JP7259815B2 - Corrosion amount prediction method and device - Google Patents

Description

The present invention relates to a corrosion amount prediction method and apparatus for predicting a future corrosion amount using actually measured corrosion amount data.

Metal materials are applied to various structures and play an important role as social capital. In particular, steel materials are used as steel structures in many applications such as construction, infrastructure, and transportation equipment. Steel structures are required to be used safely for a long period of time, but steel structures have problems such as deterioration due to corrosion. Therefore, it is important to predict the amount of corrosion of steel structures that are used for a long period of time from the viewpoint of maintenance.

For example, among the steel materials used for bridges, by adding elements such as Cu, Ni, and Cr to the steel materials, the deterioration of the load-bearing performance of the bridge does not pose an engineering problem (0.01 mm/year or less). There are weathering steels that form a "protective rust" that can inhibit corrosion up to a corrosion rate of about 10%. Since weathering steel is often used without anti-corrosion treatment such as painting, it is more important to grasp the amount of corrosion and to perform maintenance.

As a corrosion prediction technique for weathering steel, for example, Non-Patent Document 1 describes a patch-type exposure test. In the patch type exposure test, for example, a patch test piece made of weathering steel having a patch size of 50 mm×50 mm×2 mm is used. A patch test piece is installed at each part of the bridge and collected every year to measure the amount of corrosion.

By using the Wappen test described in Non-Patent Document 1 or the continuous monitoring technique of the amount of corrosion, the amount of corrosion at that time can be grasped. However, in the prediction method of weathering steel using the exposure test using the Wappen test piece described above, a long-term exposure test of 5 to 10 years is required in order to accurately predict the amount of corrosion in the future. Therefore, there is a problem that it takes a long time before the amount of corrosion becomes predictable.

The amount of corrosion in the collected patch is measured by removing rust with a pickling solution standardized by ISO8407, measuring the weight, and calculating the amount of corrosion from the difference from the initial weight. Therefore, when the exposure test time is short and the amount of corrosion is small, the change in the weight of the test piece with respect to the weight measurement resolution of the balance used is small, and the error in the amount of corrosion increases, often resulting in lack of accuracy. In addition, in order to investigate the change in the amount of corrosion over time in detail in the exposure test and to predict the amount of long-term corrosion in a short period of time, it is necessary to perform the above treatment on a large number of Wappen test pieces. For this reason, an enormous amount of work and time are required.

Therefore, there have been proposed methods for extrapolating and predicting the future corrosion amount based on the corrosion amount obtained by the patch-type exposure test (see, for example, Patent Documents 1 to 4). Patent Documents 1 to 4 describe methods for predicting the amount of corrosion of weathering steel from environmental factors such as temperature, wetting time, amount of sulfur oxides, amount of airborne salt, and material factors such as components of weathering steel. there is

Japanese Patent No. 3909057 Japanese Patent No. 4143018 Japanese Patent No. 4706279 Japanese Patent No. 5895522

"JSSC Technical Report No.73 Possibilities and New Technologies for Weathering Steel Bridges", Japan Society of Steel Construction, October 2006

The prediction methods of Patent Documents 1 to 4 are based on empirical formulas and environmental measurement results based on past exposure tests and corrosion surveys, but there is variation during measurement such as the flow of wind during measurement or the tendency for deposits to accumulate. There is a limit to the calculation of coefficients that absorb Therefore, the corrosion amount predicted by these methods has a large error compared to the corrosion amount obtained by the actual exposure test, and the long-term corrosion prediction results inevitably have a large divergence. On the other hand, in order to improve the accuracy of predicting the amount of corrosion, it is necessary to sample test data over a relatively long period of time, such as five years or more.

SUMMARY OF THE INVENTION In view of the above problems, it is an object of the present invention to provide a method and apparatus for predicting the amount of corrosion of a metal material that can accurately predict the amount of corrosion using short-term measurement results of the amount of corrosion of the metal material. It is something to do.

The present invention has the following configurations to solve these problems.
[1] A method for predicting the amount of corrosion of a metal material, which outputs a predicted value for the amount of corrosion for an exposure period longer than the measurement period using a plurality of pieces of corrosion amount data for different measurement periods,
calculating a plurality of corrosion rates during the measurement period using a plurality of the corrosion amount data;
using the plurality of calculated corrosion rates to derive a rate variation function representing the variation of the corrosion rate with respect to the measurement period;
A corrosion amount prediction method, wherein the corrosion amount in the exposure period to be predicted is output as a corrosion amount prediction value by time-integrating the derived speed variation function over the exposure period.
[2] The corrosion amount prediction value is calculated by adding the corrosion amount data in the most recent measurement period and the time integral value from the most recent measurement period to the exposure period in the speed variation function. The corrosion amount prediction method according to [1].
[3] The corrosion amount prediction method according to [1] or [2], wherein the corrosion amount data is measured using a change in electrical resistance of the metal based on the amount of reduction in the metal cross section due to corrosion.
[4] The corrosion amount prediction method according to any one of [1] to [3], characterized in that the measurement period is set according to a cycle of changes in the corrosive environment.
[5] The corrosion amount prediction method according to any one of [1] to [4], wherein the corrosion rate is calculated from two pieces of corrosion amount data whose measurement periods are separated by 30 days or more.
[6] The speed variation function is derived from a plurality of the corrosion rates calculated using a plurality of the corrosion amount data for which the measurement period is 180 days or longer [1] to [5] Corrosion amount prediction method according to any one of.
[7] A corrosion amount prediction device for a metal material that outputs a corrosion amount prediction value for an exposure period longer than the measurement period using a plurality of corrosion amount data for different measurement periods,
a corrosion rate calculation unit that calculates a plurality of corrosion rates during the measurement period using a plurality of the corrosion amount data;
a variation function derivation unit for deriving a rate variation function representing variation of the corrosion rate with respect to the measurement period using the plurality of corrosion rates calculated by the corrosion rate calculation section;
a prediction output unit that outputs the predicted corrosion amount in the exposure period as a corrosion amount prediction value by time-integrating the speed fluctuation function derived in the fluctuation function derivation unit in the exposure period;
A corrosion amount prediction device comprising:

According to the corrosion amount prediction method and apparatus of the present invention, the corrosion rate fluctuation characteristic is derived as a rate fluctuation function, and the corrosion amount prediction value for the exposure period is output using the rate fluctuation function. As a result, it is possible to minimize the influence of fluctuations due to external factors when measuring the corrosion amount data, and to accurately predict the corrosion amount based on short-term measurement results.

1 is a block diagram showing a preferred embodiment of a corrosion amount prediction device of the present invention; FIG. 4 is a graph showing an example of changes in the amount of corrosion in an exposure test of weathering steel. It is a schematic diagram which shows an example of the corrosion test apparatus for acquiring corrosion amount data. FIG. 4 is a cross-sectional view showing the AA cross section in the corrosion test apparatus of FIG. 3; 4 is a graph showing an example of a corrosion prediction function for each corrosion amount data derived in a corrosion function derivation unit; 4 is a graph showing an example of a speed variation function derived by a variation function derivation unit; 4 is a graph showing an example of changes in the amount of corrosion in an exposure test of weathering steel when the corrosive environment changes. 1 is a flow chart showing a preferred embodiment of a corrosion amount prediction method of the present invention; 2 is a graph showing an example of a velocity variation function obtained using corrosion amount data in the corrosion amount prediction apparatus of FIG. 1; FIG. 10 is a graph showing an example of adding the corrosion amount data and the time integral of the velocity variation function in FIG. 9 ; FIG.

Embodiments of the present invention will be described below. FIG. 1 is a block diagram showing a preferred embodiment of the corrosion amount prediction device of the present invention. The configuration of the corrosion amount prediction apparatus 100 as shown in FIG. 1 is constructed on a computer by executing a program stored in the computer. The corrosion amount prediction apparatus 100 in FIG. 1 predicts the corrosion amount over a long period (for example, after 50 years) based on the corrosion amount data obtained from a short-term (for example, 1 to 2 years) corrosion test by the corrosion test apparatus 1. Predict.

Corrosion behavior of corrosion-resistant steel materials represented by weathering steel in the environment where steel structures (for example, bridges using weathering steel) are constructed is that corrosion is not inhibited by the protective effect of rust, and corrosion occurs at a high corrosion rate. There are an initial period during which corrosion progresses and a period in which corrosion is inhibited by protective rust formation (hereinafter referred to as "protective rust formation period"). Weathering steel exhibits corrosion resistance through protective rust. Therefore, in order to evaluate the corrosion resistance of weathering steel, it is important to understand the corrosion rate due to the formation of protective rust rather than the initial corrosion rate.

FIG. 2 is a graph showing an example of changes in the amount of corrosion due to an exposure test of weathering steel. In FIG. 2, the horizontal axis indicates the elapsed time (years) from the start of the exposure test, and the vertical axis indicates the amount of corrosion (μm). As for time, the start of the test is 0 years (point a), and the test results are shown only up to point c, but the actual test results exist for a longer period of time.

As shown in Fig. 2, the initial period a to b in which corrosion progresses at a high corrosion rate from the start of the test (point a: time = 0 years), and the corrosion rate decreases due to the influence of the rust layer compared to the initial period a to b. There is a period b to c during which it is suppressed. This is due to the anti-corrosion function of weathering steel which is exhibited by the protection of rust formed on the surface. In other words, corrosion progresses at a predetermined corrosion rate during the initial period a to b in the process of forming stable and protective rust on the steel surface. Thereafter, the corrosion rate gradually decreases in periods bc as protective rust forms on the surface of the steel. Attenuation of this corrosion amount increase is generally represented by Fc= ^mXn . Here, Fc is the amount of corrosion (μm), X is the period (years), and m and n are coefficients.

<Corrosion test>
A corrosion test is performed on each metal material, and the time from the start of the corrosion test and the amount of corrosion during that time are measured. The corrosion test is most preferably an exposure test in an environment similar or identical to that of the metal material whose corrosion is to be predicted. This is because, when the purpose of corrosion prediction is to grasp the amount of corrosion after a long period of time in a steel structure installed in an outdoor environment, accurate prediction is possible. However, other test methods may be used as long as data on the amount of corrosion of the test material and the time from the start of the test can be obtained. If necessary, it can be selected from other known test methods (various accelerated corrosion tests, various gas corrosion tests, various corrosion resistance tests, various weather resistance tests, etc.).

On the other hand, it is most preferable to obtain the corrosion amount using an electrical resistance corrosion sensor at predetermined intervals from the start of the test. This is because the electrical resistance type corrosion sensor can detect a very small amount of corrosion and can perform corrosion sampling even if the time interval is short, so that an accurate corrosion amount can be continuously measured. However, other corrosion amount measurement methods may be used as long as the corrosion amount can be calculated accurately. Depending on the corrosion rate of the test material under the corrosion test, it can be selected from known methods for measuring the amount of corrosion (methods using various corrosion sensors such as ACn-type corrosion sensors, Wappen test, etc.).

<Corrosion monitoring technology>
FIG. 3 is a schematic diagram showing an example of a corrosion test apparatus for acquiring corrosion amount data, and FIG. 4 is a sectional view showing the AA section of the corrosion test apparatus in FIG. As shown in FIGS. 3 and 4, the corrosion test apparatus 1 uses an electrical resistance type corrosion sensor. and a reference portion 21 which is isolated from the environment. The sensor section 11 and the reference section 21 are made of the same metal material.

The sensor section 11 and the reference section 21 are arranged in parallel on one surface of a flat substrate 31 with an insulating sheet 41 interposed therebetween. Both side surfaces of the sensor section 11 and the reference section 21 are covered with an insulating resin 51 , and the upper surface of the reference section 21 is covered with an insulating cover 61 . When the corrosion test apparatus 1 is in a particularly corrosive environment, corrosion progresses in the thickness direction (arrow Z direction) of the sensor section 11 whose upper surface is exposed.

When the steel structure is installed in an environment where the temperature changes greatly due to direct sunlight or the like, a large temperature difference is generated between the steel structure and the sensor section 11, and the wet and dry environments of the steel structure and the sensor section 11 are different. There is a risk that it will not be possible to correctly evaluate the In that case, the substrate 31 is not provided, and the insulating sheet 41 is made of, for example, a thin film of insulation and high thermal conductivity such as polyimide. It is preferable to take measures to

Furthermore, the corrosion test apparatus 1 has a current source 71 and voltage measurement units 81 and 91 . The current source 71 is connected to the sensor section 11 and the reference section 21 and applies a constant current to the sensor section 11 and the reference section 21 . The voltage measuring section 81 is connected to both ends of the sensor section 11 , and the voltage measuring section 91 is connected to both ends of the reference section 21 . In such a corrosion test apparatus 1, a constant current is supplied from the current source 71 to the sensor section 11 and the reference section 21, and the voltage is measured by the voltage measurement section 81 and the voltage measurement section 91. FIG. Thereby, the electric resistance value of each of the sensor section 11 and the reference section 21 can be obtained.

In the corrosion test apparatus 1, the electrical resistance values of the sensor section 11 and the reference section 21 are measured at arbitrary fixed intervals, and the corrosion amount data CD (corrosion depth) of the sensor section 11 is obtained based on the measured electrical resistance values. is calculated. A conversion formula for converting the resistance values and the like of the sensor section 11 and the reference section 21 to the corrosion amount data CD is expressed by the following formula (1).

CD=t _init {(R _{ref_init} /R _{sens_init} )−(R _ref /R _sens )} (1)
CD: Corrosion amount (corrosion depth) [μm]
t _init : Initial thickness of sensor part 11 [μm]
R _{ref_init} : Electric resistance value [Ω] of the initial state (X=0) of the reference part 21
R _{sens_init} : Electric resistance value [Ω] in the initial state (X=0) of the sensor unit 11
R _ref : electrical resistance value [Ω] of the reference unit 21 during the measurement period X
R _sens : Electric resistance value [Ω] of the sensor unit 11 during the measurement period X

<Corrosion amount prediction device 100>
The corrosion amount prediction apparatus 100 of FIG. 1 uses a plurality of corrosion amount data CD measured by the corrosion test apparatus 1 to output the corrosion amount during an exposure period longer than the actual measurement period as a corrosion amount prediction value CPP. It has a data acquisition unit 101, a corrosion rate calculation unit 102, a variation function derivation unit 103, a prediction output unit 104, and the like.

The data acquisition unit 101 acquires a plurality of corrosion amount data CD1 to CD3 when measuring the corrosion amount of the metal material in different measurement periods X1 to X3. For ease of explanation, corrosion amount data CD1 and CD2 in three measurement periods X1, X2, and X3 of measurement periods X1 = 0.5 years, X2 = 1.0 years, and X3 = 1.5 years are shown below. , CD3 are acquired by the data acquisition unit 101. FIG. Further, the data acquisition unit 101 may store the corrosion amount data CD1 to CD3 in the database DB each time it acquires the data.

Furthermore, FIG. 1 illustrates a case where the data acquisition unit 101 acquires the corrosion amount data directly from the corrosion test apparatus 1, but the method for acquiring the corrosion amount data CD is not limited to this. For example, the data acquisition unit 101 may acquire corrosion amount data via a network, or may read corrosion amount data CD from a storage medium such as a semiconductor memory. Alternatively, the corrosion amount data CD already stored in the database DB may be obtained.

Corrosion rate calculator 102 calculates a plurality of corrosion rates V1, V2, and V3 during measurement periods X1, X2, and X3 using a plurality of pieces of corrosion amount data CD1 to CD3. Corrosion rate calculator 102 stores in advance the following equation (1) for determining the corrosion rate.
Corrosion rate (μm/year) = Corrosion amount (μm) / Measurement period (year) (1)

Specifically, the corrosion rate calculation unit 102 calculates the corrosion rate V1=CD1/X1 from the start of measurement to the measurement period X1, and the corrosion rate V2 during the measurement period X1 to X2=(CD2-CD1)/(X2- X1) and the corrosion rate V3=(CD3-CD2)/(X3-X2) during the measurement period X2 to X3 are calculated. Strictly speaking, the corrosion rate is not constant and constantly fluctuates, so the corrosion rate calculated by Equation (1) means the average value during the measurement period (corrosion period).

FIG. 5 is a graph showing an example of the corrosion rate for each corrosion amount data calculated by the corrosion rate calculator. Note that FIG. 5 exemplifies a case where measurements are made every 0.5 years, for example, X1 = 0.5 years, X2 = 1.0 years, and X3 = 1.5 years. is not limited to this, and may be any interval. Moreover, although the case where three corrosion rates are calculated is illustrated, since the corrosion rate cannot be calculated when the corrosion amount is zero, it is preferable to calculate at least two or more.

The variation function derivation unit 103 in FIG. 1 derives a velocity variation function Vp representing variation in the corrosion rate Vp during the predicted exposure period PT, using the plurality of calculated corrosion rates V1 to V3. FIG. 6 is a graph showing an example of the relationship between the measurement period and corrosion rate. In FIG. 6, since the corrosion rate means an average value as described above, it is plotted in the middle of each period for convenience, but it may be plotted in any of the beginning, middle, and end of the measurement period. As shown in FIG. 6, the corrosion rate V1 is higher than the corrosion rate V2, and the corrosion rate V2 is higher than the corrosion rate V3. This is because the corrosion process has an initial period a to b and a period b to c in which the corrosion rate is suppressed by the influence of the rust layer compared to the initial period a to b (see FIG. 2). .

Therefore, the variation function derivation unit 103 obtains an approximate relationship of time change of the corrosion rate based on the plurality of corrosion rates V1 to V3. The variation function derivation unit 103 stores an asymptotic curve represented by the velocity variation function Vp=aX ^b +c, where X is the exposure period and a, b, and c are the coefficients. Then, the variation function derivation unit 103 uses the measurement periods X1 to X3 and the plurality of corrosion rates V1 to V3 to calculate the coefficients a, b, and c by a known curve fitting technique such as the least squares method.

Although a case where the speed variation function Vp is the asymptotic curve aX ^b +c is exemplified, it is not limited to this, and various functions can be used as long as they express variation characteristics. For example, known asymptotic curves such as Vp=ae ^-bX +c, Vp=a/(X-b)+c, Vp= ^aX +c may be used as the speed variation function Vp. At this time, the speed variation function Vp to be used may be selectively used according to the type of metal material or the test environment when the test patch is performed. The method of obtaining the approximated curve of the corrosion rate change over time is not particularly limited, but it is desirable to use the one with the smallest least square error.

Then, the prediction output unit 104 time-integrates the derived speed variation function Vp over the exposure period PD, thereby outputting the predicted corrosion amount in the exposure period PD as the corrosion amount prediction value CPP. In this way, by extrapolating the approximation curve to the long-term side and obtaining the time integral of this approximation curve up to a predetermined period (equivalent to the area surrounded by the approximation curve and the test period), the metal after the lapse of the predetermined period The amount of corrosion can be predicted. Since the corrosion amount data CD3 in the measurement period X3 is known, the prediction output unit 104 outputs the sum of the corrosion amount data CD3 and the time integral of the velocity variation function Vp from the measurement period X3 to the exposure period PT. You may make it output as corrosion amount prediction value CPP.

Here, FIG. 7 is a graph showing another example of changes in the amount of corrosion due to the exposure test of weathering steel. FIG. 7 shows corrosion behavior in areas where the corrosion environment changes greatly due to seasonal variations, and there are many such areas. Unlike FIG. 2, FIG. 7 shows that the corrosion amount does not increase according to the corrosion prediction function Fc=mX ⁿ . Even in such a case, if the convergence value is determined by extrapolation of the results up to, for example, 100 days, only the data of the period when the corrosion rate is high is referred to, and the data of the period when the corrosion rate is low is reflected. Therefore, the corrosion amount prediction value CPP in the final exposure period PT may deviate from the actual measurement value.

Therefore, the measurement period X may be set according to the cycle of changes in the corrosive environment. For example, when there are four periods P1 to P4 of changes in the corrosive environment in one year, such as spring, summer, autumn, and winter, and such changes in the corrosive environment are repeated every year, one year (spring, summer, autumn, winter), which is a cycle of fluctuation, is It is set as the measurement period X. Then, the corrosion prediction function Fc is calculated using the measurement data of the measurement period X of one year. Therefore, as described above, the period for calculating the corrosion rate for deriving the velocity variation function Vp is divided by one-year variation cycles. As a result, the velocity variation function Vp can be obtained with high accuracy even in an area where the corrosive environment greatly changes due to the seasonal variations described above.

Although FIG. 7 illustrates a case in which four corrosive environment changes are repeated in a cycle of one year, such as spring, summer, autumn and winter, the present invention is not limited to this. For example, in an area where the rainy season and the dry season are alternately repeated, the measurement period X may be set with the rainy season and the dry season as one fluctuation cycle, and the measurement data for this measurement period X may be acquired.

FIG. 8 is a flowchart showing a preferred embodiment of the corrosion amount prediction method of the present invention, and the corrosion amount prediction method will be described with reference to FIGS. 1 to 8. FIG. First, corrosion amount data CD1 to CD3 for each measurement period X1 to X3 are obtained from the corrosion test apparatus 1 (see FIGS. 3 and 4) (step ST1). After that, the corrosion rates V1 to V3 are derived using the plurality of corrosion amount data CD1 to CD3 and the measurement periods X1 to X3 (step ST2), and the rate variation function Vp is derived using the plurality of corrosion rates V1 to V3. (step ST3, see FIG. 6). Then, the derived velocity variation function Vp is time-integrated during the exposure period PT, and the corrosion amount prediction value CPP is output (step ST4).

At Bridge A in a rural area located 12 km from the coast, an exposure test using a patch test and monitoring of the amount of corrosion using an electrical resistance type corrosion sensor were performed. Since Bridge A has patch test results for a maximum of 17 years, the accuracy of corrosion prediction was verified using these data and the amount of corrosion after the test period of 15 years. The amount of corrosion after 15 years already obtained in the patch test of SM490AW was 40.3 μm.

A patch test piece was obtained by processing a commercially available rolled steel SM490AW for welded structures standardized by JIS G 3114 into a size of 50×50×2 mm, washing it with ethanol, and attaching it to a bridge using a double-sided tape. Each test piece was collected for a predetermined period of time, and after removing rust with a solution standardized by ISO 8407, the weight of the test piece was measured, and the amount of corrosion was calculated from the difference from the initial weight.

As for the electrical resistance type corrosion sensor, SM490AW was processed into a sensor and used in the same manner as in the Wappen test. The measurement using the corrosion sensor was performed with an interval of 1 hour and a total measurement period of 4 years. The corrosion rate was calculated every 6 months, and the corrosion rate during each measurement period was obtained by calculation.

FIG. 9 is a graph showing the relationship between corrosion rate and test period. A velocity variation function Vp=6.9048×X ^−0.697 was derived from multiple corrosion rates obtained by the corrosion sensor. Then, the velocity variation function Vp was time-integrated from 0.5 years to 15 years, and the corrosion amount for the first 0.5 years (corrosion amount data CD1) was added. The reason for the integration from 0.5 to 15 years is that when regression is performed including the amount of corrosion from the start of the test to 0.5 years, the corrosion rate becomes infinite as it approaches 0 years, as shown in FIG. , the corrosion rate and its time integral value may not be determined.

As a result, the predicted corrosion amount CPP after 15 years was 41.0 μm. As described above, the actual corrosion amount after 15 years was 40.3 μm, which is very close to the corrosion amount obtained in the Wappen test. In this way, the long-term corrosion amount could be accurately predicted by obtaining the temporal change of the corrosion rate and obtaining the time integral of the predicted period for the approximate curve of the temporal change.

Using the results of monitoring the amount of corrosion by the electrical resistance corrosion sensor in Bridge A shown in Example 1, the amount of corrosion after 15 years was predicted under various conditions. The electrical resistance corrosion sensor converts the change in resistance value into the amount of corrosion, so it has extremely high sensitivity and can measure changes in the amount of corrosion even in a short period of time. In Example 2, the measurement period for calculating the corrosion rate was changed from 5 days to 1 year. Furthermore, the period of corrosion measurement by the sensor was changed from 60 days to 1440 days (about 4 years) to verify the accuracy of corrosion prediction. By the same method as in Example 1, the amount of corrosion of SM490AW after 15 years was predicted by obtaining the time integral of the change in corrosion rate over time.

As a result, when the measurement period for calculating the corrosion rate by the corrosion sensor is 30 days or more, the influence of changes in the corrosion rate due to weather changes can be suppressed, and the long-term corrosion amount can be predicted with higher accuracy. , was judged to be good.

In addition, when the period from the start of measurement by the sensor is 180 days or more, the effect of the season is small and the corrosion rate is sufficiently attenuated by corrosion products, so the long-term corrosion amount can be predicted with higher accuracy. It was a good judgment. Furthermore, it is preferable that the measurement period by the corrosion sensor is one year or more.

According to the above embodiment, the speed variation function Vp representing the relationship between the predicted exposure period PT and the corrosion rate V is derived, and the final corrosion amount prediction value CPP is output using this speed variation function Vp. As a result, it is possible to minimize the influence of fluctuations due to external factors when measuring the corrosion amount, and to accurately predict the corrosion amount based on short-term measurement results.

That is, when the patch test or the continuous monitoring technique for the amount of corrosion described above is used, the amount of corrosion at that time can be grasped. However, in the prediction method of weathering steel using the results of the exposure test using the patch test piece described above, the results of long-term exposure tests that span 5 to 10 years are required in order to accurately predict the amount of corrosion in the future. becomes. Therefore, there is a problem that it takes a long time before the corrosion amount becomes predictable.

The amount of corrosion in the collected patch is measured by removing rust with a pickling solution standardized by ISO8407, measuring the weight, and calculating the amount of corrosion from the difference from the initial weight. Therefore, when the exposure test time is short and the amount of corrosion is small, the skill of the tester greatly affects the removal of rust, and the calculated amount of corrosion often varies and lacks accuracy.

Furthermore, in the corrosion prediction function Fc=mX ⁿ , it is conceivable to determine the coefficients m and n in consideration of various external factors, and predict the corrosion amount using the corrosion prediction function Fc. However, for example, the flow of wind and the ease with which sediment accumulates differ for each part of a bridge. For this reason, the variation in the prediction of the corrosion amount becomes large, and it is difficult to predict the long-term corrosion amount with high accuracy.

Here, in the corrosion amount prediction apparatus 100 of FIG. 1, corrosion rates V1 to V3 are obtained as average values during the measurement period based on a plurality of pieces of corrosion amount data CD1 to CD3. As a result, even if the corrosion amount data CD1 to CD3 are affected by fluctuations due to external factors when measuring the corrosion amount data, the influence can be minimized by calculating the corrosion rate as an average value. can. Therefore, the rate variation function Vp is obtained from the corrosion rates V1 to V3 obtained as average values, and the rate variation function Vp is time-integrated over the exposure period PT, so that it is possible to accurately predict the amount of corrosion based on short-term measurement results. It can be carried out.

Embodiments of the present invention are not limited to the above embodiments, and various modifications can be made. In the above embodiment, the case where weathering steel is used as the metal material is exemplified, but the present invention is not limited to this and can be applied to prediction of corrosion of any steel material. Further, in the above embodiment, the case of calculation based on three corrosion rates V1 to V3 is illustrated, but two or more corrosion rates may be used.

1 corrosion test device 11 sensor unit 21 reference unit 31 substrate 41 insulation sheet 51 resin 61 cover 71 current source 81 voltage measurement units 81 and 91 voltage measurement unit 100 corrosion amount prediction device 101 data acquisition unit 102 corrosion rate calculation unit 104 variation function derivation Unit 105 Prediction output unit CD, CD1 to CD3 Corrosion amount data CPP Corrosion amount prediction value DB Database Vp Velocity variation function PT Predicted exposure period X, X1 to X3 Measurement period

Claims (7)

A method for predicting the amount of corrosion of a metallic material that outputs a predicted value for the amount of corrosion for an exposure period longer than the measurement period using a plurality of pieces of corrosion amount data for different measurement periods, comprising:
calculating a plurality of corrosion rates during the measurement period using a plurality of the corrosion amount data;
using the plurality of calculated corrosion rates to derive a rate variation function representing the variation of the corrosion rate with respect to the measurement period;
A corrosion amount prediction method, wherein the corrosion amount in the exposure period to be predicted is output as a corrosion amount prediction value by time-integrating the derived speed variation function over the exposure period. The corrosion amount prediction value is calculated by adding the corrosion amount data in the most recent measurement period and a time integral value from the most recent measurement period to the exposure period in the speed variation function. Item 2. The corrosion amount prediction method according to item 1. 3. The corrosion amount prediction method according to claim 1, wherein said corrosion amount data is measured using a change in electrical resistance of the metal based on a reduction amount of the metal cross section due to corrosion. The corrosion amount prediction method according to any one of claims 1 to 3, wherein the measurement period is set according to a cycle of fluctuations in the corrosion environment. 5. The corrosion amount prediction method according to any one of claims 1 to 4, wherein the corrosion rate is calculated from two pieces of the corrosion amount data whose measurement periods are separated by 30 days or more. 6. The rate variation function is derived from a plurality of the corrosion rates calculated using a plurality of the corrosion amount data for which the measurement period is 180 days or longer. Corrosion amount prediction method described in. A corrosion amount prediction device for a metal material that outputs a corrosion amount prediction value for an exposure period longer than the measurement period using a plurality of corrosion amount data for different measurement periods,
a corrosion rate calculation unit that calculates a plurality of corrosion rates during the measurement period using a plurality of the corrosion amount data;
a variation function derivation unit for deriving a rate variation function representing variation of the corrosion rate with respect to the measurement period using the plurality of corrosion rates calculated by the corrosion rate calculation section;
a prediction output unit that outputs the predicted corrosion amount in the exposure period as a corrosion amount prediction value by time-integrating the speed fluctuation function derived in the fluctuation function derivation unit in the exposure period;
A corrosion amount prediction device comprising:

Images