**1. Introduction**

The safety and reliability of equipment in the working environment has attracted the attention of scholars. However, there is no absolute method for evaluating reliability, so finding a feasible method appears to be particularly urgent. To solve that problem, lots of research has been designed and carried out, and it indicates that stress could be seen as a parameter to characterize reliability, so the method of stress evaluation was studied. In general, the method of stress evaluation can be divided into two categories, including non-destructive and destructive methods [1–7]. Because of non-destruction, low cost, high efficiency and other merits, the metal magnetic memory (MMM) technology, which is one kind of magnetic flux leakage method, becomes a commonly used method for stress evaluation of ferromagnetic material.

Based on the magnetic memory effect of ferromagnetic material in the geomagnetic field, the MMM technology was firstly proposed by a Russian scholar [8–11], and it stated clearly that the normal component of MMM signal—*H*p(*y*) signal, which was defined in the geomagnetic field, could be employed to evaluate stress. Up to now, lots of research results have been found, however lots of influential factors are found in the application of stress evaluation with this method, and its mechanism is still not elaborated clearly. Thus, lots of experimental and theoretical studies were carried out. For example, the change in the orientation of magnetic domain was explained as the reason for evaluating stress with this technology [12–14]. The results of stress effect on *H*p(*y*) signal and its influence mechanism were both clarified qualitatively [15]. The grain size effect on *H*p(*y*) signal for stress evaluation of low carbon steel was analyzed, and the relation of magnetic intensity gradient and grain size was also obtained [16]. As fatigue cycle numbers increased, the change in the maximal value of surface magnetic flux density of 304 stainless steel was discussed, and the phase transformation was seen as the main reason [17]. Besides that, the fatigue stress of ferromagnetic material was also evaluated, and the relation of fatigue cycle and *H*p(*y*) signal was discussed [18–22]. While it should be noted that defects exist inevitably in material, its effect on stress evaluation is discussed rarely, so designing the method for discussing its influence on stress evaluation and explain its results clearly are very important. As a typical defect, the influence of crack on stress evaluated with MMM technology was studied, and its influence degree affected by the crack size was explained in this study. To solve the problem, the equivalent theory was employed, and the regular rectangular groove is used to simulate cracks. After that, the *H*p(*y*) signals were collected with the uniaxial static tensile experiment, and then the parameter of *H*p(*y*) signal for characterizing stress was determined. At last, the relations between the characteristic parameter of *H*p(*y*) signal and crack size was obtained.
