Tool failure detection utilizing frequency derived, pre-characterization templates

• US 6,604,013 B1
• Filed: 10/11/2000
• Issued: 08/05/2003
• Est. Priority Date: 10/11/2000
• Status: Active Grant

• Current Assignee: Ford Motor Company
• Original Assignee: Ford Motor Company
• Inventors: Hamidieh, Youssef A., Seth, Brij B., Tata, Bhavani Shankar
• Primary Examiner(s): Picard, Leo
• Assistant Examiner(s): JARRETT, RYAN A
• Application Number:
  US09/686,612
• Time in Patent Office: 1,028 Days
• Field of Search: 700/174-177, 700/108, 700/110, 700/280, 702/34, 702/35, 702/56, 702/82, 702/84, 702/141, 702/182-185, 340/679, 340/680, 340/683, 735/77, 735/79, 735/98, 736/59
• US Class Current: 700/175
• CPC Class Codes: G05B 19/4065 Monitoring tool breakage, life or condition
• CPC Class Codes: G05B 2219/34048 Fourier transformation, analysis, fft
• CPC Class Codes: G05B 2219/37242 Tool signature, compare pattern with detected signal

Abstract

A system and method to create frequency derived, pre-characterization templates to be used by a machine operator for determining when a tool breaks in one spindle of a multi-spindle machine tool. The pre-characterization template is created by first impacting the tool or the holder and recording the time domain of the frequency pattern of vibrations. The natural frequencies of the tool/tool holder/spindle vibrations which include the vibration frequencies of the tool braking are recorded. Pre-characterization templates are created using FFT techniques to identify only those frequencies indicative of tool breakage. The machine tool operator then uses the monitoring system with pre-characterization templates during machine tool operation and is notified in real time of the breakage of a tool in a single spindle or a multi-spindle machine.

Claims

1. A method of pre-characterization a tool failure in a tool/tool holder/spindle combination for subsequent tool failure detection during operation of one or more spindles in a machine tool by creating frequency derived, pre-characterization templates for subsequent use in a machine monitoring system, the method comprising the steps of:

mounting a tool into a tool holder in one or more spindles on the machine tool;

locating at least one vibration sensor near the spindle;

connecting the vibration sensor to a computing system;

impacting the tool or the tool holder causing it to vibrate while the machine tool is not in operation, thereby simulating a tool failure during machine tool operation;

generating a frequency signature trace from the sensor having an ordinate representing amplitude and an abscissa representing frequency;

locating one or more frequency ranges on each frequency signature trace with a frequency peak representing a structural natural frequency and modal excitation representing tool failure excitation;

storing the frequency or the frequencies representing tool failure excitation in computing system memory locations; and

then

creating a frequency derived, pre-characterization template showing the ranges of frequency and modal excitation or excitations representing machine tool failure.

2. The method of claim 1 wherein the step of generating includes a frequency analyzer having a fast Fourier transform capability for generating the frequency signature trace.

3. The method of claim 1 additionally including the steps of:

mounting another tool in a tool holder of another one of the spindles;

generating another frequency signature trace of that tool; and

then

creating additional frequency templates with the ranges of frequency or frequencies and mode excitations of the associated tool failure excitations in each tool or the tool holder in the machine tool.

4. A method utilizing a frequency derived, pre-characterization template for detecting a tool failure from one or more spindles on a machine tool comprising the steps of:

securing one or more pre-characterization templates having one or more frequency ranges each representing machine tool failure with said one or more templates being derived from a frequency signature trace generated from an impact applied when the machine tool is not in operation;

placing the templates on a controller coupled to the machine tool;

placing one or more vibration sensors near the spindle for generating electrical signals responsive to tool vibration;

generating a frequency signature trace from said electrical signals;

comparing a template readings with the frequency signature trace generated from the vibration sensor during machine tool operation; and

then

generating an alarm when high level vibrations are detected and the frequency ranges of the pre-characterization template and the frequency signature trace of the machine tool during machine operation are substantially similar indicating a machine tool failure.

5. A system for generating a frequency derived, pre-characterization template for detecting a tool failure in one or more tool/tool holder/spindles in a machine tool during its cycle of operation, the system comprising:

a tool holder mounted in a spindle of the machine tool for holding a tool;

an impactor for striking the tool for generating a simulated tool failure vibration when the machine tool is not in operation;

at least one vibration sensitive sensor mounted near the spindle and responsive to said simulated tool failure vibration for generating an electrical signal representing sensed vibration;

a computing system operatively coupled to the machine tool and responsive to said electrical signal, said computing system having;

a signal conditioner for conditioning said electrical signal representing the sensed vibration;

a low pass filter operatively coupled to said signal conditioner for filtering the electrical signal;

a digitizer operatively coupled to said filter for continuously digitizing said filtered electrical signal during the operation of machine tool for creating a digitized data signal;

a fast Fourier transform algorithm for continuously transforming said digitized data signal to a frequency domain and generating a frequency signature trace continuously during the operation of the machine tool; and

memory locations responsive to said transform algorithm for storing said frequency signature trace; and

a frequency template generator responsive to said frequency signature trace for generating an operator-readable pre-characterization template of said tool/tool-holder/spindle failure.

6. A monitoring system utilizing a frequency derived, pre-characterization templates for detecting a tool failure from one or more spindles on a machine tool comprising:

a vibration sensor detecting vibrations from the one or more spindles on the machine tools;

a processor receiving the output of the vibration sensor and creating frequency domain data for comparing with the pre-recorded pre-characterization templates from an induced vibration of the tool on the machine tool when the machine tool is not in operation, the templates having one or more frequency ranges indicating machine tool failure;

a frequency and a time signature trace generator responsive to the output of the vibration sensor for generating frequency and time domain signals representative of the tool;

comparators for comparing the frequency and time domain vibration signatures with the respective frequency and time domain thresholds and templates for indication of tool failure.

7. The system of claim 6 wherein said machine tool monitor system further includes:

an energy band generator for generating energy bands responsive to said frequency signature trace;

at least one threshold detector responsive to said energy band generator for checking the energy band values against the thresholds for detecting high levels of vibration; and

a frequency domain detector responsive to said high levels of vibration detected from said frequency signature trace for detecting high levels of vibration.