volvo, looking to intelligently tap into this market, has developed the 1998 V70 XC (cross country) all-wheel-drive (AWD) vehicle.
Referring to the sometimes tricky handling and instability of these traditionally truck-based vehicles, "SUV characteristics may not be best for safety," says Peter Boisen, Volvo's business project manager. In approaching this market, Volvo is using its front-wheel-drive V70 wagon plat-form to give customers a car that is "safe and as functional as in the past but with features desired in the SUV class," says Boisen. "It can outhandle any SUV and has wagon utility," adds Helge Alten, president of Volvo Cars North America, with the fuel consumption and emissions of a car.
The AWD is key to the XC. It features: Volvo's standard front-wheel traction control, at speeds below 25 mph (for getting underway in mud or snow); a bevel-gear transfer differential to power a rear axle similar to that in rear-wheel-drive Volvo S90/V90 models; a viscous clutch from Viscodrive (Bruneck, Italy); and a locking rear-wheel differential to limit slip under 25 mph.
The silicone-oil viscous coupling normally keeps 95% of engine power on the front wheel. It changes this distribution, completely reversing it if needed, compensating for front wheel spin "in milliseconds," according to Volvo.
Freewheelin'. A freewheel unit in the drivetrain, behind the viscous clutch, disengages the rear drive during braking, engine breaking, or coasting, for better stopping action. It allows using unmodified front-wheel-drive ABS. The freewheel also locks up in reverse for four-wheel-drive traction.
The simple AWD system, except for the standard traction control, is entirely mechanical. "Without [extra] electronics, it was easier to develop," according to Krister Broo, the designer and project manager for AWD systems. He notes, "The challenge was designing the system into an existing car," such as fitting the required viscous-clutch drivetrain into the underfloor tunnel.
A rear drivetrain also dictated changes in exhaust-system and fuel-tank layouts--the lowest underbody point now being the catalytic-converter shield rather than a subframe. But rear-suspension spacers for increasing clearance also gave more "maneuver room" for larger rear wheels.
Northern muse. A recent test drive in Alaska confirmed the AWD effectiveness--even on roads so rough that at least one car-rental company forbids its customers to travel on them! When power was applied, no slipping over the gravely, rut-and-bump surfaces occurred--whether on level ground or navigating the steep, turning grades of Hatcher Pass. No changes in power distribution to the wheels were evident, and suspension dynamics smoothed out the worst of each individual hole or rock.
The XC has 6.5 inches of ground clearance, 1 inch more than the V70 wagon. "Quite respectable," Volvo says, compared to most SUVs. Other utilitarian features include: strengthened roof rails; an electrical "power point" in the cargo area; and a flip-out cover for the rear bumper so it can serve as a bench without dirtying your clothes, something appreciated first-hand. A neat option is a roof-rack-mounted, streamlined, conformal cargo pod, with gull-wing doors, that comes in the car color. More goodies: a "link-frame" rack for single-person (albeit a tall or strong one) loading of bicycles, and a dog compartment with a fan for Fido when you're out of the car fishing--or shopping.
Racking brains. Seemingly simple, the XC roof rack, stronger than those on previous Volvos, required an interdisciplinary design effort, says Ake Bengtsson, XC project manager. Design aesthetics limited the overall cross section of the thicker, stronger transverse rails. Wind-tunnel tests, however, showed that symmetric, and even aerodynamic, shapes "sounded like an organ at highway speeds," according to Bengtsson, because of the resonance from alternating vortices shed by the rails, the same way as wind-in-wires produces sound. Solution: a more square lower front edge reducing airspeed under the bars and a row of small, angled ridges along the top--like aircraft-wing vortex generators--spinning the flow streaming back, breaking up the vortices, and killing the noise. Drag is virtually unaffected, says Bengtsson.
| Viscous coupling distributes power proportioned between front and rear wheels as required by driving conditions while a locking axle limits individual rear wheel slippage.
Other impressions from two days and well over 300 miles in the state with perhaps the most rugged roads to match its scenery: This car can get you anywhere you want to go provided there is a semblance of a road. Volvo is quick to point out that the XC is not an off-road vehicle but one for unimproved tracks--as another journalist found out in committing a cardinal sin, even for an off-road vehicle: trying to turn in a rock-filled stream, and hanging-up the car! On this trek, in addition to "pushing" the XC, fuel economy respectfully averaged between 20 and 21 mpg.
In one respect, Volvo may be playing too much to the "American market" by offering only a four-speed automatic transmission and not the five-speed manual available elsewhere. The engine is a 2.5-liter, 193-hp, five-cylinder light-pressure turbo. Two well appreciated features on the XC are the separate driver and passenger air-temperature controls and a killer surround-sound system. Now, if only the instrument panel was a darker gray or black for better daytime readability.