Developing an Intelligent Condition Monitoring System

  Read in   |   Print Print

"We seamlessly integrated LabVIEW software with Nl data acquisition and real-time hardware through appropriate driver packages to reach our goals and obtain a scalable, flexible solution."

- Grigore Stamatescu, Assistant Professor , Department of Automatic Control and Industrial Informatics, University Politehnica of Bucharest

The Challenge:
Developing a flexible, scalable laboratory stand for vibration measurements and condition monitoring of mechanical systems.

The Solution:
Using NI software and hardware to integrate high-performance data acquisition and process accelerometer signals through a virtual instrumentation project from a real-time controller.

Author(s):
Grigore Stamatescu, Assistant Professor - Department of Automatic Control and Industrial Informatics, University Politehnica of Bucharest
George Sterpu - Department of Automatic Control and Industrial Informatics, University Politehnica of Bucharest
Valentin Sgârciu - Department of Automatic Control and Industrial Informatics, University Politehnica of Bucharest

The Intelligent Measurement Technologies and Transducers (iMTT) laboratory is part of the Department of Automatic Control and Industrial Informatics at the University Politehnica of Bucharest in Romania. Laboratory users address topics like fundamentals of sensors and transducers, data acquisition systems, process instrumentation, virtual instrumentation, intelligent sensors, and wireless sensor networks. An NI LabVIEW Academy has operated at iMTT since 2010. This paper discusses the development of a new laboratory stand, the Vibration Analysis System (VIBSYS), for studying and researching acceleration transducers, data acquisition, and signal processing techniques, and leading toward the implementation of an intelligent condition monitoring system for mechanical structures. We focus on usability and a quick learning curve for building applications by using NI software and hardware.

Figure 1 shows the functional system diagram. A Siemens 0.18 kW, 3000 rpm syncron rotation electric motor drives a shaft with an unbalancing flywheel through an elastic coupling. A frequency converter implements variable drive speed. We fit two bearings (RL1 and RL2) at both ends of the shaft. Our main goal was to acquire the bearing vibration signature, in different configurations of the system and with various installed parts, and to process it to determine possible faults on its components. The two integrated electronics piezo electric (IEPE) accelerometers (VIB1 and VIB2) are essential to this goal and provide a 100 mV/g signal with a 0.002 g resolution. These sensors require a constant current source, built-in or external, for operation and give the output as a modulated bias voltage. An optical sensor tachometer provides the rotation speed of the flywheel. Figure 2 shows a picture of VIBSYS installed in a laboratory setting.

 

Figure 1. VIBSYS Diagram

Figure 2. Laboratory Implementation

We found the NI 9232 dynamic signal acquisition C Series module was best suited for the design of the data acquisition subsystem. The device is a 3-channel, ±30 V analog input module that samples at 102.4 kS/s, which supports software-selectable IEPE signal conditioning and 24-bit resolution. The acceleration sensors interface with the inputs AI0 and AI1 of the module and the discrete signal of the speed sensor connects to AI2. Early implementations used the NI 9232 module in combination with an NI cDAQ-9181 1-slot Ethernet chassis.

We used the Measurement & Automation Explorer along with the NI-DAQmx driver to perform the initial set-up and configuration. Recently, we shifted to an NI cRIO-9076 chassis. The CompactRIO system offers the full capabilities of a programmable automation controller and enhances the versatility of our system since it can operate both as a complex data acquisition device and as a fully autonomous condition monitoring system with local processing and communication for integration in complex monitoring and automation scenarios. The NI 9232 module can operate in only FPGA mode; therefore, we developed a suitable application that periodically samples the three analog inputs of the device. We integrated under the same LabVIEW project.

We developed a LabVIEW software project as a companion to the hardware design. Figure 3 illustrates the main virtual instrument front panel of the Envelope Examination System (ELEOS). The application user can first configure the procedure by input of the specific bearing characteristics such as the number of rolling elements, pitch diameter, ball/roller diameter, and contact angle. The system can also acquire a working dataset from the platform. The user presets the rotation speed depending on the static frequency converter setting, but can also control the speed using an analog output signal to the converter board. Based on these parameters, we computed specific defect characteristic frequencies.

Figure 3. ELEOS Software Project

For the defect detection procedure, we implemented a custom version of the one Randall gave in Vibration-Based Condition Monitoring: Industrial, Aerospace, and Automotive Applications. We intuitively introduced the signal processing steps using a tabbed view for each of the significant operations including order analysis, linear prediction, minimum entropy deconvolution, spectral kurtosis, and envelope analysis. The application can log to NI Technical Data Management Streaming (TDMS) files, load historical data, and simulate signals when no plant is connected. For implementing the signal processing, we rely on functions from the NI Sound and Vibration Toolkit such as the Analog Tacho Processing ExpressVI, as well as custom The MathWorks, Inc. MATLAB® software scripts running in dedicated block diagram MathScript Nodes that call specific functions through the MATLAB software.

Relevant to the software design is the implementation using a queued state machine design pattern with two parallel while loops. We used a waiting queue so no states are lost in the case of a conflict, and the processor runs efficiently. We retained states in a typedef enum structure for scalability.

The bearings installed on the VIBSYS are type 6004 deep groove ball bearings with the following dimensions: D=40 cm, d=20 cm, and w=12 cm. We obtained results for NSK and CMB bearings. We consider the NSK good bearings and the CMB bad bearings. Figure 4 shows the final results of the signal processing in the form of the signal envelope. From the envelope spectrum, we can observe that, for the used bearings, the center frequency is fc=9,500 Hz with a 2,000 Hz range, and a fundamental frequency appears at 76.8 Hz along with harmonics at 153.6 Hz, 230.4 Hz, and 306.9 Hz. We defined four fundamental frequencies and associated them to places on the bearing where defects might appear. These depend on the bearing geometry, number of balls, cage and ball diameter, and the shaft rotation frequency. Given the characteristic frequency, this leads to the conclusion of a defect in the outer rim of the bearing, expressed as 

.

We can monitor peaks on the envelope spectrum to evaluate the severity of the defect.

Figure 4. Used and Working Bearing Envelope

This case study focused on the design and evaluation of an intelligent laboratory system for machine condition monitoring, VIBSYS. We seamlessly integrated LabVIEW software with Nl data acquisition and real-time hardware through appropriate driver packages to reach our goals and obtain a scalable, flexible solution. In the future, we aim to extend the system with additional sensors, build a knowledge base with identified defects, and perform a full real-time implementation of the signal processing steps on the CompactRIO platform.

Author Information:
Grigore Stamatescu, Assistant Professor
Department of Automatic Control and Industrial Informatics, University Politehnica of Bucharest
Splaiul Independenței 313
Bucharest 060042
Romania
grigore.stamatescu@upb.ro

Bookmark and Share


Explore the NI Developer Community

Discover and collaborate on the latest example code and tutorials with a worldwide community of engineers and scientists.

‌Check‌ out‌ the‌ NI‌ Community


Who is National Instruments?

National Instruments provides a graphical system design platform for test, control, and embedded design applications that is transforming the way engineers and scientists design, prototype, and deploy systems.

‌Learn‌ more‌ about‌ NI