JP4706254B2 - Steel life prediction method - Google Patents

Description

The present invention relates to a life prediction how excellent steel in structural steel especially weather resistant.

Conventionally, steel materials used in the construction and civil engineering fields include ordinary steel called SS and SM, and weather resistant steel called SMA used without painting, and recently, Ni-based high weather resistant steel and the like. The application standards for structures using these materials have so far been
(1) Regional classification by flying sea salt Weather-resistant steel
(2) Durability improvement technology for steel structure buildings Normal steel
(3) Exposure test of the steel grade Any steel grade
(4) Corrosion amount prediction formula There are weather-resistant steel, etc., and these are properly used as needed.

The above (1) and (2) summarize the conventional exposure results, and applicability determination can be made instantaneously. The amount of corrosion of steel by the exposure test (3) above is
It is known that Y = AX ^B , Y: corrosion amount, and X: years. By setting the threshold Y _1im for the amount of corrosion, this is a method in which X _1im year at that time is _used as the lifetime. The standard for determining the corrosion life is, for example, an estimated one-side thickness reduction of 100 years is 0.5 mm or less. Since A and B in this equation vary depending on the environment and the steel type, in order to determine A and B, the test piece is exposed to the actual environment or an environment close to the actual environment, and the corrosion amount of the test piece changes with time (X, A method of power approximation from Y) is used.

  By the way, many literatures show that the corrosion rate is influenced by the amount of sea salt and the amount of sulfurous acid gas. In addition, since the corrosion phenomenon is basically a chemical reaction in an aqueous solution, it also depends on the temperature, humidity, and wetting time. Therefore, it can be described as a function using these environmental factors as parameters.

Regarding the prediction method of (4) above, the Ministry of Construction (at that time) Civil Engineering Research Institute determined the above A and B from the results of the exposure test for SMA (weather-resistant steel of JIS standard) as follows. (For example, Non-Patent Document 1).
A = CS _a ^γ S _a : flying salt content, C, γ: regression coefficient B = 0.73
That is, A is correlated with the incoming salt content, and B is 0.73 regardless of the steel type, exposure direction, installation location, and the like.

In addition, with respect to the prediction method (4), for example, a method for obtaining as follows is proposed (for example, Patent Document 1). In this method, the amount of corrosion in the first year is estimated from the quadratic regression equation of the corrosivity index Z.
Z = α ・ TOW ・ exp (-κ ・ W) ・ (C + δ ・ S) / (1 + ε ・ C ・ S) ・ exp (-Ea / RT)
α = 10 ⁶ , κ = -0.1, δ = 0.05, ε = 10.0, Ea = 50 kJ / mol · K
Z: Corrosion index, R: Gas constant, C: Flying salinity, S: Sulfur oxide content,
TOW: Annual wetting time (h), W: Annual average wind speed (m / sec), T: Annual average temperature (K)
"Joint Research Report on the Application of Weathering Steel to Bridges (XVIII)" (Public Works Research Institute, Ministry of Construction, Steel Club, Japan Bridge Construction Association, published in March 1993) WO03 / 006957 pamphlet

  Among the conventional life prediction methods described above, the methods (1) and (2) described above are limited in steel grade and only determine whether the material is applicable at a certain point in time. I couldn't. On the other hand, in the above method (3), the steel type can be selected by examining each steel type. However, when a new steel having high weather resistance is developed, an exposure test must be performed at various locations for a long period of time immediately after development to create a calibration curve or formula. In the method (3), at least two points of corrosion amount data are required to determine the two unknowns A and B. In Japan, there are four seasons, and since corrosion progresses differently depending on the season, it requires a minimum of two years of measurement. Furthermore, in recent years, it has been sought to extend the life of steel for civil engineering and construction. For example, in order to accurately predict a long life such as 100 years, a long-term test such as 5 to 20 years has been conducted until it is actually used. There was a problem that it took time. Further, in the life prediction method of (4) above, although the influence of the environment of the construction site is incorporated to form the formula, a long-term exposure test is necessary. Therefore, there is a problem that the amount of incoming salt necessary for the protection of rust and the effect of steel type are not reflected.

  In the method proposed in the above-mentioned Patent Document 1, S (sulfur oxide amount) is used as a parameter of the prediction formula, but there is a significant correlation between S (sulfur oxide amount) and the thickness reduction amount. It is known not to be seen.

The present invention has been made in order to solve such problems, and provides a method for predicting the life of a steel material that makes it possible to accurately predict the long-term corrosion amount of various steel materials from short-term corrosion data. The purpose is to do.

The steel material life prediction method according to the present invention is based on the following formula: Y = AX ^B (Y: corrosion amount, X: years, A, B: coefficient depending on material and environment, power) A method of predicting the life of a steel material using the A value based on an exposure test at a construction site or a construction site of a structure, obtaining the B value as a function of the A value, The corrosion amount Y of the steel material is obtained based on the B value.

The steel material life prediction method according to the present invention is based on the following formula: Y = AX ^B (Y: corrosion amount, X: years, A, B: coefficient depending on material and environment, power) A method for predicting the life of a steel material using the A value obtained based on a test in a laboratory simulating the environment of a construction site or a construction site, and the B value as a function of the A value. The corrosion amount Y of the steel material is obtained based on these A value and B value.

In the method for predicting the life of a steel material according to the present invention, the A value is expressed by the following equation, and α, β, and γ in the following equation are obtained based on tests in the laboratory.
A = CS _a ^γ = (α · T + β) · Pw (T, H) · (S _a ^γ )
T: temperature (° C.), H: relative humidity (%), Sa: incoming salt content (mdd),
Pw (T, H): Wetting probability, α, β, γ: Coefficients set according to steel type

The method for predicting the life of a steel material according to the present invention is based on the above α, β, and γ, the amount of flying salt S _{a at} the construction site or construction site of the structure, the average temperature T, the relative humidity H, and the wetting probability Pw. Find the A value.

In the method of predicting the life of a steel material according to the present invention, the A value is specified by the following equation, and α, β, and γ in the following equation are obtained based on the test in the laboratory.
A = k (α · T + β) · TOW · (S _a ^γ )
T: temperature (° C.), S _a : incoming salt content (mdd),
TOW: Annual wetting time (h), k: coefficient,
α, β, γ: Coefficients set according to steel type

Life prediction method for a steel material according to the present invention, the alpha, and β and gamma, airborne salt amount S _a in construction areas or construction site of the _structure, the temperature T of the annual average, and annual wetting time TOW, the A Find the value.

  In the method for predicting the life of a steel material according to the present invention, the A value obtained based on the laboratory test is corrected to obtain the A value of the corrosion amount prediction formula.

  According to the present invention, the data (A value and B value) relating to the material characteristics and environmental characteristics of the corrosion amount prediction formula can be calculated by a short-term exposure test or laboratory experiment. The amount of corrosion can be predicted. Furthermore, since the amount of corrosion can be predicted in a short period of time, the reliability of selecting various steel materials can be improved, the optimum design of the structure can be achieved, and maintenance costs can be minimized.

Embodiment 1. FIG.
As Embodiment 1 of the present invention, the A value and B value of the corrosion amount prediction formula Y = AX ^B (Y: corrosion amount, X: number of years, A, B: coefficient depending on the material and the environment, power number) as the first embodiment of the present invention. A prediction method for predicting the life of a steel material obtained by a short-term exposure test will be described. Prior to that, the measurement principle of the life prediction method according to the first embodiment will be described first.
FIG. 1 is a characteristic diagram showing the relationship between the period X and the corrosion amount Y of the steel material, and the standard for determining the corrosion life is, for example, an estimated one-side plate thickness reduction amount of 100 years is 0.5 mm or less. The relationship between the period X and the corrosion amount Y of the steel material is expressed by the following equation as described above. In addition, A value has shown the corrosion resistance of steel materials itself in the environment, and B value represents the protection property of rust. If the corrosion resistance of the steel material is high, the initial inclination of the characteristic in FIG. 1 is small, and if the rust protection is high, the value of the corrosion amount Y after a long period of time is small.
Y = AX ^B , Y: Corrosion amount, X: Years (1)

  FIG. 2 is a characteristic diagram of A value obtained from test data of SMA actual exposure. In the above equation (1), when X = 1, Y = A, so the A value corresponds to the reduction in thickness (corrosion amount) for one year. A value (1 year) has shown the A value calculated | required by the thickness reduction amount for 1 year. In addition, the A value (up to 9 years) on the horizontal axis indicates the A value obtained based on the amount of decrease in sheet thickness in each elapsed year when the exposure test was conducted for 9 years. As is apparent from the characteristics of FIG. 2, the A value (A value (˜9 years)) obtained from the test data with an exposure period of 9 years and the A value obtained from the test data with an exposure period of 1 year ( A value (1 year)) is correlated, and it can be seen that the A value can be calculated by a short-term exposure test. Therefore, in Embodiment 1, the A value is calculated by a short-term exposure test.

  FIG. 3 is a characteristic diagram of B value obtained from test data of actual SMA exposure. B value obtained from test data with an exposure period of 9 years and B value obtained from test data with an exposure period of 1 year. The correlation is shown. As apparent from the characteristics of FIG. 3, the B value obtained from the test data with an exposure period of 9 years and the B value obtained from the test data with an exposure period of 1 year are not correlated, and the B value is short-term. It can be seen that it cannot be calculated by an exposure test during this period.

FIG. 4 is a characteristic diagram showing the correlation between the A value and the B value obtained from the SMA actual exposure test data (9 years). From this characteristic diagram, the B value can be regressed with the A value, and is expressed by the following cubic expression that takes the minimum value. In addition, even if A is small, rust grows and B value decreases as it becomes long-term.
When A ≦ 0.03,
B = 0.5 to 0.7, preferably 0.6 (theoretically, the parabolic law is considered to be 0.5th power, but the rust actually formed is not completely dense, so 0.5 to 0.7 fluctuate.)
When 0.03 <A <0.083,
B = −4611.3A ³ + 769.19A ² −32.421A + 1.0109
When 0.083 ≦ A,
B = 0.9 to 1.1, preferably 1 (2)
It was found that the S-shaped relationship is established.
However, for simplicity, even when A ≦ 0.03, B = −4611.3A ³ + 769.19A ² −32.421A + 1.0109
It is good.

  The above A value is a coefficient indicating the corrosion resistance of the steel material itself in that environment, but has a property that depends on the corrosive environment, and the influence of the corrosive environment is also included in the A value. In the range where the A value is small, stable rust is formed, so the B value is a constant value. However, if the corrosive environment exceeds a certain amount (for example, about A = 0.03), the rust tends to peel off and is stable. Therefore, the B value rises and B = 1, and the corrosion curve remains a straight line. However, in a range where the A value is small, a steel material with a short exposure time has a small amount of rust generation, so the protective effect is small and the B value may be high. Plotting the actual exposure test data (2 years) of steel grade 1 (1.5Ni-0.3Mo steel) and steel grade 2 (2.5Ni-very low C steel) on the SMA data, It turns out that it is in agreement with the relational expression. Therefore, it can be seen from this relational expression that the B value can be obtained from the A value regardless of the steel type.

Therefore, in the first embodiment, when the corrosion life is predicted in a certain steel type in a certain environment, the life is predicted by the following processes (a) to (d).
(A) For example, a one-year exposure test is performed at the construction site or construction site of the structure.
(B) A value is obtained from the result (corrosion amount) of the exposure test.
(C) The B value is obtained by applying the A value to the relational expression (formula (2)).
(D) The above A value and B value are applied to the corrosion amount prediction formula Y = AX ^B of the steel material to determine the corrosion amount after X years.
By treating in this way, the corrosion life can be predicted by an exposure test of about one year.

FIG. 5 shows the values of A and B for steel type 1 (1.5Ni-0.3Mo steel) and steel type 2 (2.5Ni-very low C steel) based on the results of exposure tests conducted in various environments with different salinity. It is the characteristic figure which showed the result of having calculated | required the value and applying to the corrosion amount prediction formula of steel materials, and predicting the sheet thickness reduction amount after 100 years in the environment of the amount of incoming salt. From the figure, the amount of incoming salt, that is, the salt resistance limit, at which the 100-year estimated thickness reduction amount is 0.5 mm, is obtained as follows. If this salt resistance limit is exceeded, the corrosion amount of both steel materials increases rapidly.
Steel grade 1 = 0.4 mdd (: mg / dm ² / day), Steel grade 2 = 0.6 mdd

Thus, it becomes possible to estimate the corrosive environment of the application limit of the steel type. Y = AX ^B If A = is obtained by substituting X = 100 years and Y = 0.5 mm, A = 0.03 is obtained. That is, it can also be seen that the thickness reduction amount of 0.03 mm per year in that environment is a guideline for determining applicability.

  As described above, in the first embodiment, for example, an A value is obtained by a one-year exposure test, and further, the A value is applied to the above relational expression (formula (2)) to obtain a B value. Since these A and B values can be applied to the above-mentioned steel corrosion amount prediction formula (Equation (1)), the corrosion amount Y after X years can be predicted. It is possible to predict the amount of corrosion with high accuracy.

Embodiment 2. FIG.
In the first embodiment, the example in which the A value is obtained based on the data of the actual exposure test has been described. However, in the second embodiment, the A value is mainly obtained by a test in a laboratory. Therefore, in the second embodiment, the above formulas (1) and (2) are used as they are, but the A value is expressed as the following formula (3) as a function of various environmental factors. That is, the A value is expressed by coefficients α, β, γ specific to the steel type and environmental data (Sa, T, Pw (T, H)), and α, β, γ specific to the steel type is required. For example, the A value is automatically obtained by applying the environmental data of the construction site or construction site.
Y = A · X ^B (1)
B = f (A):
When A ≦ 0.03, B = 0.6 (set within the range of 0.5 to 0.7)
When 0.03 <A <0.083
B = −4611.3A ³ + 769.19A ² −32.421A
+1.0109
When 0.083 ≦ A, B = 1 (set within a range of 0.9 to 1.1) (2)
However, even when A ≦ 0.03, to make it more convenient
B = −4611.3A ³ + 769.19A ² −32.421A
It is good.
A = CS _a ^γ = (α · T + β) · Pw (T, H) · (S _a ^γ ) (3)
X: Exposure time (y), Y: Thickness reduction (mm)
T: temperature (° C.), H: relative humidity (%), Sa: incoming salt content (mdd),
Pw (T, H): Wetting probability, α, β, γ: Coefficients set according to steel type

  In the above equation (3), the term relating to the temperature T is a linear approximation because the actual temperature range is relatively narrow. In terms of humidity H, we considered that the amount of corrosion is proportional to the wetting time, and introduced the wetting probability function proposed by Kucera et al. For ISO (Kucera, Tidblad, Mikhailov: ISO / TC156 / WG4-N314, Annex A (1999 )). The annual wetting time is 8766h × (wetting probability). The wetting probability Pw (T, H) is expressed by the wetting probability Pw (T, H) = N (T; 0; 9.96) * β (H / 100; 4.67; 1.78) according to the kucera equation. The

The wetting probability Pw (T, H) and the annual wetting time TOW are
TOW = 8766 × Pw (T, H)
Therefore, the above equation (3) is expressed as a function of the annual wetting time TOW as follows.
A = k (α · T + β) · TOW · (S _a ^γ ) (3a)
TOW: Annual wetting time (h), k: Factor

The above temperature T, relative humidity H and airborne salt amount S _a is, for example, the value of the environment in which the bridge is installed. The flying salinity Sa is _{a value} that includes the influence of the wind in the environment, the direction (to which direction), the height of the bridge girder, etc., and is measured in advance. For example, by determining the points of Japan (latitude and longitude or address) can be determined temperature T corresponding to the position information, the relative humidity H, the airborne salinity S _a. For example, the point information and the JMA data normal values of temperature T and relative humidity H are already in a database, and necessary information can be obtained by using these databases. In addition, for the amount of salinity S _a , a relational expression that aggregates measurement data at about 300 points (the relationship between the amount of sea salt in each region and the separation distance. Specifically, S _a = S _a0 · L ^-0.6 , L: rip distance) may, by using this relationship, it is possible to determine the airborne salt amount S _a in the relevant point. In addition, when these data cannot be obtained from a database, you may obtain | require by measurement in an applicable point. In this way, the temperature T at the point where applicable, it is possible to determine the relative humidity H and airborne salt amount S _a, the coefficient alpha, beta, if it is possible to determine by laboratory testing the gamma, A value The B value is a function of the A value, and Y (the thickness reduction amount) can be obtained by Y = AX ^B.

  Next, how to obtain γ will be described. The value of γ differs between A ≦ 0.03 and A> 0.03. When A ≦ 0.03, for example, for SMA, the results of the exposure test shown in FIG. 6 (“Joint Research Report on Application of Weatherproof Steel to Bridges (XVIII)” (Ministry of Construction, Civil Engineering Research Institute, Steel Club) According to the Japan Bridge Construction Association (published in March 1993), γ = 0.487 was obtained, and the inventors independently conducted exposure tests on Ni-based weathering steel. As a result, the characteristics shown in Fig. 7 are obtained, and γ = 0.487 is obtained, which indicates that the two data are consistent and have high reliability. Then, γ = 0.49 was set to a constant value regardless of the steel type.

Moreover, said γ when A> 0.03 was determined by a wet and dry repeated test in a constant temperature and humidity chamber (AGX-325 manufactured by ADVANTEC). The cycle is 24 h, and the wet and dry cycle is the following six conditions. The wet and dry transition time is 1 h, which is included in 12 h. For the adhesion of salt, an aqueous NaCl solution was added dropwise using a micropipette. The dripping amount was 40 μL / cm ^2, and the amount of adhering salt was controlled by the concentration of the aqueous solution. The test was conducted for a maximum of 52 weeks. Here, for example, three types of salt content are 0.1 mdd, 0.2 mdd, and 0.4 mdd. (This test is performed under the same conditions when α and β are obtained. However, the conditions are (1) to (6).)
(1) 13 ° C / 95% x 12h-20 ° C / 65% x 12h
(2) 20 ° C./95%×12 h-27 ° C./65%×12 h
(3) 25 ° C./95%×12 h-32 ° C./65%×12 h
(4) 20 ° C / 95% x 12h-35 ° C / 40% x 12h
(5) 25 ° C / 95% x 12h-40 ° C / 40% x 12h
(6) 13 ° C / 95% x 12h-28 ° C / 40% x 12h

  Note that the conditions of the above experiment are set because the relative humidity tends to decrease as the temperature rises in an actual environment, and the temperature range and humidity range are also within the above range. For this reason, this test is not a so-called accelerated test, and a corrosion rate in the same order as the actual environment can be obtained. It has already been confirmed that the corrosion of the girder in the bridge is well reproduced with respect to the salinity, temperature, and humidity, and hereinafter referred to as a reproducible corrosion test (or reproducibility test).

Next, a method for obtaining γ based on the data of the reproducible corrosion test will be described.
FIGS. 8A, 8B, and 8C are characteristic diagrams showing temporal changes in the corrosion amount for each amount of adhering salt of SMA, steel type 1, and steel type 2. FIG. The A value corresponding to each adhesion amount is obtained from the characteristics shown in FIG.

  9 (A), (B), and (C) are characteristic diagrams in which the horizontal axis represents the amount of the attached salt and the vertical axis represents the A value corresponding to the amount of the attached salt, and the logarithm of the logarithm is plotted. Becomes γ. A value of γ = 0.9 was obtained for any steel type.

  Next, a method for obtaining the coefficients α and β will be described. α and β are determined by applying simultaneous equations for several corrosion conditions with the same amount of deposited salt. This will be specifically described below. The amount of salt in the above formula (3) is the amount of incoming salt, but it is the amount of adhering salt in the reproducible corrosion test, and both units are mdd, but the influence is different even with the same numerical value. Therefore, α and β obtained in the reproduction test are distinguished as α ′ and β ′, but α / β is equal to α ′ / β ′.

  For the test conditions (1) to (6) with the same amount of adhering salt, the sum of the corrosion amounts of the drying step and the wet step is the corrosion amount obtained in the test. In the reproduction test, the salinity is given as the adhering salinity, but the salinity variable in the above equation (3) is the incoming salinity, and as described above, both units are mdd, but the influence is different even with the same numerical value. Since the amount corresponding to the flying salt content of the above equation (3) is not known, calculation is made here with Sa '. Since γ changes depending on whether the A value is larger or smaller than 0.03, α and β are set when the A value exceeds 0.03 or less.

For example, in the case of the test of 13 ° C./95%×12 h-20 ° C./65%×12 h with an attached salt content of 0.4 mdd, the attached salt content is 0.4 mdd and γ = 0.9 from FIG. Use this.
Substituting salinity, temperature, and humidity into A '= (α'T + β') · Pw (T, H) · (Sa ^γ )
Corrosion amount of drying step A ' _{(13 ° C / 95%, 0.4mdd, drying)} = (13α' + β ') · Pw (13, 95) (Sa' ^0.9 )
Corrosion amount of wet step A ' _{(20 ℃ / 65%, 0.4mdd, wet)} = (20α' + β ') · Pw (20, 65) (Sa' ^0.9 )
It is.
A ' _{(13 ° C / 95% × 12h-20 ° C / 65% × 12h, 0.4mdd)} = (Corrosion amount in the drying step) + (Corrosion amount in the wet step) = {(13α' + β ') · Pw ( 13, 95) + (20α '+ β') · Pw (20, 65)} (Sa ' ^0.9 )
= [{Pw (13, 95) * 13 + Pw (20, 65) * 20} α '+ {Pw (13, 95) + Pw (20, 65)} β'] (Sa ' ^0.9 )

In the reproduction test with the same salt content of 0.4 mdd, the value of Sa is unknown, but since the amount of salt content is the same, at least Sa can be considered the same. Other tests can be similarly determined as follows. From the ratio of A 'in the underlined part of the above formula
{Pw (13,95) * 13 + Pw (20,65) * 20} α '+ {Pw (13,95) + Pw (20,65)} β' = 1
{Pw (20,95) * 20 + Pw (27,65) * 27} α '+ {Pw (20,95) + Pw (27,65)} β' = 1.403
{Pw (25,95) * 25 + Pw (32,65) * 32} α '+ {Pw (25,95) + Pw (32,65)} β' = 1.661
{Pw (25,95) * 25 + Pw (40,40) * 40} α '+ {Pw (25,95) + Pw (40,40)} β' = 0.858
become that way. Here, the A value ratio (relative value to the A value of the reference condition) is taken as the corrosion amount. A similar equation is established for several test conditions with the same amount of attached salt, and a binary simultaneous linear equation is solved. Mathematically, α ′ and β ′ can be obtained with two test conditions, but α ′ and β ′ are obtained by linear regression using four test conditions in order to increase accuracy. The A value ratio is determined using the obtained α ′ and β ′. Using α 'and β' obtained for SMA, substituting the amount of salinity and yearly average temperature / humidity (closest data from the nearest Japan Meteorological Agency observation station) of the exposed areas of the 41 bridge test nationwide,
A ' _{(predicted value)} = (α'T ₁ + β') · Pw (T ₁ , H ₁ ) · (Sa ₁ ^0.9 )
FIG. 10B shows the relationship with the A value obtained from the exposure test. Both regression equations were obtained, and the ratio of the first order coefficients A '/ A = 1.895 was obtained. Pw (T ₁ , H ₁ ) ・ (Sa ₁ ^0.9 ) = K ₁
After all,
A ′ _{(predicted value) =} (α′T1 + β ′) · Pw (T ₁ , H ₁ ) · (Sa ₁ ^0.9 ) = (α′T ₁ + β ′) K ₁
A _{(actual value)} = (αT ₁ + β) · Pw (T ₁ , H ₁ ) · (Sa ₁ ^0.9 ) = (αT ₁ + β) K ₁
To hold in any T ₁ is
A '/ A = α' / α = β '/ β = 1.895 = k

  On the other hand, when A ≦ 0.03, γ = 0.487, an equation is similarly obtained from the results of the reproducible corrosion test, and α ′ and β ′ are obtained by linear regression. Furthermore, A '/ A = 4.658 was obtained from FIG. A '/ A = α' / α = β '/ β = 4.658

For steel grade 1 and steel grade 2 as well, from the results of the previous reproducible corrosion tests
When A> 0.03, γ = 0.9
When A ≦ 0.03, γ = 0.487
Then, α ′ and β ′ are obtained by linear regression, and A ′ / A = α ′ / α = β ′ / β is obtained in the same manner. .

  Table 1 summarizes the coefficients and power numbers of the corrosion amount prediction formulas of the weathering steels obtained from the above.

  If the coefficient and power shown in Table 1 are applied to the above equation (3) and the environmental data of the corresponding area are obtained, the A value is obtained, and the A value is applied to the equation (2) to obtain the B value, By applying the A value and the B value to the corrosion amount prediction formula of the steel material, the life prediction can be performed.

FIG. 11 is a characteristic diagram in which A values and predicted values thereof are plotted at 41 exposure test sites nationwide. From the figure, it can be seen that the A value in the exposure test and the predicted value A are in good agreement. Figure 12 shows exposure test results for the 9-year reduction in sheet thickness obtained by substituting the amount of airborne salinity and annual average temperature / humidity (closest data from the nearest Japan Meteorological Agency observation station) at 41 exposure test sites nationwide into the prediction formula. It is the characteristic view illustrated with the result. FIG. 13 is a characteristic diagram showing the correlation between the plate thickness reduction amount by the above prediction formula and the exposure test result.
From the characteristics of FIG. 12 and FIG. 13, the reduction in plate thickness according to the prediction formula and the exposure test result agree well, and the usefulness of the prediction method of the second embodiment can be confirmed.

FIG. 14 is obtained by regression using Y = AX ^B directly from the predicted corrosion amount after 100 years based on the above prediction formula and the values of actual exposure data for 1, 3, 5, 7, and 9 years. It is the characteristic figure which compared the value after 100 years. As shown in FIG. 14, it can be seen that the normal probability paper shows good linearity and follows a normal distribution. As a result, the standard deviation σ obtained was σ = 0.301 (≈0.3). Therefore, the accuracy of the predicted value of the present corrosion amount prediction formula for the directly regressed value can be represented by being attached to the predicted value.

  As described above, in the second embodiment, environmental conditions (temperature, humidity, amount of adhering salt) in the corresponding region are set in the laboratory, and the A value is obtained. Since they are in agreement, the A value can be obtained with high accuracy only from the laboratory test result without performing the exposure test, so the B value is also obtained by the above equation (2), and the life of the steel material is quantitatively obtained. be able to.

Embodiment 3. FIG.
In the first and second embodiments described above, an example in which an exposure test or a laboratory experiment is performed for about one year has been described in order to obtain the A value. For example, the A value can be obtained by obtaining (predicting) the corrosion amount for one year by applying an extrapolation method to a test result of about several months.

  Further, according to the above-described embodiment 1 and / or 2, for various steel materials, by obtaining the A value and B value in each region and creating a database, for example, by inputting information on the type and location of the steel material, The life of the steel material can be easily predicted.

Embodiment 4 FIG.
As described above, the life prediction can be used for designing a real structure by attaching data predicting the progress of corrosion of each part, for example, to the steel material. Note that the attachment of data includes not only paper and the like, but also the state managed as electronic data. Further, when designing a steel structure, it is possible to appropriately predict the life of the structure by appropriately selecting and designing the steel material for which such life prediction has been made.

The characteristic view which showed the relationship between the period X and the corrosion amount Y of steel materials. The characteristic figure of A value calculated | required from the test data of SMA actual exposure. The characteristic figure of B value calculated | required from the test data of SMA actual exposure. The characteristic view which showed the correlation of A value and B value which were calculated | required from the test data (9 years) of SMA actual exposure. The characteristic view which showed the result of having predicted the board thickness reduction amount 100 years after from the exposure test result implemented in the environment where various flying salinities differ. The characteristic view which showed the test data of the actual exposure test of SMA. The characteristic view which showed the test data of the test (laboratory) of Ni-type high weather resistance steel. The characteristic view which showed the time change for every adhesion salt content of SMA, steel type 1, and steel type 2. FIG. The characteristic diagram which plotted the logarithm of the A with respect to the amount of adhered salt on the horizontal axis and the A value corresponding to the amount of adhered salt on the vertical axis. The characteristic view which showed the relationship between A 'value and A value obtained from the exposure test. The characteristic figure which plotted the A value and the predicted value in 41 exposure test places nationwide. Combined with the exposure test results, the 9-year reduction in plate thickness obtained by substituting the amount of airborne salinity and annual average temperature and humidity (data from the nearest Meteorological Agency observation station) at 41 exposure test sites nationwide into the prediction formula. FIG. The characteristic view which showed the correlation with the amount of plate | board thickness reduction | decrease by said prediction formula, and an exposure test result. A corrosion amount prediction value of 100 years in accordance with the present embodiment 1, diagram the difference between the values shows the distribution when the regression of 100 years by Y = AX ^B.

Claims (7)

This is a method for predicting the life of steel using the prediction formula Y = AX ^B (Y: corrosion amount, X: years, A, B: coefficient depending on material and environment, power). And
Obtaining the A value based on an exposure test at the construction site or construction site of the structure, obtaining the B value as a function of the A value, and obtaining the corrosion amount Y of the steel material based on the A value and the B value. A method for predicting the life of a steel material. This is a method for predicting the life of steel using the prediction formula Y = AX ^B (Y: corrosion amount, X: years, A, B: coefficient depending on material and environment, power). And
The A value is obtained based on a test in a laboratory simulating the environment of a construction site or a construction site, the B value is obtained as a function of the A value, and the steel material is obtained based on the A value and the B value. A method for predicting the life of a steel material, characterized in that a corrosion amount Y of the steel is obtained. 3. The method for predicting the life of a steel material according to claim 2, wherein the A value is specified by the following equation, and α, β and γ of the following equation are obtained based on a test in the laboratory.
A = (α · T + β) · Pw (T, H) · (S _a ^γ )
T: temperature (° C.), H: relative humidity (%), S _a : incoming salt content (mdd),
Pw (T, H): Wetting probability, α, β, γ: Coefficients set according to steel type The alpha, and β and gamma, airborne salt amount S _a in construction areas or construction site of the _structure, the temperature T of the annual average, by the relative humidity H and wetting probability P _W, and obtains the value A The steel material life prediction method according to claim 3. 3. The method for predicting the life of a steel material according to claim 2, wherein the A value is specified by the following equation, and α, β and γ of the following equation are obtained based on a test in the laboratory.
A = k (α · T + β) · TOW · (S _a ^γ )
T: temperature (° C.), S _a : incoming salt content (mdd),
TOW: Annual wetting time (h), k: coefficient,
α, β, γ: Coefficients set according to steel type The alpha, and β and gamma, according to claim 5 airborne salt amount S _a in construction areas or construction site of the _structure, which by the temperature T of the annual average, and annual wetting time TOW, and obtains the value A The life prediction method of steel materials as described.   The steel material life prediction method according to any one of claims 2 to 6, wherein the A value obtained based on the test in the laboratory is corrected to obtain the A value of the corrosion amount prediction formula.

Images