Choosing the best mechanical motion components.
Machine motion means exerting force - great or small - to accomplish a particular result. But this forced movement must be controlled and measured, often down to the fractional-inch level, to accomplish automated performance needs. How are these competing requirements resolved? An answer lies in applying precise motion-control specifications first, because this can impact the design and materials decisions that follow. For you to be in control of your production, measurement, or testing processes, your machinery must have proper control of its motions.
Specifying positioners with regard to performance involves a number of concepts or terms, most of which are familiar. Resolution, for example, is sometimes confused with accuracy, but it is actually the minimum increment of motion that can be realized with a mechanical positioner. It is one of the fundamental quantities that define positioner performance, and in this context it is related to mechanical ratios in drivetrain components, the friction characteristic of a mechanism, and the capabilities of the electronic position sensor if one is used in a system. Also applicable here is the concept of linear stage resolution, which measures the rotation of an input drive shaft required to execute the linear move of a positioner over a given distance; for example, a 1-degree turn of a certain positioner's shaft might correspond to a linear move of 0.00014 inches.
Repeatability measures the degree to which a positioning device can replicate a given move, or move sequence, over a number of attempts. It defines the fundamental requirement of motion devices: precision with consistency. But in practical terms, the reference is really to "repeatability dispersion," because a mechanical positioner commanded to make a series of identical movements, or positioning trials, generally will display a degree of variation. Some moves will be larger than the commanded move, and others will be smaller, however slight the error. A mean value can be calculated, and a histogram plot will show the dispersion of individual data points about the mean value in the shape of the familiar bell curve. The width of the dispersion is a measure of system repeatability.
To carry this understanding further, in a single-axis linear or rotary device, repeatability may be measured either unidirectionally or bidirectionally; unidirectional repeatability dispersion will always be better in positioning trials, often by a factor of more than two. Bidirectional repeatability, obviously, is the more rigorous specification. To achieve a high degree of bidirectional repeatability, it is necessary to minimize sources of backlash or lost motion, such as free play in the gear mesh of drivetrain components. Recognizing that even a sloppy mechanism might still have good unidirectional repeatability, the user or engineer might well consider whether the nature of the production application absolutely requires bidirectional repeatability, or whether unidirectional could be adequate.
Since accuracy is not synonymous with resolution (minimum move size), as mentioned, we can define accuracy in motion mechanics as a dynamic for describing - or, ideally, eliminating - positioning error. It addresses the difference between the commanded target move position and the position actually attained with the mechanism, as measured at a prescribed reference point.
The accuracy of a positioner is primarily attributable to the mechanical drive system (e.g., drive screws and gears) and any measuring device it incorporates, such as an optical encoder or resistance transducer. Consider that angular errors also influence the accuracy of a positioning mechanism. These errors are caused by deficiencies in the parallelism and coplanarity of bearing systems and of their supporting structure (or by limitations in assembly techniques).
It is essential to remember, therefore, that the first requirement in selecting positioning components for an application is to define the combination of resolution, repeatability, and accuracy needed to accomplish the task.
Several other concepts, such as straight-line accuracy, orthogonality, and concentricity, are also particularly important to a positioning device specifier. Straight-line accuracy is the difference between the theoretically perfect travel path (a straight horizontal line) and the actual travel path of the positioner, measured at the defined reference point.
Orthogonality describes the precision of the right-angle relationships in multiaxis assemblies of linear positioners. As a practical example, orthogonality to better than 15 arc-seconds in alignment of a multiaxis system is equivalent to 0.000073 inches over 1 inch of travel.
Finally, concentricity, as an attribute for rotary tables, is the difference between the ideal, perfectly circular travel path and the actual path described by a point moving on the rotating carriage as measured in the horizontal plane. It is largely determined by radial tolerances in carriage support bearings plus machining tolerances in the bearing mounting surfaces.
Knowing that some aids to precise analysis are available is useful in the optimizing process. One device, the laser interferometer, is the standard for evaluating linear-positioner performance expectations. Another metrology instrument, the autocollimator, checks for angular specification conformance in accuracy and repeatability as well as pitch, roll, and yaw.
Among the subcomponents (elements) of positioners, such as carriages and drive mechanisms, bearings have a critical role in carriage and shaft movements. Like other components, bearings vary considerably in price and performance, so it is useful to examine a number of bearing types for their features and limitations.
The simplest positioner bearing is the machine slide, characterized by a square or rectangular (box-shaped) interface; the dovetail slide, a type of bearing named for the shape of the carriage-base interface, has a wider inner edge. These bearings have no rolling elements, just a lubricant film between carriage and base. To maintain straight-line accuracy, the system is preloaded by means of a gib, a plate that supplies an adjustable force along the bearing with tension screws. Straight-line accuracy is typically 0.0005 inches per foot of travel.
Machine slides are very robust, featuring high load capacity. The load-bearing characteristics of a dovetail-slide mechanism are directly proportional to the surface contact area between the carriage and the base. For example, a slide with a carriage length of 10 inches might have a capacity of 200 to 1400 pounds in horizontal translation or 150 to 800 pounds in vertical translation. Such slides are insensitive to contamination, possess high tolerance of shock or impact loads, and are generally quite economical. Common applications are in computerized-numerical-control machining centers and punch-and-press machines.
The most economical precision bearing system is the nonrecirculating linear ball bearing. A common version consists of a row of ball elements captured in a retainer to maintain their spacing as the balls roll between four hardened rods, two attached to the positioner base and two attached to the carriage. Preloading to maintain straight-line accuracy in these bearing systems is applied variously by using a screw-tensioned gib or wedges, or by machining the base to provide an elastic force when assembled with the carriage. Because the ball elements are nonrecirculating, slides incorporating this bearing provide smooth, vibration-free motion with good straight-line accuracy, typically 1.25 X [10.sup.-7] inches per inch of travel.
Of all bearing types, the linear ball bearing is the least tolerant of impact and shock loads without potential damage, since the balls make point contact with the mating surface. Also, at extremes of travel, the bearing is subjected to cantilevered moment loading and a resultant degradation of straight-line accuracy.
Linear ball slides are used typically in light, compact, quiet-running devices such as manual and motorized microscope stages, machine-vision-based inspection systems, laboratory specimen positioning, and other large-area/short-travel precision alignment systems.
Higher performance is offered in nonrecirculating systems by the cross-roller bearing. As the name implies, these bearings consist of a line of cylindrical elements that are oriented so the axes of rotation for adjacent elements are offset 90 degrees. The linear cross roller, which also offers vibration-free motion, provides normal load capacity several times that of a ball bearing with equal-size elements because its cylindrical elements make line contact with the V-ways. Linear cross rollers also have a low coefficient of kinetic friction and positioners capable of a typical maximum speed of about 18 inches per second.
Because of the precision manufacturing required with this bearing, cross-roller slides are normally specified to significantly better straight-line accuracy than that of the ball bearing, typically better than 2 X [10.sup.-8] inches per inch of travel. They are somewhat more resistant to cage creep, and therefore are more tolerant of high-duty/high-acceleration moves. Cross rollers are used in high-precision semiconductor wafer manufacturing and testing, surface interferometry, and similar applications.
These nonrecirculating varieties are non-constant-support types of bearings; their carriages and bases are of equal length, but the linear bearing is subjected to moment loading as it moves away from the center. By contrast, recirculating bearing systems incorporate a continuous closed path through which the elements roll as the system moves. They provide constant and position-invariant support of the carriage, eliminating the end-cantilever effects of nonrecirculating designs.
The most commonly used recirculating bearings are the ball sleeve (bushing) type, which consists of a cylindrical sleeve containing ball elements arranged in a number of recirculating "races" within the sleeve. The self-centering bearing sleeves are held in pillow blocks attached to the carriage, with the bearing elements riding on a hardened-steel cylindrical rail attached to the positioner's base. These bearings are available from 1/4 inch to 4 inches in diameter, the most commonly used sizes falling between 1/2 inch and 2 inches. Offering straight-line accuracy of [+ or -]0.0002 inches per inch of travel and maximum speeds up to 120 inches per second, slides based on this type of bearing are well suited for industrial positioning - parts-transfer operations, material dispensing, robotic arms, and pick-and-place applications.
The highest-capability mechanical bearing system in general use is the linear motion guide, consisting of a block that contains ball elements arranged in several horizontal recirculation tracks. The balls ride in precision grooves ground along the length of a rectangular-cross-section hardened-steel rail. The semicircular or Gothic-arch profile of the ball grooves provides a surface contact area at the interface between ball and rail, and unlike the rolling element bearings previously described, it tolerates some degree of shock and impact without suffering any damage.
This square-rail motion guide bearing, useful in heavy-duty industrial applications, provides full carriage support for the entire travel. Typical straight-line accuracy of slides using this bearing is better than [+ or -]0.0001 inches per inch; suitable applications include high-precision material working, such as laser and water-jet cutting, as well as ultrasonic or X-ray inspection systems.
Linear roller guides are roller-bearing versions of the motion guide bearing and, like the ball-bearing type, are favored for high load capability - typically from 350-pound to 2,700-pound capacity per bearing truck.
While all of these mechanical bearings satisfy the great majority of applications, a limited range of scientific applications require a level of performance provided only by noncontact bearings, principally air bearings and magnetic bearings. Air bearings employ a high-pressure laminar airflow (50 to 100 pounds per square inch) to separate the carriage and base, usually through lift and guide nozzles, so as to control straightness and flatness. The air gap is small, usually about 0.0001 inches. Air bearings can support loads to a maximum of 200 pounds. They are well suited to semiconductor wafer fabrication and testing, and X-ray and electron-beam microlithography, as well as other high-accuracy applications where their performance can justify their higher cost.
Magnetic bearings employ the repulsive forces of opposed magnetic fields generated by electromagnets in the slider and base to levitate the carriage. They achieve straight-line accuracy to 0.00004 inches per foot through a control loop that employs capacitive proximity sensors in position feedback for controlling the amplifiers that generate the magnetic fields. These bearings can support loads in excess of 1000 pounds. Like air bearings, magnetic bearings have no moving parts to wear and can be used in vacuum conditions.
DRIVES FOR LINEAR MOTION
Drive-screw and nut combinations are the most commonly used mechanisms to generate linear motion in positioners. Screws are characterized by their thread lead, defined as the amount of nut advance provided by one complete rotation of the shaft. Pitch (as used in English units as opposed to metric units) is the number of turns of a screw required to provide a 1-inch displacement of a nut relative to the screw. In general terminology, pitch is reciprocally related to lead; a 5-pitch screw (5 TPI) is the same as a 0.2-inch lead.
Among several types of mechanisms, lead nuts, which are friction devices, slide along the threads as the screw rotates. Lead-screw/nut assemblies are most commonly used in nonrecirculating ball-bearing or cross-roller-bearing positioners.
Among other drive options, rolled-thread ball-screw assemblies offer significant advantages over lead-screw assemblies, including higher axial-load capacity, greater duty-cycle capacity, and linear translation speeds exceeding 50 inches per second at accelerations of more than 1 g.
Rolled-thread ball screws are relatively economical, and are frequently used in round-rail/ball-sleeve bushing positioners. Package-labeling operations are among their high-duty/low-precision positioning applications.
Ground-thread ball screws, on the other hand, provide accuracy comparable to high-precision lead screws but with much greater speed and acceleration capacity. While more expensive, they are preferred in positioners that incorporate high-precision cross-roller or linear-motion guide bearings. Ground ball screws may be undesirable in some particularly sensitive applications because, as with recirculating element bearings, motion of the balls within the nut can cause low-amplitude vibration. Like lead nuts, ball nuts must be preloaded for high bidirectional repeatability.
Several other mechanisms - notably tangential-drive systems, which include belts/pulleys and rack-and-pinion devices - offer practical alternatives to screw/nut mechanisms in providing linear motion. The term tangential drive refers to the fact that linear motion results from a tangent contact between a cylinder rotating in the direction of travel and a linear element.
The simplest belt-drive systems are single-pass designs consisting of a pair of pulleys supported by angular contact bearings and a high-strength reinforced belt connected to a translating carriage. Belt systems can deliver repeatability to [+ or -]0.0005 inches unidirectionally and to [+ or -]0.005 inches bidirectionally, plus positioning accuracy of [+ or -]0.001 to [+ or -]0.002 inches per inch of travel. Common applications of belt positioners include high-speed shuttling, pick-and-place parts transfer, and ink-jet printing.
A rack-and-pinion mechanism provides performance somewhat between belt and ball-screw devices. It consists of a precision-machined or -ground toothed shaft and a mating spur gear. Metal-alloy spur gears are used with metal or high-strength plastic racks. Racks are made in a wide range of linear pitches to mate with spur gears of various diameters and tooth count. Accuracy, which is determined by the precision of the teeth, is specified in a manner similar to screw threads - the most precise racks being accurate within 0.0001 to 0.00005 inches tooth to tooth, and 0.0003 to 0.002 inches per foot. Racks are usually made in standard lengths from 1 foot to 5 feet, with available spacer blocks to facilitate butting pieces together for greater lengths. Thus, they offer greater travel than belts, which tend to sag over extended lengths unless they are supported with idler/tensioner pulleys.
Finally, the specifier should be acquainted with actuators, drive mechanisms that are specifically designed to generate thrust. The most common thrust actuator is called a ball-screw actuator, and consists of a rolled-thread or ground-thread ball screw enclosed in a protective housing. When packaged with an integral stepper or servo-drive motor, these units are commonly known as electric cylinders. With high force output the essential feature, screw leads for these devices are typically less than 0.2 inches with screw diameters of 0.75 to 1.5 inches. Gear boxes may be incorporated to leverage the available motor torque further and thereby increase the thrust output.
The most economical ball-screw actuators, which lack internal load support bearings, only offer thrust capability. More expensive thrust-and-load actuator variations are equipped with a thrust-shaft support bearing system that provides capacity to tolerate off-axis loads and generates high axial force. Commercial units are available with stroke lengths up to 3 feet and maximum loads exceeding 200 pounds, with moment loads exceeding 700 inches per pound. Maximum linear speeds are approximately 50 inches per second. These devices are not intended for precision positioning.
Another low-precision motion system, the screw-driven carrier actuator, has either a rolled ball-screw drive or a belt-and-pulley drive. These actuators incorporate a low-friction slide bearing or a set of plastic wheels or rollers running in a narrow gauge track to support the load-carrying translation saddle. Also commercially available are belt-driven actuators that have maximum load capability of about 500 pounds and moment capacity of 1000 to 2000 inches per pound, with linear speeds of several hundred inches per second.
DRIVE SYSTEMS FOR ROTARY MOTION
The majority of motion axes in positioning systems are linear translators. Yet rotary-motion devices are essential to applications incorporating azimuth/elevation ("pan-and-tilt") systems, plus "wrists" in robotic arms and stabilized platform systems. Payload support bearings for these devices are, basically, circular versions of ball and cross-roller bearings.
The most common drive mechanism for rotary positioners is the worm-and-ring gear. The worm shaft is a steel screw, supported by radial and thrust bearings, whose threads engage the teeth of a larger-diameter gear. Worm-and-ring gears are normally available in ratios from 24:1 to 180:1.
The mechanical advantage of this system acts both to multiply motor torque by the gear ratio and to reduce the inertia reflected to the motor by the square of the ratio, at the expense of reduced rotary positioner speed. Thus, low-torque, low-inertia motors can be used with a wide range of loads.
Tangential drives are another solution to rotary motion. Tangential belt systems, like their linear-motion counterparts, are normally used in high-speed, low-accuracy applications. Ratios between 2:1 and 10:1 with efficiencies of about 90 percent are common in these systems. For high-speed, high-accuracy, high-acceleration performance, there are direct-drive torque motors. Direct-drive rotary positioners, however, require tuning to achieve desired stiffness, and lack the hard-stop features inherent in worm-and-gear devices and demanded in material working applications.
Motor shaft couplings are essential drive components of linear and rotary positioners, providing proper stiffness to the drive train. Ideally, they offer some degree of flexibility to accommodate shaft misalignments.
In addition to compensating for shaft misalignment, a number of system design considerations influence the choice of a motor coupling: sufficient strength to endure the torques generated in the motion profile (particularly in high-rate reversing applications) and enough torsional stiffness to avoid compromising performance specifications and stability. Furthermore, because coupling inertia is directly transmitted to the motor input shaft, coupling size and mass should be minimized.
Among precision couplings used in positioners, the most economical is a helically cut cylindrical type made from aluminum or stainless steel. Variation of coil-style cuts in the coupling body, along with material composition and bore-to-outside-diameter ratio, determine the performance. For example, 1-inch-diameter couplings of this type are useful for torques up to about 500 ounces per inch. These couplings are well suited for midrange performance in both linear and rotary motion applications.
For high-precision positioning devices, stainless steel bellows-type couplings offer greater torque capacity and much lower windup, but cost about three times more. Another device, the so-called Oldham-style coupling, is a three-piece design useful for high-duty-cycle reversing applications. It consists of two hubs, usually aluminum, and a polymer torque disk. These couplings have low windup, typically in the range of 2 to 15 arc-seconds per ounce-inches. One-inch-diameter couplings with 1/4-inch bores generally offer torque capacities exceeding 620 ounces per inch.
Real-world applications also often involve operating environments that must be considered in the selection and design of positioning systems, as remote positioning becomes increasingly important in nonstandard conditions. Electronic feedback circuit elements and differential thermal effects may limit the operating-temperature envelope of positioning systems. Temperature ranges influence specifications for component materials, electrical-lead insulation, motor windings, and lubricants.
Applications in partial vacuum conditions may degrade positioner performance or, conversely, be affected by the positioner or motor as potential sources of contaminants. Chamber pressures can cause slow vaporizing or material outgassing - a concern that affects even surface finishes, which may be microscopically porous to water vapor.
Users are increasingly demanding complete motion-system engineering and integration services. Sometimes this follows an organization's downsizing, but more often it is a matter of the requisite skills and equipment simply being outside the scope of narrowly focused industries or companies. Particular points that need to be weighed in designing or specifying a positioner system include the following items.
The smaller the required motion of a positioner in a manufacturer's production application, the greater the need to grasp all the complexities of system design to fulfill total operational specifications. Any need for electromechanical motion systems to have correctly matched sensors, motors, drives, and controls places additional demands on those charged with system responsibility.
A positioner component, needing compatibility with a production machine's design, can be planned more efficiently using the same design techniques for both. Will a prospective vendor's computer-aided-design/ engineering applications interface with your product CAD/CAE system?
Beyond being capable of designing and building a positioner, having the correct skills and facilities for precision testing of accuracy and repeatability is a necessary step toward ensuring the device's conformity to operational specifications. Are proper testing facilities are available in-house or do they need to be acquired?
Investment in a new positioner, bought or built, must be analyzed to see how it can be amortized most prudently - especially whether the device planned must be dedicated to the single positioning function sought immediately, or whether versatile design standards would also make it adaptable to other comparable applications.
Costing of personnel time - from design to building, plus testing and installation - must be ensured, as does the cost of variable materials (if alternative specifications are required for clean-room or pressure-chamber conditions) in accounting for in-house costs.
The user seeking to upgrade or develop motion systems for in-plant automation almost invariably discovers that purchase of off-the-shelf motion components is the most expedient and economical choice. When motion devices are elements of an original equipment manufacturer's product, there may be economic justification for acquiring the tools and developing the requisite expertise to build precision electromechanical positioner systems. In this case, the buy-or-build decision becomes more complex and should include projected volume, percentage of total costs attributable to this subsystem, design flexibility, and competitive pressures.
As a broad rule of thumb, in-house development and manufacture of a positioning or motion subsystem is economically justified only when doing so reduces the fraction of costs attributable to the motion components (or system) to 10 or 15 percent of total product costs and when return on investment exceeds 15 to 20 percent.
However, manufacturers developing products for fast-paced competitive markets should also consider that the learning curve for attaining requisite manufacturing skills often is outpaced by concurrent technology advances elsewhere within the industry.