Ball screw cumulative error is a widely recognized issue. There are numerous sources on the web that discuss this issue in depth. If you look at a ball screw accuracy graph you will see that the lines are not horizontal and angle away from the zero axis indicating a certain amount of lead error per unit of travel. Over short distances it is typically inconsequential, but over longer distances becomes a problem forcing the use of linear encoders to get accurate positional feedback. At lengths less than 9 inches ball screws are generally as accurate as the RPS system, but beyond this the RPS is much more accurate than a ball screw. As you have seen, the cumulative error graph for the RPS is horizontal meaning there is none regardless of the distance.
As for accuracy being built in to ball screws and guiding system design effecting the RPS systems accuracy. Precision and accuracy are not the same thing and you can have one without the other. Ball screws can be precisely manufactured but are not necessarily accurate. More on this can also be found by a web search. The guiding system requirements for the RPS are not significantly different than ball screws. Typically profile guide rails are used to guide the system and dictates the precision of component installation. Ball screws cannot tolerate side loading so the parallelism requirements of the system are high in either case. If the guiding system allows the RPS to loose preload due to poor machine design or installation then accuracy would be compromised.
On the issue of backlash you say that ball screws have none. This is not entirely true of ball screws or any mechanical system. Some clearance is always needed for parts to slide or rotate, although it can be very small. Even with bearings clearance is required for ball or needle lubrication or it would be metal on metal. Yes, mechanical nut preloading takes most of the play out of a ball screw but there is still clearance for lubrication that can be squeezed out when sufficiently loaded. If you look at so called zero backlash ball screw specs in a catalog you will typically still see a backlash value. The RPS uses a preload to achieve a similar effect. Equally important is system compliance or elasticity, ball screws are much more elastic due to deformation of the ball nut and the unsupported length of the screw. The 3.2um linear backlash stated for the RPS takes into account backlash and compliance. It was measured with high precision rotary and linear encoders while applying the maximum linear force while the pinion shaft was locked. In addition to this ball screws are much more susceptible to thermal expansion due to environmental variation and heat build up in the nut due to high speeds. The RPS rack is rigidly mounted and with it's 99% efficiency very little thermal expansion is generated regardless of how hard you run it
On the issue of environmental contaminates. The RPS system is susceptible to environmental contaminates as are ball screws and traditional rack and pinion. If the rack is not oriented with teeth up and ideally downward contamination is minimized. The lubrication requirements for the RPS are much lower than ball screws. The rack gets a light greasing at installation and then typically twice a year thereafter. Harsh environments may require more frequent servicing. There are many lubrication options from food grade to dry lube that can be chosen based on the application. Lubed for life ball screws aren't and are based on clean room conditions and light loads. The thing to keep in mind is that the needle bearings that support the pinion roller pins are permanently lubricated and sealed. As the rollers engage the rack it is mostly rolling friction. Each time a pinion roller engages and disengages a rack tooth it only turns a small fraction of a revolution. This gives the RPS its 99% efficiency rating compared to a ball screws 80%. In other words very little friction or wear takes place giving the RPS much longer life in most applications. We have had customers use the RPS for life issues alone and they are very satisfied with it. If contaminates get into a ball nut they quickly eat it up due to the high amount of work the balls do in the nut and opportunities for friction. As with any drive system if shielding can be installed to protect it from contaminates it is a wise thing to do and should extend the life of the system.  |