JP2020060429A - Degradation prediction method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a deterioration prediction method capable of obtaining a prediction result in consideration of uncertainty of prediction while minimizing inspection items at one inspection. A method for predicting deterioration on the surface of a reinforced concrete structure. Then, steps S1 and S2 for acquiring information on the peeling off and cracking points on the surface of the reinforced concrete structure and information on the cover depth of the reinforcing bar, and step S3 for estimating the reinforcing bar corrosion rate by inverse analysis from the acquired information. And a step S4 of predicting the peeling-off area of the surface at a time arbitrarily set based on the estimated reinforcing bar corrosion rate. [Selection diagram] Figure 1

Description

  The present invention relates to a method of predicting deterioration on the surface of a reinforced concrete structure.

  Reinforced concrete structures constructed of reinforced concrete such as bridges, viaducts, buildings, etc., will corrode, crack and peel due to the age of the structure, the environment in which the structure is used and the construction conditions during construction. Therefore, regular inspections are conducted for maintenance.

  In addition, as disclosed in Non-Patent Documents 1 and 2, as one of suitable correction methods for predicting deformation of a reinforced concrete structure based on regular inspections, based on peeling / peeling information obtained visually. A deterioration prediction method has been proposed in which peeling and peeling are predicted using the estimated reinforcing bar corrosion rate. In this deterioration prediction method, in addition to visual inspection, the cover depth, chloride ion concentration, and neutralization depth of the reinforcing bar are measured, and the deterioration prediction is performed by using this information as well. .

Masamichi Sokabe, 3 others, "Survey of RC railroad crossing on railway viaduct and its deterioration prediction", Concrete Engineering, Vol.47, No.8, 2009, pp.16-24 Shuntaro Todoroki, 3 others, "A Study on Evaluation of Reinforcement Corrosion Rate Based on Visual Deformation of Railway RC Structures", Japan Society of Materials, Concrete Structure Repair, Reinforcement Upgrade Papers, Vol.15, 2015 Pp.41-46

  However, measuring the chloride ion concentration and the depth of neutralization requires a corresponding inspection cost and labor, which increases the inspection cost per inspection and inspection time, increasing maintenance costs. Cause In addition, if the cost is reduced, the number of inspections may be reduced, and repair may not be performed at an appropriate time.

  Furthermore, the deterioration phenomenon itself varies for each structure, its member or surface, and the places where inspection can be performed are limited, and the inspection information also has the property of depending on the inspector. These uncertainties are not explicitly taken into account.

  Therefore, it is an object of the present invention to provide a deterioration prediction method capable of obtaining a prediction result in consideration of uncertainty of prediction while minimizing the inspection items in one inspection.

  In order to achieve the above-mentioned object, the deterioration prediction method of the present invention is a deterioration prediction method on the surface of a reinforced concrete structure, wherein the peeling-off and cracking information on the surface of the reinforced concrete structure and the covering depth of the reinforcing bar are provided. And a step of estimating a reinforcing bar corrosion rate by an inverse analysis from the acquired information, and a step of predicting the peeling-off area of the surface at any time based on the estimated reinforcing bar corrosion rate It is characterized by having and.

  Here, it is preferable that the estimation of the corrosion rate of the reinforcing bar is performed based on Bayes' theorem. Further, the information on the peeled off portions and cracked portions on the surface of the reinforced concrete structure can be obtained by visual inspection. Further, the information on the cover depth of the reinforcing bar can be acquired as the measurement result by the nondestructive inspection device by the electromagnetic induction method.

  The deterioration prediction method of the present invention thus configured estimates the corrosion rate of the reinforcing bar by inverse analysis from the information on the peeling and peeling point and the cracked point on the surface of the reinforced concrete structure and the information on the cover depth of the reinforcing bar. Then, based on the estimated corrosion rate of the reinforcing bar, the peeling and peeling area at an arbitrary time is predicted.

  By just using the results of the limited inspection items such as the information that appears on the surface and the information on the depth of cover, it is possible to estimate the corrosion rate of the reinforcing bar by back analysis, which is suitable for practical use that considers the uncertainty of prediction. It is possible to obtain the predicted result.

It is a flowchart explaining the deterioration prediction method of this Embodiment. It is a figure which explains notionally the relationship between the deformation | transformation process in the surface of a reinforced concrete structure, and a corrosion rate of a reinforcing bar. It is a figure explaining the peeling off simulation result by a forward analysis. It is a figure explaining the conditions and parameter estimation result of reverse analysis, (a) is a condition list, (b) is a bar graph which showed the ratio and standard deviation of the estimated value and correct value by numerical experiment. It is a figure explaining the prediction precision of the peeling exfoliation area rate predicted by the numerical experiment, (a) is a result of Case1, (b) is a result of Case2, (c) is a result of Case3, (d) is Case4. The result of is shown. It is sectional drawing explaining the structure of the target structure which becomes an example of application to an actual structure. It is a figure explaining the example of determination of the deformation by the mesh division method and visual inspection. It is explanatory drawing which showed an example of the measurement result of the cover thickness by a reinforcing bar probe. It is a figure explaining the relationship between a cover thickness and a deformation. In the figure which illustrated the posterior distribution obtained by parameter estimation (Case 2), (a) shows the result of p ₁ , (b) shows the result of p ₂ , and (c) shows the result of p ₃ . It is a bar graph showing the ratio and the standard deviation of the estimated value of the back analysis and the correct value when applied to the actual structure. It is a figure explaining the prediction accuracy of the exfoliation exfoliation area rate predicted when applied to an actual structure, (a) is a result of Case1, (b) is a result of Case2, (c) is a result of Case3, (D) shows the results of Case 4. It is a figure explaining the measurement result and the prediction result (Case4) of the deformation distribution when applied to an actual structure, (a) is a measurement result (service 45 years), (b) is a prediction result (service 45) (Year), (c) shows the prediction result (75 years in service), and (d) shows the prediction result (100 years in service).

Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a flowchart explaining the deterioration prediction method of the present embodiment.
Structures to which the method for predicting deterioration on the surface of a reinforced concrete structure according to the present embodiment is applied are civil structures such as bridges and viaducts constructed of reinforced concrete, and building structures such as buildings. Surfaces such as walls, columns, and girders of these reinforced concrete structures are to be inspected.

  The deformation that appears on the surface of the reinforced concrete structure can be observed by visual inspection. The visual survey is a survey in which the inspector visually obtains observation information such as the position and range of deformation such as peeling and peeling spots and cracks.

  In the past, during inspection, not only visual inspection but also measurement of fog depth and neutralization depth using non-destructive inspection equipment were performed as needed. Then, based on the inspection information thus obtained and the experience of a skilled engineer, the corrosion rate and corrosion degree of the reinforcing bar have been estimated. For example, in "Maintenance Standards for Railway Structures, Commentary (Structures) Concrete Structures" (hereinafter referred to as "Maintenance Standards"), a deformation prediction model using corrosion of reinforcing bars as an index is proposed. .

  If the corrosion rate of the reinforcing bar can be estimated based on the inspection information of the reinforced concrete structure at the time of inspection (current time) and the time elapsed after construction (years), it can be linked to future prediction. In short, it becomes possible to predict the degree of deterioration (peeling and peeling area) of the reinforced concrete structure at any future time from the estimation result of the corrosion rate of the reinforcing bar.

  Here, from a general point of view, it is considered that highly-accurate deterioration prediction can be performed by making judgment by an experienced technician using a lot of inspection information. However, due to the recent shortage of maintenance engineers and technical transfer problems, it is easily predicted that the system implicitly dependent on skilled engineers will reach its limit. In order to compensate for such a shortage of engineers, labor saving in inspection work is required, and it is required to reduce the number of measurement items at the time of inspection while ensuring the prediction accuracy of the peeling and peeling area.

  Therefore, the deterioration prediction method of the present embodiment is a method that can be utilized in the maintenance cycle in practice. In short, by estimating the reinforcing bar corrosion rate in the deformation prediction model shown in the "maintenance management standard" from the exfoliation and peeling area at the time of inspection, the methodologies for predicting the exfoliation and exfoliation area at a future time require less measurement items and times. A method for efficiently predicting the peeled-off area will be described.

  Therefore, first, the examination results of evaluating the influence of the presence or absence of various measurement data necessary for the deformation prediction model on the prediction accuracy by considering the uncertainty using Bayesian estimation will be described. That is, by using the aggregated information of peeling and peeling area, even if the deformation prediction results in individual evaluation units (mesh) described later do not always match, the central limit theorem due to aggregation works It becomes possible to predict the uncertainty efficiently.

Fig. 2 shows the outline of the reinforcing bar corrosion model. The reinforcing bar corrosion model is based on the reinforcing bar corrosion rate v (= dr / dt; v ₁ -v ₅ ) in each deformation process (stage 1 to stage ₅ ), the neutralization residual c _rlim which is the limit value, and the reinforcing _bar corrosion depth Δr ( The parameters are Δr _cr , Δr _sp , and Δr _ls ). Here, the reinforcing bar corrosion depth Δr is the amount of decrease when the periphery of the reinforcing bar is uniformly corroded in the depth direction, and the reinforcing bar corrosion rate v is the amount of decrease per year.

Stages 1 to 5 correspond to the latent period, the evolution period, the early acceleration period, the late acceleration period, and the deterioration period, which are defined by the "maintenance management standard", respectively. The rate after rate Corrosion rate v ₁ is caused by the internal salt damage, the reinforcing steel corrosion rate v ₂ is Corrosion concentrated salinity with Carbonation was accelerated, Corrosion rate v ₃ is cracked or floating The speed after the corrosion of the reinforcing bar is accelerated due to the progress of moisture and local neutralization along with the generation, and the corrosion rates v ₄ and v ₅ of the reinforcing bar are the speed of the bare steel after peeling and peeling.

Particularly in an actual structure, the reinforcing bar corrosion rate v is affected by the dry and wet conditions of the member such as the temperature and the surface water content, and has a large variation and is not sufficiently quantified. Therefore, the “maintenance management standard” presupposes that the periodic inspection is appropriately modified according to the environmental conditions of the structure. Based on these reinforcing bar corrosion rates v, the reinforcing bar corrosion depth Δr is obtained by integrating the reinforcing bar corrosion rate v of each process in the time domain, and the estimated deformation S _{i, T} ^sim is the reinforcing bar corrosion depth Δr _cr and It becomes an integer value from 1 to 5 depending on the relationship with the limit reinforcing bar corrosion depth Δr _ls . Here, i represents each mesh and T represents aging.

  By the way, in the deterioration prediction method of the present embodiment, since the rebar corrosion rate is directly back-analyzed as described later, the rebar corrosion rate used in the above "maintenance standard" is neutralized, We will also investigate a reinforcing bar corrosion model that changes deterioration factors such as salt damage and complex deterioration depending on the neutralization residue. Below, the analysis result by this model is demonstrated as a result of the forward analysis for comparing with the inverse analysis result.

  FIG. 3 shows the peeling and peeling simulation result by the forward analysis. In this figure, the proportion of each transformation process is shown, and the proportion of high stages increases depending on the age, and in 20 years, the peeling and peeling point is about 10%, the cracked point is about 20%, It can be seen that 80% of the peeled and peeled parts and about 10% of the cracked parts are 100 years old.

Since the soundness of a structure is evaluated by the peeling and peeling area, when a concrete surface of 10 m ² is targeted, it will exceed one threshold in about 20 years. Therefore, the results of 20 years from now will be used for the subsequent back analysis. According to visual observation, there is no observational information in stages 1 and 2, and crack distribution is obtained in stage 3 and peeling-off distribution is obtained in stages 4 and 5, so about 20% of surface deformation is observed by 20 years. It can be inferred that it is in a state.

  Next, the inverse analysis method of this embodiment will be described. In the deterioration prediction, model parameters are identified based on the observation data and the subsequent phenomena are predicted. At this time, the Bayesian theory that can take into account the uncertainty is adopted, and the Marcov Chain Monte Carlo method (Marcov chain Monte Carlo method; hereinafter referred to as “MCMC method”) that can efficiently identify multiple parameters is applied.

  In general, it is not easy to directly measure the corrosion rate and corrosion rate of reinforcing bars in a concrete structure in service by field survey. Therefore, based on the deformations obtained by visual inspection such as cracks, peeling and peeling, the rebar corrosion rate is estimated by inverse analysis. Bayes' theorem is used for this estimation.

  According to Bayes' theorem, the prior probability of the parameter vector θ is π (θ), and the simultaneous occurrence probability π (θ | S) of the parameter vector θ when the observation information S is obtained is calculated.

Here, S is the ratio of observed deformation Si, L (S | θ) is the likelihood function, and Θ is the parameter space composed of θ.

Using the estimated deformation S _{i, T} ^sim estimated from the above model and the observed deformation S _{i, T} ^obs visually observed, the likelihood function L (S _{i, T} ^sim | θ) is It is expressed by the following equation.

Here, θ is a parameter vector, n _s is the number of samples (the number of meshes), and σ ² is the variance of the modeling error.

Note that the estimated transformation S _{i, T} ^sim and the observed transformation S _{i, T} ^obs are integer values from 1 to 5 and are discrete variables, so continuous variables are targeted for more accurate modeling. It can be said that the setting of the probability distribution for discrete variables such as Poisson distribution is more preferable than the above equation (Equation 2). However, the discrete variable distribution is not recognized as a familiar distribution that can be easily used by maintenance practitioners.Although recognizing that modeling errors occur, prioritizing applicability in practice. The above formula was adopted. As will be described in detail later, the reinforcing bar corrosion rate can be estimated with a constant accuracy even by modeling with the above equation (Equation 2).

Then, by logarithmizing the above equation (Equation 2), the logarithmic likelihood function F (S _{i, T} ^sim | θ) shown below is obtained.

Then, a parameter vector θ that minimizes the log-likelihood function F (S _{i, T} ^sim | θ) in the above equation (Equation 3) is estimated based on Bayes' theorem. The MCMC method using the Metropolis Hastings (MH) method was used for the estimation. In the MH method, random numbers that follow a Markov process are sampled in the parameter space Θ, and the acceptance decision from the maximum likelihood criterion is repeated at each step j. That is, θ ′ is sampled from the prior distribution π (θ), and θ ^{j + 1} is determined by adopting or rejecting the sampling result using the adoption probability shown in the following equation.

Here, R is a uniform random number of [0,1]. The series of samples obtained can be approximated to the posterior distribution after a certain period of time (burn-in period) in the initial stage. In the present embodiment, after the burn-in of 5000 steps, 20000 steps are used as the posterior distribution.

Next, the estimation of the reinforcing bar corrosion rates v ₁ , v ₂ , v ₃ will be described. Here, two reinforcing bar corrosion models, a quasi-standard model (Equation 5) and a direct model (Equation 6), were used for the study.

The quasi-standard model (Equation 5) is obtained by multiplying the rebar corrosion rates v ₁ ^std , v ₂ ^std of stage 1 and stage 2 defined in the above “maintenance standard” by a constant magnification χ, and the parameter vector θ is θ = ( χ). The prior distribution of χ was a normal distribution with a mean value of 1 and a standard deviation of 10. Then, the uncertainty of χ is evaluated by the inverse analysis based on the equation (1).

  As shown in Fig. 4 (a), the quasi-standard model cannot flexibly incorporate the information of the actual structure in spite of the fact that a lot of observation information is necessary, and the deterioration prediction depends on the conventional reinforcing bar corrosion model. Can be said to have done.

Direct model is intended to define the Corrosion rate directly, and the p _1, p _2, p ₃ and estimated parameters. The parameter vector θ is defined by θ = (p ₁ , p ₂ , p ₃ ), and the prior distribution of p ₁ , p ₂ , p ₃ is a normal distribution with a mean value of 0.005 and a standard deviation of 0.05.

  FIG. 4A shows a list of analysis conditions for inverse analysis. Case 1 follows the conventional method and requires information on chloride ion concentration and neutralization depth (neutralization rate), but does not use the crack distribution for observation information. Cases 2 to 4 use a direct model for the reinforcing bar corrosion model and do not require information on chloride ion concentration. Furthermore, Case3 and Case4 do not require information on the neutralization rate.

FIG. 4B shows a list of parameter estimation results by inverse analysis. When Case 2 and Case 3 which are direct models are compared, it can be seen that the accuracy of estimation of p ₁ is particularly deteriorated due to the presence or absence of the observation information of the neutralization rate α. Furthermore, comparing Case 3 with Case 4, it can be seen that the estimation accuracy of p ₁ similarly decreases depending on the presence / absence of observation information on cracks. These are considered to be due to the limited neutralization rate of v ₁ and the sensitivity of cracking. Even in Case 4 where the amount of observation information is small, it is possible to estimate p ₁ , p ₂ , and p _{3 with} an accuracy of + 70% to + 180%, -35% to + 25%, and -5% to + 35%. Can be confirmed.

  FIG. 5 shows the prediction accuracy of the peeled-off area ratio. The vertical axis represents the occurrence rate of the peeled and peeled area with respect to the entire cross section, and the thick line in the figure represents the deterioration prediction result of the forward analysis (see FIG. 3). It can be confirmed that the predicted distribution is high near the correct value at 20 years, which is the observation time for each case. In Case 2 to Case 4 using the direct model, the prediction results are close to the correct values, and even if observation information such as chloride ion concentration and neutralization rate is omitted, there is a certain degree of accuracy when errors are allowed. It shows that deterioration can be predicted. In other words, when the direct model is used, the peeling and peeling area after 100 years is 0% to + 5% when observing the neutralization rate and crack distribution (Case2), and when the neutralization rate is omitted. (Case3) can be predicted with an accuracy of 0% to + 10%, and if the crack distribution is omitted (Case4), it can be predicted with an accuracy of -5% to + 15%.

Next, the deterioration prediction method of the present embodiment will be described with reference to FIG. 1, and an application example to an actual structure will be described with reference to FIGS. 6 to 13.
First, in step S1, the inspector conducts a visual inspection of the surface of the reinforced concrete structure (visual inspection). In the visual inspection, it is possible to obtain information on the peeling and peeling points on the surface of the reinforced concrete structure and information on the cracking points.

  FIG. 6 shows an outline of a target structure which is an actual structure. The target was a box culvert about 45 years after completion, and both sides of the inner circumference were designated as survey surface A and survey surface B. On the survey surfaces A and B, the peeling and peeling of the cover concrete has become apparent, and the structure is a structure that allows comparatively close examination. As for the surrounding environment, since it is located in the mountains, it is considered that there is no influence of incoming salt content, and the annual average temperature for the past 30 years by the Japan Meteorological Agency is 16.5 ° C and the average annual precipitation is 1626 mm. Not only peeling and peeling but also many cracks directly above the reinforcing bars occur in this target structure, making it difficult to determine the area to be repaired, such as cross-section repair, and is suitable as a target for detailed examination.

  In the field survey, as factors that greatly contribute to the deterioration rate, the deformation, cover thickness, neutralization depth, and chloride ion concentration of the 19 axial reinforcing bars and their cover concrete located on the surfaces of the survey surfaces A and B investigated. Here, the diameter of the reinforcing bar is φ19 mm.

  As shown in Fig. 7, in order to introduce it into the deterioration prediction model, each surveyed surface was divided into 100 mm × 300 mm mesh units directly above the rebar, and 25 × 19 × 2 = 950 deformation parameters Si for each mesh. (i is 1 to 950). This deformation parameter Si serves as information on the peeled off portion and the cracked portion on the surface of the concrete structure.

  On the other hand, as shown in FIG. 8, the cover thickness at three locations of 0.5 m, 1.3 m, and 2.1 m in height is measured by a nondestructive inspection device by a magnetic electromagnetic induction method. Here, this nondestructive inspection device is referred to as a reinforcing bar probe (step S2).

  Then, the fogging distribution in the 2.5 m range was calculated by linear interpolation and extrapolation from the measurement result of the fogging thickness by the reinforcing bar probe. In the investigation of the neutralization depth, a hole with a bit diameter of 24 mm was used for drilling, and after cleaning with air spray, a 1% solution of phenolphthalein solution was sprayed to measure the distance from the concrete surface to the color development point. It was measured. Further, as shown in FIG. 8, about 100 mm × 100 mm was cut out at the chipping points A and B, and the corrosion state of the reinforcing bar was investigated. The chloride ion concentration was determined by a potentiometric automatic titration method using concrete powder sampled from a depth of about 60 mm to 100 mm by drilling.

According to the measurement result of fog shown in FIG. 8, the actually measured fog varies in the range of 11 mm to 53 mm, the average is 30.1 mm, and the standard deviation is 10.6 mm. The neutralization depth was 15 mm to 27 mm from the 11 samples, the average value of the neutralization rate α was 3.0 mm / √year, and the standard deviation was 0.6 mm / √year. Since the chloride ion concentration was 3.357 kg / m ^3, it was estimated that the deterioration factor of the target structure was the combined deterioration of neutralization and internal salt damage.

  FIG. 9 shows the result of summarizing the relationship between fogging and deformation. The visually confirmed deformations are cracking, floating, peeling, and peeling.Si is 1: unchanged (177), 2: unchanged (internal corrosion) (117), 3: cracked. There were floating (614 pieces), 4: peeling and peeling (64 pieces), 5: peeling and peeling (rebar corrosion degree III or more) (5 pieces). Si = 2 is based on the measured fogging and the neutralization speed, and the case where the residual neutralization is 25 mm or less is set. In 64 of all 1064 meshes, since the peeling and peeling is 4 or more in Si, the peeling and peeling area ratio is 6.5%. It can be seen from FIG. 9 that the smaller the fogging, the higher the rank of the deformation and the more the number of defects. Peeling / stripping occurs in a mesh with a fog of 25 mm or less, and cracks occur in a mesh with a fog of 40 mm or less.

Then, in step S3, parameter estimation by inverse analysis is performed based on the information acquired in the field survey. FIG. 10 shows the posterior probability density distribution of the reinforcing bar corrosion rate when the direct model is applied to the target structure (Case 2). From FIGS. 10 (a) to 10 (c), it can be seen that the corrosion rate of the reinforcing bar increases from p ₁ to p ₃ . The coefficient of variation of p ₂ is about 10% and the variation is small compared to the other two parameters, but there are many data of cracks / floats corresponding to Si = 3 in the observed data. It is thought that it depends on what you are doing. Comparing p _2, p ₃ is a reinforcing steel corrosion rate before and after the occurrence of cracks, results of rebar corrosion rate is increased by about 140% is obtained by cracking. p ₃ Although a tendency to be distributed in a wide range as compared with p ₂ seen, such as the influence of the crack width to grow from cracking occurs until peeling flaking occurs is not taken into account is considered as a factor.

FIG. 11 shows a list of parameter estimation results by inverse analysis when applied to the target structure. Here, as the standard value to be compared, using the expressions described in "Maintenance standard" assuming the internal salt damage for p _1, assuming the combined degradation for p ₂ " The formula described in "Maintenance standard" was used. Focusing on Case 1, which is a quasi-standard model, it can be seen that the average value of χ is 0.65, and the corrosion rate of the reinforcing bar is about 65% of the standard value on average.

On the other hand, paying attention from Case2 is a direct model _{Case4, p 1, p 2,} p 3 40% of the standard value, respectively, 30%, is about 70%, to be able to estimate the Corrosion rate individually in each process I understand. Regarding the influence of the amount of observed information, it can be said that the influence of the presence or absence of the neutralization rate α is small by comparing Case 2 and Case 3. This is because, in Case 3, the neutralization rate α was used as an unknown value for the estimation, but the result was that the distribution converged to a distribution close to the measured value.

Furthermore, by comparing Case 3 and Case 4, the parameter estimation accuracy decreases when there is no observation information of cracks, and the coefficient of variation p ₁ increases by about 5% and p 2 and p 3 increase by about 30%. I understand.

When the estimation parameters p ₁ , p ₂ and p ₃ are calculated in this way, the corrosion rate v ₁ , v ₂ and v ₃ of the reinforcing bar can be estimated by the above formula (Equation 6). Therefore, in step S4, the peeling-off area of the surface (survey surface A, B) of the target structure at an arbitrary time is predicted based on the estimated reinforcing bar corrosion rates v ₁ , v ₂ , v ₃ .

  FIG. 12 shows the prediction accuracy in which the peeling and peeling area when applied to the target structure is shown by the peeling and peeling area ratio. From this figure, it can be seen that in each case, the probability of occurrence near the peeling and peeling area ratio of 6.5% is large at the time of 45 years, which is the observed value, and the actual measurement and the prediction match.

Also, it can be seen that the average value of the peeled and peeled area ratio after 100 years is about 55% in Case 1, and about 65% in Case 2 to Case 4. This is consistent with the fact that the direct model shows a larger value than the quasi-standard model in the estimation result of p ₃ in FIG. Focusing on the variation in the prediction results from Case 2 to Case 4, although there is a difference in the parameter estimation accuracy in the estimation results of FIG. 11, a large difference cannot be confirmed in the prediction results of FIG. From these, it is found that even when the parameter estimation accuracy is reduced by reducing the amount of observation information as in the deterioration prediction method of the present embodiment, a certain deterioration prediction accuracy can be secured.

  Succeedingly, in a step S5, a deterioration prediction result is output. FIG. 13 shows, as an example, actual measurement results and prediction results of the deformation distribution according to Case 4. The average value obtained by estimation was used for each parameter. Since it is not possible to determine the corrosion of the internal reinforcing bar in the deformation caused by actual measurement, it can be determined that there is no deformation even in the place where the corrosion of the reinforcing bar is actually occurring.

  By comparing the actual measurement and the prediction at the time of 45 years shown in FIGS. 13A and 13B, it can be confirmed that the results are almost the same. In some areas, such as near the ground on the A side of the survey surface, peeling and flaking has occurred in the actual measurement results, but there is a portion remaining in the cracks in the prediction results. In the vicinity of the ground on the right side of the A-side surface, it has been found from a study such as photographs taken in the past that it was an environment in which moisture was likely to accumulate on the back surface due to objects kept for many years. Therefore, it is possible to make more accurate predictions by performing an inverse analysis that takes into account such environmental differences and the effects of rainwater near the entrance of the box culvert. In addition, with respect to the prediction result, the deformation process is progressing with the passage of time, and at 100 years shown in FIG. I know that

Next, the operation of the deterioration prediction method of this embodiment will be described.
The deterioration prediction method of the present embodiment configured in this way, Bayesian estimation of the corrosion rate of the reinforcing bar by inverse analysis from the information of the peeling off point and the cracked point of the surface of the reinforced concrete structure and the information of the cover depth of the reinforcing bar. To do. Then, based on the estimated corrosion rate of the reinforcing bar, the peeling and peeling area at an arbitrary time is predicted.

  In this way, by just using the results of limited inspection items such as information obtained by visual inspection and information on the depth of cover, Bayesian estimation of the corrosion rate of the reinforcing bar can be performed in practice considering the uncertainty of prediction. A suitable prediction result can be obtained.

  That is, even if the measurement of the chloride ion concentration and the neutralization rate is omitted, the corrosion rate of the reinforcing steel before and after cracking can be estimated with an accuracy of about -35% to + 35%, and the peeled and peeled area after 100 years can be estimated. It can be predicted with an accuracy of -5% to + 15%.

  Compared with the corrosion rate of internal salt damage and complex deterioration according to the above "maintenance standard", 30% to 70% is the estimation result by the deterioration prediction method of the present embodiment, and the corrosion rate of the reinforcing bar due to cracking is It was revealed that the average increase was about 140%. The predicted value of the peeled-off area at 100 years of age was 50% to 75% of the total area, independent of the amount of observed information such as neutralization rate and crack distribution.

  In addition, by utilizing the deterioration prediction result of the target structure by the deterioration prediction method of the present embodiment, as a result of carrying out a repair plan simulation that minimizes maintenance cost, which is the sum of inspection cost and repair cost, Such findings were recognized.

The optimum cross-section restoration area that minimizes the total maintenance cost in 75 years is 10 m ² , and the maintenance cost is reduced by 40% to 50% compared to the method of repairing only the area around the visible peeling and peeling area. The effect was obtained. In other words, using preventive maintenance, it is possible to reduce the total O & M cost by expanding the cross-section repair area to the surrounding area in addition to the visually peeling and peeling area as preventive maintenance. did it.

The embodiment of the present invention has been described in detail above with reference to the drawings. However, the specific configuration is not limited to this embodiment, and a design change that does not depart from the gist of the present invention is Included in the invention.
For example, in the above-described embodiment, the method of acquiring information on the peeling and peeling portion and the cracked portion of the surface of the reinforced concrete structure by the visual inspection by the inspector has been described, but the present invention is not limited to this, for example, the photographed investigation By analyzing the image data of the surface, it is possible to automatically obtain such information.

Claims (4)

A method of predicting deterioration on the surface of a reinforced concrete structure,
A step of acquiring information on the peeling off point and the cracked point on the surface of the reinforced concrete structure and the information on the covering depth of the reinforcing bar,
A step of estimating a reinforcing bar corrosion rate by inverse analysis from the acquired information;
And a step of predicting a peeling-off area of the surface at an arbitrary time point based on the estimated reinforcing bar corrosion rate.   The deterioration prediction method according to claim 1, wherein the estimation of the corrosion rate of the reinforcing bar is performed based on Bayes' theorem.   The method of predicting deterioration according to claim 1 or 2, wherein the information on the peeling off spots and cracks on the surface of the reinforced concrete structure is obtained by visual inspection.   The deterioration prediction method according to any one of claims 1 to 3, wherein the information on the cover depth of the reinforcing bar is a measurement result by a nondestructive inspection device by an electromagnetic induction method.