A six degrees of freedom mems manipulator

This thesis reports about a six degrees of freedom (DOF) precision manipulator in MEMS, concerning concept generation for the manipulator followed by design and fabrication (of parts) of the proposed manipulation concept in MEMS. Researching the abilities of 6 DOF precision manipulation in MEMS is part of the Multi Axes Micro Stage (MAMS) project in which the disciplines of precision engineering, control engineering and micro mechanical engineering are represented. Micro Electro Mechanical Systems, or MEMS technology is generally based on lithographic, deposition and etching techniques also applied in the integrated circuits (IC) industry. The basis for these microsystems is a thin polished substrate usually consisting of very pure single crystal silicon, called a wafer. Roughly, since the 1980’s a lot of mechanical devices on the micrometer scale have been developed in MEMS technology. Some strikingly resemble machines that are very common in the large world, consisting of down-scaled gears, ratchets, racks, pinions, sliders and hinge mechanisms. Others are based on flexure mechanisms resembling large scale systems used for precision position adjusting and manipulation. There are also large differences between MEMS systems and macroscopic mechatronical systems. Where actuation in macroscopic systems is dominated by electromagnetic transduction, the preferred principle of actuation in MEMS systems is far less pronounced. Efficient electromagnetic transducers are difficult to realize in MEMS while actuators based on electrostatic attraction are more simple to fabricate and generally show better performance. On the micrometer scale, large electrical fields can be obtained at relatively low voltages and the breakdown fields scales favorable due to Paschen’s law. Other actuation principles often encountered in MEMS are electrothermal and piezo-electric actuation. Another aspect in which MEMS differ from macroscopic systems is the fabrication method, characterized by a ��?2.5 dimensional��? design freedom. This largely complicates the design of 3 dimensional geometries, especially regarding structuring of devices in the direction normal to the wafer-plane. This is one of the main reasons, the design of a 6 degrees of freedom stage for precision manipulation is a large challenge. However, the potential improvements obtained by downscaling a manipulation system encourage to research the feasibility of a 6 DOF precision MEMS manipulator. A miniaturized precision system does not only benefit because of reduction in size, but thermal drift is substantially reduced as well and the primary resonance frequency increases considerably. To put the development of a 6 DOF precision MEMS stage in a practical perspective, specifications are defined for a sample manipulator in a transmission electron microscope (TEM), capable of imaging resolutions in the A° -range. This application requires a highly stable system capable of strokes of §10 mm and nanometer resolution positioning. In this thesis a concept for 6 DOF manipulation is presented, based on stacking of two 3 DOF parallel kinematic manipulators. One is used for the three planar DOFs and stacked on top of the other, manipulating the three out-of-plane DOFs. The design of the flexure based mechanisms is largely based on exact kinematic constraining of the DOFs. Limitations to fabrication of the desired mechanism rising form MEMS technology are identified and solutions are proposed. The planar 3 DOF manipulator actuated by electrostatic comb-drives is successfully fabricated in bulk SCS. Characterization shows largely linear mechanical behavior on the sub-micron scale. For actuation of the out-of-plane stage, a vertical comb-drive suspended by torsion beams is designed. To increase stability with respect to electrostatic side pull-in, the torsion beam consists of vertical and horizontal parts. A process is developed to enable fabrication of the required geometry for the torsion beams and the vertical comb-drive teeth in SCS. This process is compatible with the fabrication of the planar 3 DOF stage. Fabrication of the special 3 dimensional geometry still requires improvement and many out-of-plane devices show short circuits. However, out-of-plane actuation with a stroke of 10 mm has been observed. Increased stability with respect to side pull-in might also be obtained by a different kind of electrostatic transduction. In this case a dielectric plate or a floating conduction plate is partially inserted between to parallel electrodes. A voltage over the electrodes will result in a force pulling the plate inwards. In a first order approximation no forces in lateral (pull-in) direction are found. Finite element simulations (FEM) reveal forces in lateral direction do occur, however these are much smaller compared to the case of a comb-drive. Additionally, in case of charging, the lateral forces on a charge sheet between two parallel electrodes are modelled. The dependance of the lateral force on the lateral position of the charge sheet is found not to be highly non-linear like in the case of a comb-drive. This encourages the expectation that the alternative electrostatic actuators are far less sensitive to side pull-in than comb-drives. However, a proof of principle has not yet been realized.