Method for detecting structure-borne noise events in a roller bearing

• US 7,599,804 B2
• Filed: 01/23/2004
• Issued: 10/06/2009
• Est. Priority Date: 01/31/2003
• Status: Expired due to Fees

• Current Assignee: FAG Kugelfischer AG (INA-Holding Schaeffler GmbH & Co. KG)
• Original Assignee: FAG Kugelfischer AG (INA-Holding Schaeffler GmbH & Co. KG)
• Inventors: Pecher, Alfred
• Primary Examiner(s): Feliciano; Eliseo Ramos
• Assistant Examiner(s): Henson; Mi'schita'
• Application Number:
  US10/543,844
• Publication Number:
  US 20060150737A1
• Time in Patent Office: 2,083 Days
• Field of Search: 702 33- 36, 702/39, 702/54, 702/56, 735/72, 735/93, 734/60, 734/62, 384/1, 384/18, 384/44, 384/46, 384/48, 384 50- 59
• US Class Current: 702/39
• CPC Class Codes: F16C 19/52 with devices affected by abnormal or undesired conditions
• CPC Class Codes: G01M 13/045 Acoustic or vibration analysis

Abstract

A method for detecting structure-borne noise events in a roller bearing, wherein a measuring signal M of a pressure or dilatation sensor arranged on a roller bearing is fed to a frequency filter for filtering out undesired signal parts. A first variance value is calculated from digital values of the frequency filter output signal and at least one second variance value is calculated from other digital values of the frequency filter output signal and a weighted arithmetic average of the at least two variance values is calculated with the aid of a recursive calculation of the variance value. If the weighted arithmetic variance average value exceeds a pre-selected variance threshold value, then it is assessed that a structure-borne noise event provoked by mechanical damage to the roller bearing has occurred.

Claims

1. A method for detecting structure-borne sound events in a roller bearing, comprising:

supplying a measurement signal (M) from a pressure-sensitive or strain-sensitive sensor which is arranged on the roller bearing to a frequency filter for filtering out undesirable signal components;

a microcomputer calculating a first variance value from digital values of the frequency filter output signal;

the microcomputer determining at least one second variance value from further digital values of the frequency filter output signal;

the microcomputer calculating a weighted arithmetic mean of the first variance value and one of the at least one second variance value using a recursive calculation; and

the microcomputer assessing any exceeding of the weighted arithmetic variance mean value above a preselected variance threshold value (V_SW) as a structure-borne sound event which has been caused by mechanical damage to the roller bearing, and the frequency for exceeding the threshold can be associated with the point of origin of the damage in the bearing when a rotation speed and geometric conditions of the bearing are known;

wherein the weight used in calculating the weighted arithmetic mean of the variance values is a constant with respect to a particular recursive calculation.

2. The method as claimed in claim 1, comprising calculating the weighted arithmetic mean of the variance values using the basic recursive equation:

Ē

{X²}(k+1)=Ē

{X²}(k)+C_x²[X²(k+1)−

Ē

{X²}(k)] 

 

(Equation

     1)

where Ē

{.} represents the expected value of the weighted arithmetic mean value, k is a running variable, x is a digital sample value of the frequency filter output signal, and the weight is denoted by C_x, which is an adaptation constant.

3. The method as claimed in claim 2, wherein the adaptation constant C_x has a value which is less than unity and greater than zero.

4. The method as claimed in claim 3, wherein the adaptation constant C_x can be calculated from an equation for the adaptation rate

t={1/C_x²−

½

}*T 

 

(Equation

     2)

where t indicates how quickly the true variance can be determined with a tolerable error rate and T is the time between consecutive digital sample values of the frequency filter output signal.

5. The method as claimed in claim 2, wherein the basic recursive equation (equation 1) is expressed as a simplified recursive equation

nE=aE+aK*(nA−

aE) 

 

(Equation

     3)

in which nE represents a new result value of the variance, aE represents an old result value of the variance, aK represents the adaptation constant, and nA represents a new sample value of the frequency filter output signal.

6. The method as claimed in claim 5, wherein the simplified recursive equation is expressed as an evaluation equation

nV=aV+aK*(nA−

aV) 

 

(Equation

     4)

where nV represents the new variance, aV represents the old variance, aK represents the adaptation constant, and nA represents the new sample value of the frequency filter output signal.

7. The method as claimed in claim 6, further comprising assessing the value of the new variance as an estimated value of the true variance of the measurement signal, which marks the presence of structure-borne sound events in the roller bearing.

8. The method as claimed in claim 6, wherein a high new variance value represents a high probability of the presence of structure-borne sound events in the roller bearing.

9. The method as claimed in claim 6, wherein a high new variance value represents major damage in the roller bearing.