Ground ball-screw accuracy always beats rolled ball screws, but lead deviation, production methods, and heat treatment also play significant roles in choosing one over the other.
George A. Jaffe
Executive Vice President
Alexander F. Beck
A. Steinmeyer GmbH
Is the following true or false? — “Rolled ball screws cost less than ground ball screws and work just as well.” The first part is true. Rolled ball screws certainly cost less, sometimes 50% less than ground ball screws. But when it comes to the second part, there is no comparison. Although some manufacturers claim they run just about the same, ground-ballscrew accuracy always surpasses that of rolled ball screws, even with equal lead errors. But this doesn’t mean it’s always best to choose ground ball screws. Each type has its own advantages that help engineers choose the one that best fits the application.
Accuracy is the most common parameter for comparing ball screws. Although many designers measure ball-screw accuracy by just how precisely the nut travels the length of the screw, other factors contribute as well. In fact, at least three values are necessary to specify lead accuracy.
Factor ep represents average actual lead deviation over the full travel of the nut on the screw, thus e300 is the deviation per 300-mm interval. Designers should know whether the value is cumulative. For example, accuracy-grade 5 per DIN/ISO specification lists a maximum lead deviation of 23 μm/300 mm. However, a 900-mm-long screw has a maximum total deviation of 40 μm, not (900/300) X 23 = 69 μm. When comparing ball screws with only specified “deviation per 300 mm” to ball screws governed by DIN/ISO or JIS accuracy grades, significant accuracy differences become obvious.
Factor e2π specifies lead variation in one revolution of the screw thread. This parameter has more impact on equal load distribution among the balls than on axial-nut travel (which may affect predicted life). Because applications requiring high-lead accuracy may also be more demanding with regard to other quality issues, the DIN/ISO standard links several properties to lead accuracy. For instance, parameters of DIN/ISO standards include screw straightness, bearing-journal concentricity, preload-torque consistency, and nut squareness. When the application calls for precise slide positioning, also consider stiffness and the assembly’s frictiontorque consistency. For instance, although screws may be accurate in terms of lead error, when journals are eccentric the screws show poor torque consistency. This is often the case with rolled ball screws.
Many factors affecting performance go unnoticed when considering only lead accuracy, especially for rolled screws. Errors in manufacturing equipment, such as eccentric, worn, or wobbling rolling dies, often produce periodic lead errors. But a nut riding on several ball circles may mask periodic lead errors. Such errors, however, can distribute the load unevenly, reducing ballscrew stiffness and life.
A specific application is often the initial deciding factor for choosing rolled or ground ball screws. Rolled screws, for instance, are typically used in low-precision applications. Examples include some pick-and-place devices and actuators. Ground ball screws, on the other hand, drive precision machine tools, such as milling machines, lathes, profile grinders, and EDM machines. Ground ball screws are also recommended for precision X-Y tables, such as those used in semiconductor wafer-inspection equipment.
Rolled ball screws, however, are gaining widespread use in many applications that traditionally held ground screws. Rotary encoders on servosystems are one example. The encoders are mounted to drive motors and provide position feedback. Positioning accuracy varies with ballscrew lead error. Ground ball screws used in these applications meet ISO5 standards requiring positioning accuracy within 0.025 mm per 300 mm. Recent improvements in rolling processes, however, let manufacturers make rolled screws with lead errors meeting ISO5 standards.
On the other hand, sometimes economical designs are used when they shouldn’t, such as in linear encoders. Using rolled instead of ground screws in linear scales, for instance, doesn’t save money. The thinking is as long as nuts are preloaded for zero backlash, the scales will determine accuracy. Instead, torque varies excessively making process machines run noisily with servosystem instability, leading to inconsistent production processes.
Two fundamental differences in manufacturing rolled and ground screw threads help explain why the two types perform differently. Rolled screws are shaped by forcing round bar stock through rotating dies to form threads. Although cutting the bars before rolling may seem like an option, this produces inaccuracies near the rod ends. To prevent this, the bars are cut and heat treated after rolling. Centers and bearing journals are machined only after finishing the threads. This generally makes it difficult to maintain concentricity between ball threads and bearing journals on rolled screws. Groundscrew manufacturers, on the other hand, machine journals before grinding threads. This ensures precise concentricity with ball threads.
The second difference lies in how each production method affects accuracy. Rolling is a onestep process that involves heavy forming and significant material stresses. All inaccuracies caused during traditional rolling and heat treating remain in the finished product and can’t be corrected. Heat treatment produces stress that tends to lengthen screws, often in an inconsistent and unpredictable manner. Lead errors typically accumulate, so check the manufacturer’s definition of lead accuracy. The relatively rough finishes of rolled screws can also lead to excessive noise and vibration, especially during high-speeds. Rolled-screw finishes also generate high torque variations due to variable preloads as nuts traverse the screws. Grinding ball screws, on the other hand, finishes functional surfaces (the ball-thread OD and the thread itself) one at a time. Grinding can maintain concentricity and squareness because the same centers are used to grind all critical surfaces. Grinding also lets manufacturing partition the job into smaller steps and insert procedures, such as QC, restraightening, or heat treating to release material stresses. The order of steps can also be adjusted. For example, threads can be finished after all rough machining. This improves overall accuracy and addresses the various specifications one at a time. Performing in-process inspection before finishing also reduces scrap, improves quality consistency, and reduces costs.
Ground screws also give designers greater flexibility than rolled screws. Rolled screws often cannot provide full shoulders or oversized journals unless the screws comprise several joined parts, such as shrunk collars or frictionwelded journals. In general, however, “assembling” ball screws using welded journals is not recommended.
Although grinding is still widely believed to be more expensive than rolling, this is not always true. Recent manufacturing developments combine grinding with special prerolling processes to gain advantages from both. Manufacturers preroll screws, which partially forms threads, and then touch up the screws on thread-grinding machines to eliminate inaccuracies from rolling. The process also minimizes stresses released from heat-treated material. The idea is to remove less material during grinding to achieve more consistent results in terms of lead error, straightness, and thread profile. The accuracy of such screws equals that of conventional ground screws, yet prerolling maintains the economical advantages of rolling.