JPH067096B2 - Indentation type hardness tester - Google Patents

Description

The present invention relates to an indentation type hardness tester such as Vickers hardness and Brinell hardness.

[Conventional technology]

Vickers hardness and Brinell hardness are used to judge the mechanical properties of metallic materials. Especially, the Vickers hardness has a facing angle of 13
It is said that a 6 ° diamond pyramid indenter is pressed against the surface of the test piece and the pressing load is divided by the surface area of the indentation formed by pressing.

A conventional indentation type hardness tester is disclosed, for example, in Japanese Examined Patent Publication No. 63-10379. This is as shown in FIG.
The indenter 2 is pressed against the surface of the test piece 1 with a predetermined load, and the indentation 3
And a microscope 5 provided in the testing machine 4
And an A / D converter 8 for performing analog / digital conversion of the optical image information 7 of the indentation 3 projected on the imager 6 with, for example, a gray level of 256 gradations,
A digital image processing unit 20 for extracting a point where the grayscale level digital image signal 9 from the A / D conversion unit 8 changes abruptly as an end of the indentation 3, and an indentation obtained by the digital image processing unit 20. The indentation size measuring unit 21 that measures the size of the indentation from the position of the end portion, and the hardness is calculated by the size of the indentation 3 measured by the indentation size measuring unit 21 and the pressing load of the indenter 2, and the printer 22 or the like. It is mainly configured by a hardness calculation unit 18 that outputs to.

The feature of the above-mentioned indentation type hardness tester is that if there is a matrix structure, a scratch, or a corroded part that is as bright as the indentation, the end of the indentation 3 becomes unclear or measurement is not possible. In order to prevent such a situation from occurring,
It is processed as a digital image in which the grayscale level is converted to analog / digital and the brightness at each pixel is replaced by a numerical value.

The digital image processing unit 20 processes and analyzes the digital image signal 9 thus analog-to-digital converted at the gray level so as to eliminate the influence of noise due to matrix tissue, scratches, etc., and obtains the gray level curve obtained. Points P1 and P2 at which 26 'rapidly changes are deeply searched along the measurement axis A as shown in FIG. 16 (b) and extracted as the end of the indentation 3. FIG. 16 (b) is a schematic trace of FIG. 16 (a).
The vertical axis of the diagram in the figure is the gray level and the horizontal axis is the position.

[Problems to be Solved by the Invention]

There are the following problems (1) and (2) as problems of the indentation type hardness tester in the prior art.

(1) A line B connecting the opposite corners of the indentation as shown in FIG.
Is not on the measurement axis A, the indentation size measuring unit is the true value l.
_An erroneous value is output as a measured value l ₁ ′ for ₁ .

(2) As shown in FIG. 18, the matrix structure of the background,
When the corners of the indentation are particularly unclear due to scratches, uneven brightness due to rust, etc., the indentation size measurement unit can understand and correct the factor and perform measurement so as to match the human measurement value l _2. Is difficult and outputs an incorrect value like l ₂ ′.

As described above, the measurement error is a serious problem because the quality characteristics of the steel material are erroneous. It should be noted that the above (1) occurs when there is a slight mechanical backlash when moving the sample after indentation.

[Means for Solving the Problems]

The present invention advantageously solves the drawbacks of the prior art, and provides a tester capable of forming an indentation by pressing a pyramid indenter on the surface of a test piece with a predetermined load, and a microscope equipped with the tester. Connected imager, A / D converter for analog / digital conversion of optical image information of the indentation projected on the imager, and processor for digitizing digital image signal from the A / D converter From the digitized signal from the processing unit,
An indentation apex detection unit having a function of detecting an indentation apex from the intersection of the linear approximation expressions of pixels of the indentation piece and a function of directly searching an indentation corner to detect the indentation apex, and the indentation apex detection unit obtained 2 An indentation vertex selection unit that selects an appropriate one of the two indentations of each type of indentation vertices, an indentation size measurement unit that obtains an indentation size from the vertex coordinates selected by the indentation vertex selection unit, and A hardness calculation unit that calculates hardness from the indentation size measured by the indentation size measurement unit is provided. The present invention will be described below with reference to the drawings.

〔Example〕

 FIG. 1 shows an embodiment of an embedded hardness tester according to the present invention.

A tester 4 capable of forming an indentation 3 by pressing a quadrangular pyramid indenter 2 on the surface of a test piece 1 with a predetermined load, an imager 6 connected to a microscope 5 provided in the tester 4, and the imager Of the digital image signal 9 from the A / D conversion unit 8 that performs analog / digital conversion of the optical image information 7 of the indentation 3 projected on the image 6 on the gray level of 256 gradations, for example. With respect to the image focusing control unit 11 that controls the autofocus unit 10 so as to maximize the degree of increase or decrease in the gray level of the indentation boundary portion, and the digital image signal 12 adjusted by the image focusing control unit 11, An indentation is made by a binarization processing unit 13 that automatically obtains a threshold level that distinguishes the indentation from the background and binarizes the binary image created by the binary image signal 14 from the binarization processing unit 13. Line approximation formula An indentation apex detection unit 15 having a function of detecting an indentation apex from an intersection point and a function of detecting a linear indentation corner portion near the indentation center to detect an indentation apex, and two types of indentation apex obtained by the indentation apex detection unit. For each vertex, the indentation vertex selection unit 16 that selects an appropriate one of the two, and the four selected by the indentation vertex selection unit
From the indentation size measuring unit 17 for obtaining two diagonal lines of a quadrangle connecting the coordinates of one vertex, the average value of the two diagonal lengths of the indentation measured by the indentation size measuring unit 17, and the pressing load of the indenter, the hardness is obtained. Mainly composed of a hardness calculator 18 for calculating and outputting Reference numeral 19 denotes a table (sample stage) on which the test piece is placed, which can be moved vertically by the autofocus unit 10 and can be moved laterally.

The function of the image focusing control unit 11 will be described with reference to FIGS. 2 and 3.

FIG. 2 (-A direction, the same applies hereinafter) and FIG. 3 show a flow for focusing the boundary portion of the indentation 3. The threshold level is SHi as high side and S as low side.
L ₀ is set in advance, and the absolute value of the difference between _{x 1} and _{x 2} when the indicates the position of the pixel at that time and _x 1, _{x 2,} respectively _(| x 1 _-x 2 |) is preset The sample stage 19 is moved upward or downward from the autofocus unit 10 so that it becomes smaller than the set value δ or becomes the minimum value. Note that [1] in FIG. 2 shows before focusing, and [2] in FIG. 2 shows after focusing.

The optical image information 7 after being focused is converted into an A / D converter 8
The contents of processing in the binarization processing unit 13 that divides the digital image signal 12 that has been analog-to-digital converted in step 1 into an indentation and a background and obtains a binary image signal will be described using the processing flow of FIG.

Generally, an automatic push-type hardness meter stores the pressing position of the indenter and brings the lens of the microscope to that position at the time of measurement. Therefore, the indentation is approximately in the center of the field of view of the imager.

Therefore, as shown in FIG. 5, digital image signals of an arbitrary number of lines (three lines in FIG. 5) 31 and 32 are extracted around a line passing through the vertical and horizontal centers of the field of view of the image pickup device. . As a result, an effective signal including both the indentation and the background can be extracted. In this case, the signal of either the horizontal line 31 or the vertical line 32 may be used.

Another frequency distribution of gray level (hereinafter referred to as a histogram) is created by the signal thus extracted. that time,
It is more appropriate to use not a histogram for all gray levels but a histogram in which frequencies for gray levels, for example, about 5 gray levels are obtained. FIG. 6 shows an example of a histogram in which the abscissa represents the gray level, which is a total of 256 gradations and is divided into 5 gradations.

FIG. 6 (a) shows the case where the sample surface is a mirror surface, and FIG. 6 (b) shows the case where the matrix texture surface. The mountain on the left is due to the indentation,
The mountain on the right is due to the background, and in general, the mountain in the background is white in many cases in the mirror surface sample and becomes steep, whereas the pattern in the matrix texture surface becomes gentle due to the pattern. Therefore, as shown in FIG. 7 (a), a certain frequency N is set as a threshold level, and a mountain interval S is obtained. If S is larger than a predetermined gradation difference C, as shown in FIG. 7 (b), The mountain is divided by the specified ratio d, and the gradation TH
1 is the binarization threshold level. Also S <
In the case of C, the gradation TH2 which is the minimum value between the peaks is set as the binarized threshold level as shown in FIG. 7 (c).

By binarizing the digital image processing signal 12 with the binarized threshold level thus obtained, a binary image signal 14 capable of distinguishing the indentation from the background can be obtained for any sample. The N, S, and
The values of C and d have appropriate values depending on the brightness of the light source, the sensitivity of the image pickup device, and the like, and are calculated in advance for each hardness meter.

Next, the indentation apex detection unit 15 that detects four apexes of an indentation using a binary image created by the binary image signal will be described with reference to FIGS. 8 to 12.

FIG. 8 shows a binary image formed by the binary image signal 14,
The curves (SG _X , SG _Y ) in which the values obtained by integrating the binary image for each line in the Y-axis direction and the X-axis direction are plotted are shown.
The _SG x, the threshold level of the accumulated value corresponding to the impression that has been set in advance and SG _{y T} x, _{T y}
The indentation portions D _x and d _y are determined by the above. Approximately 2 for linear C _x is X direction indentation in the Y-axis direction passing through the middle point of the D _x
Similarly, the straight line C _y in the X-axis direction, which passes through the midpoint of D _y , divides the indentation in the Y direction in almost equal parts.

Therefore, by searching in the right and left of each pixel on the C _x as FIG. 9 (a) to the starting point, first pixel or one before the pixel was from 0 to 1 selected as pixels representing the indentation boundary. Further, as shown in FIG. 9 (b), each pixel on C _y is searched up and down, and the first pixel from 0 to 1 or the pixel immediately before is selected as the pixel representing the indentation boundary. As a result, the pixel at the indentation boundary is obtained as shown by the dotted line in FIG. 10 (a). Then, those pixels are given by C _x and C _y .
It is divided into four zones, a, b, c and d, and each zone is excluding the end, that is, the vicinity of the corner of the indentation.
It is (b). Therefore, if these pixels are linearly approximated by the least squares method for each zone, A, B, and B in FIG.
There are four straight lines like C and D. Furthermore, the straight line intersections H _1a , H _1b , H _1c , and H _{1d are calculated} from the approximation formulas of the four straight lines.
Then, these correspond to the vertices of the indentation. Also,
An intersection point 0 of a straight line connecting H _a and H _b and a straight line connecting H _c and H _d corresponds to almost the center of the indentation. Next, as shown in FIG. 12 (a), a line from the start point to the end point is searched by searching in the X direction with the center O as the start point, and setting the first pixel from 0 to 1 or the previous pixel as the end point. Find the length of the minute (number of pixels). The same operation is performed for arbitrary m pixels while shifting the Y coordinate one by one in the + direction and in the one direction with the same center O and X coordinate, and selects the longest one of all line segments, The end point at that time is the indentation vertex H _2a .

Similarly, for the other three vertices, FIG. 12 (b) (c) (d)
By operating as described above, the indentation vertices H _2b , H _2c and H _2d can be obtained.

Here, the number m to be searched is determined for each hardness meter according to test conditions such as sample type, load, optical system, and the like. Further, instead of the center O, the coordinates of the intersection point of C _x and C _y obtained previously may be used.

From the above, the indentation vertices H ₁ (H _1a , H _1b , H _1c , obtained from the intersections of the linear approximation formulas of the pixels on the indentation side,
H _1d ) and the indentation vertices H ₂ (H _2a , H _2b , H _2c , obtained by directly searching the indentation corners from the vicinity of the center of the indentation,
Although H _2d ) is detected, an indentation vertex selecting unit 16 that selects one of two types of indentation vertices for each of the four vertices will be described with reference to FIG.

[1] When the average value of the deviation between the linear approximation formula used for H ₁ detection and the actual pixel is large, H ₁ is set as a large error and H ₂ is adopted.

[2] The length of 0～H ₁ is longer than the length of the 0～H _2, employing the _{H 2} (FIG. 13 (a)).

[3] If the length of 0～H ₁ is shorter than the length of the 0~H _2, longer than H 1 to H ₂ of length l is the length previously determined l '(l> l ′) If H ₁ is short (l <
l ') to adopt a _{H 2.}

With the above determination method, two types of indentation vertices (H _1a , H _1b ,
H _1c , H _1d ) and (H _2a , H _2b , H _2c , H
From the _2d), selected indentation vertex for each vertex _{(H a, H b, H} c, H d) a person is very close to the results determined by visual observation.

The indentation size measuring unit 17 uses the H _a obtained as described above.
The -H _b -H _c -H _d confirmed indentations make measurements of the diagonal _l 1, _{l 2.} When calculating the Vickers hardness Hv, for example, the hardness calculator 18 calculates Hv from the pressing load P and the average value l of l ₁ and l ₂ by the following equation.

Hv = P / S …… (1) ＝ 2Psin 68 ° / l ² = 1.854P / l ² (kg / mm ² ) …… (2) The explanation so far is an image of a general Vickers hardness tester. Although the case of FIG. 14 (a) has been performed, it is also possible to realize the case where the field of view of the image pickup device and the impression are arranged in the same manner as shown in FIG. 14 (b). Figures 5,8,9,10,11,12 (respectively -A) are respectively 2-B, 5-B, 8-B, 9-B, 10-B, 11-.
This is as shown in Figures B and 12-B. In addition, the configuration shown in FIG. 14 (b) has an advantage that a larger indentation can be processed than in the case of FIG. 14 (a) when the field of view of the image pickup device is the same. Here, in order to obtain the image of FIG. 14 (b), FIG. 14 (a)
From this state, the image pickup device 6 may be rotated by 45 °.

〔The invention's effect〕

As described above, according to the present invention, the side of the indentation is obtained by the linear approximation formula, the indentation vertex is obtained from the intersection obtained by the calculation from the approximation formula, and the indentation vertex is obtained by directly searching the indentation corner portion, Since an appropriate one of the two types of indentation vertices is selected, the reliability of the measurement can be improved without being influenced by the surface property of the test piece.

[Brief description of drawings]

FIG. 1 is a schematic diagram of an indentation type hardness tester of the present invention, and FIG. 2 is before focusing [1] and after focusing [2] of the boundary line of the indentation.
FIG. 3, FIG. 3 is an image processing flow chart for focusing the boundary line of the indentation, FIG. 4 is a diagram showing a processing flow of the binarization processing unit, and FIG. 5 is an arbitrary view from the visual field of the image pickup device. FIG. 6 (a) shows an example of a histogram when the sample surface is a mirror surface, and FIG. 6 (b) shows a case where the sample surface is a matrix tissue surface. FIG. 7 is a diagram showing an example of a histogram, FIG. 7 is a diagram showing the logic of binarization processing, FIG. 8 is a diagram showing a method of dividing an indentation into two substantially equal parts in the X direction and the Y direction, and FIG. 9 is a boundary of the indentation. Fig. 10 (a) is a diagram showing pixels at the boundary of the indentation, and Fig. 10 (b) is a diagram excluding pixels near the corner of the indentation among pixels at the boundary of the indentation. , FIG. 11 is a diagram showing a method of obtaining an indentation apex from a linear approximation formula of pixels on the side of the indentation, and FIG. FIG. 13 is a diagram showing a method for directly obtaining an indentation apex from the vicinity of the center to obtain an indentation apex, FIG. 13 is a diagram showing a method for selecting an indentation apex, FIG. FIG. 16 is a schematic diagram of a conventional indentation type hardness tester, FIG. 16 is a diagram showing image processing by a conventional technique, FIG. 17 is a diagram showing image processing when the measurement axis A and the diagonal axis B are deviated, The figure is a diagram showing the image processing when the corners of the indentation are not clear. 1: Test piece, 2: Indenter, 3: Indentation, 4: Tester, 5: Microscope, 6: Imager, 7: Optical image information, 8: A / D converter, 9, 12: Digital image signal, 10 : Auto focus unit, 11: Image focusing control unit, 13: Binarization processing unit, 14:
Binary image signal, 15: Indentation vertex detection unit, 16: Indentation vertex selection unit, 17: Indentation size measurement unit, 18: Hardness calculation unit, 19: Sample stage, 20: Digital image processing unit, 21: Indentation size measurement unit , 22: printer, 23: control unit, 2
4: Auto stage, 26: Gray level curve, 31, 32: Extraction line for histogram creation.

Claims (1)

[Claims] 1. A tester capable of forming an indentation by pressing a pyramidal indenter on the surface of a test piece with a predetermined load, an imager connected to a microscope provided in the tester, and an imager projected on the imager. A / D conversion section for converting analog / digital conversion of optical image information of the indentation, a processing section for digitizing the digital image signal from the A / D conversion section, and an indentation from the digitized signal from the processing section. Of the indentation apex having a function of detecting an indentation apex from the intersection of the direct approximation formulas of the pixels of one side and a function of detecting an indentation apex by directly searching an indentation corner, and two types obtained by the indentation apex detection For each apex of the indentation vertices, the indentation apex selecting unit that selects an appropriate one of the two indentations, an indentation size measuring unit that obtains an indentation size from the vertex coordinates selected by the indentation apex selecting unit, and the indentation Measured by the size measurement unit An indentation type hardness tester, comprising: a hardness calculation unit that calculates hardness from a determined indentation size.