Method for measuring and plotting corrosion amount of metal material, method for selecting metal material, and device for measuring and plotting corrosion amount of metal material

Abstract

A corrosion amount mapping method of a metal material, using corrosion amount data including a corrosion amount of a metal material, a plurality of environmental parameters, position coordinates of the environmental parameters on a map, topographic data of the map, and a period of use of the metal material, to create a corrosion amount prediction map, includes: an environmental map creation step of creating an environmental map at an arbitrary grid interval from the plurality of environmental parameters, the topographic data of the map, and the position coordinates of the environmental parameters; a prediction request point input step; a similarity calculation step; a corrosion amount prediction step; and a corrosion amount prediction map creation step of creating a corrosion amount prediction map by coloring and marking a prediction result of the corrosion amount at the prediction request point of the grid on the map.

Description

Technical Field

[0001] This invention relates to a method for mapping the corrosion of metallic materials, a method for selecting metallic materials, and a device for mapping the corrosion of metallic materials. Background Technology

[0002] As shown in Non-Patent Document 1, the corrosion amount of metallic materials in atmospheric corrosion environments has historically been expressed by the following formula (1) as an empirical formula.

[0003] [Mathematical Expression 1]

[0004] Y = AX B …(1)

[0005] Here, in the above equation (1), Y is the corrosion amount of the metallic material, X is the service period of the metallic material, A is a parameter representing the corrosion amount of the metallic material over a period of one year from the initial stage, and B is a parameter representing the attenuation of the corrosion rate caused by the effect of the rust layer formed by corrosion. The values ​​of these parameters A and B vary depending on the type of metallic material and the atmospheric corrosion environment. Therefore, currently, when predicting long-term corrosion, most methods are used to extrapolate the time-dependent changes in corrosion amount by exposing the metallic material to the atmospheric corrosion environment to be targeted for multiple periods and using the above equation (1).

[0006] However, the corrosion rate of metallic materials is determined by the complex interplay of the metal's corrosion resistance and atmospheric corrosion environmental factors, such as temperature, relative humidity, wetting time, rainfall, air salinity, and SO2 concentration. Therefore, as shown below, a technique for predicting the corrosion rate of metallic materials is proposed by formulating the above environmental factors.

[0007] For example, in Non-Patent Literature 2, for carbon steel, zinc, copper and aluminum, the logarithm of corrosion is calculated by summing the terms obtained by performing a multiple regression of the logarithms of temperature, relative humidity and fly salt with the terms obtained by performing a multiple regression of the logarithms of temperature, relative humidity and SO2.

[0008] In addition, in Patent Document 1, the annual wetting time, annual average wind speed, air salt content, sulfur oxide content, activation energy of corrosion reaction and temperature are used as parameters to represent the corrosion index Z as shown in the following formula (2), and the quadratic function of the corrosion index Z is used to calculate the long-term corrosion amount.

[0009] [Mathematical Expression 2]

[0010]

[0011] Here, in the above equation (2), TOW is the annual wet time (h), W is the annual average wind speed (m / s), C is the air salinity (mdd), S is the sulfur oxide content (mdd), and E is the annual average wind speed (m / s). α Let be the activation energy of the corrosion reaction (J / mol), R be the gas constant (J / (K / mol)), T be the annual average temperature (K), and α, κ, δ, and ε be constants. Furthermore, [mdd] above represents the amount of NaCl collected per unit number of days and per unit area, expressed as [mg NaCl·dm]. －2 ·day －1 ] is an abbreviation for .

[0012] Furthermore, Patent Document 2 proposes a technique for predicting air salinity C. Additionally, Patent Document 3 proposes a corrosion prediction technique that uses temperature, relative humidity, air salinity, and wetting probability as parameters to calculate parameter A of the aforementioned equation (1), which is known as an empirical formula, using the following equation (3), and calculates parameter B as a function of parameter A through laboratory testing.

[0013] [Mathematical Expression 3]

[0014] A=(α·T+β)·Pw(T,H)·(Sa Y (3)

[0015] In equation (3) above, T is temperature (°C), H is relative humidity (%), and Sa is air salinity (mg / dm³). 2 / day(=mdd)), Pw(T,H) is the wetting probability, and α, β, γ are coefficients set according to the steel grade.

[0016] In addition, Patent Document 4 proposes a corrosion prediction technique that uses temperature, wetting time and air salinity as parameters in an outdoor atmospheric corrosion environment, calculates the parameter A of the above formula (1) as an empirical formula using the following formula (4), and sets the parameter B in the range of 0.3 to 0.6.

[0017] [Mathematical Expression 4]

[0018] A = kT α TOW β ·SA Y …(4)

[0019] Here, in equation (4) above, T is the temperature (°C), TOW is the wetting time (h), and Sa is the air salinity (mg / dm³). 2 / day(=mdd)), α, β, γ are coefficients.

[0020] Furthermore, in Patent Document 5, a technique is proposed to calculate parameter A of the above formula (1), which is known as an empirical formula, according to the following formula (5), and to calculate parameter B according to the following formula (6) when predicting the reduction in the thickness of steel under atmospheric conditions.

[0021] [Mathematical Expression 5]

[0022] A=(CR0+CR1)÷2…(5)

[0023] [Mathematical Expression 6]

[0024] B=2CR1÷(CR0+CR1)…(6)

[0025] In equations (5) and (6) above, CR0 is a function representing the initial corrosion rate of the steel immediately after manufacturing, taking environmental factors as parameters, and CR1 is a function representing the corrosion rate of the steel one year after manufacturing, taking environmental factors as parameters. Furthermore, the environmental factors here refer to the annual average temperature (°C), annual average humidity (%), annual average wind speed (m / sec), and air salinity (mg / dm³). 2 / day (=mdd)), sulfur oxide content (mg / dm 2 / day(=mdd)).

[0026] Furthermore, in Patent Document 6, when predicting the corrosion rate of metallic materials, a multiple regression analysis is performed with corrosion rate as the target variable and environmental and topographical factors affecting the corrosion rate as explanatory variables. Moreover, in this multiple regression analysis, at least one of the explanatory variables is a hypothetical wetting time weighted according to relative humidity from 0% to 100%. Furthermore, a method is proposed whereby the hypothetical wetting time is calculated by summing the products obtained by multiplying the weighting coefficients (which vary according to the relative humidity) by the time corresponding to the varying relative humidity, and a corrosion rate estimation formula is created based on the measured corrosion rate of the metallic material using multiple regression analysis.

[0027] Furthermore, Patent Document 7 proposes the following technique: as a technique for predicting the corrosion rate of a metal material by mapping (mapping) the corrosion amount, a multivariate regression analysis optimized by using a stepwise method is proposed to predict the corrosion rate of the metal material, and the corrosion rate of the metal material is mapped by clustering.

[0028] [Existing Technical Documents]

[0029] [Patent Literature]

[0030] [Patent Document 1] Japanese Patent No. 3909057

[0031] [Patent Document 2] Japanese Patent No. 4143018

[0032] [Patent Document 3] Japanese Patent No. 4706254

[0033] [Patent Document 4] Japanese Patent No. 5895522

[0034] [Patent Document 5] Japanese Patent No. 5066160

[0035] [Patent Document 6] Japanese Patent No. 5066955

[0036] [Patent Document 7] Japanese Patent No. 5684552

[0037] [Non-patent literature]

[0038] [Non-Patent Document 1] [Joint Research Report on the Application of Weather-Resistant Steel in Bridges (XVIII)], Ministry of Construction Civil Engineering Research Institute, (Company) Steel Club, (Company) Japan Bridge Construction Association, March 5, 2008

[0039] [Non-patent document 2] ISO9223: 1992 "Corrosion of metals and alloys-Corrosivity of atmospheres-Classification, determination and estimation" Summary of the Invention

[0040] [The problem the invention aims to solve]

[0041] In patent documents 1-5, environmental parameters were selected as explanatory variables and formulas were created by evaluating the relationship between corrosion amount and corrosion rate and various environmental parameters. However, corrosion amount, corrosion rate, and various environmental parameters have complex correlations. For example, the relationship between corrosion amount and temperature is non-linear, and air salinity is approximately correlated with SO2 concentration. Given such relationships, high-precision predictions cannot be expected when formulas are created as in patent documents 1-5.

[0042] Furthermore, Patent Document 6 is characterized by weighting based on wetting time. While this weighting improves accuracy, even when weighting only wetting time among multiple environmental parameters, a significant improvement in accuracy cannot be expected. Moreover, the method in Patent Document 6 can only perform corrosion prediction during the period of current data retention, and cannot perform long-term corrosion prediction.

[0043] Furthermore, in Patent Document 7, data is classified through clustering, and multiple environmental parameters are selected from multiple environmental parameters for each class. Repeated multivariate regression is then performed to obtain the optimal formula, thereby improving prediction accuracy. However, the relationship between corrosion amount and corrosion rate cannot be simply represented by a linear expression for each environmental parameter obtained from multivariate regression. Therefore, a significant improvement in accuracy cannot be expected from the method in Patent Document 7.

[0044] Furthermore, the method disclosed in Patent Document 7 obtains environmental parameters for the prediction location from publicly available data (such as agricultural meteorological data provided by agricultural research institutions), for example, obtaining environmental parameters for each 1km square grid in the case of Japan. However, such data does not include all environmental parameters such as air salinity and SO2 levels. Moreover, the method disclosed in Patent Document 7 does not explicitly provide a method for supplementing insufficient environmental parameters. Therefore, it is believed that even using the method disclosed in Patent Document 7, clustering and corrosion prediction using only known environmental parameters can only produce a low-accuracy corrosion prediction map.

[0045] The present invention was made in view of the above circumstances, and its purpose is to provide a method for measuring the corrosion amount of metallic materials, a method for selecting metallic materials, and a device for measuring the corrosion amount of metallic materials that can accurately measure the corrosion amount of metallic materials in an atmospheric corrosion environment.

[0046] [Methods used to solve problems]

[0047] To address the aforementioned issues and achieve the objectives, the present invention provides a method for mapping the corrosion of metallic materials. This method uses corrosion data to predict the corrosion amount of metallic materials and creates a corrosion prediction map. The corrosion data includes: the usage period of the metallic material; multiple environmental parameters known on a map representing the usage environment of the metallic material during the usage period; the location coordinates of the environmental parameters on the map; the topographic data of the map; and the corrosion amount of the metallic material during the usage period. The method for mapping the corrosion of metallic materials is characterized by including an environmental map creation step, whereby, based on the multiple environmental parameters, the topographic data of the map, and the location coordinates of the environmental parameters on the map, an environmental map is created for each environmental parameter at an arbitrary grid interval. The process includes: a map; a prediction request point input step, whereby a prediction request point is input, which includes the usage period of the metal material for creating the corrosion prediction map and multiple environmental parameters in the corrosion data; a similarity calculation step, whereby the similarity between the multiple environmental parameters in the corrosion data and the multiple environmental parameters in the prediction request point is calculated; a dimensionality compression step, whereby the dimensionality of the multiple environmental parameters in the corrosion data is compressed into latent variables, taking into account the similarity; a corrosion prediction step, whereby the corrosion of the metal material at the prediction request point in the grid is predicted based on a prediction formula constructed using the latent variables and the similarity; and a corrosion prediction map creation step, whereby the corrosion prediction map is created by color-coding the predicted corrosion results at the prediction request point in the grid on a map.

[0048] Furthermore, the corrosion measurement method for metallic materials of the present invention is characterized in that, in the above scheme, the environmental map production step includes: a first step, inputting the plurality of environmental parameters into the location coordinates on the map; a second step, correcting the environmental map to an environmental map with an altitude of 0m based on the topographic data of the map; a third step, supplementing the plurality of environmental parameters in the environmental map with an altitude of 0m; and a fourth step, correcting the environmental map with supplemented environmental parameters with an altitude of 0m back to the original altitude environmental map based on the topographic data of the map.

[0049] Furthermore, the corrosion measurement method for metallic materials of the present invention is characterized in that, in the above scheme, the plurality of environmental parameters include air salinity, and in the third step, when supplementing the air salinity of the grid near the coastline, the air salinity is supplemented in a manner that does not exceed a preset upper limit value.

[0050] Furthermore, the corrosion measurement method for metallic materials of the present invention is characterized in that, in the above scheme, the corrosion prediction step includes: an initial corrosion prediction step, predicting a first parameter representing the corrosion amount of the metallic material during a specified period; an attenuation prediction step, predicting a second parameter representing the attenuation of the corrosion rate of the metallic material; and a long-term corrosion prediction step, predicting the corrosion amount of the metallic material during a period longer than the specified period based on the service period of the metallic material, the first parameter, and the second parameter.

[0051] Furthermore, the corrosion measurement method for metallic materials of the present invention is characterized in that, in the above scheme, in the initial corrosion prediction step, the corrosion amount of the metallic material under the environmental parameters of the prediction request point during a specified period is predicted based on the prediction formula constructed using the latent variables and the similarity.

[0052] Furthermore, the corrosion mapping method for metallic materials of the present invention is characterized in that, in the above scheme, in the attenuation prediction step, the second parameter is predicted based on the usage period of the metallic material in which the corrosion prediction map is made, multiple environmental parameters in the corrosion data, and the similarity.

[0053] Furthermore, the corrosion measurement method for metallic materials of the present invention is characterized in that, in the above scheme, the plurality of environmental parameters include at least one of temperature, relative humidity, absolute humidity, wetting time, and rainfall; and air salinity, SO₂, etc. X Concentration and NO X At least one of the concentrations.

[0054] Furthermore, the corrosion measurement method for metallic materials of the present invention is characterized in that, in the above scheme, the metallic material is steel.

[0055] Furthermore, the corrosion measurement method for metallic materials of the present invention is characterized in that, in the above scheme, the corrosion prediction map is a map formed by coloring and marking each grid according to the corrosion value predicted in the corrosion prediction map making step.

[0056] In order to solve the above-mentioned problems and achieve the objectives, the method for selecting metal materials of the present invention is characterized by using the above-mentioned method for measuring the corrosion amount of metal materials to select metal materials that are appropriate for the use environment.

[0057] To address the aforementioned issues and achieve the objectives, the present invention provides a metal material corrosion measurement device, comprising: a database storing corrosion data, the corrosion data including: the service period of the metal material, multiple environmental parameters known on a map representing the service environment of the metal material during the service period and located on a map representing the region where the metal material is used, the location coordinates of the environmental parameters on the map, topographic data of the map, and the corrosion amount of the metal material during the service period; an environmental map generation unit that generates an environmental map for each environmental parameter based on the multiple environmental parameters, the topographic data of the map, and the location coordinates of the environmental parameters on the map, at arbitrary grid intervals; and an input unit that inputs prediction request points. The prediction request point includes: the usage period of the metal material for creating the corrosion prediction map and multiple environmental parameters in the corrosion data; a similarity calculation unit that calculates the similarity between the multiple environmental parameters in the corrosion data and the multiple environmental parameters at the prediction request point; a dimensionality compression unit that considers the similarity to compress the dimensionality of the multiple environmental parameters in the corrosion data into latent variables; a corrosion prediction unit that predicts the corrosion amount of the metal material at the prediction request point of the grid based on a prediction formula constructed using the latent variables and the similarity; and a corrosion prediction map creation unit that creates the corrosion prediction map by coloring the predicted corrosion amount at the prediction request point of the grid on a map.

[0058] This invention is based on the above insights, and its main points are as follows.

[0059] [1] A method for mapping the corrosion of metallic materials, which uses corrosion data to predict the corrosion of metallic materials and creates a corrosion prediction map, wherein the corrosion data includes: the service period of the metallic material; multiple environmental parameters known on a map representing the service environment of the metallic material during the service period and on a map representing the region where the metallic material is used; the location coordinates of the environmental parameters on the map; the topographic data of the map; and the corrosion of the metallic material during the service period, wherein the method for mapping the corrosion of metallic materials is characterized by including:

[0060] The environmental map creation steps involve creating an environmental map for each environmental parameter using arbitrary grid intervals, based on the multiple environmental parameters, the terrain data of the map, and the location coordinates of the environmental parameters on the map.

[0061] The prediction request point input step involves inputting a prediction request point, which includes the usage period of the metallic material used to create the corrosion prediction map and multiple environmental parameters from the corrosion data.

[0062] The similarity calculation step calculates the similarity between multiple environmental parameters in the corrosion data and multiple environmental parameters in the prediction request point;

[0063] The dimension compression step considers the similarity to compress the dimension of multiple environmental parameters in the corrosion data into latent variables;

[0064] The corrosion prediction step, based on a prediction formula constructed using the latent variables and the similarity, predicts the corrosion amount of the metallic material at the prediction request point on the grid; and

[0065] The steps for creating a corrosion prediction map are as follows: the predicted corrosion amount at the prediction request points of the grid is colored and marked on the map, thereby creating a corrosion prediction map.

[0066] [2] According to the method for measuring the corrosion of metallic materials described in [1], wherein,

[0067] The environmental map creation steps include:

[0068] The first step is to input the multiple environmental parameters into the location coordinates on the map;

[0069] The second step is to correct the environmental map to an environmental map with an altitude of 0m based on the terrain data of the map.

[0070] The third step is to supplement environmental parameters among the multiple environmental parameters in the environmental map at an altitude of 0m.

[0071] The fourth step is to correct the environmental map with the elevation of 0m, which has been supplemented with environmental parameters, to the original elevation based on the terrain data of the map.

[0072] [3] According to the method for measuring the corrosion of metallic materials described in [2], wherein,

[0073] The environmental parameters include air salinity.

[0074] In the third step, when replenishing the air salinity of the grid near the coastline, the air salinity is replenished in a manner that does not exceed a preset upper limit.

[0075] [4] The corrosion measurement method for metallic materials according to any one of [1] to [3], wherein the corrosion prediction step includes:

[0076] The initial corrosion prediction step predicts a first parameter representing the corrosion amount of the metallic material over a specified period;

[0077] The attenuation prediction step predicts a second parameter representing the attenuation of the corrosion rate of the metallic material; and

[0078] The long-term corrosion prediction step predicts the corrosion amount of the metal material during a period longer than the specified period, based on the service life of the metal material, the first parameter, and the second parameter.

[0079] [5] According to the method for measuring the corrosion of metallic materials described in [4], wherein,

[0080] In the initial corrosion prediction step, the corrosion amount of the metallic material during a specified period under the environmental parameters of the prediction request point is predicted based on the prediction formula constructed using the latent variables and the similarity.

[0081] [6] According to the method for measuring the corrosion of metallic materials described in [4], wherein,

[0082] In the attenuation prediction step, the second parameter is predicted based on the usage period of the metal material in which the corrosion prediction map is made, multiple environmental parameters in the corrosion data, and the similarity.

[0083] [7] A method for measuring the corrosion of metallic materials according to any one of [1] to [6], wherein,

[0084] The environmental parameters include at least one of the following: temperature, relative humidity, absolute humidity, wetting time, and rainfall; and at least one of air salinity, SOX concentration, and NOX concentration.

[0085] [8] A method for measuring the corrosion of metallic materials according to any one of [1] to [4], wherein,

[0086] The metal material is steel.

[0087] [9] The method for measuring the corrosion of metallic materials according to any one of [1] to [8], wherein,

[0088] The corrosion prediction map is a map created by coloring and marking each grid based on the corrosion value predicted in the corrosion prediction map creation step.

[0089]

[10] The method for selecting metallic materials is characterized by using the corrosion measurement method of any one of [1] to [8] to select metallic materials that correspond to the environment of use.

[0090]

[11] A device for measuring the corrosion of metallic materials, comprising:

[0091] The database stores corrosion data, which includes: the service period of the metal material, multiple environmental parameters known on a map representing the region where the metal material is used during the service period, the location coordinates of the environmental parameters on the map, the topographic data of the map, and the corrosion amount of the metal material during the service period.

[0092] The environmental map production department produces an environmental map for each of the environmental parameters based on the multiple environmental parameters, the terrain data of the map, and the location coordinates of the environmental parameters on the map, at arbitrary grid intervals.

[0093] The input unit inputs prediction request points, which include: the usage period of the metallic material for creating the corrosion prediction map and multiple environmental parameters in the corrosion data;

[0094] The similarity calculation unit calculates the similarity between multiple environmental parameters in the corrosion data and multiple environmental parameters at the prediction request point;

[0095] The dimension compression unit takes into account the similarity to compress the dimension of multiple environmental parameters in the corrosion data into latent variables;

[0096] A corrosion prediction unit, based on a prediction formula constructed using the latent variables and the similarity, predicts the corrosion amount of the metallic material at the prediction request point of the grid; and

[0097] The corrosion prediction map production department creates the corrosion prediction map by coloring and marking the predicted corrosion amount at the prediction request points of the grid on a map.

[0098] [The effects of the invention]

[0099] According to the present invention, in an atmospheric corrosion environment, the corrosion amount of metallic materials can be measured with high precision, and the optimal metallic material with corrosion resistance corresponding to the use environment can be selected. Attached Figure Description

[0100] [ Figure 1 ] Figure 1 It is a graph showing the relationship between temperature (annual average) and corrosion amount (over-year) in an atmospheric corrosion environment.

[0101] [ Figure 2 ] Figure 2 This is a graph showing the simulated correlation between SO2 concentration (annual average) and air salinity (annual average) in an atmospheric corrosion environment.

[0102] [ Figure 3 ] Figure 3This is a block diagram illustrating the structure of a metal material corrosion measurement device according to an embodiment of the present invention.

[0103] [ Figure 4 ] Figure 4 This is a flowchart illustrating the method for measuring the corrosion of metallic materials according to an embodiment of the present invention.

[0104] [ Figure 5 ] Figure 5 It is a corrosion prediction map that represents the predicted corrosion amount of a metal material one year later, obtained through existing methods for measuring the corrosion amount of metallic materials.

[0105] [ Figure 6 ] Figure 6 This is a corrosion prediction map representing the predicted corrosion value of a metallic material fifty years later, obtained through the corrosion measurement method of the present invention. Detailed Implementation

[0106] Hereinafter, with reference to the accompanying drawings, the method for measuring the corrosion amount of metallic materials, the method for selecting metallic materials, and the apparatus for measuring the corrosion amount of metallic materials according to embodiments of the present invention will be described. It should be noted that the present invention is not limited to the following embodiments.

[0107] (Supplementary environmental parameters)

[0108] There are complex correlations between metallic materials, corrosion levels, and various environmental parameters in the atmosphere. In the international standard ISO 9223 mentioned in non-patent literature 2, based on annual average temperature, relative humidity, air salinity, and SO₂... X Concentration is used to formulate the annual corrosion amount. However, it is difficult to obtain these environmental parameters on a map with narrow grid intervals of about 1 km or 2 km. Therefore, it is necessary to supplement the environmental parameters with the location coordinates of each environmental parameter based on the location coordinates of the available environmental parameters.

[0109] At this point, for example, temperature varies not only with longitude and latitude but also with elevation. Furthermore, the manner in which inhaled salts are transported is also affected by obstacles such as mountains or hills. Therefore, the influence of such terrain needs to be considered when supplementing environmental parameters.

[0110] However, it is difficult to reflect terrain data and supplement environmental parameters between location coordinates in a single calculation. Therefore, in this embodiment, multiple environmental maps plotted with known environmental parameters are first corrected into multiple environmental maps converted to an elevation of 0m. Next, environmental maps with supplemented environmental parameters between location coordinates and an elevation of 0m are created. Then, by combining actual terrain data to correct the supplemented environmental parameters, environmental maps at the original elevation are generated. It should be noted that the aforementioned "environmental map" refers to a map plotted with environmental parameters. The specific method for supplementing each environmental parameter is described below.

[0111] (Temperature supplement)

[0112] In the temperature supplementation process, the available temperature data is first plotted on the map. Next, based on the rule that "if the elevation decreases by 100m, the temperature rises by 0.6℃" (temperature attenuation rate), topographic data is used to correct the temperature values ​​at each plotted point to a value equivalent to 0m elevation. Then, the temperatures between the data points are supplemented at arbitrary grid intervals. It should be noted that the grid interval can be determined based on the number of known environmental parameters on the map, the specifications of the machine used for calculation, etc.

[0113] Various methods can be used to supplement data, such as linear supplementation and polynomial supplementation. However, any supplementation method can be used as long as the accuracy required by the user in this implementation method is achieved through full cross-validation when predicting the environment. This also applies to the supplementation of environmental parameters other than temperature. Then, based on the rule that "if the elevation increases by 100m, the temperature decreases by 0.6℃", the temperature values ​​of each plotted point on the environmental map at 0m elevation are corrected to the values ​​converted to the original elevation using topographic data, thereby creating an environmental map (temperature map) of the original elevation temperature.

[0114] (Supplement to absolute humidity)

[0115] Absolute humidity can be calculated based on environmental parameters such as temperature and relative humidity. In this context, the atmospheric corrosion environment is characterized by "absolute humidity remaining roughly constant even with temperature changes." Therefore, utilizing this characteristic, it is not necessary to reflect topographic data such as elevation; instead, the absolute humidity at each grid point is calculated based on temperature and relative humidity, and the data on the map are supplemented to generate an environmental map of absolute humidity (absolute humidity map).

[0116] (Supplementing relative humidity)

[0117] In the supplement to relative humidity, the atmospheric corrosion environment characteristic of "absolute humidity remaining roughly constant even with temperature changes" is utilized. Relative humidity is calculated at each grid point of the temperature and absolute humidity environment map to create an environmental map of relative humidity.

[0118] (Supplement to wetting time)

[0119] Regarding wetting time, the international standard defines it as "the time when the relative humidity is above 80%". Therefore, based on the hourly and daily relative humidity variation maps obtained using the above method, the annual wetting time is calculated by accumulating the time when the relative humidity is above 80% at each grid point, thereby creating an environmental map of wetting time (wetting time map).

[0120] (Supplementary rainfall data)

[0121] Rainfall is not affected by topography, so there is no need to reflect topographic data such as elevation. Instead, an environmental map of rainfall (rainfall map) is created by supplementing the rainfall data on the map.

[0122] (air salinity, SO2) X Concentration and NO X (Concentration supplement)

[0123] In terms of air salinity and SO2 X Concentration and NO X In the concentration supplementation, data are supplemented by Euclidean distances based on longitude, latitude, and altitude, thereby creating air salinity and SO₂ concentrations. X Concentration and NO X Environmental map of concentrations (air salinity, SO2) X Concentration and NO X Concentration map).

[0124] Here, the specific concentration of air salinity is expressed by the model formula "y = ax －b "Calculation. In this model formula, x: distance from the shore (km), y: air salinity (mdd), a, b: coefficients. Furthermore, the distance from the shore x is the minimum distance between each point and the shape data of the coastline. It should be noted that, in addition to the above model formula, existing corrosion prediction formulas such as the Cole model and the mesoclimate model can also be used to calculate air salinity."

[0125] Furthermore, when supplementing air salinity, it is preferable to set an upper limit for air salinity to avoid abnormally high salinity levels when grid points are close to the coastline. For example, an upper limit for air salinity could be set to a value such as "1.0mdd = 62.3mmd".

[0126] (Prediction of corrosion)

[0127] In predicting the corrosion amount of metallic materials, the corrosion amount is predicted according to each grid of the environmental map created as described above. The inventors of this application have discovered that by separately predicting parameter A (first parameter), which represents the corrosion amount of metallic materials under various atmospheric corrosion environments over a period of one year from the initial point, and parameter B (second parameter), which represents the attenuation of the corrosion rate caused by the rust layer, and by making predictions based on data obtained by weighting each environmental parameter, the accuracy of corrosion prediction is improved.

[0128] The corrosion rate of metallic materials generally decreases over time. This is due to the protective effect of corrosion products (such as rust) formed on the surface of the metal. Furthermore, this protective effect varies greatly depending on the surrounding environment and the type of metal. Thus, the corrosion rate of metallic materials is interrelated with numerous factors, including various environments and the corrosion resistance of the metal itself. Therefore, it is very difficult to accurately predict the corrosion rate of metallic materials in any environment and period based on the principles of the relationship between environmental parameters and corrosion rate. For example, statistical prediction using data sets of corrosion rates and various environmental parameters stored in a database is realistic and also relates to improving accuracy.

[0129] On the other hand, generally speaking, long-term data is scarce in datasets of accumulated corrosion and various environmental parameters. For example, if the predicted corrosion amount is over a long period of several decades, directly predicting the corrosion amount by including the period as a variable would result in reduced accuracy because the corrosion amount over any given long period is predicted based on data far removed from that period. Therefore, in this invention, improved accuracy is achieved by separately predicting parameter A, which represents the corrosion amount of the metallic material over a one-year period from the initial point, and parameter B, which represents the decay of the corrosion rate caused by the rust layer.

[0130] Here, key environmental parameters related to the corrosion rate of metallic materials over any given period include, for example, temperature, relative humidity, absolute humidity, wetting time, rainfall, air salinity, and SO₂. X Concentration and NO X Concentration, etc. Among these environmental parameters, there are, for example... Figure 1 The relationship between temperature and corrosion rate shown is a non-linear environmental parameter. Furthermore, there are also, for example... Figure 2 The relationship between air salinity and SO2 concentration, as shown, reveals a multicollinearity among environmental parameters. Besides the presence of multiple environmental parameters affecting the corrosion of metallic materials, these two factors are the main reasons why it is difficult to accurately predict the corrosion rate of metallic materials in any given environment and period.

[0131] Regarding cases where environmental parameters exhibit a non-linear relationship with corrosion rates, prediction accuracy can be improved by weighting each sample based on its similarity to the desired environment and time period, and then performing localized multiple regression analysis. It should be noted that the "samples" mentioned above refer to the data sets of corrosion rates and various environmental parameters stored in the database (corrosion rate data described later).

[0132] Furthermore, the issue of multicollinearity among environmental parameters can be addressed by compressing the dimensionality of each environmental parameter into independent parameters, thereby generating new parameters. Moreover, as one method to simultaneously achieve these goals, the "Locally Weighted Partial Least Squares (LW-PLS)" method, as shown in Reference 1, is an example.

[0133] Reference 1: Jin Shanghong, Okajima Ryota, Kano Manabu, Hasebe Shinji, “Sample Selection for Constructing High-Precision Local PLS Models,” 54th Conference of the Federation of Automatic Control, 54 (2011), p. 1594

[0134] In this invention, for parameter A, representing the corrosion amount of metallic materials under various atmospheric corrosion environments over a one-year period from the initial point, and parameter B, representing the decay of the corrosion rate caused by rust, the following prediction method is used: For each sample, the similarity (similarity score) with the point of prediction is calculated; this similarity score is then weighted, and prediction is performed through local regression. Furthermore, new parameters (derived latent variables) are generated by dimensionality compression of each environmental parameter, serving as explanatory variables for the local regression. Additionally, at this point, latent variables are determined by maximizing the inner product of the weighted latent variable based on similarity with the target variable, and local multiple regression is performed. Specific embodiments of the invention will now be described with reference to the accompanying drawings.

[0135] (Corrosion Measurement Device)

[0136] Reference Figure 3 The structure of the corrosion measurement device for metallic materials according to an embodiment of the present invention will be described. The corrosion measurement device 1 includes an input unit 10, a database 20, a calculation unit 30, and a display unit 40.

[0137] The input unit 10 is implemented, for example, by an input device such as a keyboard, mouse pointer, or number keys. As will be described later, the prediction request point, which will be described later, is input to the arithmetic unit 30 via the input unit 10.

[0138] Database 20 stores corrosion data as actual values ​​of corrosion amount for metallic materials. The corrosion data includes the service period of the metallic material (e.g., steel), the corrosion amount of the metallic material during that service period, multiple known environmental parameters representing the service environment of the metallic material during that service period and on a map representing the region where the metallic material is used, the location coordinates of the environmental parameters on the map, and the topographic data of the map.

[0139] The aforementioned "multiple environmental parameters" include at least one of temperature (air temperature), relative humidity, absolute humidity, humidification time, and rainfall; air salinity, SO2, etc. X Concentration and NO X At least one of the concentrations. Furthermore, these environmental parameters are, for example, annual average data. Additionally, corrosion data for each steel grade is stored in database 20.

[0140] Specifically, the arithmetic unit 30 is implemented by a processor consisting of a CPU (Central Processing Unit), a DSP (Digital Signal Processor), an FPGA (Field-Programmable Gate Array), etc., and a memory (main storage unit) consisting of RAM (Random Access Memory), ROM (Read Only Memory), etc.

[0141] The calculation unit 30, for example, loads a program stored in a storage unit (not shown) into the working area of ​​the main storage unit and executes it. Through program execution, it controls various component units, thereby achieving a function consistent with the intended purpose. The calculation unit 30 functions as the map creation unit 31, the similarity calculation unit 32, the dimension compression unit 33, and the erosion prediction unit 34 through program execution. It should be noted that detailed descriptions of the map creation unit 31, the similarity calculation unit 32, the dimension compression unit 33, and the erosion prediction unit 34 will be provided later.

[0142] The display unit 40 is implemented by a display device such as an LCD display or a CRT display. Based on the display signal input from the arithmetic unit 30, it displays a corrosion prediction map as a prediction result of the corrosion amount of the metallic material. It should be noted that the aforementioned "corrosion prediction map" refers to, for example, a map that displays the predicted corrosion amount using color (see reference...). Figure 5 and Figure 6 ).

[0143] (Corrosion Measurement Methods)

[0144] Reference Figure 4The corrosion mapping method for metallic materials according to embodiments of the present invention will be described. The corrosion mapping method includes the following steps: environmental map creation, prediction request point input, first similarity calculation, first dimension compression, initial corrosion prediction (first parameter prediction), second similarity calculation, second dimension compression, attenuation prediction (second parameter prediction), long-term corrosion prediction, and corrosion prediction map creation. It should be noted that the corrosion mapping method of this embodiment, when applied to metallic materials, especially weather-resistant steels, can predict and map corrosion with higher accuracy.

[0145] In the environmental map creation step, the map creation unit 31 supplements the data between the available environmental parameters on the map with arbitrary grid intervals to create an environmental map for each environmental parameter (step S1).

[0146] In the environmental map creation process, based on multiple known environmental parameters on the map, the map's terrain data, and the location coordinates of the environmental parameters on the map, an environmental map for each environmental parameter is created at arbitrary grid intervals. More specifically, the environmental map creation process involves the following four steps.

[0147] First, input the known environmental data into the location coordinates on the map (Step 1). Next, based on the map's terrain data, correct the environmental map to an elevation of 0m (Step 2). Then, in the 0m elevation environment map, supplement the known environmental parameters (Step 3). Finally, based on the map's terrain data, correct the 0m elevation environment map with supplemented environmental parameters back to the original elevation environment map (Step 4).

[0148] Here, in the third step described above, as mentioned, when replenishing the air salinity at grid points near the coastline, it is preferable to replenish the air salinity in a manner that does not exceed a preset upper limit (e.g., 1.0 mgd). By setting an upper limit for the air salinity, it is possible to prevent the air salinity at grid points near the coastline from becoming abnormally high.

[0149] In the prediction request point input step, the environmental parameters of the grid points, i.e., the prediction request points, are input to the calculation unit 30 via the input unit 10 (step S2). The prediction request points include the usage period of the metal material whose corrosion amount is to be predicted (the usage period of the metal material for which the corrosion amount prediction map is made), and multiple environmental parameters (multiple environmental parameters in the corrosion amount data) representing the annual average of the usage environment of the metal material during the usage period.

[0150] Next, in the first similarity calculation step, the similarity calculation unit 32 calculates the similarity between multiple environmental parameters in the corrosion data of metallic materials stored in the database 20 for a period of one year and multiple environmental parameters at the prediction request point (step S3). In this step, the similarity calculation unit 32 calculates the similarity, for example, using the formula (8) described later. It should be noted that specific examples of this step will be described in the embodiments described later.

[0151] Next, in the first dimension compression step, the dimension compression unit 33, taking into account the similarity calculated in the first similarity calculation step, compresses the dimensions of multiple environmental parameters (explanatory variables) in the corrosion data into latent variables (step S4). In this step, the dimension compression unit 33 calculates the latent variables, for example, using equation (7) described later. It should be noted that specific examples of this step will be described in the embodiments described later.

[0152] Next, in the initial corrosion prediction step, the corrosion prediction unit 34 predicts the corrosion amount of the metallic material over a period of one year from the initial point, based on a prediction formula constructed using the latent variables calculated in the first dimension compression step and the similarity calculated in the first similarity calculation step (step S5), under the environmental parameters of the prediction request point. The corrosion amount of the metallic material over a period of one year from the initial point refers to parameter A (the first parameter) in the above formula (1). In this step, the corrosion prediction unit 34, for example, constructs the prediction formula shown in formula (10) described later, and predicts the corrosion amount of the metallic material over a period of one year from the initial point based on this prediction formula. It should be noted that specific examples of this step will be described in the embodiments described later.

[0153] Next, in the second similarity calculation step, the similarity calculation unit 32 considers the usage period of the metal material whose corrosion amount is to be predicted and calculates the similarity between multiple environmental parameters in the corrosion amount data of the metal material and multiple environmental parameters at the prediction request point (step S6). In this step, the similarity calculation unit 32 calculates the above similarity, for example, using the formula (13) described later. It should be noted that a specific example of this step will be described in the embodiments described later.

[0154] Next, in the second dimension compression step, the dimension compression unit 33, taking into account the similarity calculated in the second similarity calculation step and the usage period of the metal material for which the corrosion amount is to be predicted, compresses the dimensions of multiple environmental parameters (explanatory variables) in the corrosion amount data into latent variables (step S7). In this step, the dimension compression unit 33 calculates the aforementioned latent variables, for example, using equation (12) described later. It should be noted that specific examples of this step will be described in the embodiments described later.

[0155] Next, in the attenuation prediction step, the corrosion amount prediction unit 34 predicts a parameter representing the attenuation of the corrosion rate of the metallic material based on a prediction formula constructed using the latent variables calculated in the second dimension compression step and the similarity calculated in the second similarity calculation step (step S8). The parameter representing the attenuation of the corrosion rate of the metallic material refers to parameter B (the second parameter) in the above formula (1). In this step, the corrosion amount prediction unit 34 constructs, for example, the prediction formula shown in formula (16) described later, and predicts the parameter representing the attenuation of the corrosion rate of the metallic material based on this prediction formula. It should be noted that a specific example of this step will be described in the embodiments described later.

[0156] Next, in the long-term corrosion prediction step, the corrosion prediction unit 34 predicts the corrosion amount (long-term corrosion amount) of the metal material for a period longer than one year based on the service life of the metal material, parameter A calculated in the initial corrosion prediction step, and parameter B calculated in the decay prediction step (step S9). In this step, the corrosion prediction unit 34 predicts the long-term corrosion amount based on the above formula (1). It should be noted that specific examples of this step will be described in the embodiments described later.

[0157] Next, in the corrosion prediction map creation step, the map creation unit 31 colors and marks the predicted corrosion amounts at the prediction request points of the grid on the map, thereby creating a corrosion prediction map (step S10). This corrosion prediction map is a map created by coloring and marking each grid based on the corrosion amount values ​​predicted in the corrosion prediction map creation step (for example, see the description below). Figure 5 and Figure 6 ).

[0158] As described above, according to the corrosion measurement method for metallic materials using the corrosion measurement apparatus 1 of this embodiment, long-term corrosion prediction of metallic materials can be performed with high accuracy in atmospheric corrosion environments. Furthermore, the corrosion amount of metallic materials can be measured with high accuracy, allowing for the selection of the optimal metallic material with corrosion resistance appropriate to the operating environment.

[0159] Furthermore, if environmental parameters are selected as explanatory variables and formulas are created as in existing technologies, such as Patent Documents 1-5, the prediction accuracy will vary depending on the region where the corrosion amount is to be predicted, resulting in higher prediction accuracy in some regions and lower prediction accuracy in others. On the other hand, according to the corrosion prediction method for metallic materials of this embodiment, since corrosion amount data similar to the corrosion amount data of the region where the corrosion amount is to be predicted exists in the database 20, prediction can be made with high accuracy regardless of the region.

[0160] [Example]

[0161] (Example 1)

[0162] The present invention will be described in more detail below with examples. In this example, in Vietnam, the annual corrosion amount of steel under eaves (corrosion amount over one year from the initial value) is predicted, and a corrosion prediction map is created. Here, the content of this example is equivalent to the environmental mapping steps, prediction request point input steps, first similarity calculation steps, first dimension compression steps, initial corrosion prediction steps, and corrosion prediction map creation steps in the corrosion mapping method described above.

[0163] In this embodiment, annual corrosion data and annual average environmental parameters (corrosion data) of steel from 55 regions around the world, stored in a database, are used to predict annual corrosion rates. In this embodiment, temperature (°C), relative humidity (%), and air salinity (mmd / Cl) are used as environmental parameters. － The four are: (conversion), SO2 concentration (mmd (SO2 conversion)). Here, "mmd" refers to the amount of Cl collected per unit of day and per unit area. － Or the amount of SO2, expressed in mg·m －2 ·day －1 The abbreviation of "".

[0164] The environmental map for each environmental parameter was created using the method described above. In this embodiment, linear complementation was used for calculation as a method to supplement the data. Figure 4 The “Environmental Map Production Steps”). In the corrosion prediction at the grid points of the obtained environmental map, as mentioned above, when the explanatory variables are correlated with each other, it is known that the prediction accuracy deteriorates due to multicollinearity. In the LW-PLS mentioned above, since the environmental parameters used as explanatory variables are correlated, the correlation is eliminated by dimension compression to generate new parameters (latent variables). Here, LW-PLS is calculated according to the steps shown in Reference 1 above. The formula for compressing the four environmental parameters into latent variables (parameter t) can be shown in Equation (7) below.

[0165] [Mathematical Expression 7]

[0166] t=w1T+W2RH+w3Cl+w4SO2…(7)

[0167] In equation (7) above, T is temperature (°C), RH is relative humidity (%), and Cl is air salinity (mg / m³). 2 / day(=mmd)(Cl － (Conversion) SO2 is the concentration of SO2 (mg / m³) 2 / day(=mmd)(SO2 conversion)), w1~w4 are coefficients. It should be noted that in the above formula (7), only four environmental parameters are used as an example, but in practice, it is preferable to include all environmental parameters that are expected to be related to corrosion in the region where the corrosion amount is to be predicted.

[0168] In this embodiment, the prediction request point is first input into the calculation unit of the corrosion measurement device. Figure 4 The “prediction point input step” contains the service period of the steel for which the corrosion level is to be predicted and several environmental parameters representing the annual average of the service environment of the steel during that service period.

[0169] Next, the similarity ω between the environmental parameters of the predicted corrosion point to be determined and the environmental parameters i of the corrosion data referenced for predicting corrosion is calculated using the Euclidean distance shown in equation (8). i ( Figure 4 (The "first similarity calculation step"). It should be noted that the values ​​were standardized because the range of data varies depending on various environmental parameters.

[0170] [Mathematical Expression 8]

[0171]

[0172] Here, ω i For localization parameters, the environmental parameter with the lower right subscript q is the environmental parameter of the prediction request point where the corrosion amount is to be predicted, and the environmental parameter with the lower right subscript i is the environmental parameter of the corrosion amount data referenced from the database. σ is the standard deviation of the following equation (9). In addition, φ is an adjustment parameter, for example, its value is determined by appropriate adjustment based on φ = 1.

[0173] [Mathematical Expression 9]

[0174]

[0175] Next, based on the similarity ω calculated from equation (8) above... i The environmental parameters and corrosion amount of the corrosion data were analyzed according to the steps shown in Reference 1 (Chapter 2.1), with a similarity ω. i The coefficients w1 to w4 of equation (7) are determined by maximizing the inner product of the associated latent variable and the target variable (corrosion amount). Then, using the determined coefficients w1 to w4, the latent variables of each environmental parameter are calculated through equation (7). Figure 4 (The "first dimension compression step").

[0176] Next, a prediction formula for corrosion amount is constructed using local regression, as shown in equation (10). Based on equation (10), the annual corrosion amount of steel (corrosion amount over one year from the initial point) under the environmental parameters of the prediction request point is predicted. Figure 4 The "Initial Corrosion Prediction Steps").

[0177] [Mathematical Expression 10]

[0178] Y=αt…(10)

[0179] Here, Y is the predicted value of corrosion, and α is the coefficient (regression coefficient). It should be noted that the description is omitted in the above equation (10), but a constant term or multiple latent variables can also be included in the above equation (10).

[0180] Thus, in this embodiment, whenever a prediction request point for predicting corrosion is input, the similarity between the prediction request point and each corrosion data point is calculated, the coefficients of latent variables are calculated, and the prediction formula is constructed. The predicted annual corrosion values ​​for each grid point are then used to create a map, for example... Figure 5 The corrosion prediction map shown ( Figure 4 The “Corrosion Prediction Map Creation Steps” are displayed through the display unit 40.

[0181] (Example 2)

[0182] In this embodiment, in Vietnam, the long-term (fifty-year) corrosion of steel in an environment under eaves is predicted, and a corrosion prediction map is created. Here, the content of this embodiment is equivalent to all the steps of the corrosion mapping method described above.

[0183] In this embodiment, taking the environment under the eaves as the object, using data sets (corrosion data) of corrosion amount and annual average environmental parameters for 55 regions of the world over 1 year, 55 regions over 3 years, 39 regions over 5 years, 38 regions over 7 years, and 38 regions over 9 years stored in the database, the parameter A representing the corrosion amount of steel over one year from the beginning was calculated for the steel using the same method as in Example 1 (refer to the above formula (1)). Figure 4 The steps include environmental map creation, prediction request point input, first similarity calculation, first dimension compression, and initial corrosion prediction.

[0184] Next, the results of the long-term corrosion test over 7 years were weighted to calculate the parameter B, which represents the attenuation of the corrosion rate caused by the rust layer. The corrosion amount of the steel after 50 years was then predicted using the above formula (1). In the prediction of the corrosion amount after 50 years, the above formula (1) was first transformed into the following formula (11). The difference between the logarithm of the corrosion amount in the first year from the beginning and the logarithm of the corrosion amount after 50 years was calculated using the parameter B and the usage period X.

[0185] [Mathematical Expression 11]

[0186] logY-logA=BlogX…(11)

[0187] LW-PLS is used in the prediction on the left side of the above equation (11). Specifically, new environmental parameters that take into account the time factor are created by taking the product of the four environmental parameters and the logarithm of the usage period X, and these new environmental parameters are compressed into latent variables (parameter u) as shown in the following equation (12).

[0188] [Mathematical Expression 12]

[0189] u=(v1T+v2RH+v3Cl+v4SO2)logX…(12)

[0190] In this equation (12), T is the temperature (°C), RH is the relative humidity (%), and Cl is the air salinity (mg / m³). 2 / day(=mmd)(Cl － (Conversion) SO2 is the concentration of SO2 (mg / m³) 2 / day(=mmd)(SO2 conversion)), v1~v4 are coefficients.

[0191] Next, the similarity ω between the environmental parameters of the predicted corrosion point to be determined and the environmental parameters i of the corrosion data referenced for predicting corrosion is calculated using the Euclidean distance shown in equation (13). i ( Figure 4 (The "second similarity calculation step"). It should be noted that the values ​​were standardized because the range of data varies depending on various environmental parameters.

[0192] [Mathematical Expression 13]

[0193]

[0194] Here, ω iFor localization parameters, the environmental parameter with the lower right subscript q is the environmental parameter of the prediction request point where the corrosion amount is to be predicted, and the environmental parameter with the lower right subscript i is the environmental parameter of the corrosion amount data referenced from the database. σ is the standard deviation of the following equation (14). In addition, φ is an adjustment parameter, which is determined by adjusting appropriately, for example, with φ = 1 as a reference.

[0195] [Mathematical Expression 14]

[0196]

[0197] Next, based on the similarity ω calculated from equation (13) above... i The environmental parameters and corrosion amount of the corrosion data were analyzed according to the steps shown in Reference 1 (Chapter 2.1), with a similarity ω. i The coefficients v1 to v4 of equation (12) are determined by maximizing the inner product of the associated latent variable and the target variable (corrosion amount). Then, using the determined coefficients v1 to v4, the latent variables of each environmental parameter are calculated through equation (7). Figure 4 (The "second dimension compression step").

[0198] Next, the prediction formula for corrosion amount shown in Equation (15) is constructed by local regression, and the difference in logarithms of corrosion amount in the environmental parameters to be predicted is calculated.

[0199] [Mathematical Expression 15]

[0200] logY-logA=βu…(15)

[0201] Here, in equation (15) above, β is the coefficient (regression coefficient). It should be noted that the term is omitted in equation (15) above, but a constant term or multiple latent variables can also be included in equation (15) above.

[0202] Thus, in this example of the invention, whenever a prediction request point for predicting corrosion amount is input, the similarity between the prediction request point and each corrosion amount data is calculated, the coefficients of the latent variables are calculated, and the prediction formula is constructed.

[0203] Next, based on equations (12) and (15) above, the parameter B of equation (11) above is calculated using equation (16) below. Figure 4 The "attenuation prediction step").

[0204] [Mathematical Expression 16]

[0205]

[0206] Next, using the calculated parameters A and B, the corrosion amount Y after 9 years is calculated using the above formula (1). Figure 4 (The long-term corrosion prediction steps). The predicted long-term corrosion values ​​for each grid point are plotted on a map and used as an example. Figure 6 The corrosion prediction map shown ( Figure 4 The “Corrosion Prediction Map Creation Steps” are displayed through the display unit 40.

[0207] The corrosion measurement method, metal material selection method, and metal material corrosion measurement device of the present invention have been specifically described above through embodiments and methods for carrying out the invention. However, the spirit of the present invention is not limited to the above description and should be interpreted broadly based on the claims. Furthermore, various modifications and alterations made based on the above description are also included in the spirit of the present invention.

[0208] In the above-described embodiments, the corrosion amount of the metallic material over a period of one year from the beginning (parameter A) and the parameter representing the decay of the corrosion rate of the metallic material (parameter B) are predicted separately. The long-term corrosion amount is predicted based on the corrosion amount over a period of one year from the beginning, but the basis for predicting the long-term corrosion amount is not limited to the corrosion amount over a period of one year from the beginning.

[0209] That is, in the initial corrosion prediction step, the corrosion amount of the metal material during any predetermined period (initial period) can be predicted, and in the long-term corrosion prediction step, the long-term corrosion amount can be predicted based on the corrosion amount during the predetermined period.

[0210] For example, given as the initial corrosion amount and the corrosion amount after 1.5 years as A', it is assumed that the prediction formula for the corrosion amount after X years can be extended from the above equation (1) as described in the following equation (17).

[0211] [Mathematical Expression 17]

[0212]

[0213] If we generalize this, and set the corrosion amount for an initial period of X0 years as A', and set the attenuation parameter based on X0 years as B', we can obtain the following equation (18). By using equation (18) as the corrosion amount based on the period of X0 years, we can calculate the corrosion amount for the period X>X0.

[0214] [Mathematical Expression 18]

[0215] Y = A′X′ B =A′(X / X0) B′ …(18)

[0216] The corrosion amount A' and attenuation parameter B' of the metallic material in any initial period are predicted separately, as shown in equation (18) above. The long-term corrosion amount after the initial period can be predicted by exponentiating the attenuation parameter B' with respect to the number of years X' after the initial period. However, the initial corrosion amount A in equation (1) above is based on the corrosion amount over a one-year period. Therefore, for the period X0 in equation (18) above, we do not assume a large deviation from one year, and consider a practical range of about six months to two years.

[0217] [Explanation of reference numerals in the attached figures]

[0218] 1 Corrosion Measurement Device

[0219] 10 Input Section

[0220] 20 databases

[0221] 30 Computational Unit

[0222] 31 Map Production Department

[0223] 32 Similarity Calculation Department

[0224] 33-dimensional compression section

[0225] 34 Corrosion Prediction Department

[0226] 40 Display Section

Claims

1. A method for mapping the corrosion of metallic materials, which uses corrosion data to predict the corrosion of metallic materials and create a corrosion prediction map, wherein the corrosion data includes: the service period of the metallic material; multiple environmental parameters known on a map representing the service environment of the metallic material during the service period and on a map representing the region where the metallic material is used; the location coordinates of the environmental parameters on the map; and the topographic data of the map. And the corrosion amount of the metallic material during the period of use, wherein the method for measuring the corrosion amount of the metallic material is characterized by comprising: The environmental map creation steps involve creating an environmental map with the environmental parameters marked on each environmental parameter, based on the multiple environmental parameters, the terrain data of the map, and the location coordinates of the environmental parameters on the map, at arbitrary grid intervals. The prediction request point input step involves inputting a prediction request point, which includes the usage period of the metallic material used to create the corrosion prediction map and multiple environmental parameters from the corrosion data. The similarity calculation step calculates the similarity between multiple environmental parameters in the corrosion data and multiple environmental parameters in the prediction request point; The dimension compression step considers the similarity to compress the dimension of multiple environmental parameters in the corrosion data into latent variables; The corrosion prediction step, based on a prediction formula constructed using the latent variables and the similarity, predicts the corrosion amount of the metallic material at the prediction request point on the grid; and The steps for creating a corrosion prediction map involve color-coding the predicted corrosion amounts at the requested points of the grid onto the map, thereby creating the corrosion prediction map. The environmental map creation steps include: The first step is to input the multiple environmental parameters into the location coordinates on the map; The second step is to correct the environmental map to an environmental map with an altitude of 0m based on the terrain data of the map. The third step is to supplement environmental parameters among the multiple environmental parameters in the environmental map at an altitude of 0m, based on the characteristics of each environmental parameter. The fourth step involves correcting the environmental map (which had been supplemented with environmental parameters and had an elevation of 0m) back to its original elevation, based on the terrain data of the map. The environmental parameters include at least one of the following: temperature, relative humidity, absolute humidity, wetting time, and rainfall; and air salinity, SO2, etc. X Concentration and NO X At least one of the concentrations.

2. The method for measuring the corrosion of metallic materials according to claim 1, characterized in that, The environmental parameters include air salinity. In the third step, when replenishing the air salinity of the grid near the coastline, the air salinity is replenished in a manner that does not exceed a preset upper limit.

3. The method for measuring the corrosion of metallic materials according to claim 1 or 2, characterized in that, The corrosion prediction step includes: The initial corrosion prediction step predicts a first parameter representing the corrosion amount of the metallic material over a specified period; The attenuation prediction step predicts a second parameter representing the attenuation of the corrosion rate of the metallic material; and The long-term corrosion prediction step predicts the corrosion amount of the metal material during a period longer than the specified period, based on the service life of the metal material, the first parameter, and the second parameter.

4. The method for measuring the corrosion of metallic materials according to claim 3, characterized in that, In the initial corrosion prediction step, the corrosion amount of the metallic material during a specified period under the environmental parameters of the prediction request point is predicted based on the prediction formula constructed using the latent variables and the similarity.

5. The method for measuring the corrosion of metallic materials according to claim 3, characterized in that, In the attenuation prediction step, the second parameter is predicted based on the usage period of the metal material in which the corrosion prediction map is made, multiple environmental parameters in the corrosion data, and the similarity.

6. The method for measuring the corrosion of metallic materials according to any one of claims 1 to 3, characterized in that, The metal material is steel.

7. The method for measuring the corrosion of metallic materials according to any one of claims 1 to 6, characterized in that, The corrosion prediction map is a map created by coloring and marking each grid based on the corrosion value predicted in the corrosion prediction map creation step.

8. A method for selecting metallic materials, characterized in that, Using the corrosion measurement method for metallic materials according to any one of claims 1 to 7, a metallic material corresponding to the usage environment is selected.

9. A device for measuring the corrosion of metallic materials, characterized in that, include: The database stores corrosion data, which includes: the service period of the metal material, multiple environmental parameters known on a map representing the region where the metal material is used during the service period, the location coordinates of the environmental parameters on the map, the topographic data of the map, and the corrosion amount of the metal material during the service period. An environmental map production department creates an environmental map with the environmental parameters marked on each environmental parameter, based on the multiple environmental parameters, the terrain data of the map, and the location coordinates of the environmental parameters on the map, at arbitrary grid intervals. The input unit inputs prediction request points, which include: the usage period of the metallic material for creating the corrosion prediction map and multiple environmental parameters in the corrosion data; The similarity calculation unit calculates the similarity between multiple environmental parameters in the corrosion data and multiple environmental parameters at the prediction request point; The dimension compression unit takes into account the similarity to compress the dimension of multiple environmental parameters in the corrosion data into latent variables; A corrosion prediction unit, based on a prediction formula constructed using the latent variables and the similarity, predicts the corrosion amount of the metallic material at the prediction request point of the grid; and The corrosion prediction map creation department creates the corrosion prediction map by color-coding the predicted corrosion amounts at the requested points of the grid on a map. The environmental map production department. Input the multiple environmental parameters into the location coordinates on the map; Based on the terrain data of the map, the environmental map is corrected to an environmental map with an altitude of 0m; In the environmental map at an altitude of 0m, environmental parameters are supplemented among the multiple environmental parameters according to the characteristics of each environmental parameter; Based on the terrain data of the map, the environmental map with an elevation of 0m, supplemented with environmental parameters, was corrected to the original elevation. The environmental parameters include at least one of the following: temperature, relative humidity, absolute humidity, wetting time, and rainfall; and air salinity, SO2, etc. X Concentration and NO X At least one of the concentrations.

Images