汽轮机叶片

检测的过程是对汽轮机叶片的凸面进行检测，从蒸汽的进口和出口开始。图1是汽轮机叶片检测的方案示例图。

使用动力诊断金属磁记忆应力集中磁检测仪专用扫描装置进行检测时，从叶根段到外围段的方向或相反的方向进行扫描。

建议定期对所有叶片执行检测工作。

在检测过程中，金属磁记忆检测仪的屏幕显示了沿着检测长度的磁场（Hp）分布图，检测结果存储在仪器的ROM中，以便在计算机上进一步处理。

The blades opera根据检测结果确定可以在最大应力条件下运行的叶片。这些叶片具有横向于叶片翼的应力集中线（SC线），并且具有最大的磁场（Hp）梯度值：Kin=dHp/dx。表征叶片残余应力水平的基准梯度值Kin的值由操作人员的输入到仪器并显示在屏幕上。提交给客户的叶片检测技术确定了裂纹形成前的梯度极限值。

The scheme of turbine blades inspection — 图1：使用金属磁记忆检测仪和2型扫描装置检测汽轮机叶片的方案：1, 2 - 磁通门传感器； 3 - 长度编码器；4 - 连接电缆； 5 - 金属磁记忆检测仪；6 - 汽轮机叶片。

图2是使用金属磁记忆检测法对Konakovo TPS K-300-240汽轮机低压转子（LPR）第33级的31、32和33号汽轮机叶片进行检测的结果。记录在31号叶片上的磁场（Hp）分布以及相应的磁场梯度值Kin(dHp/dx)（见图2a）表征了其令人满意的状态。从图2b和图2c中可以看出，在距离根部180-200mm处的32号和33号的叶片上记录了磁场（Hp）的突然变化，以及磁场梯度值Kin(dHp/dx)的突然变化。根据现有经验，此类梯度值Kin接近极限值。需要指出的是，梯度值Kin的数值在一定程度上也取决于所用检测仪器的灵敏度。

The distribution of the field. Blade #31. — Fig.2a. The distribution of the Hp field. Blade #31.

The distribution of the field. Blade #32. — Fig.2b. The distribution of the Hp field. Blade #32.

The distribution of the field. Blade #33. — Fig.2c. The distribution of the Hp field. Blade #33.

Fig.3 shows the stress concentration zones with maximal value of the field gradient dH/dx (Kin). Stress concentration zones are detected in the sites of sign alternation of Hp field. Shown Hp=0 lines are detected during the more datailed inspection of the blades #32 and #33. In these stress concentration zones development of the blades damaging should be expected. It is was verified by the metallographic investigation.

SC lines detected on the surface of the blades #32 and #33. — Fig.3. SC lines detected on the surface of the blades #32 and #33: 1 - SC lines.

The pronounced corrosion-fatigue damaging of the metal, corresponding to the pre-failure state of the blades #32 and #33, can be seen on the structure fragments (magnification ×500) shown in fig.4,a and 4,b.

It should be noted that the maximum damaging of the metal's structure was detected in the course of investigations exactly in the zones of the maximum Hp field gradient recorded during the inspection near the SC line. The metal's structure is satisfactory at a distance of just 100 micrometers from the SC line (for example, see the structure fragment shown in fig.4, c for the blade #33).

Of the total number of more than 800 LPR blades, inspected using the metal magnetic memory method, only three blades in the pre-failure state were detected. The non-failure operating time of these blades was ~ 200000 hours at the date of inspection.

In the course of the industrial verification of the considered inspection method Energodiagnostika Co. Ltd. accumulated the data bank of the numerical Kin values for the blades in the pre-failure state for the last stages of the PT-60-130, K-100, K-200 and K-300 turbines.