JPH09145340A - Detection apparatus for damage of gas-turbine stationary blade - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a detection apparatus by which an operating amount regarding the detection and the measurement of a crack in an inspection is reduced and whose reliability is enhanced by automating an operation by a method wherein a detector which is used to detect the crack is driven to the blade length direction and the circumferential direction of a stationary blade and the image of the detected surface of the stationary blade is processed. SOLUTION: A detector 1 which is used to detect a crack is attached to a driving gear 2. The driving gear 2 is position-controlled by a position control device, and a camera mounting part and a mounting part to an inner wall 9 are provided with electrically-driven motors. A movement in a blade length direction is performed by the motor at the camera mounting part, and a movement in a circumferential direction is performed by the motor on the side of the inner wall 9. In addition, because of the structure of a stationary blade 8, the interval between the inner wall 9 and an outer wall is changed little on the rear side and the surface side of the blade. As a result, both the rear side and the surface side can be observed by the same driving gear. Information, on the surface of the stationary blade 8, which is taken into by the detector 1 is processed by an image processing device 4, and a crack is detected.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a stator vane damage detecting device in a gas turbine, which is damaged by fatigue and creep in a high temperature combustion gas and a large number of cracks are generated.

[0002]

2. Description of the Related Art In a stationary blade of a gas turbine, a large number of cracks are generated on the surface of the member during use due to repeated thermal strains associated with starting and stopping the plant, which are the main damages. Therefore, in order to prevent the plant from being stopped for a long time due to an accident or failure, the plant is stopped at an appropriate time and the members are inspected in order to maintain the safety of operation and the reliability of the equipment. Although the state of crack generation is investigated there, the detection is often performed visually after a penetrant flaw detection test, and the length of all detected cracks is measured to determine the degree of damage. I was evaluating.

[0003]

As described above, at present, the amount of work involved in crack detection during inspection is enormous. On the other hand, there is a demand for shortening the inspection period in order to increase the plant operating rate. In the case of visually detecting and measuring cracks, the data greatly depends on the voluntaryness of the measurer. The arbitrariness also greatly affects the subsequent damage assessment.

The object of the present invention is to reduce the amount of work relating to crack detection / measurement at the time of inspection of a gas turbine stationary blade, and to automate the work to provide highly reliable data that does not include the voluntaryness of the measurer. It is an object of the present invention to provide a detection device and a data processing device capable of obtaining the same.

[0005]

SUMMARY OF THE INVENTION The present invention is directed to a crack detecting device for detecting a crack generated on the surface of a gas turbine stationary blade, and the detecting device includes a detecting device for detecting a crack and a blade length direction and a circumference of the stationary blade. Drive device for driving in the direction, a position control device for controlling the position of the drive device, an image processing device for processing an image of the detected vane surface, and measurement of cracks from the image processed by the image processing device. It is characterized by comprising an arithmetic processing unit for performing a damage degree.

Alternatively, a recording device for recording the shape of the stationary blade as data is provided, and the actual length of the crack generated in the curved portion is calculated from the projected length, the position of the detector, and the shape of the stationary blade. It is characterized in that it has an arithmetic unit for obtaining it.

Using a small camera capable of scanning the surface of the stationary blade and its driving device, the crack generated can be captured as an image, and the image can be processed to automate the detection of the crack and the measurement of the crack length. .

[0008]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing the configuration of the device of the present invention, which is a detector (camera) 1 for detecting a crack, a drive device 2 for driving the detector, a position control device 3 for controlling the position of the drive device, and a detector. The image processing device 4 that performs image processing of the surface information of the stationary blade input by the device and the arithmetic processing device 5 that measures the crack length from the processed image and obtains the damage degree such as the maximum crack length or the sum of the crack lengths. It is composed of the stationary blade shape data 6 in which the blade shape is recorded.

FIG. 2 is a view showing an example in which the apparatus of the present invention is actually applied, and is a view showing observation of the ventral side of the stationary blade. The detector 1 is attached to the drive device 2.
The drive device 2 has an electric motor in the camera mounting portion and in the mounting portion to the inner wall 9, respectively. The movement in the wing length direction is performed by the camera mounting portion motor, and the circumferential movement is performed by the inner wall side motor. Inner wall 9 of static wings
Since the outer wall 7 and the outer wall 7 are concentric, the motor can be provided in either one of the mounting portions, and the bearing with low resistance can be mounted on the other side to drive in the circumferential direction without any problem. In addition, since the distance between the inner wall and the outer wall is almost the same on the blade back side and the ventral side due to the structure of the stationary blade, both the back side and the ventral side can be observed with the same driving device.

During operation, the vane is exposed to a high temperature combustion gas, so that an oxide film is formed. The oxide film often makes it difficult to detect cracks. It is desirable to input the image by removing the oxide film by polishing, but the work amount at the time of inspection will greatly increase. Therefore, when the oxidation is so severe that the image cannot be input, the crack is easily detected by a penetration test or the like before the image is captured.

The surface information of the stationary blade taken in by the detector 1 is processed by the image processing device 4 to detect cracks.
Basically, binarization and differential processing are performed to determine cracks. The problem in the damage evaluation of the vane is a macroscopic crack that has grown to the extent that it opens, and a relatively small crack of about several mm is often ignored in the damage evaluation. Of the cracks that occur parallel to each other at close positions, either one of them often stays, and this is considered to be one crack in evaluation, and there is often no problem. By inputting such a judgment criterion in advance, only the cracks, which are problematic in damage evaluation, are left as an image in the processing by the image processing apparatus.
From the crack image obtained as described above, the arithmetic processing unit 5 obtains parameters for evaluating damage such as the maximum crack length and the total sum of crack lengths.

Since the shape of the stationary blade has many curved portions, it is often difficult to actually input it as an image and obtain the crack length. However, moving the detector to an optimum position for all cracks and inputting an image requires a considerable amount of work, and the purpose of labor saving of inspection work cannot be achieved. As shown in FIG. 3, when the image of the crack generated in the curved portion is directly input, the crack length at the time of input is a straight line A- in the figure.
It will be recognized as a projection length as shown by A '. Therefore, by recording the shape of the stationary blade and providing the program for obtaining the actual crack length from the position of the detector at the time of image input and the direction of the visual field and the input projection length, Address the problem. Specifically, FIG.
, The cracks on the detected image are appropriately divided, and the coordinates c1, c2, c3, ... Of the divided points are obtained,
The actual crack length is obtained by projecting the coordinates c1, c2, c3 ... On the actual shape model from the observation position and the shape data of the stationary blade in the arithmetic processing unit. By adopting this method, the moving range of the detector can be reduced and the inspection work amount can be reduced. Further, when observing the wing portion, the detector is attached so that the inclination of the detector is perpendicular to the wing length direction. Since cracks occur not only in the wings but also in the walls, it is necessary to detect cracks in the walls as well. Since the wall part is almost perpendicular to the blade length direction, cracks cannot be detected at the detector mounting position as it is. Therefore, the detector is tilted to some extent to observe the wall part. When you do so, what is detected as the crack length is
It is the projection length in the direction of the tilted angle of the detector. Therefore, the crack length generated in the wall portion is evaluated by converting the projected length into the actual crack length from the shape of the stationary blade and the angle of the detector in the procedure described above.

Since all of the above work is automatically performed by the device of the present invention, there is no room for the operator's voluntary operation.
The resulting data will be objectively reliable and the original objectives achieved.

[0014]

EFFECTS OF THE INVENTION According to the present invention, crack detection / measurement work for inspection of a gas turbine stationary blade can be saved, and the operating rate of the plant can be improved. Also, by automating the detection and measurement, it is possible to perform highly reliable damage evaluation without the voluntaryness of the measurer.

[Brief description of the drawings]

FIG. 1 is a block diagram of an apparatus of the present invention.

FIG. 2 is an explanatory view of the device of the present invention.

FIG. 3 is an explanatory diagram showing an actual crack length and a detected length.

FIG. 4 is a block diagram showing a procedure for obtaining an actual crack length from a projected length.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Detecting device, 2 ... Detecting device drive device, 3 ... Detecting device position control device, 4 ... Image processing device, 5 ... Arithmetic processing device, 6 ... Stationary blade shape data, 7 ... Outer wall, 8 ...
Static wings, 9 ... Inner wall, 10 ... Cracks.

Claims (2)

[Claims] 1. A detection device for a crack generated on the surface of a gas turbine stationary blade, a detection device for detecting a crack, and a drive device for driving the detection device in a blade length direction and a circumferential direction of the stationary blade. A position control device for performing position control of the drive device,
Damage to a gas turbine vane characterized by comprising an image processing device for processing an image of a detected vane surface and an arithmetic processing device for measuring a crack from an image processed by the image processing device to obtain a damage degree. Detection device. 2. The apparatus according to claim 1, further comprising a recording device for recording the shape of the stationary blade as data, and regarding a crack generated in a curved portion, its projected length, the position of the detector, and the shape of the stationary blade. A damage detection device for a gas turbine vane, which has an arithmetic unit that calculates the actual length from the