Linear servo motors are helping motion system designers keep pace with industry demands
Form, fit, and function, the engineering mantra made public in the 1990s, is usually applied in reference to consumer goods such as cars and cell phones. But the concept also holds in the design of motion systems, particularly with respect to linear actuators.
The use of servo control has, in a sense, spoiled industry. Now, everyone expects the same sort of improvements along the “linear axis” that they’ve become accustomed to on the “rotary axis.” Never mind that the available envelope of space keeps shrinking. If you’re a “good engineer,” you’ll find a way to actuate that big, clumsy load more quickly and precisely than before – so they say.
It is precisely for this reason that researchers are delving ever deeper into the realm of electromagnetics, electronics, and control dynamics, hoping to nudge the technological envelope just enough to satisfy the demands of industry one more day. So far, one of the more promising developments is a device in the form of an electric motor, but with the fit and function of a pneumatic or hydraulic cylinder.
To see one of these mechanical specimens in the lab, you’d think you were looking at an ordinary electric motor with, perhaps, a somewhat fat rotor. But when the rotor fails to turn and instead begins to extend smoothly along its axis, your instincts are likely to take over, telling you there’s a pressure cavity somewhere inside filling with fluid. Most people are inclined to believe the mechanism is hydraulic until they see the actuator run full bore, cycling back and forth at more than 100 in./sec.
It’s that kind of dynamic response that makes electromagnetic actuation a natural for linear motion. It’s also the reason why researchers have spent years experimenting with rotary motors, manipulating copper, iron, and magnets like pieces of a Rubik’s cube. Retracing the twists and turns from rotary motor to linear actuator is, in fact, the best way to get a handle on the construction and operation of today’s moving-magnet, or tubular, linear motors.
Imagine starting with a brushless dc motor. After removing the rotor, you open it up and lay it flat. (If you stopped here, you’d essentially have a conventional linear motor, but you’re not done yet.) Next, you wrap the sides together – which were originally at opposite ends of the rotor – curling the assembly into a tube. The permanent magnets, initially oriented along the length of the rotor, now form a stack of rings of alternating magnetic polarity.
In a tubular motor, the magnetic shaft rides on an integral bearing system, two sliding-contact bearings incorporated in the endbells. The shaft is suspended in, and passes through, a column of currentcarrying coils held in slots along an iron core. A dozen or more coils separated by poles or teeth are arranged by phase (A, B, and C) in repeated succession down the length of the stator. The electromagnetic circuit thus formed bathes the entire surface area of the shaft in magnetic flux, achieving maximum linear force per unit of volume.
Naturally, the more powerful the magnets – neodymium-iron-boron is the standard today – the more force the actuator can produce. Force is also a function of the length and diameter of the motor. Usually, it’s easier and more economical to change the length of the motor, rather than the diameter. In fact, in the lab as well as in the field, motors are frequently assembled in pairs, end-to-end, to double force output.
There’s another advantage in increasing the length of the motor. The longer the shaft, the longer the stroke. Although any length is feasible, a 12-in. stroke is usually plenty for most applications.
New technologies like tubular linear motors typically emerge to fill a gap or shortcoming in existing art. In this case, the gap (in linear actuation technology) is fairly sizable.
The fitness of a linear actuator is primarily a matter of accuracy, force, footprint, programmability, environmental considerations, and of course, cost. There are other considerations as well; but the point is, a linear actuator must satisfy several (often-conflicting) requirements. It’s no surprise that so many actuation technologies exist today – hydraulic, pneumatic, mechanical, electromagnetic – and it’s a safe bet it will be that way for years to come.
Continue on page 2
Hydraulic actuators historically have been the workhorse of industry. Their appeal is their ability to deliver high force from a small cylinder, a nearly ideal component in terms of fit and function. Hydraulic linear actuators are also durable and can withstand high shock loads and trying environments. These strengths are tempered, however, by the requirement for external power supplies that pump environmentally unfriendly oil through networks of valves and tubing susceptible to leakage. Hydraulic actuators also tend to be a bit sluggish and they’re difficult to program should an application call for a particular motion profile.
Another linear actuation technology used throughout industry is pneumatics. Fueled by the expansion of factory floor automation and the demands for increased production, pneumatic technology is booming, and for good reason. Pneumatic systems are generally rugged, durable, and compact – at least at the point of actuation. But the technology does have limitations in terms of speed, programmability, and maintenance.
Mechanical solutions provide yet another route for linear motion. Compared to fluid power devices, mechanisms such as ballscrews, cams, belts, and pulleys are quieter, more efficient, and at least an order of magnitude more accurate. They also withstand higher temperatures and are more environmentally friendly.
On the down side, mechanical devices often incorporate many parts that rub against each other, making them subject to wear. In ballscrews, for example, wear is an enemy of precision. Not only does it increase play and (usually) backlash, it can clutter the ballscrew track with contaminants that impede the balls, causing jerky motion.
The combination of wear and a large number of parts usually adds up to a higher failure rate. This in turn means higher maintenance costs and increased downtime. Shock loads, particularly in drives coupled by linkages and gears, can be equally devastating.
Lay it on the line
Linear motors, the new kids on the block, come in several shapes. Some are tubular, and some are flat. Although both forms lend themselves to a programmable solution offering an optimum blend of accuracy, resolution, and speed, some of the flat designs may be a little tougher to work with.
For one thing, many flat types require an external bearing system to support and position the moving member in relationship to the magnets. It’s not uncommon for the additional components – rails, slides, bushings, and so on – to cost as much as the motor. And that doesn’t include the associated design time.
Contained bearings – without which tubular motors wouldn’t even work – not only save space, but they also put the actuator closer to the load. What’s more, because thrust is in-line with the shaft, there are no cantilevered forces to deal with. This means longer bearing life.
Other concerns that crop up with flat motors often have to do with moving coils. Many motors require flexible trays to carry conduit to these coils. The weight of the tethers cannot always be ignored, as it may influence the dynamics of the system. In addition, because the leads are constantly flexed during motion, the connections may become stressed and the cable compromised.
Moving coils, suspended in space, are also difficult to cool because they typically have access to less surface area. Stationary coils, encased in steel, can dissipate heat over the entire surface of the motor, where cooling is more efficient. Keeping coils cool is the key to higher continuous force in linear motors and greater overload (high-force transient) capability. In general, the higher the heat capacity, the more force you can expect. This is why some people shun natural convection, resorting to forced-air or even water-cooling.
A lot of potentially good technology dies on the vine because of obstacles to implementation. Getting tubular motor technology into the field was no small challenge. It had to be integrated into a complete closed-loop programmable servo positioning system before designers would give it serious consideration.
Like any closed-loop system, a tubular motor requires a position sensor. As the information link between the stator phase current and shaft position, the sensor has to be precise and reliable. If the phase current is off-target even a little, the magnetic field in the stator-shaft gap won’t be right, nor will the resulting motion of the shaft.
Continue on page 3
When commanded to move, the controller calculates the required phase current and then monitors the actual shaft position, comparing it to the desired motion. In this way, the motor can be directed to any position within its stroke, following any velocity or acceleration profile. The controller makes sure that the current in the stator coils stays in synch with the position of the shaft.
Position as well as force and jerk are fully programmable through the servo controller. And, because there’s no backlash or compressibility to compromise accuracy, the shaft moves on command with no dead time or torque wind-up. Travel is smooth and controllable even at low speeds.
Installing the device is a matter of bolting up the end flanges on the stator and shaft. No additional supports or mechanical connections are necessary. That a magnetic field is the only thing between the moving and stationary parts means that the motor can survive high shock loads; in other words, the shaft can hit hard without damaging anything inside. The bearing is actually the only part that can wear out and is easily replaced.
Although the bearings are not actually exposed to the environment, they may see some contamination dragged in by the shaft. Bellows can be used to fully enclose the shaft in extremely dirty environments where bearing life could be compromised. Ordinarily, bearing life expectancy is a minimum of 100 million inches, which should get your machine anywhere it needs to go.
Put it on paper
Although tubular motor technology is still relatively new, it is presently at work in many applications around the globe. One area where movingmagnet linear motors have been particularly helpful is in paper mills.
At one mill, for example, engineers turned to the new motors to control a showerhead. Prior to the switch, a rotary dc motor was employed to drive a ballscrew through a planetary gearbox. A quadrature encoder closed the loop. Problems included frequent breakdowns and a lack of programmability.
The wet, high-pH environment of the mill makes life very rough for electromechanical components. Gearboxes and encoders seem to suffer the most. Mill engineers also believe that reliability problems are made worse by continual metal-tometal contact and the resulting wear.
A solution based on tubular motor technology has apparently solved the problem. Not only is it programmable, but its single moving part greatly reduces wear and extends life. Day in and day out, the motor delivers a continuous force of more than 500 lb, achieving the type of oscillation profile needed for the best shower coverage. Velocities range from 0.5 to 60 in./min and are smooth throughout.
Q. What’s the problem?
A. Need to accelerate and decelerate a 70-lb load at least 60 times per minute over a distance of 1/8 in. in 10 msec.
Q. What didn’t work?
A. Cams were too slow and took two weeks to modify. Ballscrews couldn’t handle rapid reversals.
Q. What did work?
A. Two tubular motors electronically geared to 0.005 in. reduced cycle time by 20%, handling 5 g’s while extending and retracting the 70-lb load.
Chuck Schultz is vice president of sales and marketing at California Linear Devices Inc., Carlsbad, Calif.