The Challenge:

Building a robust test system for testing ball bearings, which, can perform various analysis such as Fast Fourier Transform, Order analysis, Orbit plot, Power Spectrum, Power Spectral Density, Micro vibration analysis, Octave, calculate Leq level of sound signal, Force and Unbalance measurements for preventing catastrophic failure.

The Solution:

Developing a flexible, rugged and cost effective “Bearing Evaluation Test System” using National Instrument PXI controller and Dynamic signal acquisition card along with Lab VIEW development environment.

Introduction:

Detecting mechanical faults in bearings and machinery has long been recognized as being important for preventing catastrophic failure and effective maintenance planning. The human senses of sound and touch were the first mechanisms used to detect machinery problems. Electronic sensors have since offered the ability to feel and listen to machinery with more precision, at more locations, and over more time than was ever before possible. Interpretation of the electronic signals delivered by these sensors has provided the maintenance engineer with the diagnostic information necessary to pinpoint bearing faults, thus enabling a more efficient and predictable maintenance effort. However, skilled and trained personnel have been required to effectively interpret this diagnostic information. Developed system delivers the user with direct diagnostic information.

Bearing Geometry

Test System Setup:

Bearing is connected to the brushless dc motor which is driven by the DC servo drive. Speed profile generated by the DAQ card and is set as a set point to the drive. User has the provision to set the PID parameter in the drive based on the bearing geometry.

The test system uses National Instrument PXI-DSA card for measuring accelerometer and microphone signals. Speed of the motor is measured through PXI-6624 counter card. Force measurement from the dynamometer is done through SCXI-1125. Axial and Radial displacements and bearing temperature is measured using SCXI-1102B. PXI-6704 is used for the profile generation. The test system generates a trapezoidal profile for testing the bearing. User has the provision to set different run up and run down rates.

National instrument PXI platform enabled us to implement the timing and synchronization of multiple measurements easily through the PXI start trigger and RTSI.

Diagnosis through signal processing:

System uses the following signal processing, which gives vital and critical information about the bearing element which, in turn, are used to determine the defect with better precision and accuracy.

Processing is carried through the steady state period of trapezoidal speed profile. A base frequency spectrum of the newly overhauled bearing is recorded at its operating speed and this spectrum is used as a reference for diagnosing the bearing under test.

Fast Fourier Transform and Power Spectrum

This processing helped us in finding the amplitude and power of the different bearing component vibration i.e. the cage rotating frequency, the unbalance frequency, the inner race rotation and outer race rotation frequency. Frequency of these bearing components are calculated from the speed measurement and bearing geometry. Bearing component vibration amplitude is compared with the base spectrum with the tolerance for the diagnostics. Similarly the acoustic signal from microphone is also analyzed. All the calculated components are clearly annotated in the graph which makes it easier for our client to identify which component of the bearing constituted to misbehavior during the test.

Order Tracking

This processing helped us in obtaining the frequency response of the different bearing component for the entire operating speed. This processing of accelerometer signal is done during the run up and run down of the bearing test. Order tracking processing enables us to identify the resonance of the each bearing component. Frequency response of the each order is also compared against the base order frequency response for the diagnostic purpose.

Balancing through force measurement

Rotating components experience significant quality and performance improvements when balanced. Balancing is the process of aligning a principal inertial axis with the geometric axis of rotation through the addition or removal of mass. By doing so, the centrifugal forces are reduced, minimizing vibration, noise and associated wear.

Unbalance in the bearing is measured by adding a balanced fly wheel in to it and rotating the wheel at low speed. Static and Dynamic unbalance is measured through Force, Moment and Vibration measurements. First Wheel with the bearing is kept on the dynamometer platform then its rotational axis is aligned with the X axis of Platform. Forces due to the self weight are nullified in the dynamometer. Wheel with the bearing is allowed to rotate at very low speed and following force, moment, vibration and reference position, speed signals are measured. Unbalance mass caused by individual bearing component is calculated by taking FFT of the force and acceleration signals. Unbalance angle is calculated from the reference position measurement and the individual bearing component time domain signal. Axial and radial displacement measurement also reveals the qualitative information about the unbalance.

Orbit Plot

The Orbit plot shows an X vs. Y plot of vibratory displacement of the shaft, the shape of Orbit plot and the position of the tachometer probe “dot” indicate the quantitative measures of faults such as unbalance and /or misalignment. The Orbit plot can be unfiltered or filtered for selected order component.

FFT/Power Spectrum processing of accelerometer and microphone signal enabled us to diagnose the bearing faults like spalling (Pits on the surface), deformed raceways, damaged cage, and cracked race. Force measurement, orbit plot and order tracking enabled us to diagnose the unbalance and misalignment caused by the bearing components. Figure 4 and 5 shows the configuration and the speed profile generation of the Bearing Evaluation Test System.

Conclusion:

Using National Instruments Lab VIEW and the NI PXI platform, we were able to implement a high-performance, flexible and rugged Bearing Test and Evaluation System within a moderate budget. Use of Lab VIEW tools such as sound and vibration and order analysis, advance signal processing and GOOP enhanced the productivity and reduced the time and complexity involved in delivering the software.