Overview of Ball Screw/Lead Screw Systems for Linear Movement 

A bit of a late response, but a good question(with a long winded answer). In case anyone else has the same question, the answer is that it's handled through software using the feedback from each stage's encoder. This may be different for Nippon's systems, but for higher end systems used in defense, aerospace and semiconductor production(systems I generally deal with for work), the software is always designed with this functionality in mind. Generally when you have two stages that you want synced to perform a specific function, you set one stage as the master and the other stage as a slave axis. The master is issued the command and that command is simultaneously issued to the slave(High end controllers often store the programs in RAM after the computer starts the program, so network delay to specific controllers cannot cause issue with timing). The software will use the encoders on the stages to constantly monitor for position error(how far from the expected position in the move it is), and constantly correct for any deviations from this expected path with the master being the primary handler of the coordinated motion. If you are running a dual encoder system, you will also have velocity error on separate feedback lines that will also read how far off from the expected velocity you are during accel and decel motions. With two linear/ballscrew stages a bit of error is not usually too critical, as the stages are usually not mechanically connected, or if they are, a bit of play is intentionally designed into the system. This prevents a little bit of position error on one stage from causing twisting and binding to a stage parallel to it in a gantry configuration. However say you have 2 rotary stages that are one either side of a heavy load that are meant to precisely and quickly rotate that central load. If you didn't have one axis specified as a master and you had both stages individually running the same commands, this configuration will likely cause oscillation and eventually failures. Running the stages separately, but having each stage trying to maintain it's position causes the 2 stages to often either fight against each other if they are not synced properly, or to over adjust when trying to get back in position. If each individual stage believes it needs 1.4A of current at a specific electrical angel of the motor to get to a precise position, and both stages try to command the same move, they can apply too much force causing the load to barely overshoot the position it's supposed to be moved to. This causes oscillation when the stages are constantly pulling the load just past where it wants it, and they end up indefinitely repeating the error. The system will start to vibrate and eventually fail. Lowing the servo tuning of the system will cause higher position errors, but can prevent oscillation and the stages will be less stiff, thus not as forceful or quick in the correction movements. This is obviously not an ideal solution for precision, but for systems where 0.004mm or say 0.04° of error during a move is acceptable this is fine. Having one stage as the master(if done correctly) removes this problem as that one stage takes priority on feedback and correction of the movements. Again, I don't know Nippon's software specifically, but with the systems I use, that is how that is handled

A flexible coupling is used in most cases where multiple bearings are in line with each other - this helps adjust for misalignment. A solid coupling may cause premature wear on one or more bearings in the mechanism or motor.

The purpose of the bearings is to act as a support for the ball/lead screw so that it doesn't bow out of its rotation radius (keeps it perfectly in line). The support bearings also prevent the ball/lead screw from grinding on the support itself or a weaker bearing such as plastic.

It is attached to the workpiece.