JP3207038B2 - Estimation method of quenching thermal shock critical temperature difference - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for estimating a critical temperature difference due to quenching thermal shock of a brittle material such as ceramics.

[0002]

2. Description of the Related Art When a structural material such as ceramics is used for various purposes, it is important to design a dimension according to its strength. For that purpose, it is necessary to accurately and quickly grasp the mechanical and thermal properties of the material.

One of the characteristics required for such a member design is durability against thermal shock due to rapid cooling. A quenching thermal shock critical temperature difference ΔTc is generally used as an index indicating the durability. This indicates to what degree of temperature difference the member can withstand rapid cooling.

As a method for measuring the critical temperature difference between thermal shocks, JISR1615 is known. This JISR1
The method of 615 applies a thermal shock to each of a plurality of test pieces that do not form a pre-crack at different temperature differences, and statistically obtains an expected value of ΔTc from measurement results having variations.

[0005]

In this method, since the measurement results vary from one test piece to another, a large number of test pieces are required to obtain an accurate thermal shock critical temperature difference. Usually, the required number of test pieces is 30 or more even when the test is performed efficiently.

Further, since the obtained ΔTc depends on the size of the fracture source of the test piece, even if the same material is used, it varies depending on the manufacturing method, the processing state of the surface, or the size (effective volume) of the structure. Therefore, each time the fracture source size was different, it had to be measured.

[0007]

Means for Solving the Problems The present inventors have found that there is a relation of ΔTci = bCi− ^a between a critical temperature difference ΔTci at which a precrack starts to propagate due to thermal shock and the precrack length Ci. Using the fact that a plurality of pre-cracks are formed in one test piece and that the variation of ΔTci is small to break from the pre-crack, the above relational expression is obtained from a small number of test pieces. Further, they have found that the critical temperature difference ΔTc can be estimated with high accuracy by obtaining the fracture source size of a test piece into which no pre-crack is introduced as an equivalent crack length and substituting it into the relational expression.

That is, according to the present invention, a plurality of test specimens having pre-cracks having different lengths are tested at a plurality of temperature differences, and the pre-crack length Ci and the temperature difference at which the pre-cracks start to propagate are obtained. Two or more combinations of Tci (Ci, △ Tci) are measured, and constants a and b included in the relational expression are obtained. Further, from the fracture toughness value, the effective volume or effective surface area in the Weibull coefficient and the strength test and the effective volume or effective surface area of the quenching treatment, to estimate the equivalent crack length Co corresponding to the fracture source size of the specimen without pre-crack, From the equivalent crack length Co and the above relational expression, the critical temperature difference ΔT at which the test material is broken by thermal shock
c is estimated.

Hereinafter, the present invention will be described in more detail.

According to the method of the present invention, first, (Ci, ΔTci) is measured by the following quenching thermal shock test. First, a plurality of cylindrical test pieces made of a material to be measured are prepared. Then, pre-cracks having different lengths are introduced into each test piece. For example, as shown in the schematic diagram of the test piece in FIG. 1, the pre-crack is introduced at a plurality of locations using a Knoop hardness measuring indenter around a cylindrical test piece 1. The length of the pre-crack can be changed by changing the press-fit load of the indenter. When introducing precracks having different lengths, a precrack having the same length is introduced into one test piece, and a plurality of test pieces having different precrack lengths are prepared. Alternatively, different lengths of precracks can be introduced in one test piece depending on the size of the test piece.

After the test piece 1 into which the pre-crack 2 having the length Ci has been introduced as described above is heated to a predetermined temperature T1, it is rapidly cooled to a temperature T2 lower than the temperature T1.
As a method of performing the rapid cooling treatment, a test piece heated by a heater or the like may be dropped into a liquid such as water or solder set at T2. In addition, it is desirable to attach a conical cap 3 to the tip of the test piece 1 so that air does not get caught in the test piece at the time of dropping.

There are the following methods for obtaining (Ci, △ Tci) by the above test. One method is to simultaneously quench test specimens having pre-crack lengths C1, C2,... Cn at a certain temperature difference ΔT, and grasp the pre-crack in which the crack has propagated. In general, the smaller the length of the pre-crack is, the less the crack propagates.Therefore, among the pre-cracks in which crack growth was observed in pre-cracks of different lengths, Let the short pre-crack and the temperature difference at that time be (Ci, △ Tci). A similar test was performed by changing ΔT, and a plurality of (Ci, ΔTc
Obtain i).

Another method is to use ΔT for a given Ci.
Is gradually increased to obtain ΔTci.
First, using only one specimen into which a pre-crack of C1 was introduced, a test was performed with a small temperature difference such that the pre-crack did not propagate.
The pre-crack introduced into the test piece 1 is observed, and it is checked that no crack has propagated. Next, increase the temperature difference a little and investigate the presence of pre-crack propagation.If the pre-crack propagation is confirmed,
The temperature difference at this time is ΔT1c, and ΔT is further increased when the pre-crack has not developed. Such a test is repeated to obtain ΔT1c. Similarly, pre-crack length C
, TC3,..., TCn are also determined for the test pieces of 2, C3,.

On the other hand, the pre-crack length Ci and the temperature difference ΔTci are expressed by the following equation (1).

[0015]

(Equation 1)

The quenching thermal shock test (C)
i, △ Tci), the constants a and b in Equation 1 are determined by the least square method or the like.

Here, even when only one set of (Ci, △ Tci) is known, a = １ ／ derived from fracture mechanics
Is used, ΔTc can be estimated. However, since a physical property value generally has temperature dependency, it is difficult to accurately estimate when a = 1/2 is used.

Next, the equivalent crack length of the material to be measured is determined according to the following method. First, the effective volume VE of the thermal shock test is determined. The effective volume VE varies depending on the shape, size, thermal characteristic value and the like of the test piece, but can be obtained by numerical calculation or may be given by an approximate expression.

Next, a strength test is performed using test pieces of the same material, and a strength σ _p , a Weibull coefficient m, and an effective volume V _p for the strength test are determined.

The average strength σ _ts of the test piece under thermal shock is calculated using the following _equation (2).

[0021]

(Equation 2)

Next, the equivalent crack length Co, assuming that the natural defect of the specimen without a pre-crack is a semi-elliptical crack, is calculated based on the following equation (3).

[0023]

(Equation 3)

Here, K _IC is the fracture toughness value obtained by the CSF method or the like, Y is the shape factor of the crack, the equivalent crack length is small (100 μm or less), and the depth of the crack is 0% of the crack length. .
In the case of an eight-fold semi-elliptical shape, Y = 1.28.

Then, the critical temperature difference ΔTc of the material to be measured can be estimated by substituting the equivalent crack length Co obtained as described above into Ci in the above equation (1).

[0026]

According to the JIS standard, it has been necessary to increase the number of test pieces because the strength variation directly affects the variation in the critical temperature difference.
Since the critical temperature difference at which the pre-crack introduced into the test piece propagates is measured, the variation is small, and since multiple pre-cracks can be introduced into one test piece, the number of test pieces to be subjected to the thermal shock test Is reduced. Also, since the relationship between the crack length and the critical temperature difference can be known, the thermal shock critical temperature difference can be estimated even when the equivalent crack length changes.

[0027]

【Example】

Example 1 A test was performed using a silicon nitride sintered body as a test specimen to be measured. In the test, as shown in FIG. 1, six cylindrical test pieces having a diameter of 8 mm and a length of 70 mm were prepared. A load of 5 kgf, 20 was applied to the six test pieces by a Knoop hardness measuring indenter.
Pre-cracks were introduced at 2 kgf and 50 kgf, respectively.
Pre-cracks were introduced at eight locations with the same load in one test piece. After the introduction of the pre-crack, the residual stress was removed by heat treatment. When the average pre-crack length after introduction was measured, it was 5 kg.
The f-load pre-crack test piece was 125 μm, the 20 kgf pre-crack test piece was 284 μm, and the 50 kgf pre-crack test piece was 506 μm.

Next, a cap as shown in FIG. 1 was adhered to the above six test pieces, and three test pieces having different crack lengths were heated to 850 ° C. in the air, and then melted at 200 ° C. Dropped into solder and quenched. After the quenched test piece was allowed to cool to room temperature, the crack length after the test was measured. As a result, in the 5 kgf load pre-crack test piece, the average pre-crack length was 1
The cracks extended from 25 μm to 136 μm, and the cracks extended to the entire test piece in the 20 kgf load pre-crack test piece and the 50 kgf load pre-crack test piece, exceeding the measurement limit. From this result, it was found that the temperature difference was
It turned out that 50 degreeC becomes a critical temperature difference.

Similarly, when the remaining three test pieces heated to 600 ° C. were simultaneously quenched, no crack growth was observed in the 5 kgf load pre-crack test piece and the 20 kgf load pre-crack test piece. In the 50 kgf pre-cracked test piece, the average crack length had grown from 506 μm to 517 μm. Therefore, it was found that the critical temperature difference was 400 ° C. for the 506 μm pre-crack. From this result, (C ₁ , ΔT ₁ c
) = (125,650), (C ₂ , ΔT ₂ c) = (50
6,400). From this value, a
And b, a = 0.347 and b = 3471.7.
Becomes However, the unit of Ci was calculated as μm.

On the other hand, the JI of the silicon nitride sintered body to be measured is
When the four-point bending strength based on SR1601 was measured,
802.0 MPa, the fracture toughness (KIc) by the CSF test was 7.2 MPa · m ^1/2 , and the Weibull coefficient by the four-point bending strength test was 22.0. Co was determined.

Firstly, the effective volume V _E in the thermal shock test was determined by the following equation (4). Equation 4 is an approximate expression for a cylindrical test piece having a diameter of 8 mm.

[0032]

(Equation 4)

In the above formula 4, m is the Weibull coefficient, Vo is the actual volume of the test piece, Bi is the bio coefficient, and Bi is obtained by the following formula 5.

[0034]

(Equation 5)

In the above equation 5, L is the representative length of the test piece, λ is the thermal conductivity of the test piece, and h is the heat transfer coefficient of the cooling medium.

When molten solder is used as the cooling medium, the heat transfer coefficient h is 128000 W / km ² . The representative length L is the radius of the cylinder L = 4 × 10 ⁻³ m, and the thermal conductivity of the test piece λ = 1
From the value of 8.2 W / m · k, Bi = 28.1 from the expression 5 and the actual volume Vo = π4 ² × 50 mm ³ , the effective volume VE was calculated based on the expression 1, and as a result, VE = 54.05 mm ^3. Was. However, the actual volume was the volume of a portion to which thermal stress was applied.

On the other hand, the effective volume Vp of the strength test can be obtained from the following equation 6 in the case of the four-point bending test.

[0038]

(Equation 6)

In the above equation 6, L is the distance between the lower fulcrums in the four-point bending test, w is the width of the test piece, t is the thickness, and l is the distance between the stress applying points when applying stress from the upper part. .

In this embodiment, the shape of the test piece is l = 10 m
Since m, L = 30 mm, w = 4 mm, and t = 3 mm, Vp = 2.84 mm ³ . The average strength of the thermal shock test piece is σ _ts = 701 MPa from the above equation (2). From the above results, the equivalent crack length Co was calculated based on the above Equation 3, and as a result, was 64 μm.

The equivalent crack half-length Co obtained by the above calculation is Co =
As a result of estimating ΔTc from 64 μm using Equation 1, a critical temperature difference ΔTc = 820 ° C. was obtained. Regarding the accuracy of the above estimation method, using the same silicon nitride sintered body, JIS
When the critical temperature difference was determined based on R1615, it was 800
° C, and the estimation according to the present invention was found to have high accuracy.

Example 2 Five test pieces to be measured having the same shape and material as in Example 1 were prepared. A load of 5 kgf, 10 kgf, 20 kgf, 30 kgf, and 5 kgf was applied to the five test pieces by a Knoop hardness measuring indenter.
A pre-crack was introduced at 50 kgf. The number and position of the pre-cracks are the same as in the first embodiment. In the test, first, a 5 kgf load test piece was subjected to ΔT = 580 ° C., 600 ° C., 620 ° C., 640 ° C.
The quenching treatment was repeatedly performed in the order of, and as a result of observing the length of the pre-crack each time, the average temperature difference at which the crack began to propagate was 65
It was found to be 5 ° C. Similarly, 10kgf, 20
kgf, 30 kgf, and 50 kgf load test pieces were respectively measured at 430 ° C, 470 ° C, 560 ° C, and 620 ° C.
Table 1 shows the average of the lowest temperature differences at which the cracks propagated.

[0043]

[Table 1]

When the relationship between Ci and ΔTci is plotted on a log-log graph, it can be seen that there is a linear relationship as shown in FIG. As a result of obtaining a and b in Equation 1 by the least square method, a = 0.351 and b = 3555.3 were obtained.
However, the unit of the pre-crack length was μm. Estimating ΔTc of the test material using Co obtained in Example 1, ΔTc = 8
26 ° C. It can be seen that this is in good agreement with the measured value (ΔTc = 800 ° C.).

Further, by obtaining the above relational expression, even when the equivalent crack length Co changes, ΔTc can be easily estimated only from the result of a simple strength test.

[0046]

As described in detail above, according to the estimation method of the present invention, the behavior of small pre-cracks formed on the test specimen is observed, so that the number of test specimens subjected to the thermal shock test is small. Since it breaks from a pre-crack, the dispersion of measured values is small and the critical temperature difference can be predicted with high accuracy. further,
Since you can know the relationship between the crack length and the critical temperature difference,
Even when the equivalent crack length changes, the critical temperature difference in the quenching thermal shock can be estimated.

[Brief description of the drawings]

FIG. 1 is a schematic view of a test piece used in the present invention.

FIG. 2 is a diagram showing a relationship between a pre-crack length Ci and a critical temperature difference (ΔTci).

[Explanation of symbols]

 1 Test piece 2 Pre-crack

────────────────────────────────────────────────── ─── Continued on the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) G01N 3/00-3/62 JICST file (JOIS)

Claims (1)

(57) [Claims] When a plurality of test specimens having pre-cracks having different lengths are quenched at a plurality of temperature differences ΔT, the length of the pre-cracks Ci and the lowest temperature at which the pre-cracks propagate. Difference △ Tc
The relational expression between Ci and ΔTci is obtained by measuring two or more combinations (Ci, △ Tci) of i, and the strength, fracture toughness, Weibull coefficient, effective volume or effective surface area in the strength test, and the test during the quenching treatment Estimate the equivalent crack length from the effective volume or effective surface area of the piece, the critical temperature difference of quenching thermal shock characterized by estimating the critical temperature difference that causes the material to break by thermal shock from the equivalent crack length and the relational expression Estimation method.