Without disassembly and inspection, it is very important to identify or predict whether the bearing is faulty in operation, which is very important to improve productivity and economy.
1) Recognition by sound
Recognition by sound requires a wealth of experience. Must be adequately trained to recognize bearing and non bearing sounds.
Therefore, special personnel should be assigned to carry out this work as far as possible. The sound of the bearing can be clearly heard by sticking the listener or listening stick to the shell.
2) Identification by operating temperature
This method is a comparative identification method, which is only used in the situation where the running state does not change much. For this purpose, continuous temperature recording must be carried out.
In case of failure, the temperature will not only rise, but also change irregularly.
This method should be combined with voice recognition method.Fault Identification Method of Bearing
3) Identification by lubricant status
The lubricant shall be sampled and analyzed to judge whether it is polluted or mixed with foreign matters or metal powder.
This method is especially effective for bearings that cannot be close to observation or large bearings.
This data set is more used to detect the remaining useful life (rul) of the bearing than to detect the bearing fault.Fault Identification Method of Bearing
It mainly provides real accelerated aging bearing data under different conditions. The sensors that collect data are rotor sensor and pressure sensor.
While collecting vibration data at high frequency, it also monitors the temperature data of the bearing, which is convenient for the neural network to determine the rul of the bearing from multiple physical characteristics.
Lubricating oil plays the role of lubrication, cooling, cleaning and sealing. The quality of lubricating oil is directly related to the service life of the engine. Therefore, do not blindly choose engine oil. New vehicles have very strict requirements for lubricating oil. What lubricating oil to choose, when to replace and how to replace should be carried out in accordance with the regulations of the manufacturer.
Engine oil works under harsh high temperature conditions, which is easy to oxidize and deteriorate, produce acidic substances and stack them into polymer deposits, so as to destroy the lubrication. In addition, the combustion exhaust gas channeling into the crankcase will also promote the oxidation and deterioration of the engine oil. Therefore, after a certain period of use, the engine oil will usually blacken and deteriorate and lose its due lubrication.
However, with the continuous improvement of lubricating oil quality, especially the popularization and use of thickened engine oil, the blackening of engine oil does not necessarily mean the deterioration of engine oil. This is because the clean dispersant added to the thickened engine oil disperses the deposits on the parts into the oil, making the engine oil black.
Whether it has deteriorated or not, the test of various indicators shall be carried out, and the oil shall be changed regularly according to the quality to reduce unnecessary waste. For example, the engine oil added with yyl02 and other engine protective agents will also blacken after being used for a certain period of time. Therefore, the blackening of engine oil does not mean the deterioration of engine oil.
1. Sliding: the current fault detection is based on the assumption that there is no sliding between the rolling part and the bearing raceway, but in reality, this kind of sliding is not uncommon
2. Frequency interaction: when multiple faults occur at the same time, they will affect each other, thus blurring the information frequency of a single fault.Fault Identification Method of Bearing
3. External vibration: in the real environment, in addition to the detected bearing, there are likely to be other vibrations to interfere, while the anti noise ability of ordinary ml is not strong.Fault Identification Method of Bearing
4. Observability of fault: when some faults occur, they will not produce a separate vibration frequency. Some faults have no obvious impact on the vibration frequency. Ml, which relies on the vibration frequency as the detection subject, is not comprehensive enough for bearing fault detection.Fault Identification Method of Bearing
5. Sensitivity: under different environmental conditions, the sensitivity of each fault type is also different. Before being put into use, the sensitivity of each fault type under different environments needs to be collected.Fault Identification Method of Bearing
Due to the characteristics of ML, such as difficult interpretation, unstable accuracy and weak anti noise robustness, ml was not put into use in large quantities and was gradually replaced by later dl.Fault Identification Method of Bearing
Copyright notice: This article is the original article of CSDN blogger “xd_onmyway”, which follows the CC 4.0 by-sa copyright agreement. For reprint, please attach the original source link and this notice.Fault Identification Method of Bearing