Method of anticipating machine failure

Claim 2

In a method of operating a machine producing body waves in a machine part and in which a measurement is made of a first body-wave parameter at a relatively high-frequency range of a frequency spectrum of said body waves and a second body-wave parameter is measured in a low-frequency range of said frequency spectrum, the improvement which comprises: generating an auto-spectral density value G 11 corresponding to the measurement of said first parameter; generating a second auto-spectral density value G 22 from said measurement of said second body-wave parameter; generating a cross spectral density value G 12 from both of the measurements; automatically forming the coherence function γ 2 =G 12 2 /G 11 ·G 22 ; and detecting an increase in the value of γ 2 as an anticipation of impending failure of said machine.

Claim 3

The improvement defined in claim 1, further comprising the step of cutting off operating of said machine upon a rise in the value of γ 2 .

Claim 4

The improvement defined in claim 1, further comprising the step of generating a product γ 2 ·G 22 and detecting an increase in the value of said product as representing anticipatory failure of said machine.

Claim 5

The improvement defined in claim 1, further comprising the step of terminating operation of said machine upon an increase in said product.

Claim 6

In a system for monitoring a machine generating body waves in operation for anticipation of failure of said machine, wherein a first body-wave parameter is measured on said machine in a relatively high-frequency range of a frequency spectrum of said body waves and a second body-wave parameter is measured on said machine at a relatively low frequency of said frequency spectrum, the improvement which comprises: means for deriving respective high-frequency and low-frequency values from said measurements; a frequency analyzing computer connected to said means for generating from said values an auto-spectral density function G 11 from said high-frequency value, a second auto-spectral density function G 22 from said low-frequency value, and a cross spectral density function G 12 from both values; and a function generator connected to said computer for generating the function γ 2 =G 12 2 /G 11 ·G 22 , where an increase in the value of γ 2 anticipates a defect in said machine.

Claim 7

In a system for monitoring a machine generating body waves in operation for anticipation of failure of said machine, wherein a first body-wave parameter is measured on said machine in a relatively high-frequency range of a frequency spectrum of said body waves and a second body-wave parameter is measured on said machine at a relatively low frequency of said frequency spectrum, the improvement which comprises: means for deriving respective high-frequency and low-frequency values from said measurements; a frequency analyzing computer connected to said means for generating from said values an auto-spectral density function G 11 from said high-frequency value, a second auto-spectral density function G 22 from said low-frequency value, and a cross spectral density function G 12 from both values; and a function generator connected to said computer for generating the function γ 2 =G 12 2 /G 11 ·G 22 , where an increase in the value of γ 2 anticipates a defect in said machine.

Claim 8

The system defined in claim 5, further comprising means responsive to an increase in a value of γ 2 for automatically cutting off operation of said machine.

Claim 9

The system defined in claim 5 wherein the means for generating said high-frequency value includes a demodulator.