Takahito Sagane

Modeling, Motion and Vibration Control for Flexible Rotor Supported by Active Magnetic Bearings

This paper proposes a new modeling technique and control system design for flexible rotors using active magnetic bearings (AMB) to pass through many critical speeds and fulfill high-speed rotation. To achieve this purpose, it is necessary to control not only motion but also many modes of bending vibration. For the purpose, an extended reduced order physical model that is able to express simultaneously the motion and bending vibration of the flexible rotor, is proposed. Furthermore, a new controller combined PID with LQ control is adapted to control the flexible rotor. Effectiveness of the proposed modeling and control approach for the flexible rotor is verified through simulations and experiments.Copyright

A New Modeling Technique and Control System Design of a Flexible Rotor Supported by Active Magnetic Bearings for Motion and Vibration Control

This paper proposed a new modeling technique and control system design of a flexible rotor using active magnetic bearings (AMB) for motion and vibration control. The purpose of the research was to pass through a critical speed and achieve high-speed rotation. To achieve this, it is necessary to control both vibration and motion. Even though reduced order physical model [1] that we used before is available technique in expressing vibration, this technique cannot express motion. Thus we propose an extended reduced order physical model [2] that can simultaneously express motion and vibration. Further, by using the model we apply the design of a new controller that combined proportional integral derivative (PID) with linear quadratic (LQ) control to a flexible rotor. The procedure we propose is verified by simulations as being effective for a flexible rotor.Copyright

Technique to Handle Local Constraints in Multibody Dynamics

Sometimes we have a system that can be expressed basically by independent variables, with some redundant variable groups to decide only small parts of the system respectively. If the constrains of these redundant variables in each group are not coupled with each other, namely these are no common variables, we call them as local constraints. Example of local constrains are Euler parameter constraint and constrains of link mechanism of car suspension system. The necessity of redundant variables can also be limited for position level. Within the simple nonholonomic system we can always select the appropriate independent velocities to express the other velocity level variables. If the system has only local constraints, it is inefficient to use the typical DAE formulation, which handled all the constraints simultaneously. The technique we explain in this paper has advantage in calculation time and also in the sense of constrain stabilization. This paper gives a basic idea of the technique and its general formulation. Also three examples are explained which we used to confirm the effectiveness of the technique. We got a good result of constrains stabilization. More detailed examination about the calculation time and the constraint stabilization is planned in near future before we proceed to construct a simulation program of complex elastic vehicle model.Copyright

Implementing the Extended Reduced Order Physical Model Into Commercialized Software Packages: Implementing Modal Constraints Into Software Packages

In this paper, the extended reduced order physical model, which is developed to represent dynamics of flexible structures in multibody systems, is implemented to general multibody formulations. The elastic deformations are described using vibration modes associated with the rigid bodies that are effectively defined using dynamic properties of the system in this reduced order modeling method. The recursive formulation and the augmented formulation are used to develop the constraint equations associated with the connections between the reference body and the several rigid bodies in the modal representations used to modal elastic deformations. For this purpose, the n-body constraint is developed for the recursive formulations, while these constraint equations are transformed into the form that is suitable for the augmented formulation using the singular value decomposition or QR decomposition. The use of the formulations in this paper allows for the implementation of the extended reduced order physical model into the commercial multibody codes such as SIMPACK, ADAMS, and DADS.Copyright