CN105488336B - A kind of method of measure 9Cr ferritic heat-resistant steel hardness inhomogeneities - Google Patents

Abstract

The invention discloses a method for measuring the hardness inhomogeneity of 9Cr ferritic heat-resistant steel, which comprises the following steps: ① Measuring the microhardness of the steel at equal intervals; Three stages of low, medium and high microhardness are selected; ③ Starting from the lowest hardness, gradually increase the data points until the slope of the linear regression equation reaches the minimum value that meets the requirements to obtain the microhardness range of ferrite as the main section; The descending curve of the curve and gradually increase the data points starting from the highest hardness until the slope of the linear regression equation reaches the minimum value that meets the conditions, so as to obtain the microhardness range of martensite as the main segment; ④ The obtained ferrite is mainly segment, ferrite plus martensite mixed segment, the fractional value corresponding to the boundary hardness data of the martensite-based segment is the proportion of the above-mentioned typical segments. The method of the invention realizes the identification and quantification of the structure and hardness of the microhardness inhomogeneity of the 9Cr ferritic heat-resistant steel.

Description

A Method for Measuring Hardness Inhomogeneity of 9Cr Ferritic Heat-resistant Steel

technical field

The invention relates to the measurement of hardness inhomogeneity of 9Cr ferritic heat-resistant steel, discloses a method for measuring the hardness inhomogeneity of 9Cr ferritic heat-resistant steel, in particular, relates to the measurement of ferrite in the measured area of a steel sample The main segment, ferrite plus martensite as the main segment and the proportion of the three typical segments of martensite as the main segment.

Background technique

According to the requirements of ASME SA335-SA335M, the hardness of P91 material pipes does not exceed 250HB. However, in the inspection process of P91 materials, whether it is imported materials or domestic materials, the low hardness is already a very common problem. In the newly promulgated DL/T438-2009 "Metal Technical Supervision Regulations for Thermal Power Plants" issued by the power industry, it is clearly stated that the hardness of P91 materials should not be lower than the requirement of 180HB. However, there are still a large number of low-hardness phenomena in the P91 pipes and headers that have been in operation, which has brought serious safety hazards to the long-term stable operation of the unit. Metallographic analysis of low hardness P91 pipe is extremely important. When a variety of typical structures appear in the material, the proportion of these typical structures cannot be quantified by traditional metallographic methods. How to use the microhardness equidistant measurement data to calculate the boundary hardness data between each typical area, so as to determine the proportion of each typical area, is a key point of the present invention.

Contents of the invention

In order to overcome the above defects, the object of the present invention is to provide a method for measuring the hardness inhomogeneity of 9Cr ferritic heat-resistant steel, which is simple, accurate and easy to implement.

In order to achieve the above object, the technical solution adopted in the present invention is: a method for measuring the hardness inhomogeneity of 9Cr ferritic heat-resistant steel, comprising the following steps:

① Acquire equidistant microhardness measurement data of relevant steel samples;

② According to the equidistant microhardness measurement data, the corresponding microhardness data normalized ascending order and descending order distribution curves are made respectively;

③In the normalized ascending distribution curve of microhardness data, gradually increase the data points starting from the lowest hardness until the slope of the linear regression equation reaches the minimum value that meets the conditions. There are at least 3 consecutive data points decreasing on the left side adjacent to the point, and at least 3 consecutive data points increasing on the right side adjacent to the minimum value, so as to obtain the upper boundary hardness data of ferrite as the main segment;

④In the normalized descending distribution curve of microhardness data, gradually increase the data points starting from the highest hardness until the slope of the linear regression equation reaches the minimum value that meets the conditions. There are at least 3 consecutive data points decreasing on the left adjacent to the point, and at least 3 consecutive data points increasing on the right adjacent to the minimum value point, so as to obtain the lower boundary hardness data of the martensite as the main segment;

⑤ Based on the obtained upper boundary hardness data of ferrite as the main segment and lower boundary hardness data of martensite as the main segment, calculate the ferrite as the main segment in the measured area of the steel sample, and ferrite plus martensite as The proportion of the three typical segments of the main segment and martensite as the main segment.

There is also the following step between step ① and step ②: removing abnormal data points.

In step ①, the preparation steps of the steel sample are: grinding, polishing, and eroding the observation surface of the test sample according to the conventional metallographic sample preparation method.

In step ⑤, the proportion F% of ferrite as the main segment is determined according to the upper boundary hardness data of the main segment of ferrite, and the proportion of the main segment of martensite is determined according to the lower boundary hardness data of the main segment of martensite. Accounting for a proportion of M%, the proportion of ferrite plus martensite as the main segment is finally calculated (F+M)%=100%-F%-M%.

In step ③, in the normalized ascending order distribution curve of microhardness data, gradually increase the data points starting from the lowest hardness to perform linear regression y=a _Ⅰ +b _Ⅰ f _s , a _Ⅰ is a constant, y is arranged in order According to the n microhardness data, find the slope b _Ⅰ corresponding to each set of data below. The slope b _Ⅰ includes b _Ⅰ 3, b _Ⅰ 4, b _Ⅰ 5...b _Ⅰ n, fs1, fs2, fs3... The abscissas 1/n, 2/n, 3/n... of the normalized ascending distribution curve of microhardness data, y1, y2, y3... are the abscissas of the normalized ascending distribution curve of microhardness data The vertical coordinates corresponding to fs1, fs2, fs3... are as follows:

Then use the obtained slope value b _Ⅰ 3, b _Ⅰ 4, b _Ⅰ 5...b _Ⅰ n as the slope distribution curve Ⅰ, the abscissa is y3, y4, y5...yn, and the ordinate corresponds to b _Ⅰ 3, b _Ⅰ 4, b _Ⅰ 5...b _Ⅰ n, find the first qualified minimum value b _Ⅰ i from left to right on the slope distribution curve Ⅰ (there are at least 3 consecutive data points decreasing on the left, and at least There are 3 consecutive data points increasing), b _Ⅰ i is one of b _Ⅰ 3, b _Ⅰ 4, b _Ⅰ 5...b _Ⅰ n, at this time the yi corresponding to b _Ⅰ i is the ferrite as the main segment Upper boundary hardness data.

In step ④, in the normalized descending order distribution curve of microhardness data, gradually increase the data points starting from the highest hardness for linear regression y=a _Ⅱ +b _Ⅱ f _s , a _Ⅱ is a constant, y is arranged in order n pieces of microhardness data, and calculate the slope b _Ⅱ corresponding to each set of data, the slope b _Ⅱ includes b _Ⅱ (n-2), b _Ⅱ (n-3)...b _Ⅱ 1, fsn, fs(n -1), fs(n-2)... are the abscissas n/n, (n-1)/n, (n-2)/n..., yn of the normalized descending distribution curve of microhardness data , y(n-1), y(n-2)... are respectively the abscissa fsn, fs(n-1), fs(n-2)... The corresponding vertical coordinate, the specific operation is as follows:

Then use the obtained slope values b _Ⅱ 1, b _Ⅱ 2...b _Ⅱ (n-3), b _Ⅱ (n-2) as the slope distribution curve Ⅱ, and the abscissa is y1, y2...y(n- 3), y(n-2), the ordinate corresponds to b _Ⅱ 1, b _Ⅱ 2...b _Ⅱ (n-3), b _Ⅱ (n-2), find the right to left on the slope distribution curve Ⅱ The first qualified minimum value b _Ⅱ j (there are at least 3 consecutive data points decreasing on the left, and at least 3 consecutive data points increasing on the right), b _Ⅱ j is b _Ⅱ 1, b _Ⅱ 2...b _Ⅱ (n-3), b _Ⅱ (n-2), at this time, yj corresponding to b _Ⅱ j is the lower boundary hardness data of martensite as the main segment.

Compared with prior art, the beneficial effect that the present invention has is:

① The present invention utilizes the fact that traditional metallographic methods cannot identify and quantify the proportions of three typical tissue segments in 9Cr steel, and proposes to use equidistant microhardness measurement data to calculate the boundary hardness data of each typical area, thereby determining the proportion of each typical area .

② The calculation process of the present invention is simple, fast, and the result is accurate and reliable;

3. The calculation result of the present invention can guide the test, reducing the blindness of judging the microstructure in the metallographic examination;

④ In addition to being applied to 9Cr ferritic heat-resistant steel, the present invention can also be used to measure and calculate the proportions of typical structures in other steels and alloys, and has broad application prospects.

In a word, the calculation process of the present invention is simple and fast, and can be widely used in the determination of hardness inhomogeneity of 9Cr ferritic heat-resistant steel.

Description of drawings

Figure 1 is the microhardness measurement grid line;

Fig. 2 is the normalized ascending order distribution curve of microhardness data;

Fig. 3 is the normalized descending order distribution curve of microhardness data;

Fig. 4 is slope distribution curve 1;

Fig. 5 is slope distribution curve 2;

Figure 6 is the normalized ascending curve of the finally obtained microhardness data.

detailed description

The present invention will be further described below in combination with specific embodiments and accompanying drawings.

The method of the present invention comprises the following steps:

① Get data

The present invention first needs to obtain hardness data through a microhardness test. Select a 10mm×10mm area on the metallographic observation surface of the sample to be tested, and conduct the microhardness test (HV0.1) according to the grid lines shown in Figure 1.

② Remove abnormal data points, and make normalized ascending and descending distribution curves of microhardness data according to the obtained data. This step is to eliminate the data points that deviate too much from the beginning and end of the entire curve. If not, it will affect the calculation results. Specifically, remove points with a large distance (for example, remove the distance greater than or equal to 8).

In the normalized ascending order distribution curve of microhardness data, the abscissa (fs) is 1/n, 2/n, 3/n...(n-1)/n, n/n, and the corresponding ordinate y These n microhardness data are arranged in ascending order in turn.

In the normalized descending distribution curve of microhardness data, the abscissa (fs) is 1/n, 2/n, 3/n...(n-1)/n, n/n, and the corresponding ordinate y These n microhardness data are arranged in descending order in turn.

Among them, the abscissa (fs) represents the counting frequency after the total number of measuring points of each pipe sample is normalized, that is, the proportion of the number of measuring points less than or greater than a certain value of y in the total number of measuring points.

③Calculation of upper boundary hardness data of ferrite as the main segment

In the normalized ascending order distribution curve of the microhardness data, starting from the lowest hardness, gradually increase the data points for linear regression (y=a _Ⅰ +b _Ⅰ f _s ), a _Ⅰ is a constant item, which is also the equation on the Y axis The intercept, b _Ⅰ is the regression coefficient, which is also the slope of the linear regression equation y=a _Ⅰ +b _Ⅰ f _s on the coordinate axis, y is the n microhardness data arranged in order, and the corresponding data of each set of the following The slope b _Ⅰ (including b _Ⅰ 3, b _Ⅰ 4, b _Ⅰ 5...b _Ⅰ n), fs1, fs2, fs3... are the abscissas 1/n, 2/n, 3/n..., y1, y2, y3... are the vertical coordinates corresponding to fs1, fs2, fs3... in the normalized ascending order distribution curve of microhardness data, the specific operation is as follows:

Then use the obtained slope values (b _Ⅰ 3, b _Ⅰ 4, b _Ⅰ 5...b _Ⅰ n) as the slope distribution curve I, the abscissa (fs) is (y3, y4, y5...yn), and the vertical The coordinates correspond to (b _Ⅰ 3, b _Ⅰ 4, b _Ⅰ 5...b _Ⅰ n), find the first qualified minimum value b _Ⅰ i from left to right on the slope distribution curve Ⅰ (the left side has at least continuous 3 data points are decreasing, and there are at least 3 consecutive data points increasing on the right), b _Ⅰ i is one of b _Ⅰ 3, b _Ⅰ 4, b _Ⅰ 5...b _Ⅰ n, at this time b _Ⅰ i corresponds to yi That is, the upper boundary hardness data of ferrite as the main segment.

④Calculation of lower boundary hardness data of martensite as the main segment

In the normalized descending order distribution curve of microhardness data, starting from the highest hardness, gradually increase the data points for linear regression (y=a _Ⅱ +b _Ⅱ f _s ), a _Ⅱ is a constant term, which is also the equation on the Y axis The intercept, b _Ⅱ is the regression coefficient, which is also the slope of the linear regression equation y=a _Ⅱ +b _Ⅱ f _s on the coordinate axis, y is the n microhardness data arranged in order, and the corresponding value of each group of data is calculated Slope b _Ⅱ (including b _Ⅱ (n-2), b _Ⅱ (n-3)...b _Ⅱ 1), fsn, fs(n-1), fs(n-2)... are microhardness data The abscissas of the normalized descending distribution curve are n/n, (n-1)/n, (n-2)/n..., yn, y(n-1), y(n-2)...respectively It is the vertical coordinate corresponding to the abscissa fsn, fs(n-1), fs(n-2)... in the normalized descending order distribution curve of the microhardness data, and the specific operation is as follows:

Then use the obtained slope values (b _Ⅱ 1, b _Ⅱ 2... b _Ⅱ (n-3), b _Ⅱ (n-2)) as the slope distribution curve II, and the abscissa (fs) is (y1, y2 ...y(n-3), y(n-2)), the ordinate corresponds to (b _Ⅱ 1, b _Ⅱ 2...b _Ⅱ (n-3), b _Ⅱ (n-2)), and the slope Find the first qualified minimum value b _Ⅱ j from right to left on the distribution curve II (at least 3 consecutive data points on the left are decreasing, and at least 3 data points on the right are increasing), b _Ⅱ j is b _Ⅱ 1. One of b _Ⅱ 2...b _Ⅱ (n-3), b _Ⅱ (n-2), at this time, yj corresponding to b _Ⅱ j is the lower boundary hardness data of martensite as the main segment.

⑤ Calculate the proportion of each typical segment.

The proportion of ferrite as the main segment M% = i/n

The proportion of martensite as the main segment F% = j/n

The proportion of ferrite and martensite as the main segment (F+M)%=1-i/n-j/n

Example 1: Select a "main steam pipe size head" pipe fitting to process a metallographic sample in an area with uneven local hardness, select a 10mm×10mm area on the metallographic observation surface, and equidistantly space the grid lines as shown in Figure 1 A microhardness test (HV0.1) was carried out. The microhardness values are shown in Table 1.

Table 1: Equidistant microhardness test results (HV0.1)

First, the data points in Table 1 are integrated, and the abnormal data points (144HV0.1, 164HV0.1) are removed. At this time, n=98, and the normalized ascending and descending distribution curves of microhardness data are drawn respectively.

According to step ③, gradually increase the data points starting from the lowest hardness to perform linear regression (y=a _Ⅰ +b _Ⅰ f _s ), find the slope corresponding to each group of data, and then use the obtained slope value as the slope distribution curve Ⅰ , find the first qualified minimum value b _Ⅰ i from left to right on the curve (there are at least 3 consecutive data points decreasing on the left, and at least 3 consecutive data points increasing on the right), and its corresponding hardness is 188HV0 .1 (i=36, n=98 at this time), that is, 188HV0.1 is the upper boundary hardness data of ferrite as the main segment.

According to step ④, gradually increase the data points starting from the highest hardness to perform linear regression (y=a _Ⅱ +b _Ⅱ f _s ), find the slope corresponding to each group of data, and then use the obtained slope value as the slope distribution curve Ⅱ , find the first qualified minimum value b _Ⅱ j from right to left on the curve (the left side has at least 3 consecutive data points decreasing, and the right side has at least 3 consecutive data points increasing), its corresponding hardness is 220HV0 .1 (at this time j=16, n=98), that is, 220HV0.1 is the lower boundary hardness data of martensite as the main segment.

Finally, based on the above results, the segmented microhardness curves of the main steam pipe samples with small and large heads are obtained, as shown in Figure 6, where:

172～188HV0.1, ferrite is the main segment, and the proportion is F%=36/98=36.7%;

221～230HV0.1, martensite is the main segment, and the proportion M%=16/98=16.3%;

191～220HV0.1, ferrite plus martensite as the main segment, the proportion (F+M)%=100%-36.7%-16.3%=47.0%.

The ferrite determined by the method of the invention is the main section, the ferrite plus martensite mixed section, and the microhardness of the martensite main section is in good agreement with the microhardness measured result of the control metallographic picture.

The embodiments of the present invention described above are not intended to limit the protection scope of the present invention. Any modifications, equivalent replacements and improvements made within the spirit and principle of the present invention shall be included in the protection scope of the claims of the present invention.

Claims (3)

1. a kind of method of measure 9Cr ferritic heat-resistant steel hardness inhomogeneities, comprises the following steps：

1. obtain the equidistant micro-hardness measurement data of related steel sample；

2. micro-hardness data normalization ascending order and descending point according to corresponding to equidistant micro-hardness measurement data are made respectively Cloth curve；

3. in micro-hardness data normalizes ascending order distribution curve, gradually increase data point using lowest hardness as starting point to linear The slope of regression equation reaches qualified minimum, and condition refers in slope distribution curve, adjacent with minimum point At least continuous 3 data points in the left side are successively decreased, and at least continuous 3 data points in the right adjacent with minimum are incremented by, so as to Obtain the coboundary hardness data that ferrite is principal piece；

Step 3. in, micro-hardness data normalize ascending order distribution curve in, gradually increase number by starting point of lowest hardness Strong point carries out linear regression y=a_Ⅰ+b_Ⅰf_s, a_ⅠFor constant, y is the n micro-hardness data arranged by ascending order, is obtained following every Slope b corresponding to one group of data_Ⅰ, slope b_ⅠIncluding b_Ⅰ3、b_Ⅰ4、b_Ⅰ5......b_ⅠN, fs1, fs2, fs3...fsn are microhardness The abscissa of data normalization ascending order distribution curve, abscissa value are 1/n, 2/n, 3/n......n/n, y1, y2, y3...yn Ordinate, tool corresponding to abscissa fs1, fs2, fs3...fsn difference in ascending order distribution curve are normalized for micro-hardness data Gymnastics is made as follows：

The slope value b that will be tried to achieve again_Ⅰ3、b_Ⅰ4、b_Ⅰ5......b_ⅠN makees slope distribution curve I, abscissa y3, y4, Y5......yn, ordinate correspond to b_Ⅰ3、b_Ⅰ4、b_Ⅰ5......b_ⅠN, found on curve from left to right first it is eligible Minimum b_ⅠI, b_ⅠI is b_Ⅰ3、b_Ⅰ4、b_Ⅰ5......b_ⅠOne in n, now b_ⅠYi corresponding to i is that ferrite is principal piece Coboundary hardness data；

4. in micro-hardness data normalizes descending distribution curve, gradually increase data point using maximum hardness as starting point to linear The slope of regression equation reaches qualified minimum, and condition refers in slope distribution curve, adjacent with minimum point At least continuous 3 data points in the left side are successively decreased, and at least continuous 3 data points in the right adjacent with minimum point are incremented by, from And obtain the lower boundary hardness data that martensite is principal piece；

Step 4. in, micro-hardness data normalize descending distribution curve in, gradually increase number by starting point of maximum hardness Strong point carries out linear regression y=a_Ⅱ+b_Ⅱf_s, a_ⅡFor constant, y is the n micro-hardness data arranged in descending order, obtains each group Slope b corresponding to data_Ⅱ, slope b_ⅡIncluding b_Ⅱ(n-2)、b_Ⅱ(n-3)......b_Ⅱ1, fsn, fs (n-1), fs (n-2) ... fs1 For micro-hardness data normalize descending distribution curve abscissa, abscissa value be n/n, (n-1)/n, (n-2)/ N......1/n, yn, y (n-1), y (n-2) ... y1 are respectively abscissa in micro-hardness data normalization descending distribution curve Fsn, fs (n-1), ordinate corresponding to fs1, concrete operations are as follows by fs (n-2) ...：

The slope value b that will be tried to achieve again_Ⅱ1、b_Ⅱ2......b_Ⅱ(n-3)、b_Ⅱ(n-2) slope distribution curve II is made, abscissa is Y1, y2......y (n-3), y (n-2), ordinate corresponds to b_Ⅱ1、b_Ⅱ2......b_Ⅱ(n-3)、b_Ⅱ(n-2), looked on curve Turned left first qualified minimum b to from the right side_ⅡJ, b_ⅡJ is b_Ⅱ1、b_Ⅱ2......b_Ⅱ(n-3)、b_Ⅱ(n-2) one in It is individual, now b_ⅡYj corresponding to j is the lower boundary hardness data that martensite is principal piece；

5. the coboundary hardness data and martensite using the ferrite obtained as principal piece are counted for the lower boundary hardness data of principal piece It is principal piece to calculate ferrite in tapping sample tested region, and ferrite adds that martensite is principal piece and martensite is this 3 typical segments of principal piece Proportion；Step 5. in, determine that ferrite is principal piece proportion according to the coboundary hardness data that ferrite is principal piece F%, determine that martensite is principal piece proportion M% according to the lower boundary hardness data that martensite is principal piece, final calculate is tapped a blast furnace Ferritic adds ratio (F+M) %=100%-F%-M% of martensite shared by principal piece；Ferrite is principal piece proportion M%= i/n；Martensite is principal piece proportion F%=j/n；Ferrite adds martensite to be principal piece proportion (F+M) %=1-i/n- j/n。

2. a kind of method of measure 9Cr ferritic heat-resistant steel hardness inhomogeneities according to claim 1, its feature exist In, described step 1. and step 2. between also have following steps：Remove exceptional data point.

3. a kind of method of measure 9Cr ferritic heat-resistant steel hardness inhomogeneities according to claim 1, its feature exist In, step 1. in, the preparation process of steel sample is：Routinely tested sample sightingpiston is polished, polished, invaded by metallographic sample preparation method Erosion.