December 9, 1996

Project Title: Dynamic Performance Characterization and Enhancement of High-Speed Machine Tool Drives - Implementation on an Industrial Open Architecture CNC System

Principal Investigator: Tsu-Chin Tsao

Background/Description: The advances of tooling materials such as carbide, ceramic, polycrystalline diamond and cubic boron nitride has dramatically increased metal removal rate potential. This has made the machine tool performance, in particular dynamic rigidity and motion control accuracy, the limiting factor in the exploitation of higher cutting speeds. To increase feed rate, there is a trend toward using direct-drive ball screw slides and direct-drive linear motors. Current CNC servo control technologies are based on low to medium bandwidth motion of the de-coupled prime drives and hence problems arise when they are used to control the high bandwidth motion of the process coupled direct-drives. Servo control is thus a critical factor that has limited current direct-drive machine tool performance. The purpose of this project is to address these problems at the servo control system level. Three aspects of motion control that can benefit from improvements in high speed servo control are point to point control, axis tracking, and multi-axis coordinated motion. Based on our previous and present research results on system dynamics identification and advanced motion controls, and with the proposed effort in software and hardware integration of a PC based open architecture CNC system, this project will demonstrate that advanced motion control can be implemented on an industrial machine with substantial performance improvement over the existing servo control system. The research results can thus be implemented on industrial CNC systems and be used by the industry.

Research Objectives: This projects research objectives are to:

· integrate our system identification tools and advanced motion control algorithms into the front end PMAC-PC CNC system,

· establish the PMAC's benchmark servo control performance on the Bridgeport Machine tool drives and the linear motor XY table for comparison with our advanced control performance, and

· demonstrate on the PMAC that our advanced motion control can be implemented on an industrial machine with substantial performance improvement over existing industrial servo control system.

Expected Results/Deliverables: This project will result in the development of software and hardware interfacing to allow open-system architecture at the servo control level. Advanced servo control methods for high speed motion control in machining applications will be developed on a industrial platform and in a format that can be combined with existing servo control systems without changing the user interface. A benchmark of industrial servo control performance will be created, enabling valid evaluation of these advanced control methods.

Progress to Date:

PC based machine tool controller: A PMAC motion control board has been interfaced with our laboratory hardware (i.e. Bridgeport and linear motor X-Y table) and several open architecture capabilities have been developed using it.

· User-written servo control - servo control algorithms such as multi-axis PID feedback control and feedforward control have been written in assembly language and implemented, replacing the nominal PMAC servo control algorithms with no visible change to the PMAC user.

· Intelligent control - a supervisory controller has been developed and implemented that controls feed rate override on the PMAC in real time, changing the feed rate according to measured cutting conditions such as cutting force and tracking error. A fuzzy logic based approach is being developed.

· Event driven servo controller - a controller synchronizing the internal controller with external events such as encoder pulses is being developed using the PMAC's position capture and compare capability.

Dynamic performance characterization: A system identification procedure, running on the PMAC PC-based CNC system, has been developed for characterizing dynamic behavior of machine tool feed drives. This algorithm is being tested at Motorola.

Rapid tracking motion control algorithms: A frequency response method for designing feedforward controller without having to resort to a parametric model has been developed on a TMS320C30 digital signal processor and is now being tested on a PC-based CNC system using a PMAC motion control board

Member Company Benefits:

· Advanced motion control algorithms have been implemented on an industrial CNC platform, without changing the user interface, demonstrating the applicability of these algorithms to industry.

· The system identification algorithm package is available and can be used on a PC-based CNC system to characterize the dynamics of an industrial feed drive.

· New rapid tracking motion control algorithms could be useful for high-speed high-precision motion control of machine tool feed drives.