introduction
High-speed gearboxes, with their advantages of high power, high speed, light weight, and low noise, have become key components in large-scale equipment in industries such as petroleum, chemical, power plants, and metallurgy. Their performance directly affects the normal operation of the entire equipment. Currently, domestic production of high-speed gearboxes has reached a considerable scale, with power outputs up to 36,000 kW and linear speeds up to 152 m/s, playing a significant role in various industries.
After the high-speed gearbox is assembled, it is necessary to perform necessary performance tests. Therefore, it is essential to develop a high-speed gearbox performance testing system [1] . The testing system introduced in this paper realizes the functions of on-site real-time performance testing, remote real-time monitoring, and viewing historical tests. It can effectively test the performance of high-speed gearboxes.
1 System Overall Design
The overall system architecture is shown in Figure 1 .
The entire system is divided into two parts: ① Measurement and control subsystem ② Remote monitoring subsystem.
Figure 1 System Overall Architecture
The measurement and control subsystem comprises two parts: testing and control, enabling real-time performance monitoring on-site. The test is divided into three cycles: a running-in test cycle, an no-load test cycle, and a type test cycle. The measurement and control subsystem controls each cycle to operate according to a pre-set test curve, during which data is acquired, analyzed, and processed. It calculates the corresponding vibration and noise values, and plots temperature rise and transmission efficiency curves, among others.
The remote monitoring subsystem includes a network service subsystem and a client monitoring subsystem. The network service subsystem mainly consists of a web server and a database server, primarily running the monitoring system program and the database program. The client monitoring subsystem is mainly used for remote real-time monitoring of high-speed gearbox performance test results by clients within a local area network (LAN) or wide area network (WAN), as well as displaying historical test results.
2 Measurement and Control Subsystem
2.1 System Control Principle Diagram
The system control principle diagram is shown in Figure 2. First, the industrial control computer communicates with the frequency converter, PLC, torque meter and electric load through the serial port. The high-speed gearbox device is controlled to run according to the preset curve. Due to the large noise interference on site, RS-485 serial port is used. Temperature, noise and vibration signals are collected during the test. Data acquisition is the key. The quality of signal acquisition directly affects the test results. Therefore, before the signal is acquired, it is first processed by the signal preprocessing card for signal amplification and filtering, etc., in order to obtain a suitable signal [2] .
2.2 Layout of measuring points
(1) Arrangement of temperature measurement points
During gearbox operation, the temperature of key components effectively reflects the gearbox's performance. For example, friction between gears causes the lubricating oil temperature to rise; normally, the gearbox lubricating oil temperature should be kept within allowable limits. However, under abnormal operating conditions (such as design or manufacturing errors), the gearbox temperature can quickly exceed its limits. Therefore, temperature is a critical test parameter. The measurement point arrangement is shown in Table 1.
(2) Arrangement of measuring points for box vibration measurement
During gearbox operation, gear meshing, friction between parts, and load all induce internal stress, which is then transmitted to the gearbox housing through the bearings, causing vibration. Therefore, when measuring gearbox vibration, a triaxial accelerometer is installed at the bearing to simultaneously measure vibration in three directions. The measurement point arrangement is shown in Table 2.
(3) Arrangement of measuring points for shaft vibration and online clearance measurement
During gearbox operation, shaft vibration reaches its maximum when the gearbox speed reaches the critical speed. At this point, the deflection increases significantly, causing severe vibration of the supports, resonance, and potentially affecting the entire unit and plant, leading to destructive accidents. Therefore, shaft vibration signals are collected during the gearbox's acceleration process. A curve is then plotted with rotational frequency as the horizontal axis and the relative displacement in the vertical direction of the shaft as the vertical axis to identify the critical speed, aiming to avoid this speed during normal operation. Simultaneously, the shaft's online clearance is measured during the experiment to analyze its relationship with temperature. Eddy current sensors are used for non-contact measurement of relative displacement. The measurement point layout is shown in Table 3.
2.3 Noise Test
Noise testing employs the sound intensity method, which calculates sound power using sound intensity as the measured quantity. There are two methods: the discrete point method and the scanning method. The scanning method is chosen because it offers shorter testing time and higher accuracy.
(1) Noise test procedure
The noise test consists of six steps, as shown in Figure 3. First, the reference body and measurement plane are determined based on the shape and dimensions of the gearbox. Then, parameters such as the scanning method, speed, and route are set to meet the requirements of GB/T16404.2—1999. Finally, the sound power level is calculated to obtain the noise level.
(2) Noise test hardware design
The hardware required for noise measurement using the scanning method includes: a sound intensity probe assembly, a data acquisition unit, and a sound intensity calculation module.
2.4 Software Design
Figure 4 shows the software design block diagram of the measurement and control subsystem, which mainly includes: data acquisition, data analysis, serial communication, and database.
The entire measurement and control software is programmed based on the LabVIEW platform. LabVIEW uses a graphical programming method, which is simple, intuitive and easy to understand. It simplifies the development process of measurement and control systems, shortens the system debugging and development cycle, and can quickly realize the display of real-time data such as pressure, flow rate, speed and power using virtual instruments [4,5] .
The software adopts a modular structure, mainly composed of modules for parameter setting, serial communication, data analysis, and database. The modules are described below:
(1) Data collection
During the testing process, temperature, noise, and vibration values at the measurement points need to be collected in real time to prepare for subsequent data analysis. The data acquisition module developed based on LabVIEW consists of four parts: AI Config, which configures sampling parameters, device parameters, buffer size, etc.; AI Start, which sets the sampling frequency and starts data acquisition; AI Read, which reads the acquired data array from the buffer for analysis; and AI Clear, which releases system resources after acquisition is complete.
(2) Data Analysis
Vibration and noise levels are important test parameters for evaluating the performance of a high-speed gearbox under test. The program uses the Sound and Vibration toolkit provided by NI, applying its functions to calculate the vibration RMS value and the equivalent sound intensity of the noise, respectively. Correct use of these parameters requires configuring the channel info, including sensor sensitivity and signal preprocessing card gain information.
(3) Serial communication
Multithreading is one of the core technologies in advanced programming and a major way to improve application efficiency and performance. By applying multithreading, the operating system can handle multiple tasks simultaneously.
Multithreading can also be used in LabVIEW. In the main program framework, one While loop corresponds to one thread. If each serial port communication corresponds to one thread, plus data acquisition, the program would need six threads, which is obviously inefficient. To optimize the program, multiple serial port communications are placed in one thread, with each serial port performing read-then-write operations sequentially, the write operation determined by a condition.
(4) Database
This system uses two Oracle databases: one to store gearbox parameters, test curves, and performance testing standards; and the other to store test information and results. The databases can be easily accessed using the LabSQL toolkit within LabVIEW.
3 Remote Monitoring Subsystem
3.1 Implementation Mode of Monitoring System
We adopt the B/S model as the system development model. The developed monitoring system program and database are both running on the server. The client connects to the server via the Internet to realize remote monitoring. With the B/S model, the system development cost and maintenance cost are low, the system has good portability, and the system has high security [6] .
3.2 System Functional Modules
The remote monitoring subsystem primarily enables remote real-time monitoring and display of historical monitoring results. It comprises three main functional modules: system settings, real-time monitoring, and historical data recording. The system functional module block diagram is shown in Figure 5.
The functional modules are described below:
(1) System settings
It mainly manages user information and login status of the login system, including two sub-modules: user management and system logout.
(2) Real-time monitoring
This module allows the client to read real-time temperature, noise, and vibration data via the network. After data processing, the required monitoring interface is generated.
(3) Historical testing shows
This functional module primarily enables users to query monitoring results for a specific period in the past. The client reads historical data stored in the database via the network, and after data processing, generates the required monitoring interface.
4. Conclusion
(1) The system can not only display signal data and curves in real time on site, but also enable remote monitoring of test data by the client.
(2) The acquisition, analysis and processing of field data are based on LabVIEW programming. LabVIEW uses the G language, which is simple to program and speeds up the development cycle.
(3) Remote monitoring is based on B/S structure, which has low requirements for the client and is easy to maintain and upgrade the system. At the same time, it utilizes the advantages of Java language to make the system have good portability.
