JP2013079917A - Remaining life diagnosis apparatus for metal in which creep damage develops - Google Patents

Abstract

PROBLEM TO BE SOLVED: To facilitate remaining life diagnosis processing when diagnosing a remaining life of metal in which creep damage develops by a D parameter method.SOLUTION: A remaining life diagnosis apparatus includes a diagnosis computer 3 that performs the following: acquiring grain boundary image data corresponding to a plurality of grain boundaries on a metal surface (S3); acquiring grain boundary data including a start point coordinate and an end point coordinate of each of the grain boundaries from the grain boundary image data (S4); setting reference direction data indicating a reference direction perpendicular to a stress direction of piping in the grain boundary image data (S5); extracting perpendicular grain boundary data pieces within a predetermined angle range centering on the reference direction from the plurality of boundary data pieces on the basis of the reference direction data (S5), determining whether the perpendicular grain boundary data is damaged grain boundary data corresponding to damaged grain boundaries on the basis of a voids formation state on grain boundaries corresponding to the perpendicular grain boundary data (S6); and diagnosing a remaining life of the metal by using a D parameter method on the basis of the number of the grain boundaries corresponding to the perpendicular grain boundary data and the number of grain boundaries corresponding to the damaged grain boundary data (S8).

Description

  The present invention relates to an apparatus for diagnosing the remaining life of a metal that is subject to creep damage.

  In plant equipment such as boilers and turbines in power generation facilities, high-temperature and high-pressure fluid such as power steam is handled. When used for a long time in units of years, the metals (for example, chromium molybdenum steel and stainless steel) constituting the boiler and turbine are subject to aging damage such as creep damage. Excessive damage leads to breakage of the equipment. Therefore, in order to use the plant equipment safely for a long period of time, it is required to diagnose the remaining life of the metal and perform maintenance at an appropriate time.

  Here, if the maintenance cycle is set short, the safety factor increases, but the number of maintenance increases, so the cost increases. On the contrary, if the maintenance cycle is set to be long, the number of maintenance can be reduced and the cost can be suppressed, but the safety factor is lowered. For this reason, it is important to improve the diagnosis accuracy of the remaining life and to appropriately determine the maintenance cycle.

  In order to improve diagnosis accuracy, various remaining life diagnosis methods have been developed. One of the diagnostic methods is the D parameter method. As shown in Non-Patent Document 1, in the D-parameter method, a vertical grain boundary in a direction perpendicular to the stress direction and a damaged grain boundary having a high degree of damage are determined using a micrograph obtained by photographing the grain boundary. The remaining life is obtained using the ratio of the total number and the total number of damaged grain boundaries as an index.

Kenji Kikuchi, Yoshiyuki Kaji, Materials, 44-505 (1995), 1244

  Since the D parameter method focuses on the degree of damage at the vertical grain boundaries perpendicular to the stress direction, there is an advantage that a highly accurate diagnosis can be performed. On the other hand, it is necessary for an operator to determine whether or not the grain boundary is a vertical grain boundary for many grain boundaries.

  The present invention has been made in view of such circumstances, and an object thereof is to provide a remaining life diagnosis apparatus capable of easily performing a remaining life diagnosis by the D parameter method.

  In order to achieve the above object, the present invention is a remaining life diagnostic apparatus for diagnosing the remaining life of a metal subject to creep damage by a computer, and obtained by taking a microscope image of a replica of the metal on the computer. Grain boundary image data obtaining step for obtaining grain boundary image data corresponding to a plurality of grain boundaries on the metal surface, and obtaining grain boundary data including start point coordinates and end point coordinates of each grain boundary from the grain boundary image data. Based on the reference direction data, the reference direction data setting step for setting the grain boundary image acquisition step, the reference direction data indicating the reference direction perpendicular to the stress direction of the metal to the grain boundary image data, and the plurality of the reference direction data. Vertical grain boundary data extraction step for extracting vertical grain boundary data belonging to a predetermined angle range centered on the reference direction from the grain boundary data of And a damaged grain boundary determining step for determining whether the vertical grain boundary data is damaged grain boundary data corresponding to the damaged grain boundary based on a void formation state on the grain boundary corresponding to the vertical grain boundary data; A remaining life diagnosis step of diagnosing the remaining life of the metal using a D-parameter method based on the number of grain boundaries corresponding to the vertical grain boundary data and the number of grain boundaries corresponding to the damaged grain boundary data; It is characterized by making it carry out.

  According to the present invention, the vertical grain boundary data is extracted in the vertical grain boundary data extraction step by comparing the reference direction data set in the reference direction data setting step and the grain boundary data obtained in the grain boundary data obtaining step. It can be done easily. Further, the determination in the damaged grain boundary determination step can be facilitated based on the extracted vertical grain boundary data and void information. As a result, when performing the D parameter method using the number of grain boundaries corresponding to the vertical grain boundary data and the number of grain boundaries corresponding to the damaged grain boundary data, the remaining life diagnosis process can be facilitated.

  In the above remaining life diagnosis apparatus, it is preferable that the reference direction data is repeatedly set while changing the position in the stress direction in the reference direction data setting step. Thereby, vertical grain boundary data can be reliably extracted from a plurality of grain boundary data.

  In the above-mentioned remaining life diagnosis device, in the damaged grain boundary determination step, whether the vertical grain boundary data is the damaged grain boundary data based on a ratio between the grain boundary length and the void length corresponding to the vertical grain boundary data. It is preferable to determine whether or not. Thereby, it can be judged reliably whether it is damage grain boundary data.

  In the aforementioned remaining life diagnostic apparatus, in the grain boundary image data acquisition step, a plurality of grain boundary image data is acquired by performing microscopic imaging while changing the imaging target position, and in the reference direction data setting step, Preferably, the reference direction data is set for each of a plurality of grain boundary image data, and the vertical grain boundary data is extracted for each of the plurality of grain boundary image data in the vertical grain boundary data extraction step. Thereby, it is not necessary to perform tiling for laying and joining images taken with a microscope, and work efficiency can be improved.

  In the above-mentioned remaining life diagnosis apparatus, in the remaining life diagnosis step, a master curve indicating a correlation between the D parameter and the time fraction is acquired in advance, the number of grain boundaries corresponding to the vertical grain boundary data, and It is preferable to diagnose the remaining life of the metal by applying a D parameter based on the number of grain boundaries corresponding to the damaged grain boundary data to the master curve. Thereby, the remaining life can be easily diagnosed.

  According to the present invention, the remaining life diagnosis process can be facilitated when diagnosing the remaining life of a metal that is damaged by creep by the D parameter method.

It is an external view which shows the structure of a remaining life diagnosis system. It is a block diagram which shows the structure of a remaining life diagnosis system. It is a flowchart explaining the remaining life diagnosis procedure. It is a conceptual diagram explaining the some SEM image image | photographed by the matrix form. It is a conceptual diagram explaining the one part superimposed SEM image. It is a figure explaining the duplication part of each SEM image. It is a conceptual diagram explaining the SEM image which performed grain boundary interpolation and dust removal. It is a conceptual diagram explaining extraction of a vertical grain boundary. It is a figure explaining the vertical grain boundary extracted in the example of FIG. It is a figure explaining the vertical grain boundary extracted in a plurality of SEM images. (A)-(e) is a figure explaining the relationship between a grain boundary and a void for every degree of damage, respectively. It is a figure explaining the damage grain boundary extracted in a plurality of SEM images. It is a figure explaining an example of a master curve.

  Hereinafter, embodiments of the present invention will be described. First, a remaining life diagnosis system as a remaining life diagnosis apparatus will be described. As shown in FIG. 1, the remaining life diagnosis system 1 of the present embodiment includes a scanning electron microscope 2 (hereinafter simply referred to as SEM 2) and a diagnostic computer 3.

  The SEM 2 is an apparatus that obtains image data of a measurement target by squeezing an electron beam and irradiating the measurement target (replica described later) as an electron beam and detecting secondary electrons emitted from the measurement target. For example, as shown in FIG. 2, the SEM 2 includes an electron gun 21, a magnetic lens 22, a sample stage 23, and a secondary electron detector 24.

  The electron gun 21 is a part that emits an electron beam. For example, thermal electrons are accelerated by a potential difference and emitted. The magnetic lens 22 is a part that focuses the electron beam, and includes, for example, a coil that generates a magnetic field. The intensity of the magnetic field generated can be changed by adjusting the current flowing through the coil, and the degree of focusing of the electron beam emitted from the electron gun 21 can be adjusted. The sample stage 23 is a portion on which a sample (replica) to be measured is placed, and the direction and height of the measurement target can be adjusted. Thereby, an electron beam can be irradiated to the desired position in a measuring object. The secondary electron detector 24 is a part that detects secondary electrons emitted from the measurement object by irradiation with an electron beam. An image signal corresponding to the detected amount of secondary electrons is transmitted to the diagnostic computer 3.

  The diagnostic computer 3 is a device that communicates with the SEM 2 to control the operation of the SEM 2 and obtains SEM image data from an image signal received from the SEM 2. Then, the remaining life diagnosis is performed based on the obtained SEM image data. As shown in FIG. 2, for example, the diagnostic computer 3 includes a computer main body 31, a display unit 32, and an input unit 33.

  The computer main body 31 is a part that performs various controls by the diagnostic computer 3, and includes a CPU 34, a memory 35, and a data storage unit 36. The CPU 34 is a central part of control and operates according to an operation program stored in the memory 35. The memory 35 is configured by various storage devices such as an SRAM and a hard disk, and stores an operation program and secures a work area. The data storage unit 36 is also configured by various storage devices such as SRAM and hard disk, and stores various image data, data serving as a basis for a diagnostic master curve (see FIG. 13), and the like. The display unit 32 is a part that displays various images such as an SEM image and an operation image for the SEM 2, and is configured by a liquid crystal display, for example. The input unit 33 is a part that receives various operations of the worker, and is configured by, for example, a keyboard 33a and a mouse 33b.

  In the remaining life diagnosis system 1, the operation of the SEM 2 is controlled through the diagnosis computer 3. Then, the diagnostic computer 3 acquires sample SEM image data based on the image signal transmitted from the SEM 2 and diagnoses the remaining life. The remaining life diagnosis process will be described below.

  FIG. 3 is a flowchart for explaining the remaining life diagnosis processing. In this diagnosis processing, replica collection processing (S1), SEM observation processing (S2), observation image acquisition processing (S3), image adjustment processing (S4), reference line drawing / vertical grain boundary extraction processing (S5), damaged grain boundary The recognition process (S6), the D parameter calculation process (S7), and the remaining life diagnosis evaluation (S8) are sequentially performed.

  Hereinafter, each process will be described. In this embodiment, a case will be described in which the remaining life of a weld heat affected zone (HAZ zone) of a low alloy steel pipe such as 2.25% Cr-1% Mo steel widely used in boiler equipment is diagnosed.

  The replica collection process (S1) is a process for transferring a tissue to a plastic film and collecting a replica as a measurement object for a HAZ portion where damage to a measurement object (in this embodiment, a steel pipe) is predicted. In this replica collection processing, the HAZ portion is polished to a mirror surface with a grinder or the like and then etched. Next, a plastic film is attached to the HAZ portion to transfer the tissue, thereby collecting a replica (plastic film to which the tissue has been transferred).

  The SEM observation process (S2) is a process of observing the collected replica with the SEM2. In this SEM observation process, gold or platinum is vapor-deposited as a pretreatment, and a vapor deposition film such as gold is formed on the surface of the replica. If a vapor deposition film is formed, observation by SEM2 is performed. For example, a replica on which a deposited film is formed is placed on the sample stage 23, and the diagnostic computer 3 is operated to observe the surface shape of the replica.

  In this case, the diagnostic computer 3 generates replica SEM image data based on the image signal transmitted from the SEM 2 and causes the display unit 32 to display the replica SEM image data. The operator operates the input unit 33 to determine the display magnification so that the grain boundary can be visually recognized on the display unit 32. In addition, the operator operates the sample stage 23 by operating the input unit 33 to determine the observation direction of the replica. Specifically, the observation direction of the replica is determined so that the direction of stress acting on the steel pipe as the measurement object (the direction estimated as the stress direction) is a predetermined direction. In this embodiment, the observation direction is determined so that the stress direction is the vertical direction (direction parallel to the side edge of the rectangular screen) on the display screen of the display unit 32.

  In the observation image acquisition process (S3), the diagnostic computer 3 acquires SEM image data corresponding to the observation image photographed by the SEM2. For example, when imaging is instructed by an operator operating the input unit 33, the diagnostic computer 3 stores the generated SEM image data in the data storage unit 36. At this time, the diagnostic computer 3 controls the SEM 2 to capture a plurality of SEM image data in a matrix (matrix) with respect to the planar direction. That is, a plurality of images are acquired by causing the replica to be photographed with a microscope a plurality of times while changing the photographing target position. Thereby, a plurality of SEM image data is acquired over the plane range necessary for calculating the D parameter. The SEM image data is image data in a horizontally long rectangular range corresponding to the display screen on the display unit 32. For this reason, like the display image, the vertical direction corresponds to the stress direction (the direction estimated as the stress direction).

  In the present embodiment, as illustrated in FIG. 4, a case where 42 pieces of image data GR including the A column to the G column and the first row to the sixth row are acquired will be described as an example. For convenience, in the following description, the SEM image data in the first row of the A column is represented as image data GR (A1), the SEM image data in the first row of the B column is represented as image data GR (B1), and so on. The SEM image data in the sixth row is represented as image data GR (G6).

  Each of the image data GR (A1) to GR (G6) includes grain boundary image data L, void image data V, and dust image data N. Among these image data, the grain boundary image data L is acquired as, for example, straight line image data having start point coordinates and end point coordinates. The void image data V and the dust image data N are acquired as, for example, circle, ellipse, and polygon image data along the outer periphery of the void and dust. That is, the image data is acquired as a combination of a center and radius, a major axis and a minor axis, a vertex coordinate of a polygon and a short straight line.

  Further, as partly shown in FIGS. 5 and 6, each of the image data GR (A1) to GR (G6) is the data of the edge portion ED indicated by hatching with the other adjacent image data GR. Sharing. As a result, the plane range required for calculating the D parameter is fully covered.

  In the image adjustment process (S4), the diagnostic computer 3 performs various adjustments on the acquired image data GR (A1) to GR (G6). For example, as shown in FIG. 4, the image data GR obtained from the SEM 2 includes a dust image N and an extra grain boundary image LX. Further, as indicated by the symbol LY, there is a lack between the grain boundary images. If processing is performed using such image data GR as it is, the accuracy of the obtained A parameter is impaired, and consequently the remaining life diagnosis result is affected. Therefore, in the image adjustment process, the dust image N is removed, the excess grain boundary image LX is removed, and the missing grain boundary images LY are interpolated.

  Since the dust image N shown in FIG. 4 is not located on the grain boundary image L, it can be distinguished as being different from the void image V. Note that the dust image N can be distinguished from the void image V even when the shape is elongated or significantly larger than the void image V.

  Further, in the image data GR (D3), GR (E4), and GR (F5) in FIG. 4, an extra grain boundary image LX exists. These grain boundary images LX have no corresponding grain boundary image on the extension line. For this reason, it can be judged that it is an extra grain boundary image. On the other hand, in the image data GR (D1) to (E1), GR (D2), GR (E3), GR (E5), there are grain boundary images LY that are interrupted in the middle. For these grain boundary images LY, there is a corresponding grain boundary image LY on the extension line, or there is an end point PL of another grain boundary image L. Therefore, one grain boundary can be obtained by interpolating the image. Can be treated as

  When the dust removal process, the removal process of the excess grain boundary image LX, and the interpolation process of the broken grain boundary image LY are performed on each of the image data GR (A1) to GR (G6), FIG. As shown, the dust image N and the excess grain boundary image LX are removed from each image data GR, and the broken grain boundary image LY is interpolated. Thereby, clear image data suitable for diagnosis is obtained. The diagnostic computer 3 acquires grain boundary data including the start point coordinates and the end point coordinates of each grain boundary for all grain boundaries existing in the plane range to be observed for each processed SEM image data. .

  Thus, in this embodiment, this image adjustment process (S4) is SEM image data (grain boundary image) corresponding to a plurality of grain boundaries on the metal surface, which is obtained by microscopic imaging of a steel pipe (metal) replica. This corresponds to a grain boundary image data acquisition step in which the diagnosis computer 3 is acquired. This also corresponds to a grain boundary data acquisition step in which the diagnostic computer 3 acquires grain boundary data including the start point coordinates and end point coordinates of each grain boundary from the grain boundary image data.

  The grain boundary image data acquired by this image adjustment process (S4) and the grain boundary data of all the grain boundaries are stored in the data storage unit 36 and referred to in subsequent processing. Similarly, the void image data V is stored in the data storage unit 36. There are two types of void image data V, circular data and elliptical data. In the case of circular data, data that can specify the position and size of a circle, such as center coordinate data and radius data, is stored. In the case of elliptical data, data that can specify the position, size, and inclination of the ellipse, such as focal point coordinate data, long axis length data, and short axis length data, is stored.

  As described above, the image data GR (A1) to GR (G6) share the data of the edge portion ED indicated by hatching with the adjacent image data GR (see FIG. 5). If the data of the edge portion ED (overlapping portion) is assigned to a plurality of image data GR, it is a different grain boundary image L or void image V even though it is the same grain boundary image L or void image V. There is a risk of being counted. Therefore, in the image adjustment process, the edge portion ED of each image data GR is assigned to one image data GR, and the other image data GR is set as a non-target region.

  For example, as shown in FIG. 6, an inverted L-shaped non-target region EX is set from the right end portion to the lower end portion in each image data GR. As a result, the overlapping portion ED1 in which the four of the image data GR (A1), the image data GR (B1), the image data GR (A2), and the image data GR (B2) overlap is allocated to the image data GR (B2). . The overlapping portion ED2 extending in the horizontal direction (long side direction) from the overlapping portion ED1 is assigned to the image data GR (A2), and the overlapping portion X7 extending in the vertical direction (short side direction) from the overlapping portion X5 is image data. Assigned to GR (B1). As a result, it is possible to prevent a problem that the same grain boundary image L and void image V existing in different image data GR are counted repeatedly.

  In addition, when one grain boundary exists across a plurality of image data GR, the diagnostic computer 3 links the grain boundary data existing in each image data GR so that it can be handled as one grain boundary. Yes.

  In the example of FIG. 6, the grain boundary data LA exists across the image data GR (A1) and the image data GR (B1). Therefore, the diagnostic computer 3 individually stores the data LA1 that belongs to the image data GR (A1) of the grain boundary data LA and the data LA2 that belongs to the image data GR (B1) of the grain boundary data LA. The attribute data indicating that the data LA1 and the data LA2 are data of the same grain boundary is assigned. The diagnostic computer 3 recognizes that the data LA1 and the data LA2 are data of the same grain boundary LA based on the attribute data.

  Similarly, the grain boundary data LB exists across the image data GR (A1) and the image data GR (A2). Therefore, the diagnostic computer 3 individually stores the data LB1 that belongs to the image data GR (A1) of the grain boundary data LB and the data LB2 that belongs to the image data GR (A2) of the grain boundary data LB. The attribute data indicating that the data LB1 and the data LB2 are data of the same grain boundary is assigned. Based on the attribute data, the diagnostic computer 3 recognizes that the data LB1 and the data LB2 are data of the same grain boundary LB.

  When the above image adjustment processing is completed, reference line drawing / vertical grain boundary extraction processing (S5) is performed. In this reference line drawing process, the diagnostic computer 3 draws a reference line indicating a reference direction perpendicular to the stress direction of the steel pipe (metal) on the display unit 32. In other words, the reference direction data indicating the reference direction is set in the image data GR. In the present embodiment, as shown in FIG. 8, the stress direction is set to the short side direction of the image data GR. Therefore, the diagnostic computer 3 sets the reference direction data RL in the long side direction of the image data GR. This reference direction data RL is used for specifying the grain boundary data to be determined among the grain boundary data included in the image data GR in the vertical grain boundary extraction process. That is, the grain boundary data L of the grain boundary formed at a position intersecting with the reference line drawn by the reference direction data RL is a determination target.

  In this case, the diagnostic computer 3 repeatedly sets the reference direction data RL while changing the position in the stress direction at a predetermined pitch. This is because the diagnostic computer 3 recognizes all the grain boundary data L included in the image data GR. In the example of FIG. 8, the image data GR (A1) includes five grain boundary data LA to LE. If only the reference direction data RLA located approximately at the center in the short side direction is set, only the grain boundary data LC, LA, LE of the grain boundary intersecting the reference line of the data RLA is determined. If the grain boundary data LB and LD are not determined by the other image data GR, they are leaked from the determination target. In order to prevent such a problem, the reference direction data RL is repeatedly set.

  In the vertical grain boundary extraction process, the diagnostic computer 3 determines whether each grain boundary data L included in the image data GR is grain boundary data of vertical grain boundaries formed in a direction substantially perpendicular to the stress direction. Then, the grain boundary data of the vertical grain boundary is extracted as the vertical grain boundary data L ′. Therefore, the diagnostic computer 3 extracts the grain boundary data L of the grain boundary that intersects the reference line, obtains the angle of the grain boundary from the start point coordinates and the end point coordinates of the extracted grain boundary data L, and refers to the obtained angle. By comparing with the angle of the line, it is determined whether or not the extracted grain boundary data L is vertical grain boundary data L ′. In the present embodiment, the grain boundary data L of the grain boundary existing within an angle range of ± 30 ° with the angle of the reference line as the center (0 °) is set as the vertical grain boundary data L ′.

  This vertical grain boundary extraction process is performed every time the reference direction data RL is switched. That is, the reference line drawing process and the vertical grain boundary extraction process are alternately performed. In the vertical grain boundary extraction process, no judgment is made on grain boundaries that have already been judged. For this reason, the data storage unit 36 stores information about whether or not the vertical grain boundary data L ′ for each grain boundary data L. If these processes are performed for each data from the image data GR (A1) to the image data GR (G6), the reference line drawing / vertical grain boundary extraction process (S5) is terminated. As a result, for example, as shown in FIGS. 9 and 10, a part of the plurality of grain boundary data L (dotted line) is extracted as vertical grain boundary data L ′ (solid line). In these drawings, the void image V is not shown.

  As described above, in the present embodiment, the reference line drawing / vertical grain boundary extraction process (S5) uses the diagnostic computer 3 to generate the reference direction data RL indicating the reference direction perpendicular to the stress direction of the pipe (metal). This corresponds to a reference direction data setting step set in (grain boundary image data). Further, this corresponds to a vertical grain boundary data extraction step for extracting, from the plurality of grain boundary data L, vertical grain boundary data L ′ belonging to a predetermined angle range centered on the reference direction based on the reference direction data.

  If the vertical grain boundary data L ′ is extracted, damaged grain boundary recognition processing (S6) is performed. In this damaged grain boundary recognition process, the diagnostic computer 3 uses the vertical grain boundary data L ′ stored in the data storage unit 36 and an image of voids formed at the vertical grain boundaries corresponding to the vertical grain boundary data L ′. Read data V. Then, the ratio of voids in the vertical grain boundaries is obtained, and it is determined whether or not the damage grain boundary data L ″ (see FIGS. 11 and 12) corresponding to the damaged grain boundaries.

  FIGS. 11A to 11E are diagrams showing the relationship between grain boundaries and voids for each degree of damage. That is, FIG. 11A shows a state where no void is generated, and FIG. 11B shows a state where a void having a diameter of 0.5 μm or more is generated. FIG. 11C shows a state where a void having a diameter of 0.5 μm or more is saturated, FIG. 11D shows a state where the void is growing, and FIG. 11E shows a state where the void is connected. .

  As shown in these figures, the damage due to the voids at the grain boundaries starts from the occurrence of extremely small voids on the grain boundaries (state shown in FIG. 11B). When the number of voids on the grain boundary is saturated (the state shown in FIG. 11 (c)), each void grows and becomes large (the state shown in FIG. 11 (d)), and adjacent voids are united (FIG. 11 ( d) state). In this embodiment, it is determined that the grain boundary is damaged when the number of voids is saturated. In order to make this determination, the diagnostic computer 3 acquires the length LL of the vertical grain boundary from the vertical grain boundary data L ′. Further, the void width W is acquired from the void image data V. Here, when a plurality of voids exist on the grain boundary, the width W of the void is obtained by adding the widths W1 to Wn of each void (n = 4 in the example of FIG. 11). Then, the ratio of the void width W to the length LL of the vertical grain boundary is calculated. If this ratio is equal to or greater than the criterion value (50% or more in the example of FIG. 11), the vertical grain boundary is the damaged grain boundary. Is determined. In this case, information indicating that the grain boundary data L ″ is the damaged grain boundary data L ″ is stored in the corresponding grain boundary data L in the data storage unit 36.

  If this process is performed for all the extracted vertical grain boundary data L ′, the damaged grain boundary recognition process (S6) is terminated. Thereby, for example, as shown in FIG. 12, a part of the plurality of vertical grain boundary data L ′ (solid line / thin line) is extracted as damaged grain boundary data L ″ (solid line / thick line).

  Thus, in this embodiment, the damaged grain boundary recognition process (S6) is based on the void formation state on the grain boundary corresponding to the vertical grain boundary data, and the vertical grain boundary data corresponds to the damaged grain boundary. This corresponds to a damaged grain boundary determination step in which the diagnosis computer 3 determines whether the data is boundary data.

In the subsequent D parameter calculation process (S7), the diagnostic computer 3 calculates the D parameter based on the number of vertical grain boundaries corresponding to the vertical grain boundary data and the number of damaged grain boundaries corresponding to the damaged grain boundary data. To do. Here, the D parameter is calculated based on the following equation (1). That is, it is calculated by dividing the number of damaged grain boundaries in the observation region (range of the observation image) by the total number of damaged grain boundaries in the observation region. The diagnostic computer 3 stores the calculated value of the D parameter in the data storage unit 36 and ends this process.

  In the remaining life diagnosis evaluation (S8), the diagnostic computer 3 reads the master curve from the data storage unit 36 and applies the calculated D parameter value to obtain the time fraction (that is, the remaining life) of the steel pipe (metal). To do. Here, for example, as shown in FIG. 13, the master curve is a graph showing the correlation when the horizontal axis is the fraction of time and the vertical axis is the D parameter, which is acquired in advance by a creep acceleration test of the steel pipe. It is. The time fraction indicates the elapsed time as a ratio when the elapsed time at the time of fracture is the value “1”. For this reason, it is also a parameter | index showing the remaining life of a steel pipe. Therefore, it can be said that this master curve shows a correlation between the D parameter and the remaining life of the steel pipe.

  Thus, the diagnostic computer 3 diagnoses the remaining life of the steel pipe by applying the calculated D parameter value to the master curve. Therefore, this remaining life diagnosis evaluation corresponds to a remaining life diagnosis step for causing the diagnosis computer 3 to diagnose the remaining life of the steel pipe (metal) using the D parameter method. When it is determined that the replacement time is based on the remaining life, the operator is notified by displaying the fact on the display unit 32. Further, even when there is a sufficient remaining life, the remaining life is displayed on the display unit 32 to notify the operator.

  When this remaining life diagnosis evaluation is completed, a series of diagnosis processing is completed, and the diagnosis for the next replica is repeatedly performed according to the above-described procedure.

  As described above, according to this embodiment, the reference direction data indicating the reference direction perpendicular to the stress direction of the pipe is set in the grain boundary image data in the reference line drawing / vertical grain boundary extraction process (S5). In addition, vertical grain boundary data belonging to a predetermined angle range centered on the reference direction is extracted from a plurality of grain boundary data based on the reference direction data. Field data can be extracted quickly and reliably. In the damaged grain boundary recognition process (S6), when determining whether the grain boundary data is damaged based on the extracted vertical grain boundary data and void information, the determination process can be facilitated. it can. As a result, the remaining life diagnosis process can be facilitated when diagnosing the remaining life of a pipe that is subject to creep damage by the D-parameter method.

  Further, when setting the reference direction data, the reference direction data is repeatedly set while changing the position of the stress direction, so that the vertical grain boundary data can be reliably extracted from the plurality of grain boundary data. .

  Further, when determining whether or not the data is damaged grain boundary data, the ratio between the grain boundary length LL and the void width (length) W corresponding to the vertical grain boundary data is used. Certainty can be increased.

  Further, when acquiring SEM image data (grain boundary image data), the microscope is imaged a plurality of times while changing the imaging target position, and when setting the reference direction data, each SEM image data is targeted, In addition, when extracting the vertical grain boundary data, each SEM image data is targeted, so that it is not necessary to perform tiling for laying and joining images taken with a microscope, thereby improving work efficiency. it can.

  Further, when diagnosing the remaining life, a master curve indicating the correlation between the D parameter and the time fraction is acquired in advance, and the D parameter is applied to the master curve, so that the remaining life can be easily diagnosed. .

  The above description of the embodiment is for facilitating the understanding of the present invention, and does not limit the present invention. The present invention can be changed and improved without departing from the gist thereof, and the present invention includes equivalents thereof. For example, you may comprise as follows.

  With regard to the remaining life diagnosis apparatus according to the present invention, in the above-described embodiment, the remaining life diagnosis system 1 in which the SEM 2 is provided integrally with the diagnosis computer 3 is exemplified, but the remaining life diagnosis system 1 may be configured only by the diagnosis computer 3. In this case, SEM image data obtained by the SEM 2 may be input to the diagnostic computer 3.

  Moreover, regarding the microscope used for imaging | photography, although SEM2 was illustrated in the above-mentioned embodiment, another type of electron microscope (for example, a transmission electron microscope) may be sufficient, and an optical microscope may be sufficient. In short, any material can be used as long as it has the ability to confirm the structure (grain boundaries and voids) transferred to the replica.

  Regarding the object to be diagnosed, in the above-described embodiment, the steel pipe used for the power pipe or the like is exemplified, but the present invention is not limited to this. For example, a partition wall of a boiler or a turbine may be used. In short, the present invention can be applied to any metal that is exposed to a high temperature and high pressure environment and whose remaining life can be diagnosed by the D parameter.

1 ... Remaining life diagnosis system 2 ... Scanning electron microscope (SEM)
DESCRIPTION OF SYMBOLS 21 ... Electron gun 22 ... Magnetic lens 23 ... Sample stage 24 ... Secondary electron detector 3 ... Diagnostic computer 31 ... Computer main body 32 ... Display part 33 ... Input part 34 ... CPU
35 ... Memory 36 ... Data storage part GR ... Single image data L ... Grain boundary data L '... Vertical grain boundary data L "... Damage grain boundary data LL ... Length of vertical grain boundary V ... Void image data W ... Void width (void length)
N: Dust image data ED: Edge portion EX of image data EX: Non-target region RL of image data Reference direction data (reference line image)

Claims (5)

A remaining life diagnostic device for diagnosing the remaining life of a metal subject to creep damage by a computer,
In the computer,
Grain boundary image data acquisition step for acquiring grain boundary image data corresponding to a plurality of grain boundaries on the metal surface, obtained by microscopic imaging of the metal replica,
From the grain boundary image data, a grain boundary data acquisition step for acquiring grain boundary data including start point coordinates and end point coordinates of each grain boundary,
Reference direction data setting step for setting reference direction data indicating a reference direction perpendicular to the stress direction of the metal in the grain boundary image data,
A vertical grain boundary data extraction step for extracting vertical grain boundary data belonging to a predetermined angle range centered on the reference direction from the plurality of grain boundary data based on the reference direction data;
A damaged grain boundary determination step for determining whether the vertical grain boundary data is damaged grain boundary data corresponding to the damaged grain boundary based on a void formation state on the grain boundary corresponding to the vertical grain boundary data;
A remaining life diagnosis step of diagnosing the remaining life of the metal using a D parameter method based on the number of grain boundaries corresponding to the vertical grain boundary data and the number of grain boundaries corresponding to the damaged grain boundary data; A remaining life diagnosis apparatus characterized by being performed.   The remaining life diagnosis apparatus according to claim 1, wherein in the reference direction data setting step, the reference direction data is repeatedly set while changing the position of the stress direction.   In the damaged grain boundary determination step, it is determined whether or not the vertical grain boundary data is the damaged grain boundary data based on a ratio between a grain boundary length and a void length corresponding to the vertical grain boundary data. The remaining life diagnosis apparatus according to claim 1 or 2. In the grain boundary image data acquisition step, a plurality of grain boundary image data is acquired by performing microscopic imaging while changing the imaging target position,
In the reference direction data setting step, the reference direction data is set for each of the plurality of grain boundary image data,
4. The remaining life diagnosis apparatus according to claim 1, wherein in the vertical grain boundary data extraction step, the vertical grain boundary data is extracted for each of the plurality of grain boundary image data. 5.   In the remaining life diagnosis step, a master curve indicating the correlation between the D parameter and the time fraction is acquired in advance, and the number of grain boundaries corresponding to the vertical grain boundary data and the damaged grain boundary data are handled. The remaining life diagnosis apparatus according to claim 1, wherein the remaining life of the metal is diagnosed by applying a D parameter based on the number of grain boundaries to the master curve.

Images