Design, Optimization, and Experimental Characterization of a Novel Magnetically Actuated Finger Micromanipulator

Abstract

The ability of external magnetic fields to precisely control micromanipulator systems has received a great deal of attention from researchers in recent years due to its off-board power source. As these micromanipulators provide frictionless motion, and precise motion control, they have promising potential applications in many fields. Conversely, major drawbacks of electromagnetic micromanipulators, include a limited motion range compared to the micromanipulator volume, the inability to handle heavy payloads, and the need for a large drive unit compared to the size of the levitated object, and finally, a low ratio of the generated magnetic force to the micromanipulator weight. To overcome these limitations, we designed a novel electromagnetic finger micromanipulator that was adapted from the well-known spherical robot.

The design and optimization procedures for building a three Degree of Freedoms (DOF) electromagnetic finger micromanipulator are firstly introduced. This finger micromanipulator has many potential applications, such as cell manipulation, and pick and place operations. The system consists of two main subsystems: a magnetic actuator, and an electromagnetic end-effector that is connected to the magnetic actuator by a needle. The magnetic actuator consists of four permanent magnets and four electromagnetic coils that work together to guide the micromanipulator finger in the xz plane. The electromagnetic end-effector consists of a rod shape permanent magnet that is aligned along the y axis and surrounded by an electromagnetic coil.

The optimal configuration that maximizes the micromanipulator actuation force, and a closed form solution for micromanipulator magnetic actuation force are presented. The model is verified by measuring the interaction force between an electromagnet and a permanent magnet experimentally, and using Finite Element Methods (FEM) analysis. The results show an agreement between the model, the experiment, and the FEM results. The error difference between the FEM, experimental, and model data was 0.05 N.

The micromanipulator can be remotely operated by transferring magnetic energy from outside, which means there is no mechanical contact between the actuator and the micromanipulator. Moreover, three control algorithms are designed in order to compute control input currents that are able to control the position of the end-effector in the x, y, and z axes. The proposed controllers are: PID controller, state-feedback controller, and adaptive controller. The experimental results show that the micromanipulator is able to track the desired trajectory with a steady-state error less than 10 µm for a payload free condition.

Finally, the ability of the micromanipulator to pick-and-place unknown payloads is demonstrated. To achieve this objective, a robust model reference adaptive controller (MRAC) using the MIT rule for an adaptive mechanism to guide the micromanipulator in the workspace is implemented. The performance of the MRAC is compared with a standard PID controller and state-feedback controller. For the payload free condition, the experimental results show the ability of the micromanipulator to follow a desired motion trajectory in all control strategies with a root mean square error less than 0.2 mm. However, while there is payload variation, the PID controller response yields a non smooth motion with a large overshoot and undershoot. Similarly, the state-feedback controller suffers from variability of dynamics and disturbances due to the payload variation, which yields to non-smooth motion and large overshoot. The micromanipulator motion under the MRAC control scheme conversely follows the desired motion trajectory with the same accuracy. It is found that the micromanipulator can handle payloads up to 75 grams and it has a motion range of ∓ 15 mm in all axes.