| Engineers have put significant efforts into improving the precision of machining processes. Toward year 2000, the precision of an ultra-precision machine would be around one nanometer. The major obstacles to attaining such high precision are temperature stability and vibration. In the Stanford Quiet Hydraulics Lab, we attempt to maintain uniform temperature distribution by using hydraulic components, while reducing or eliminating vibration from hydraulics by applying laminar flow design.;In carriage system integration, we applied a new sensor and a novel actuation design. The sensor is submersible with satisfactory dynamic range. Laminar flow devices are chosen to provide quiet actuation. We explored the characteristics of laminar flow actuation through mathematical modeling and experimentation. Its application in ultra-precision machining has been proven by the test at system level. The integration result shows that with the no-overshoot criterion, we have successfully extended the bandwidth of the system to 5 Hz and we can further increase it by stiffening the linkage of the actuation to the carriage.;With closed-loop control, we can explore the error sources of carriage positioning. The standard deviation of the steady-state error is 12 nm. The major source of error is vibration in the structure. We built an active dynamic absorber for reducing the vibration. The model of the structure is derived and verified in experiments. With the model conforming to the experimental results, we concluded that the fault in the metrology frame is the main cause of error. The supports of the metrology frame may have been damaged in the earthquake in 1989.;In this research, we have successfully integrated a novel submersible high-precision sensor into our quiet hydraulics testbed. The application of laminar flow actuation is first implemented and verified at system level. Furthermore, the most important work in a precision machine design is to identify the sources of error. We found that vibration in the metrology frame accounts for more than 50% of the error in carriage positioning. We strongly believe that by fixing the faulty supports we could achieve nanometer level precision in the positioning of the carriage motion. |
