In order to improve the production efficiency of gear machining and ensure the surface integrity of the gear, the gear machining is usually divided into rough machining and finish machining. For the common gear machining methods. During rough machining, the feed rate is large and the production efficiency is high. It is mainly to finish the rough outline required before forming. The requirement for surface integrity is not high. It is mainly to control the surface layer within the final size requirement of forming process without introducing structural defects and excessive residual tensile stress. The measures adopted are high-speed fast grinding of grinding wheel and high-speed cooling and lubrication; However, the feed rate and grinding speed are very small in finishing, which means that the typical “grinders” can produce fine work, and it is necessary to ensure that the surface residual stress is in a very small value or residual compressive stress state. No matter in rough machining or finish machining, when the contact surface is large and the cooling and lubrication conditions are not good, the surface is prone to burn, which will have a negative impact on the bearing and service life of the gear, especially the critical heavy-duty gear surface is not allowed to have burn defects, so it is very important and necessary for the on-line detection and monitoring of surface integrity in the machining process.
In order to ensure the surface integrity of gears, the machining process parameters must be controlled; in order to ensure that there is no burn or excessive residual tensile stress in the machining process, the parameters of surface integrity must be detected online and the machining process must be controlled by feedback.
The surface defects of gears are easy to occur in the process of machining. The common surface defects of machined gears are surface decarburization or oxidation, surface burn or soft spot, surface micro crack and so on. Decarburization or oxidation of the surface and burn of the surface may lead to the appearance of black tissue, the decrease of hardness or soft spots, the decrease of surface strength and the deterioration of anti bending and contact fatigue performance. In serious cases, the failure phenomena such as broken teeth or falling off of the tooth surface may occur. When the surface grinding process parameters are improper or the cooling and lubrication are not good, on the one hand, when the burn is not serious, the surface layer may have large residual tensile stress, on the other hand, when the burn is serious, micro or macro cracks will be formed, which is called grinding burn cracks. In order to ensure the surface integrity of the gear, it is not only necessary to test the surface integrity of the parts after final processing and forming, but also need to carry out on-line detection in the process of machining. The on-line monitoring of the gear is very important, for which a complete detection equipment system and method as shown in Figure 6 have been established in foreign countries.
From the current available methods of gear surface integrity detection, it can be divided into two levels: the detection technology that can be detected in industrial production and has been mature and applied, and the detection technology that needs to be carried out in the laboratory and has been mature and available.
The common methods to detect surface integrity damage in industry include visual inspection, crack inspection, corrosion (nitric acid acetic acid solution is commonly used for gear steel materials), acoustic emission, power consumption, Barkhausen, etc.; the methods to detect surface integrity damage in laboratory include X-ray analysis, hardness test, metallurgical analysis, etc.
In order to facilitate the detection of the surface integrity of gear parts, a practical and simple Barkhausen analysis method  has been developed abroad (such as Finland), and has been successfully applied in industrial production, as shown in Figure 7, manual detection, automatic detection and robot online detection. According to the results of many tests, the processing technology can be evaluated and analyzed correctly and timely according to the properties and values of residual stress.