JP2010217076A - Method of evaluating inclusion - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method of evaluating inclusions for stably evaluating a nonmetal inclusion included in a metal material quickly at low costs. <P>SOLUTION: A tensile test is performed to a tensile test piece made of a metal material where hydrogen has penetrated, and destruction with a nonmetal inclusion affected by hydrogen as a starting point is generated in the tensile test piece. Then, the type of the nonmetal inclusion starting destruction existing in a risk volume of the tensile test piece is identified and the dimensions are measured. The dimensions of the nonmetal inclusion are obtained by calculating the square root of a projection area in the direction of a tension axis in the tensile test. By analyzing the square root of the projection area by a method of extreme value statistics, the dimensions of the maximum nonmetal inclusion is predicted from among nonmetal inclusions existing in the metal material. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a method for evaluating inclusions contained in a metal material.

  It is said that about 80% of destruction accidents in machinery, structures, etc. are caused by fatigue failure, and it is extremely important to prevent destruction accidents due to fatigue failure. In high-strength steels with a hardness of HV400 or higher, it is known that defects and non-metallic inclusions contained in the material are the starting points of fatigue failure, so investigate non-metallic inclusions contained in the material. Is very important.

Various methods for evaluating non-metallic inclusions have been proposed around the world. For example, the American ASTM method, the German VDEh method, and the Japanese JIS (Japanese Standards Association) point calculation method.
However, it has been pointed out that there is no clear correlation between the inclusion index defined by these evaluation methods and the fatigue strength (see Non-Patent Documents 1 and 2). This is because the above-described method for evaluating non-metallic inclusions has not been proposed based on detailed observation results of fatigue phenomena. Detrimental to fatigue strength is the largest non-metallic inclusions contained in the material, and the chemical composition, number, and distribution density of non-metallic inclusions are essentially unrelated.

As a method for predicting the size of the largest non-metallic inclusion contained in the material, there is an inclusion evaluation method based on an extreme value statistical method (see Patent Document 1). The extreme value statistical method is a method of predicting the size of the largest nonmetallic inclusion among all the nonmetallic inclusions contained in the material from the distribution of the nonmetallic inclusion in a part of the material.
In addition, as a method for quantitatively evaluating the cleanliness of a metal material, a method of extracting and evaluating non-metallic inclusions from a metal material by acid dissolution (see Patent Documents 2 and 3), an electron beam melting method (EB) There is a method (see Patent Document 4) in which a metal material is dissolved by a melting method) and nonmetal inclusions that have floated are observed with a microscope.

  On the other hand, if a fatigue test is performed, it is considered that the largest non-metallic inclusion among non-metallic inclusions contained in the dangerous volume is the starting point and breakage occurs (see Non-Patent Document 3). Therefore, an actual fatigue test was performed to cause a test specimen made of a metallic material to break starting from nonmetallic inclusions, and the dimensions of the nonmetallic inclusions were measured. A method for predicting the size of the largest nonmetallic inclusion among the nonmetallic inclusions to be performed is disclosed (see Patent Document 5).

By the way, it has long been known that hydrogen reduces the strength of metal materials, and this phenomenon is called “hydrogen embrittlement”, “hydrogen assisted cracking”, or “delayed fracture”, and many studies have been conducted. ing. However, at present, the mechanism by which hydrogen lowers the strength of a metal material has not been fully clarified.
Fukui et al. Used a low-alloy steel with different tempering temperatures to calculate the ratio of bending stress to static bending strength at which delayed fracture does not occur for more than 30 hours in an aqueous hydrogen chloride solution (hereinafter referred to as delayed fracture strength ratio). It has been found that if the hardness is HRC40 (HV375) or more, the delayed fracture strength ratio is significantly reduced (see Non-Patent Document 4).

Japanese Patent Laid-Open No. 11-230961 JP-A-9-125199 JP 9-125200 A JP-A-6-192790 Japanese Patent No. 3944568

J. Monnot, B. Heritier, J. Y. Cogne, "Relationship of Melting Practice, Inclusion Type, and Size with Fatigue Resistance of Bearing Steel", ASTM STP, 987 (1988), p.911. Akira Adachi, Hideo Shoji, Ikuo Kuwahara, Yoshiyuki Inoue, “Rotating Bending Fatigue of Bearing Steel”, Electric Steel Making Vol. 46, no. 3 (1975), p. 176-182 Takayoshi Murakami, “Effects of Metal Fatigue and Small Defects and Inclusions”, Yokendo, 1993, p94 Shoichi Fukui, “Effect of Tempering on Delayed Fracture Properties of Low Alloy Steel”, Iron and Steel, Nippon Steel & Steel Association Vol. 55, no. 12 (1969), p. 151-161 Zhou Seei, Takayoshi Murakami, Yoshihiro Fukushima, Stefano Beretta, "Extreme statistical evaluation of non-metallic inclusions based on three-dimensional observations", Iron and Steel, Nippon Steel & Steel Association Vol. 87, no. 12 (2001), p. 748-755 R. Takahashi and M. Shibuya, "Prediction of maximum size in Wicksell's corpuscle problem", Annals of the Institute of Statistical Mathematics, Vol. 50, No. 2 (1998), p361-377 R. Takahashi and M. Shibuya, "The maximum size of the planar sections of random spheres and its application to metallurgy", Annals of the Institute of Statistical Mathematics, Vol. 48, No. 1 (1996), p127-144 T. Fujita and Y. Yamada, "Stress corrosion cracking and hydrogen embrittlement of iron base alloys", (1977), p736, NZCE-5 Keizo Onishi and Kaga Kotoshi, “Room-temperature hydrogen gas embrittlement of structural steel”, Summary of the 105th Lecture Meeting Discussion, Japan Iron and Steel Institute, p. A136-A139

The evaluation of nonmetallic inclusions by the extreme value statistical method as in Patent Document 1 quantifies the size of the largest nonmetallic inclusion (square root √area _max of the projected area of the largest nonmetallic inclusion) based on microscopic observation. It is a predictable method and is used by many researchers.
However, when a SCM435 as the metal material, when predicting the size of the largest non-metallic inclusions as a starting point of fatigue fracture in a relatively small reference area S ₀ by microscopic observation, that there is an error may increase It has been pointed out (see Non-Patent Document 5). The reason for this is that the type (component) of the non-metallic inclusions that are the starting point of fatigue fracture differs from the non-metallic inclusions observed with a microscope.

Further, in order to accurately evaluate the type of non-metallic inclusions that are the starting point of fatigue fracture by microscopic observation, it has been pointed out that the limit value S _crit of the inspection standard area is about 10,000 mm ² or more in SCM435. Yes. Therefore, it is not always easy to find a large non-metallic inclusion as a starting point of fatigue fracture by microscopic observation, and it is considered to be expensive.

  Next, the evaluation methods described in Patent Documents 2 and 3 are methods for extracting nonmetallic inclusions from a metal material by acid dissolution, but when a plurality of nonmetallic inclusions are connected to form a cluster. The maximum non-metallic inclusion in the inspection volume may not be measured accurately. Furthermore, since nonmetallic inclusions may be dissolved in acid, it was not the best method for evaluating various nonmetallic inclusions.

  The evaluation method described in Patent Document 4 is a method in which a metal material is melted by an electron beam. However, when a plurality of nonmetallic inclusions are connected to form a cluster shape, Metal inclusions may not be measured accurately.

Furthermore, the evaluation method described in Patent Document 5 is to evaluate non-metallic inclusions using a fatigue test, but when evaluated using an ultrasonic fatigue tester, if the repetition rate is 20 kHz, Since the stress repetition number of about 10 ⁷ is about 10 minutes and the stress repetition number of about 10 ⁸ is completed in about 100 minutes, rapid evaluation is possible. However, as described in Patent Document 5, the ultrasonic fatigue tester has a problem that an hourglass-type test piece has to be used. When the ultrasonic fatigue testing machine described in Patent Document 5 is used, the dangerous volume of the test piece is small, 33 mm ³ . In the case of a fatigue test piece of a hydraulic servo fatigue tester to be described later, the dangerous volume is 1272 mm ³ . Since the data obtained from the specimen having a large dangerous volume is highly reliable, further inspection is desired for the inspection of non-metallic inclusions using an ultrasonic fatigue tester.

In the hydraulic servo type fatigue tester, it is possible to use a test piece having a parallel portion, and therefore, it is possible to use a test piece having a dangerous volume 10 times or more that of an ultrasonic fatigue tester. However, the repetition rate of the hydraulic servo fatigue tester is slower than that of the ultrasonic fatigue tester and is about 20 to 1000 Hz. When the repetition rate is 20 Hz, it takes about 6 days for a stress repetition number of about 10 ⁷ times and about 58 days for a stress repetition number of about 10 ⁸ times. Even when the repetition rate is 500 Hz, a stress repetition number of about 10 ⁷ requires about 6 hours and a stress repetition number of about 10 ⁸ requires about two days. Further, the axial load fatigue testing machine has a problem that the mounting of the test piece is complicated because the bending stress tends to act on the test piece.

  Then, this invention makes it a subject to solve the trouble which the above prior arts have, and to provide a simple and quick inclusion evaluation method.

  As a result of intensive studies by the present inventors in order to solve the above-mentioned problems, if hydrogen is allowed to penetrate into a metal specimen and a tensile test is performed in the presence of hydrogen in the specimen, the influence of hydrogen As a result, it was found that the breakage starting from the largest non-metallic inclusion in the dangerous volume of the test piece was likely to occur, and the test piece was broken at a low load. As a result, it was found that the evaluation of non-metallic inclusions can be performed in a short time and at low cost, and stable evaluation is possible. Furthermore, when the dimension of the nonmetallic inclusion which became the starting point of destruction was analyzed by the extreme value statistical method, it discovered that the dimension of the largest nonmetallic inclusion in a metal material was predictable.

  That is, in the inclusion evaluation method of the present invention, a tensile test is performed on a test piece made of a metal material into which hydrogen has been infiltrated, and breakage starting from a non-metallic inclusion affected by the hydrogen is detected as the test piece. Then, the type of the non-metallic inclusion which is the starting point of the destruction existing in the dangerous volume of the test piece is identified and the dimension is measured, and the distribution function of the dimension of the non-metallic inclusion is obtained. The cleanliness of the metal material is evaluated by the distribution function. In addition, the cleanliness of the metal material can be evaluated in the same manner as described above by performing a tensile test while allowing hydrogen to penetrate into a test piece made of a metal material.

In such an inclusion evaluation method of the present invention, if the projected area in the tensile axis direction of the tensile test is measured for the nonmetallic inclusion, and the square root of the projected area is analyzed by an extreme value statistical method, the metal The size of the largest nonmetallic inclusion among the nonmetallic inclusions present in the material can be predicted.
The amount of hydrogen in the test piece is preferably 0.2 ppm or more, more preferably 0.3 ppm or more, and further preferably 0.6 ppm or more. Further, the hardness of the metal material is preferably HV400 or more, more preferably HV450 or more, and further preferably HV500 or more.

  The inclusion evaluation method of the present invention is a simple method, and according to the inclusion evaluation method, stable evaluation can be performed in a short time and at low cost.

It is a figure which shows the shape and dimension of a tensile test piece. It is the extreme value statistical graph which analyzed the data of the dimension of the nonmetallic inclusion obtained by metallurgical microscope observation by the extreme value statistical method. It is a figure which shows the shape and dimension of a fatigue test piece. It is the extreme value statistical graph which analyzed the data of the dimension of the nonmetallic inclusion obtained by the fatigue test by the extreme value statistical method. It is the graph which converted the extreme value statistical graph obtained by the fatigue test of FIG. It is a graph which shows the hydrogen release characteristic of a tensile test piece. It is a graph which shows an example of a nominal stress-nominal strain curve. It is a graph which shows an example of a nominal stress-nominal strain curve. It is a graph which shows an example of a nominal stress-nominal strain curve. It is a graph which shows an example of a nominal stress-nominal strain curve. It is a graph which shows an example of a nominal stress-nominal strain curve. 2 is an enlarged view of a microscope observing a fracture surface of a tensile test piece of Example 1. FIG. It is the microscope enlarged view which observed the fracture surface of the tensile test piece of Example 14. 6 is an enlarged view of a microscope observing a fracture surface of a tensile test piece of Comparative Example 2. FIG. 10 is an enlarged view of a microscope observing a fracture surface of a tensile test piece of Comparative Example 6. FIG. It is the extreme value statistical graph which analyzed the data of the dimension of the nonmetallic inclusion obtained by the tension test by the extreme value statistical method. It is a graph which shows the relationship between tensile strength and hydrogen content. It is a graph which shows the relationship between tensile strength ratio and the hardness of a tensile test piece.

Embodiments of the inclusion evaluation method according to the present invention will be described in detail with reference to the drawings.
First, a tensile test piece made of a metal material is prepared. The type of the metal material is not particularly limited, but bearing steel (for example, high carbon chromium bearing steel SUJ2), structural steel (for example, SCM435), spring steel (for example, SUP12), tool steel, Stainless steel or the like can be used.

  Next, hydrogen is allowed to penetrate into the tensile test piece (hereinafter, referred to as hydrogen charge). The method of hydrogen charging is not particularly limited, but a method in which a tensile test piece is immersed in an acidic aqueous solution (for example, an ammonium thiocyanate aqueous solution), a method in which the tensile test piece is exposed to hydrogen gas, and a tensile test piece in an electrolytic solution. (For example, a method of applying an electric current while being immersed in an aqueous solution of sodium chloride and ammonium thiocyanate or an aqueous solution of sulfuric acid and arsenous acid). The tensile test piece may be subjected to hydrogen charge, but the metal material before being processed into the shape of the tensile test piece may be subjected to hydrogen charge to produce the tensile test piece from the hydrogen-charged metal material. .

  Then, a tensile test (static test) is performed on the tensile test piece subjected to hydrogen charge, and the tensile test piece is broken. Tensile specimens that have been charged with hydrogen are susceptible to hydrogen and are susceptible to fracture starting from the largest non-metallic inclusions in the dangerous volume of the tensile specimen. Fractures with a lower load than those without. At this time, it is preferable that hydrogen exists uniformly in the dangerous volume of a tensile test piece.

  The hydrogen charge may be performed before the tensile test, but the tensile test piece may be charged with hydrogen during the tensile test. That is, a method in which a tensile test piece is immersed in an acidic aqueous solution (for example, an ammonium thiocyanate aqueous solution), a method in which the tensile test piece is exposed to hydrogen gas, or a tensile test piece in an electrolytic solution (for example, sodium chloride and ammonium thiocyanate). A tensile test may be performed while hydrogen charging the tensile test piece by a method of immersing in an aqueous solution or an aqueous solution of sulfuric acid and arsenous acid.

  Next, the nonmetallic inclusions in the fractured tensile test piece are analyzed. That is, the type of non-metallic inclusions that are present in the critical volume of the tensile test piece and that is the starting point of fracture is identified, the dimensions of the non-metallic inclusions are measured, and the distribution function of the dimensions of the non-metallic inclusions is determined. Ask. The cleanliness of the metal material can be evaluated by the distribution function thus obtained. According to such an evaluation method, non-metallic inclusions can be evaluated in a short time and at low cost, and stable evaluation is possible. Also, without using a special testing machine such as an ultrasonic fatigue testing machine, non-metallic inclusions that are the starting point of fatigue fracture can be easily and accurately evaluated by using a general tensile testing machine. Can do. Therefore, the present invention can be applied to quality assurance and quality improvement evaluation of metal materials such as steel materials. In particular, it is suitable for the evaluation of rolling bearing steel and precision product steel that require high cleanliness.

The dangerous volume means the volume of a portion where high stress acts on the test piece, which can be a starting point of fracture. In a test piece usually used in a tensile test or a fatigue test (dynamic test) described later, the volume of a parallel portion located between both ends to which stress is input corresponds to a dangerous volume. When the parallel part is a cylindrical test piece, the dangerous volume V _s can be expressed by 0.25πd ² l. Here, d is the diameter of the parallel part, and l is the length of the parallel part.

  Furthermore, it is preferable that the dimension of the non-metallic inclusion that is the starting point of the destruction is the square root of the projected area of the non-metallic inclusion. That is, for the non-metallic inclusion that is the starting point of the fracture, the projected area when measured from the tensile axis direction of the tensile test is measured, the square root √area of the projected area is calculated, and the distribution of this value The cleanliness of the metal material can be evaluated by the function. At this time, if the square root √area of the projected area is analyzed by the extreme value statistical method, the dimension of the largest nonmetallic inclusion among the nonmetallic inclusions existing in the metallic material can be predicted.

  The square root √area of the projected area area corresponds to the equivalent defect size. In the case of a fatigue test to be described later, the projected area of the non-metallic inclusion that is the starting point of fracture when viewed from the principal stress direction of the fatigue test is measured, and this is taken as the projected area. Moreover, when observing a nonmetallic inclusion by metal microscope observation mentioned later, the area of a nonmetallic inclusion is measured from the obtained microscope image (for example, micrograph), and this is made into the projection area area.

  Further, the extreme value statistical method means a method of analyzing the extreme value distribution. The extreme value distribution means a distribution that is followed by the maximum value or the minimum value of each set when a set of a certain number of data is extracted from a data group according to a certain basic distribution function. Even if the basic distribution is a normal distribution or an exponential distribution, the extreme value distribution is different from the basic distribution.

Hereinafter, the present invention will be described more specifically with reference to examples, comparative examples, and conventional examples.
[Evaluation of non-metallic inclusions by metal microscope observation]
A conventional example in which a nonmetallic inclusion contained in a metal material is evaluated by observation with a metal microscope will be described. This evaluation method is the method described in Patent Document 1 described above.
The kind of the metal material is the same high-carbon chromium bearing steel SUJ2 as that used in the conventional example (fatigue test) and the example (tensile test) described later. The composition ratio of the alloy components is 1.00% by mass of carbon, 1.44% by mass of chromium, 0.26% by mass of silicon, 0.36% by mass of manganese, 0.03% by mass of molybdenum, 0.011% by mass of phosphorus, 0.007% by mass of sulfur, 0.10% by mass of copper, 0.07% by mass of nickel, 0.002% by mass of titanium, 6 ppm by mass of oxygen, and the balance is iron.

  Using such a metal material, a test piece for observing a metallographic microscope was produced as follows. A parallel portion of the tensile test piece of FIG. 1 to be described later is cut so that a plane orthogonal to the longitudinal direction (tensile axis direction) of the tensile test piece becomes a cut surface, and has a diameter of 4.5 mm and a thickness of 2.0 mm. A cylindrical specimen for observation with a metal microscope was cut out. This tensile test piece is an unused one that has not been subjected to a tensile test.

A cylindrical metal microscope observation specimen is embedded in a resin, and only one of the upper and lower flat surfaces of the metal microscope observation specimen is polished using emery paper and surface-finished to # 2000. (Hereinafter, surface finishing using such emery paper is referred to as emery polishing), and buffing was performed. The inspection reference area S ₀ is the entire plane of the test piece for metal microscope observation, and is 15.9 mm ² (4.5 × 4.5 × π / 4). There are 40 observation points (number of inspection visual fields: 40) by a metal microscope.

The metal micrograph of the observation surface is image-processed by an image processing apparatus, and the dimension of the largest non-metallic inclusion among the non-metallic inclusions in the inspection reference area S ₀ (the square root √area of the projected area of the largest non-metallic inclusion) ) Was measured. Thereafter, non-metallic inclusions were identified with an energy dispersive X-ray analyzer (EDX) attached to a scanning electron microscope (SEM). The acceleration voltage in SEM observation and EDX analysis is 20 kV.

Thus, the data of the dimension of the largest nonmetallic inclusion obtained by observation with a metallographic microscope was analyzed by the extreme value statistical method. An extreme value statistical graph is shown in FIG. In this graph, y is a standardized variable (y = 0.711 × √area−4.984), F is a cumulative frequency, and T (= 1 / (1-F)) is a recursion period.
In the case of the two-dimensional inspection, by adding a virtual thickness h to the inspection reference area S ₀ , it can be approximately regarded as a three-dimensional inspection for the inspection reference volume V ₀ (= S ₀ × h). A specific method is to obtain an average value (= 7.77 × 10 ⁻³ mm) of the dimension √area _max of the maximum non-metallic inclusion measured by two-dimensional inspection, and use that value as the virtual thickness of the observation area. Let h. Therefore, the inspection reference volume V ₀ corresponding to the two-dimensional inspection is calculated by S ₀ and h and becomes 0.124 mm ³ . The validity of this approximate method is supported by the theoretical analysis of Takahashi et al. (See Non-Patent Documents 6 and 7).

  Further, when the chemical component analysis (elemental analysis) was performed on the largest nonmetallic inclusion by analysis by EDX, aluminum, calcium, and sulfur were detected in 14 out of 40 nonmetallic inclusions. Aluminum and sulfur were detected in the non-metallic inclusions. Regarding the remaining 25 non-metallic inclusions, the chemical constituents could not be analyzed because the non-metallic inclusions were missing.

[Evaluation of non-metallic inclusions by fatigue test]
A conventional example in which a non-metallic inclusion contained in a metal material is evaluated by a fatigue test will be described. This evaluation method is the method described in Non-Patent Document 3 described above.
The kind of the metal material is exactly the same as that used in the case of the above-mentioned metal microscope observation and in the case of the tensile test described later. Using this metal material, a fatigue test piece having a shape and dimensions as shown in FIG. 3 was produced. That is, a round bar having a diameter of 25 mm was machined so that the machining allowance of the parallel part was 1.0 mm per diameter and the machining allowance of the other part was 0.6 mm per diameter, and then heat treatment was performed. The contents of the heat treatment are quenching (oil cooling) at 840 ° C. and tempering (furnace cooling) at 240 ° C.

The critical volume V _s that becomes the starting point of the fracture of the fatigue test piece is 477 mm ³ . Further, when the Vickers hardness was measured at a measurement load of 2.94 N at 20 points from the surface to the center of the parallel part, there was almost no variation in hardness from the surface to the center, and was within ± 2%. . Moreover, the average value of the hardness of each fatigue test piece was HV682.
Next, the heat treatment-treated fatigue test piece was machined, emery polished, and buffed to finish the parallel portion. Some fatigue test pieces were charged with hydrogen by immersing them in an aqueous solution of ammonium thiocyanate having a concentration of 20% by mass at 40 ° C. for 48 hours. Then, the emery polishing and the buff polishing were performed on the parallel part, thereby removing the corrosive layer formed by the hydrogen charge and finishing.

  The fatigue test pieces (seven types) thus obtained were subjected to a fatigue test (dynamic test) to break the parallel part. This fatigue test is performed in the atmosphere using a hydraulic servo type tensile compression fatigue tester, the repetition rate f is 50 to 75 Hz, and the stress ratio R is −1 (tensile stress and compressive stress are equal). In the tensile compression fatigue test, a slightly lower fatigue strength tends to be obtained when bending stress acts on the test part. Therefore, attach a strain gauge at a position that equally divides the circumference near the gripping part into four parts for each specimen, and carefully adjust the specimen so that bending stress does not act on the specimen when stress is applied. Attached to the testing machine.

  For the fatigue test piece that was charged with hydrogen, the time from the end of the hydrogen charge to the start of the fatigue test was 3 hours. Further, immediately after the fatigue test was completed, the amount of hydrogen in the fatigue test piece was measured. The amount of hydrogen was measured by temperature programmed desorption gas analysis (TDS) using a quadrupole mass spectrometer. That is, a disc-shaped sample having a diameter of 7.0 mm and a thickness of 2.0 mm is cut out from the parallel portion of the fatigue test piece, heated to 573 K at a heating rate of 0.5 K / s, and the amount of desorbed hydrogen is determined. It was measured. The results are shown in Table 1.

  When the parallel part of the fatigue test piece was broken by the fatigue test, the entire surface perpendicular to the main stress direction (tensile stress direction) of the fatigue test was observed. That is, the fracture surface was observed with an SEM, and the dimension of the nonmetallic inclusion that was the starting point of the fracture (the square root √area of the projected area when viewed from the principal stress direction of the fatigue test) was measured. Thereafter, the nonmetallic inclusions were identified by EDX attached to the SEM. The acceleration voltage in SEM observation and EDX analysis is 20 kV.

Load stress in fatigue tests, rupture life (number of stress repetitions until the fatigue test piece breaks), type of non-metallic inclusions that are the starting point of fracture (in parentheses are elements constituting non-metallic inclusions), non-metallic Table 1 shows the square root √area of the projected area of inclusions and the amount of hydrogen present in the fatigue specimen after the fatigue test. The test piece No. No. 2 ^* is a test piece No. 2 that was non-ruptured with an additional stress of 850 MPa and a stress repetition number of 2 × 10 ⁷ times. This is a result of subjecting the fatigue test piece 2 to a fatigue test at an additional stress of 950 MPa.

Regardless of the presence or absence of hydrogen charge, the fatigue test specimens were destroyed starting from non-metallic inclusions. A typical fish eye was observed on the fracture surface of the fractured fatigue specimen, and a non-metallic inclusion was present in the center. The shape of the fish eye was almost circular. Since mainly aluminum and calcium were detected as chemical components of the nonmetallic inclusions, the nonmetallic inclusions are considered to be Al ₂ O _3. (CaO) _x series.

Thus, the data of the dimension (square root √area of the projected area) of the non-metallic inclusion that was the starting point of the fracture obtained by the fatigue test was analyzed by an extreme value statistical method. FIG. 4 shows an extreme value statistical graph (plotted with square marks). In this graph, y is a standardized variable (y = 0.194 × √area−3.030), F is a cumulative frequency, and T is a recursion period. The inspection reference volume V _{0 in} this case is the volume of the parallel part of the fatigue test piece. That is, the dangerous volume V _s of the fatigue test piece is 477 mm ³ .

In addition, this graph also shows the extreme value statistical graph (circled plot) obtained by observation with the metallographic microscope shown in FIG. The inspection reference volume V _{0 in} the case of observation with a metal microscope is 0.124 mm ³ . Thus, the extreme value statistical data when the numerical value of the inspection reference volume V ₀ is different can be converted and evaluated based on the inspection reference volume V ₀ of the reference sample. Therefore, according to the procedure described in Non-Patent Document 5, when the non-metallic inclusions that are the starting points of the fracture by the fatigue test are converted and evaluated based on the inspection reference volume V ₀ , a graph of FIG. 5 is obtained. That is, the straight line of the extreme value statistical graph is a straight line translated by ln (V _s / V ₀ ), and in this case, ln (V _s / V ₀ ) is 8.236.

The reason why the slopes of the two extreme statistical graphs are different is because the nonmetallic inclusions that are the starting point of the fracture in the fatigue test are different from the types (components) of nonmetallic inclusions that are observed by metallurgical microscopy. It is thought that. It is also pointed out in Non-Patent Document 5 that the slopes of both straight lines are different. In order to predict non-metallic inclusions that will be the starting point of fracture in fatigue tests, it is necessary to observe a large area with an inspection reference area S ₀ of 15.9 mm ² or more in the evaluation by metal microscope observation. Therefore, finding a large non-metallic inclusion as a starting point of destruction by observation with a metal microscope is not always an easy operation and is considered to increase the cost.

[Evaluation of non-metallic inclusions by tensile test]
Examples of the present invention in which non-metallic inclusions contained in a metal material are evaluated by a tensile test will be described.
The kind of the metal material is exactly the same as that used in the case of the above-mentioned metal microscope observation and fatigue test. Using this metal material, a tensile test piece having a shape and dimensions as shown in FIG. 1 was produced. That is, after machining a round bar with a diameter of 25 mm so that the allowance for the parallel part is 1.0 mm per diameter and the allowance for the other part is 0.6 mm per diameter, Heat treatment was performed under any of the conditions.

Condition A: After holding at 840 ° C. in an RX gas atmosphere, quenching was performed with oil cooling, and then holding at 240 ° C. for 2 hours, followed by furnace cooling and tempering.
Condition B: After holding at 840 ° C. in an RX gas atmosphere, quenching was performed with oil cooling, and then holding at 300 ° C. for 2 hours, followed by furnace cooling and tempering.
Condition C: After holding at 840 ° C. in an RX gas atmosphere, quenching was performed with oil cooling, and then holding at 370 ° C. for 2 hours, followed by water cooling and tempering.
Condition D: After maintaining at 840 ° C. in an RX gas atmosphere, quenching was performed with oil cooling, and then maintaining at 470 ° C. for 2 hours, followed by water cooling and tempering.
Condition E: After holding at 840 ° C. in an RX gas atmosphere and quenching with oil cooling, holding at 550 ° C. for 2 hours, followed by oil cooling and tempering.

The critical volume V _{s that} is the starting point of the fracture of these tensile test pieces is 477 mm ³ . Further, when the Vickers hardness was measured at a measurement load of 2.94 N at 20 points from the surface of the parallel part to the center part, there was almost no variation in hardness from the surface to the center part, and was within ± 3%. . Furthermore, the average value of the hardness of each tensile test piece is HV678 (variation is within ± 2%) when the heat treatment condition is condition A, and HV611 (variation is within ± 2%) when the heat treatment condition is condition B. ) HV559 (variation is within ± 3%) when the heat treatment condition is condition C, HV447 (variation is within ± 2%) when the heat treatment condition is condition D, and HV346 when the heat treatment condition is condition E (The variation was within ± 3%).

  Next, machining, emery polishing, and buffing were performed on the heat-treated tensile test pieces to finish parallel portions. Some of the tensile test pieces were charged with hydrogen by immersing them in an aqueous solution of ammonium thiocyanate having a concentration of 1% by mass or 20% by mass at 40 ° C. for 48 hours. The concentration of the aqueous ammonium thiocyanate solution may be referred to as the hydrogen charge concentration). Then, the emery polishing and the buff polishing were performed on the parallel part, thereby removing the corrosive layer formed by the hydrogen charge and finishing.

  A tensile test (static test) was performed on the tensile test pieces (24 types) thus obtained, and the parallel part was broken. This tensile test was performed at room temperature in the atmosphere, and the tensile speed (crosshead speed V) was any one of 0.05, 1 and 100 mm / min. Since the tensile test is performed at room temperature in the atmosphere, the amount of hydrogen contained in the tensile test piece varies depending on the time from the end of the hydrogen charge to the start of the tensile test. It is. Therefore, the time from the end of the hydrogen charge to the start of the tensile test (standing time) is unified to 2 hours, and the amount of hydrogen contained in the tensile test piece at the start of the tensile test is almost the same for each tensile test piece. I did it.

In addition, from the end of the hydrogen charge to the start of the tensile test, the hydrogen amount is changed by leaving it at room temperature in the atmosphere for 24 hours, 72 hours, 150 hours, or 200 hours to change the tensile strength and The relationship between fracture mode and hydrogen content was investigated.
With respect to the tensile test piece subjected to hydrogen charge, the amount of hydrogen present in the tensile test piece was measured as soon as the tensile test was completed. The amount of hydrogen was measured by temperature programmed desorption gas analysis (TDS) using a quadrupole mass spectrometer. That is, a disk-shaped sample having a diameter of 4.5 mm and a thickness of 2.0 mm was cut out from the parallel portion of the tensile test piece, heated to 573 K at a temperature increase rate of 0.5 K / s, and the amount of hydrogen desorbed It was measured. The results are shown in Table 2.

  Here, the hydrogen release characteristics of the tensile specimens charged with hydrogen were investigated. A round bar test piece (diameter: 4.5 mm, length: 100 mm) made of the metal material was used for investigation of the hydrogen release characteristics. Then, after heat treatment and hydrogen charging as described above, the amount of hydrogen was measured by a temperature programmed desorption gas analysis method using a quadrupole mass spectrometer. The temperature increase rate was 0.5 K / s, and the temperature was increased to 573 K. The hydrogen charge concentration is 20% by mass.

  In order to examine the hydrogen release characteristics at room temperature, a chip-like sample having a thickness of 2.0 mm was cut out from a hydrogen-charged round bar test piece every time a predetermined time passed, and the amount of hydrogen was measured. This chip sample was cut out from a position 4.5 mm or more away from the end face of the round bar test piece. The amount of hydrogen of the round bar test piece under the heat treatment condition A is shown in the graph of FIG. As can be seen from this graph, the amount of hydrogen decreased monotonously as time elapsed from the end of hydrogen charging. In addition, the hydrogen amount of the round bar test piece which is not hydrogen-charged is 0.02 ppm.

  When the parallel part of the tensile test piece was broken by the tensile test, the entire surface perpendicular to the tensile axis direction (tensile stress direction) was observed. That is, the fracture surface was observed with an SEM, and the dimension of the non-metallic inclusion that was the starting point of fracture (the square root √area of the projected area when viewed from the tensile axis direction of the tensile test) was measured. Thereafter, the nonmetallic inclusions were identified by EDX attached to the SEM. The acceleration voltage in SEM observation and EDX analysis is 20 kV.

  Tensile speed, tensile rupture strength, strain, hydrogen charge concentration, standing time (time from the end of hydrogen charge to the start of the tensile test), the amount of hydrogen present in the tensile test piece after the tensile test, Table 2 shows the types of metal inclusions (elements constituting nonmetallic inclusions in parentheses) and the square root √area of the projected area of the nonmetallic inclusions. For Comparative Examples 1, 2, 8, and 9, since the starting point of fracture was not a non-metallic inclusion but a matrix or surface of a tensile test piece, the type of non-metallic inclusion that was the starting point of fracture was In the column, “matrix” or “surface”, which is a starting point of destruction, is described.

  Here, an example of the nominal stress-nominal strain curve in Examples and Comparative Examples is shown in FIGS. In this case, the time from the end of the hydrogen charge to the start of the tensile test is 2 hours, and the tensile speed is 1 mm / min. In each figure, heat treatment conditions and the same hardness are compared, but the curves of the comparative examples in FIGS. 7 to 10 whose hardness is HV678,611,559,447 are broken in the plastic region after yielding. However, the curve of the example broke in the elastic region. Therefore, the tensile breaking strength of each Example is lower than the tensile breaking strength of the comparative example which is not hydrogen-charged. On the other hand, the curves of both comparative examples in FIG. 11 having a hardness of HV346 did not break in the elastic region, the tensile breaking strength was almost the same, and the strength was not reduced by hydrogen charging.

As can be seen from Table 2, Examples 1 to 15 that were charged with hydrogen had a hardness of HV400 or more, and were destroyed starting from non-metallic inclusions. As chemical components of the non-metallic inclusions, mainly aluminum and calcium were detected, so the non-metallic inclusions are considered to be Al ₂ O _3. (CaO) _x series.
FIG. 12 shows an enlarged microscopic view of the fracture surface of the tensile test piece of Example 1, and FIG. 13 shows an enlarged microscopic view of the fracture surface of the tensile test piece of Example 14. (A) of each figure is a figure of the whole fracture surface, (b) is the figure which expanded the origin of destruction and its peripheral part among (a).

From these figures, it can be seen that the destruction spread radially starting from the nonmetallic inclusion. Further, from FIG. 12B and FIG. 13B, it is possible to measure the dimension of the non-metallic inclusion that is the starting point of destruction, that is, the largest non-metallic inclusion in the dangerous volume.
Although the tensile test piece of Comparative Example 1 was charged with hydrogen, the tensile test piece caused cup-and-cone fracture, and the observation result of the fracture surface identified the nonmetallic inclusion that was the starting point of the fracture. I couldn't.

  Moreover, the tensile test pieces of Comparative Examples 2 to 9 are not charged with hydrogen, and the amount of hydrogen in the tensile test pieces is 0.03 ppm or less. Of Comparative Examples 2 to 9, 5 were broken starting from non-metallic inclusions, and 1 was broken starting from the surface of the tensile test piece, and broken starting from the base structure (matrix). There were 3 cases. Since Ti was detected as a chemical component of the nonmetallic inclusion, it is considered that the nonmetallic inclusion is TiN. In FIG. 14, the microscope enlarged view which observed the fracture surface of the comparative example 2 destroyed from the surface of the tension test piece as the starting point is shown. FIG. 15 is an enlarged view of a microscope observing the fracture surface of Comparative Example 6 in which the Ti-based non-metallic inclusion is broken as a starting point.

Data on the dimensions (square root of projected area √area) of non-metallic inclusions that were the starting points of fracture obtained by tensile tests of the tensile test pieces of Examples 1 to 15 were analyzed by an extreme value statistical method. An extreme value statistical graph is shown in FIG. Triangle marks are examples, and x marks are comparative examples. In this graph, y is a normalization variable, F is a cumulative frequency, and T is a recursion period. The normalization variable in the example is y = 0.1730 × √area−2.4879. The normalization variable of the comparative example is y = 1.15777 × √area-11.326. In the example and the comparative example, the components of the nonmetallic inclusions at the starting point of tensile fracture are different, so the extreme value statistical graphs do not match. The inspection reference volume V _{0 in} this case is the volume of the parallel part of the tensile test piece. That is, in the tensile test piece, the dangerous volume V _s is 477 mm ³ . This graph also shows an extreme value statistical graph (plotted with square marks) obtained by the fatigue test shown in FIG.

As can be seen from the graph in FIG. 16, the extreme value statistical graph obtained by the tensile test and the extreme value statistical graph obtained by the fatigue test almost coincide. Of the nonmetallic inclusions in the dangerous volume V _s (= 477 mm ³ ) of the test piece, using the recurring period T (N) in the graph of FIG. 16 and the distribution line obtained by the tensile test, When the maximum size √area _max of the non-metallic inclusion was predicted, it was 27.7 μm in the case of 10 test pieces and 41.0 μm in the case of 100 test pieces.

  On the other hand, when the prediction is performed in the same manner using the recursion period T (N) and the distribution straight line obtained by the fatigue test, 27.6 μm when the number of test pieces is 10 and 39. It was 5 μm. From these results, if a tensile test is performed on a tensile test piece that has been charged with hydrogen, and analysis is performed using the data of the dimensions of the nonmetallic inclusion that is the starting point of fracture, the result obtained in the fatigue test, that is, the metal material It can be seen that the dimension of the largest nonmetallic inclusion among the nonmetallic inclusions present therein can be accurately predicted.

  Next, for Examples 1 to 7, 11, and 12 and Comparative Examples 2 to 4, the relationship between tensile strength and hydrogen content was examined. The condition for heat treatment of these tensile test pieces is Condition A. In addition, the tensile test pieces of Examples 1 to 7, 11, and 12 were charged with hydrogen, and the tensile test pieces of Comparative Examples 2 to 4 were not charged with hydrogen. Further, in Examples 1 to 7, 11, and 12, the amount of residual hydrogen in the tensile test piece is changed by variously changing the time (standing time) from the end of the hydrogen charge to the start of the tensile test. .

From the graph of FIG. 17, it can be seen that the tensile strength is reduced due to hydrogen charging, and from this, it is presumed that the breakage starting from the nonmetallic inclusion is likely to occur due to the influence of hydrogen.
Onishi and Kaga performed creep type low-load tests under high-pressure hydrogen atmosphere using AISI 4340 steel with a strength of 1160 to 1570 MPa by changing the tempering temperature, and the fracture strength decreased with increasing hydrogen pressure, It shows that the lower the breaking strength is, the higher the strength is. Further, the amount of hydrogen was investigated, and it was shown that the hydrogen pressure was 0.14 ppm when the pressure was 3 MPa, 0.13 ppm when the pressure was 5 MPa, and 0.23 ppm when the pressure was 10 MPa (see Non-Patent Document 9).

  In the present invention, when a tensile test is performed in a state where 0.32 ppm of hydrogen is present, fracture starts from non-metallic inclusions, resulting in a decrease in strength. Therefore, considering based on the reports of Onishi and Kaga, when a tensile test is performed in a state where hydrogen is present in a tensile test piece in an amount of 0.2 ppm or more, fracture starts from non-metallic inclusions, resulting in a decrease in strength. it is conceivable that. The amount of hydrogen in the tensile test piece is preferably 0.2 ppm or more, more preferably 0.3 ppm or more (or 0.32 ppm or more), and 0.6 ppm or more (or 0.64 ppm or more). More preferably.

  For Examples 1 to 3, Example 13, Example 14, and Comparative Example 1, the ratio of tensile strength before and after hydrogen charging (tensile strength after hydrogen charging / tensile strength before hydrogen charging) was calculated, The relationship between the tensile strength ratio and the hardness of the tensile specimen was investigated. The hardness of the tensile test piece is adjusted by the heat treatment conditions.

  From the graph of FIG. 18, it can be seen that the tensile strength decreases as the hardness of the tensile test piece increases. Fukui et al. Used a low-alloy steel with a different tempering temperature and a bending stress ratio at which delayed fracture does not occur in an aqueous hydrogen chloride solution for 30 hours or more with respect to static bending strength (hereinafter referred to as delayed fracture strength ratio) When the hardness of the tensile test piece is HRC40 (HV375) or more, the delayed fracture strength ratio is significantly reduced.

  Moreover, Fujita et al. Calculated the tensile fracture strength when immersed in water for 100 hours using low alloy steel, and when the hardness of the tensile test piece was HV400 or more, tensile fracture when immersed in water for 100 hours. It has been reported that the strength is lower than the tensile fracture strength in the atmosphere (see Non-Patent Document 8).

  In the present invention, the influence of hydrogen differs between a tensile test piece having a hardness of HV366 and a tensile test piece having a hardness of HV447 or more. That is, in the tensile test piece having a hardness of HV366, the tensile fracture strength is almost the same regardless of the presence or absence of hydrogen charge. On the other hand, in a tensile test piece having a hardness of HV447 or more, hydrogen promotes fracture starting from non-metallic inclusions, and the tensile fracture strength decreases.

  Therefore, based on the reports of Fukui et al. And Fujita et al., If the hardness of the tensile test piece is HV400 or more, it is considered that breakage starting from nonmetallic inclusions is likely to occur due to the influence of hydrogen. In order to perform stable evaluation of metal inclusions, the tensile test piece preferably has a hardness of HV450 or more, more preferably HV500 or more.

Claims (8)

  After performing a tensile test on a test piece made of a metal material into which hydrogen has penetrated and causing the test piece to break starting from the non-metallic inclusions affected by the hydrogen, the danger of the test piece The type of non-metallic inclusion that is the starting point of the fracture existing in the volume is identified and the dimension is measured to obtain a distribution function of the dimension of the non-metallic inclusion, and the clean function of the metallic material is obtained by this distribution function. Inclusion evaluation method characterized by evaluating degree.   The projected area in the tensile axis direction of the tensile test is measured with respect to the non-metallic inclusions, and the square root of the projected area is analyzed by an extreme value statistical method to determine the largest non-metallic inclusions present in the metallic material. The inclusion evaluation method according to claim 1, wherein a dimension of the nonmetallic inclusion is predicted.   The inclusion evaluation method according to claim 1 or 2, wherein the amount of hydrogen in the test piece is 0.2 ppm or more.   The inclusion evaluation method according to claim 1 or 2, wherein an amount of hydrogen in the test piece is 0.3 ppm or more.   The inclusion evaluation method according to claim 1 or 2, wherein the amount of hydrogen in the test piece is 0.6 ppm or more.   The inclusion evaluation method according to any one of claims 1 to 5, wherein the metal material has a hardness of HV400 or more.   The inclusion evaluation method according to any one of claims 1 to 5, wherein the metal material has a hardness of HV450 or more.   The inclusion evaluation method according to any one of claims 1 to 5, wherein the metal material has a hardness of HV500 or more.

Images