A Large Range XY flexure stage for nanopositioning

This paper presents the design and test results of an XY flexure stage with large ranges of motion and substantially small error-motions. The flexure topology is conceived by means of a systematic and symmetric arrangement of double parallelogram flexure modules. Finite Element Analysis is performed to quantify the flexure stage performance, which is validated by means of experimental measurements. The prototype flexure stage of size 300mm x 300mm exhibits a 5mm x 5mm range with cross-axis errors less than 10microns, and motion stage yaw errors within 5 microradians. Introduction Large range XY flexure stages are important in several applications such as semiconductor mask and wafer alignments [1], scanning interferometry and atomic force microscopy [2-3], micromanipulation and microassembly [4], highdensity memory storage [5], molecular experiments, and MEMS devices [6]. Since these applications generally require nanometric positioning and pose space limitations, flexure-based motion stages are the optimal bearing choice. Despite several XY stage designs that exist in the literature [7-9], achieving large range of motion has been a challenge. Recently, some new parallel kinematic XY flexure designs have been proposed [10-11], that are based on a systematic assembly of common flexure building blocks in a fashion that does not overconstrain the primary motions. An understanding of the properties of the building blocks and a symmetric layout results in improved performance measures such as cross-axis coupling, parasitic yaw motions, and actuator isolation. Because of it large motion and high degree of symmetry, one particular configuration was chosen for the purpose of design, fabrication and testing, and is presented in this paper. Design and Analysis The XY flexure design presented in Fig. 1 is based on a constraint arrangement that is realized by utilizing the double parallelogram flexure module. The constraint arrangement includes four basic rigid stages: ground, motion stage, and intermediate stages 1 and 2. Intermediate stage 1 is connected to ground by means of a flexure module which only allows relative X translation, and motion stage is connected to intermediate stage 1 such that only a relative Y translation is allowed. Similarly, the flexure module connecting intermediate stage 2 to ground only allows relative Y translation, and that connecting motion stage to intermediate stage 2 only allows a relative X translation. Thus, in any deformed configuration of the mechanism, intermediate stage 1 will always have only an X displacement with respect to ground while Intermediate Stage 2 will have only a Y displacement. Furthermore, the motion stage inherits the X displacement of Proc. of 5 euspen International Conference Montpellier – France May 2005 Intermediate Stage 1 and the Y displacement of Intermediate Stage 2, thus acquiring two translational degrees of freedom that are mutually independent. Since the Y and X displacements of the motion stage do not influence intermediate stage 1 and stage 2, respectively, the latter provide ideal locations for actuation. Without causing overconstraint, the design is further enhanced by making insightful use of symmetry, which involves adding intermediate stages 3 and 4, and repeating constraint the arrangement described above. Since all the connecting flexure modules are stiff in planer rotation, the rotation of the motion stage is also constrained with respect to ground, and therefore no active motion stage yaw compensation is necessary. Double parallelogram flexure modules are inherently thermally stable and therefore result in a stable overall XY mechanism. Modules with tilted beams may also be consider for improved inline stiffness between the actuator and motion stage.