M.A. Elbestawi

A simulation system for improving machining accuracy in milling

Abstract The end milling process is used extensively in a variety of manufacturing areas, especially in the aerospace industry, where many of the machined parts are pockets and thin webs. Many of the end milling applications involve machining of thin webs using long and flexible cutters. This leads to the fundamental problem of surface accuracy in milling. The superposition of the tool and the workpiece deflections, which is known as effective system deflection, will cause dimensional errors and irregularities on the finished surface. In the case of thin web machining, using conventional cutting speeds (i.e. 30 to 60 m/min) the surface generation is mainly influenced by the workpiece vibration and to a lesser extent by the tool deflection. The severity of these vibrations depend on the cutting conditions used, as well as material and geometry of the workpiece being cut. In this paper, a simulation system for finish milling of flexible structures is developed. This system includes an improved model for the prediction of the cutting force. This enhanced dynamic model takes into account the deflections of the tool and the workpiece in the calculation of the chip load. A dynamic, 3-D finite element model of the workpiece, which takes into account the effect of material removal, is used in the simulation. An Automatic Mesh Generation program is developed to simulate the material removal during cutting. The simulation results are then compared with the experimental results obtained on a vertical milling machine. Several finishing cuts are performed and the resulting workpiece surface dimensions are compared with the predicted errors.

Knowledge-based adaptive computer control in manufacturing systems: a case study

A knowledge-based system approach for designing an adaptive controller is introduced. The scheme has been used successfully in designing a self-tuning controller for force regulation in a computer numerically controlled (CNC) milling machine. In this scheme, frames are used for knowledge representation and rules of logic for reasoning. This synergistic combination of frames and rules provides the environment for intelligent control. As a consequence of representing knowledge in frames, the large amount of logic that goes along with most conventionally designed adaptive controllers to ensure safe operation is considerably reduced. Procedural attachments to the slots in the frame replace the extra logic elements in the knowledge-based controller. The self-tuning controller for the CNC milling machine is implemented on a 32-b microprocessor-based computer running at 20 MHz. The knowledge representation and the reasoning process are implemented in Prolog, whereas the numerical algorithms are written in C. Simulations and experimental results are provided that demonstrate the usefulness of this approach. >

Some advanced control strategies for modern machine tools

Abstract Several adaptive control strategies are developed for metal cutting machine tools. A Self Tuning Controller is applied to the problem of cutting force regulation in milling. The performance characteristics of the controller are examined using real-time control experiments and the results are shown to be quite satisfactory. Geometric adaptive control strategies are next developed for both turning and milling processes. The control objective is to maintain the geometric accuracy of the workpiece in the presence of significant workpiece-tool deflections and random disturbances. It is shown, using computer simulations, that the proposed control systems result in significant improvement in geometric accuracy of the workpiece in contouring over conventional NC controllers.

Adaptive control for geometric tracking in turning

Abstract An adaptive tracking algorithm is adopted to develop an adaptive control system for contouring operations in turning. The control objective is to maintain the geometric accuracy of the finished workpiece in the presence of significant workpiece-tool deflection error as well as random and periodic disturbances. The controlled variable is the position command. A model incorporating the NC servo loop, the cutting process and the machine tool-workpiece dynamics is used to demonstrate the effectiveness of the algorithm. It is shown, using computer simulations, that the proposed control system results in significant improvement in geometric accuracy of the workpiece in contouring over conventional NC controllers.

Adaptive tracking in pure-feedback nonlinear systems

A method of adaptive tracking for pure-feedback nonlinear systems is presented for the case where linear parameterization conditions are met. The resulting adaptive controllers can be viewed as a state-space model reference adaptive controller with updating of both the state diffeomorphism and the control variables. The tracking problem is considered with respect to parametric and dynamic uncertainties, while prior results have been limited to the parametric uncertainty. Restrictive assumptions on the unknown parameters are removed, overparameterization is avoided, computation time is considerably reduced, and an error model is derived for formulation of the adaptive tracking problem. The robustness of the adaptive controller with respect to unmodeled dynamics is analyzed, and simulation results are presented.<<ETX>>

Surface accuracy control in end milling using adaptive state tracking

Abstract A geometric adaptive control system for the end milling process is developed based on adaptive state tracking. The control specification is given by an external reference trajectory, which represents the desired geometry of the finished workpiece. The machining system is represented by a state-space model which includes the flexibilities of both the cutter and the workpiece. The Recursive Prediction Error (RPE) method is used for joint state and parameter estimation of the state innovations model and the necessary adaptive state tracking control law is derived. In this study, on-line assessment of workpiece geometry is considered not realizable in practice, and only the cutting force is assumed measurable. The desired control is achieved by manipulating the position command for the servo loop so that the actual tool position follows the desired workpiece geometry. Simulations are obtained for the proposed geometric adaptive control system and the results are shown to be quite satisfactory.

Simulation of robust dynamic controller for redundant manipulators

The robustness of dynamic control for redundant manipulators is investigated. A robust controller based on Lyapunovs stability analysis is derived. Asymptotic tracking of the end effector in Cartesian space is thus achieved. Using the proposed redundancy resolution measures at the acceleration level, a subtask with singularity avoidance can be accomplished. Simulation results confirm the effectiveness of the proposed schemes. The simulations show that the proposed control scheme guarantees tracking of a desired Cartesian space trajectory while seeking a better manipulability gesture even in the presence of parameter uncertainties.<<ETX>>

Simulation of adaptive controller for flexible joint manipulators during constrained motion task execution

An adaptive control scheme for flexible joint manipulators constrained by contact with the environment is presented. Based on a singular perturbation model of constrained manipulators, an estimate of the range of mu in which stability is guaranteed is given. Simulation results for a two-link flexible joint manipulator demonstrate the effectiveness of the proposed control algorithm. The proposed method has the following attractive features. First, fundamental passivity properties of rigid robot dynamics are used to design the adaptive controller for flexible joint manipulators. Second, the implementation of the full controller requires only joint and link position and velocity information. Robustness to parametric uncertainty is achieved without the need for acceleration and jerk measurements. Finally, this control algorithm is uncomplicated for practical applications.<<ETX>>