1. A method for predicting the corrosion life of weather-resistant bridge steel is characterized by comprising the steps of adopting corrosion data of a weather-resistant bridge steel corrosion test as an original data column, carrying out initialization processing on the original data column by a grey theory method, establishing a corrosion data prediction model, and calculating the corrosion rate of the weather-resistant bridge steel at a specific time point by using the corrosion data prediction model;

wherein the corrosion data prediction model satisfies the following mathematical relationship: x¹(m)＝[X⁰(1)-b/a]e^-a(m-1)+b/a；

In the mathematical relation, a and b are specific model parameters;

X¹(m) predicted corrosion rate value in year m; m is a positive integer;

X⁰(1) is the actual corrosion rate value in the first year.

2. The method for predicting the corrosion life of the weather-resistant bridge steel according to claim 1, wherein a and b in the mathematical relationship can be obtained by a least square mathematical formula.

3. The method for predicting the corrosion life of weather-resistant bridge steel as claimed in claim 1, wherein the raw data column is obtained by using an on-site coupon corrosion test, and for the area where timely on-site coupon cannot be carried out, the X is calculated by measuring the comprehensive environmental corrosivity factor Rn of the weather-resistant bridge steel in the service area⁰(1)；

And said X⁰(1) And the Rn satisfies the following mathematical relation:

X⁰(1)＝AlgRn+B；

in the above mathematical relation, A and B are the specific coefficients of the weather-resistant bridge steel.

4. The method of predicting the corrosion life of a weathering bridge steel as claimed in claim 3, wherein A and B in the mathematical relationship are calculated by measuring the corrosion rate of the weathering bridge steel in a known area.

5. The method for predicting the corrosion life of the weather-resistant bridge steel according to claim 3 or 4, wherein the environmental corrosivity comprehensive factor Rn satisfies the following mathematical relation:

Rn＝R1+R2；

r1 is a corrosion factor and satisfies the following mathematical relation:

R1＝[P_SO2+P_Cl ^-]^0.5wherein P is_SO2Is mean of SO per year₂Deposition rate, P_Cl ^-Is mean Cl of year^-The deposition rate;

the R2 is an environmental factor and satisfies the following mathematical relation:

r2 τ/(T × T), where τ is the time at relative humidity RH > 80% in h;

t is the annual average temperature in units of ℃;

t is the annual illumination time in h.

Background

The weather-resistant bridge steel is exposed in the natural environment for a long time, and corrosion damage of different degrees can be generated on the steel structural member due to corrosion of different degrees, so that the safety reliability and the durability of the weather-resistant bridge steel are influenced. Therefore, the prediction research of the corrosion full life of the steel structural part must be solved with great effort. The method has very important significance for protecting and reasonably utilizing various resources, building a conservation-oriented society, realizing sustainable development, strengthening corrosion protection of materials and prolonging the safe service life of a steel structure.

Corrosion is a slow changing process caused by material and environmental effects, often in years or even longer, and the randomness of corrosion phenomena, so that corrosion research based on experimental observation takes a long time and a great deal of repeated labor. The material corrosion prediction uses short-term corrosion data to predict long-term corrosion behavior, uses local sample characteristics to predict general corrosion rules of large samples, uses simple-condition indoor corrosion data to predict actual environmental material corrosion behavior and the like, and the two main pillars of corrosion research are formed by the material corrosion prediction and corrosion tests. In addition, metal corrosion is affected by a variety of factors, such as moisture content, resistivity, temperature, humidity, time, and the like. The factors are mutually influenced to form an abnormal complex corrosion system, and the conventional method for predicting the corrosion rate by linear fitting and extrapolation through experimental data is difficult to directly establish a clear functional relation, so that the prediction accuracy and the practicability are not ideal.

Therefore, for the corrosion prediction problem with ambiguity and complexity, the classical prediction method is difficult to work, a proper method is found to establish a corrosion rate prediction model, and the prediction of the corrosion life of the weather-resistant bridge steel is very important.

Disclosure of Invention

In order to solve the technical problems, the invention provides a method for predicting the corrosion life of weather-resistant bridge steel, wherein a corrosion prediction model is established by a proper mathematical method, and the prediction model can accurately predict the corrosion life of the weather-resistant bridge steel in a long period under the actual application environment.

In order to achieve the purpose, the invention discloses a method for predicting the corrosion life of weather-resistant bridge steel, which comprises the steps of adopting corrosion data of a weather-resistant bridge steel corrosion test as an original data column, carrying out initialization processing on the original data column by a grey theory method, establishing a corrosion data prediction model, and calculating the corrosion rate of the weather-resistant bridge steel at a specific time point by using the corrosion data prediction model;

wherein the corrosion data prediction model satisfies the following mathematical relationship: x¹(m)＝[X⁰(1)-b/a]e-^a(m-¹⁾+b/a；

In the mathematical relation, a and b are specific model parameters;

X¹(m) predicted corrosion rate value in year m; m is a positive integer;

X⁰(1) is the actual corrosion rate value in the first year.

Further, a and b in the mathematical relation can be calculated by a least square mathematical formula.

Further, the original data is listed as obtained by adopting an on-site coupon corrosion test, and for the area where the on-site coupon cannot be carried out in time, the X is calculated by measuring the environmental corrosivity comprehensive factor Rn of the weather-resistant bridge steel in the service area⁰(1)；

And said X⁰(1) And the Rn satisfies the following mathematical relation:

X⁰(1)＝AlgRn+B；

in the above mathematical relation, A and B are the specific coefficients of the weather-resistant bridge steel.

Further, A and B in the mathematical relationship may be calculated by measuring the corrosion rate of the weathering bridge steel in known areas.

Further, the environmental corrosivity comprehensive factor Rn satisfies the following mathematical relation:

Rn＝R1+R2；

r1 is a corrosion factor and satisfies the following mathematical relation:

R1＝[P_SO2+P_Cl-]^0.5wherein P is_SO2Is mean of SO per year₂Deposition rate, P_Cl-is the annual average Cl-deposition rate;

the R2 is an environmental factor and satisfies the following mathematical relation:

r2 τ/(T × T), where τ is the time at relative humidity RH > 80% in h;

t is the annual average temperature in units of ℃;

t is the annual illumination time in h.

The beneficial effects of the invention are mainly embodied in the following aspects:

the method designed by the invention can find out the intrinsic rule from the disordered discrete phenomenon and scientifically predict the service life of the material, and on one hand, the method has higher accuracy and precision, and on the other hand, the method provides scientific basis for overhauling and maintenance by predicting the corrosion life of the metal material.

Detailed Description

The corrosion test of the weather-resistant bridge steel in the practical application environment is the most common test method for researching metal corrosion, the obtained corrosion data is the most accurate, and the corrosion behavior characteristics of the practical environment can be reflected most. However, depending on the test cycle conditions, long-cycle test data may not be obtained in a short time or a hanging test may not be performed in time in some areas. Therefore, a method for predicting the corrosion life of the weather-resistant bridge steel is urgently needed, and the corrosion prediction model is established by using the existing corrosion test data and the measurement of the service environment parameters through a proper mathematical method, so that the corrosion life of the metal material is well predicted.

Specifically, for the area where the field coupon corrosion experiment can be carried out, a corrosion life prediction model is constructed, and the process is as follows:

(1) field coupon corrosion experiments: the weather-resistant bridge steel is used as a test material, and an atmosphere exposure test is carried out according to GB/T14165-general requirements of atmospheric corrosion test field tests of metals and alloys in different application environments, specifically, the test material is arranged with the front facing south and forms an angle of 45 degrees with the ground for testing. After the samples exposed in different periods are taken back, derusting is carried out by using a derusting liquid, the samples are washed by distilled water after derusting is finished, the samples are dried by cold air and then weighed, and the corrosion weight loss amount is calculated, so that the corrosion rate is obtained;

(2) constructing a corrosion data prediction model:

the corrosion rates measured in the actual tests were listed as raw data:

X⁰＝{X⁰(1)，X⁰(2)，X⁰(3)，…，X⁰(n) } formula I; wherein, X is⁰(n) actual corrosion rate in year n;

the formula I corresponds to different time sequences t ═ t₁，t₂，t₃…t_nFormula II; wherein, t_nThe nth year;

the new sequence was generated by the grey theory method: x¹(m)＝{X¹(1)，X¹(2)，X¹(3)，…，X¹(n) } formula III; wherein, X¹(m) predicted corrosion rate value in year m; m is a positive integer;

x of the formula III¹(m) the differential equation established is: dX¹/dt+aX¹B is formula IV;

in formula IV, a and b are to-be-determined model parameters, which can be calculated by using a least square mathematical formula: [ a, b ]]^T＝(B^TB)-¹B^TY is formula V;

thereby obtaining X¹(m) further satisfies the following mathematical relationship:

X¹(m)＝[X⁰(1)-b/a]e-^a(m-¹⁾+ b/a is of formula VI;

wherein X in the formula VI⁰(1) For actual rotting in the first yearThe etch rate value.

And for some areas, the coupon test cannot be carried out in time, and the X is calculated by measuring the environmental corrosivity comprehensive factor Rn of the weather-resistant bridge steel in the service area⁰(1)；

And said X⁰(1) And the Rn satisfies the following mathematical relation:

X⁰(1) AlgRn + B formula VII;

in the mathematical relation, A and B are the specific coefficients of the weather-resistant bridge steel; and a and B can be calculated by measuring the corrosion rate of the weathering bridge steel in known areas.

The environment corrosivity comprehensive factor Rn satisfies the following mathematical relation:

rn ═ R1+ R2 formula VIII;

r1 is a corrosion factor and satisfies the following mathematical relation:

R1＝[P_SO2+P_Cl-]^0.5formula IX wherein P_SO2Is mean of SO per year₂Deposition rate, P_Cl-is the annual average Cl-deposition rate;

the R2 is an environmental factor and satisfies the following mathematical relation:

r2 ═ τ/(T × T) formula X, where τ is the time at relative humidity RH > 80%, in units of h;

t is the annual average temperature in units of ℃;

t is the annual illumination time in h.

Further, for the mathematical relationship satisfied by the corrosion data prediction model: x¹(m)＝[X⁰(1)-b/a]e-^a(m-¹⁾The precision of + b/a is further checked as follows:

the residual error test is as follows:

in the formula XI, X⁰(m) is the original sequence of numbers,data column predicted for model, S₁，S₂Respectively, the mathematical statistics of the variance values are shown, and the posterior difference is C ═ S₂/S₁The probability P of the small error index of the posterior error index is defined asAnd judging the accuracy of the prediction model according to C and P, wherein the smaller the prediction accuracy grade is, the better the prediction effect is.

In order to better explain the invention, the following further illustrate the main content of the invention in connection with specific examples, but the content of the invention is not limited to the following examples.

Example 1

The weather-resistant bridge steel corrosion test is carried out for 10 years in Wuhan region. The test steel is 09CuPTi weathering steel, and the chemical composition is shown in Table 1. The test method is carried out according to GB/T14165-general requirements of metal and alloy atmospheric corrosion test field tests, the rust removing liquid is used for removing rust after the tests are carried out, the rust removing liquid is washed by distilled water after the rust removing is finished, the weight is weighed after the distilled water is dried by cold air, and the corrosion weight loss is calculated and is shown in Table 2.

TABLE 1 chemical composition (wt%)

TABLE 2 weathering steel corrosion test data

Let X⁰(t) is the corrosion rate of the weathering steel, the original sequence is:

X0(t)＝{52,28.45,16.69,10.67}。

carrying out initial value processing on the original data to obtain X¹(t) {52,80.45,97.14,107.81}, and substituting [ a, b }]^T＝(B^TB)-¹B^TThe coefficients of the calculated model Y are:

a＝0.4935，b＝60.9641。

then the equations satisfied by the model are predicted to be: x¹(m)＝-71.5259e-0.4935^(m-¹⁾+ 123.5259; further, predicted corrosion rate values at different times can be derived, see table 3.

TABLE 3 prediction data parameters

Precision analysis of the prediction, X⁰(m) is the original sequence of numbers,for the data column of the model prediction,. epsilon. (m) is the residual number column, S₁，S₂Respectively, the mathematical statistics variance values of the original data column and the residual data column. Calculating to obtain:

mean of raw dataStandard deviation S₁＝15.81；

Mean of residual dataStandard deviation S₂＝0.3363；

The index C is equal to S₂/S₁0.021 < 0.35, it can be seen that the prediction accuracy is good. According to small error index probabilityP is calculated to be much larger than 0.95, thus indicating that the prediction accuracy is good.

For the areas where the film hanging cannot be timely carried out, according to the existing test data, corrosion data of Q235 and W450QN steel in the first year are selected to be compared with the field film hanging data of the China society for corrosion and protection in the western sand ocean environment, and the reliability of the estimated model is verified, as shown in the following table 4. The verification result shows that the corrosion prediction model has certain reliability in predicting the corrosion rate of the steel grade in the relevant sea area. And further correcting the atmospheric corrosivity comprehensive factor N by different weighting coefficients at the later stage so as to realize more accurate prediction of the corrosion of the metal material.

TABLE 4 Corrosion prediction model verification

Steel grade	Actual data (mm/a)	Prediction data (mm/a)	Error (%)
				Q235	0.087	0.070	19
W450QN	0.063	0.0464	26

As can be seen from the table 4, the estimation model designed by the invention can better estimate the corrosion rate of the steel grade in the relevant sea area, the error is in a controllable range, and the subsequent predicted value is more accurate by further correcting the weighting coefficient of the corrosion rate increased by the corrosion years.

The above examples are merely preferred examples and are not intended to limit the embodiments of the present invention. In addition to the above embodiments, the present invention has other embodiments. All technical solutions formed by adopting equivalent substitutions or equivalent transformations fall within the protection scope of the claims of the present invention.