RMRL
Robotics and Mechatronics Research Laboratory
|
|
|
RMRL: Research
Micro/Nano Precision Manipulation
The broad aim of this fundamental research program is to establish new methodologies and concepts for analysing, characterising, and optimising a complex micro/nano manipulation system based on flexure joints and parallel actuation.
It also aims to establish joint facility between Monash and Deakin University for micromanipulation research. The proposed facility consists of complex/intelligent micromanipulation systems capable of motion accuracy of 10-100 nanometers.
The project boasts its micro/nano research facilities including various unique designs of micro/nano manipulators, precision stages (PI hexapod, Thorlab flexure stages), various precision sensors: capacitive sensors (PI), strain gauges, laser interferometry systems (Zygo and API), and precision force/torque sensor (ATI), all with hard real-time closed-loop control implementation. Soon available in the lab are a high magnification optiscal microscopy (in setting up phase) and confocal microscopy for micro manipulation application. This project is funded through the Australian Research Council (ARC) Discovery Grant.
|
Flexure based, piezo-actuated precision XYZ stage for micro/nano positioning research.
A pair of capacitive sensors from PI, measurement range 50mm with sub-nano resolution.
|
Flexure joint formed by a cross section area that is much smaller than that over the rest of the mechanism. |
Flexure Jointed Parallel Mechanisms
Conventional joint articulation methods are limited in its application by various nonlinear dynamics, such as friction and backlash. For a small range of motion, flexure joints offer an alternative that almost eliminates all these effects. Flexure joints are created by reducing the cross section of the links in the area where a revolute joint is to be placed in the mechanism. Prismatic joints can also be created through springlike designs that would allow compliance in the direction of motion and stiffness in the orthogonal directions. As such, the mechanism is created out of a single monolithic structure, eliminating any needs (and the resulting effects) of part joining and motion between multiple parts - producing the above-mentioned advantages of the design. Parallel configuration is a necessary for the strength of the structure due to the small cross section of the joints.
|
Modeling, Analysis, and Control of Piezoelectric Actuator
Piezoelectric actuators (PEA) possess unlimited resolutions, with high stiffness and fast response. It is very suitable for applications requiring small range of motion with high precision. However, it has its drawbacks in the form of nonlinear behaviour such as hysteresis and creep, between the command voltage and the displacement.
Varous control strategies are formulated to obtain a robust control over the PEA and the manipulation tasks. Due to its nonlinear behaviours, conventional PID approach is not suitable to cover the working range, hence adaptive and robust control strategies are formulated, to adapt to the changing parametric conditions or to treat such behaviours as disturbances to the system and provide allowances for such parametric uncertainties.
Modeling of the nonlinear behaviour of the PEA is also investigated to better understand the characteristics of the material. Various system identification methods were employed to identify dynamic / lump parameters of the PEA model. Neural network approach is also explored with promising results.
|
Another piezo actuated flexure based experimental setup (a Scott-Russell mechanism).
Piezoelectric actuator from PI, with a pre-loaded housing and built-in strain gauge sensor.
|
Return to Research Topics |
|
