Data processing: generic control systems or specific application – Specific application – apparatus or process – Product assembly or manufacturing
Reexamination Certificate
2000-10-11
2003-08-05
Picard, Leo (Department: 2125)
Data processing: generic control systems or specific application
Specific application, apparatus or process
Product assembly or manufacturing
C700S174000
Reexamination Certificate
active
06604013
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to a system and method for detecting tool failure in general and more particularly to a system using frequency derived pre-characterization templates based on spindle/tooling/gearbox dynamic characteristics.
2. Description of the Related Art
There are a number of tool monitoring systems available utilizing different sensing strategies and algorithms. The benefits of such systems when successful, are improving machine up-time, lowering tooling costs, lowering scrap rates, lowering maintenance and improving quality. However, the performance of these systems has been inconsistent and may work relatively well in one application but is not reliable enough to work in another application. In addition they may require extensive set-up and tuning of the system.
U.S. Pat. No. 5,407,265 issued to Seth et al. on Sep. 2, 1997 and entitled “System and Method for Machining Process Characterization Using Mechanical Signature Analysis” is assigned to a common assignee and is incorporated herein by reference. This patent provides a vibration signature analysis based system and a methodology for machinery characterization that can be used to verify machine build to design intent and can also be utilized for identification of root cause of various anomalies in existing machinery. However, this machine condition detection system/strategy is not suitable for real-time tool monitoring applications.
The article entitled “Monitoring End-mill Wear and Predicting Tool Failure Using Accelerometers” by Roth et al. provides a vibration based tool wear monitoring scheme for end-milling whereby the modal energies of the first and second multiples of the cutter tool pass frequency are utilized for monitoring. This system is limited since in a multi-spindle configuration or with a multi-tool spindle cluster there would be several tools with different configurations. Also, no pre-characterization is performed.
In particular, the performance of these systems is deficient for spindle clusters. The commercially available systems utilize time domain monitoring only for detection of tool failures. U.S. Pat. No. 5,407,265 issued to Hamidieh et al. on Apr. 18, 1995 and entitled “System and Method for Detecting Cutting Tool Failure” is assigned to a common assignee and is incorporated herein by reference. This patent outlines a system that utilizes both time and frequency domain monitoring.
This invention takes the teaching of U.S. Pat. No. 5,407,265 a step further to the difficult applications in machining such as spindle clusters or multi-tool stations. This invention provides additional tool monitoring analytical capability by frequency derived pre-characterization templates resulting in identification of spindle/tooling/gear box natural frequencies and dynamic response characteristics.
Subsequently these frequencies and response characteristics are registered in pre-characterization templates and utilized for reliable monitoring in difficult-to-detect machining applications. If in a particular machining cycle, high vibration levels are encountered, they would be discarded when they may be due to sources such as a vibration being transmitted from a neighboring station or slide motion. The high vibration levels that are registered as tool failure will result with the machine tool being stopped only if the frequencies and response characteristics match those recorded in the pre-characterization templates for tool failures.
SUMMARY OF THE INVENTION
It is therefore a principal advantage of this invention to have a reliable system for determining tool failures in machine tools.
It is yet another advantage to have a system to determine tool failures in multi-spindle machine tools.
It is still another advantage to create frequency derived pre-characterization templates on machine tool monitors, to have a reliable real time indication of a tool failure allowing the machine tool to be stopped and the quality of the work output maintained at high levels.
These and other advantages will be found in the following method and system for creating frequency derived pre-characterization templates for the various tools in each spindle of a multi-spindle machine tool. The method has the steps of first mounting a tool into a tool holder in one of the spindles. A vibration sensor is then located and mounted on the spindle housing to sense vibrations induced from the tool and other sources. The sensor is connected to a computing system having an algorithm for measuring the vibration signals. With the tool not rotating, an impact tool strikes the tool or the tool holder causing it to vibrate to simulate tool failure during machine tool operation. The impacting generates a time domain signature trace. The time domain signature trace is supplied to a fast Fourier transform for generating a frequency spectrum signature trace having an ordinate representing amplitude and an abscissa representing frequency of the vibrations. One or more frequency ranges are located on each frequency signature trace with frequency peaks representing structural natural frequencies associated with modes similar to tool failure excitation in a machining operation.
The frequency ranges that represent tool failure excitation are stored in the computing system memory locations or recorded by the machine tool monitor setup operator. These frequency ranges are used to create operator-readable pre-characterization templates showing the ranges of frequencies and the natural frequency and mode excitation or excitations representing machine tool failure.
These and other advantages are described and claimed in the following specification and drawings.
REFERENCES:
patent: 4326257 (1982-04-01), Sata et al.
patent: 4429578 (1984-02-01), Darrel et al.
patent: 4758964 (1988-07-01), Bittner et al.
patent: 4764760 (1988-08-01), Bedard et al.
patent: 4918427 (1990-04-01), Thomas et al.
patent: 5159563 (1992-10-01), Miller et al.
patent: 5407265 (1995-04-01), Hamidieh et al.
patent: 5579232 (1996-11-01), Tong et al.
patent: 5602757 (1997-02-01), Haseley et al.
patent: 5663894 (1997-09-01), Seth et al.
patent: 6308138 (2001-10-01), Jones et al.
Hamidieh Youssef A.
Seth Brij B.
Tata Bhavani Shankar
Coppiellie Raymond L.
Ford Motor Company
Jarrett Ryan
Picard Leo
LandOfFree
Tool failure detection utilizing frequency derived,... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Tool failure detection utilizing frequency derived,..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Tool failure detection utilizing frequency derived,... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3123194