2.1. Bridge Condition Data and Processing Method
According to the Standards for Technical Condition Evaluation of Highway Bridges (JTG/T H21-2011) [
28], when the bridge is new and fully functional, the technical condition score is [95,100] and the technical condition class is 1; when the bridge has minor defects that do not affect the use function, the score is [80,95) and the class is 2; when the bridge has moderate defects but is still able to maintain normal use function, the score is [60,80) and the class is 3; When the main components of the bridge have large defects that seriously affect the bridge function or load-bearing capacity and cannot guarantee normal use, the score is [40,60) and the class is 4; when the main components of the bridge have serious defects and cannot be used normally, endangering the safety of the bridge and putting the bridge in a dangerous state, the score is [0,40) and the class is 5. The technical conditions of nearly 20,000 concrete girder bridges in various regions of Shaanxi Province were assessed using this method. Scatter diagrams were drawn and are presented in
Figure 1.
As bridge deterioration is influenced by multiple factors, such as the bridge’s structure type and age, the traffic load, environmental factors, and maintenance, bridge technical condition data are discrete, and especially for older bridges that have been maintained and repaired several times cannot accurately reflect the deterioration characteristics of the bridge structure in its natural state. In this section, a simply supported slab bridge with a span of 8 m and a load rating of highway class I [
29] is used as the object. The technical condition score of the bridge is used as the performance deterioration index, and the deterioration curve of each part of the bridge structure is obtained using mathematical regression analysis. The information from the successive bridge inspection and maintenance history can be used as a data basis for predicting the deterioration of the technical conditions of bridges. Considering the uncertainty of the data, it is necessary to process them in advance, as follows:
- (a)
During service, bridges are inevitably maintained and repaired. Bridges that have taken reinforcement or renovation are considered newly built bridges, and their ages are recorded as 0. For example, the technical condition score is 75 when the bridge is 29 years old, and the score is 94 after maintenance or reinforcement when the bridge is 30 years old. The age is corrected to 0 from 30 years;
- (b)
Assume that the bridge deteriorates gradually and that the technical condition score cannot be higher in the latter than in the former. If the latter is higher than the former, the relevant data are excluded;
- (c)
The data should be evenly distributed across the various bridge ages. If there is a period of a lack of data, then the relevant bridge technical condition data can be supplemented by calculating the transition probability matrix for the first two bridge ages and using Markov chain prediction methods;
- (d)
Considering that the process of assessing the technical conditions of bridges is influenced by the subjective factors of the inspectors and there is a certain amount of error, it is assumed that the technical conditions of bridges of the same bridge type, span, and age obey a normal distribution. Then, the abnormal values can be removed according to the PauTa criterion [
30].
2.2. Deterioration in Different Parts of Bridge
Bridges consist of three parts: the deck system (including deck pavement, drainage system, expansion joints, and parapets), superstructure (including main girders and bearings), and substructure (including piers, piles, and foundations) [
28]. A total of five fitting functions (linear, quadratic polynomial, cubic polynomial, trigonometric, and exponential) were used to fit the technical condition scores of 108 bridges. The fitted curves for the deck system are shown in
Figure 2.
The evaluating indicators for the regression models are shown in
Table 1. Among the five models, the exponential model had the smallest error sum of squares (2813), and its square of the correlation coefficient was close to 0.8. Thus, the exponential model fit best among the five models.
According to the above method, the superstructure and substructure technical condition data are processed by the exponential fitting method to obtain their deterioration models, as shown in
Figure 3.
Figure 4 shows the deterioration curves of various parts. The deterioration rate of the deck system is the fastest, followed by the superstructure and substructure, and it is close to the existing results [
31]. The deck system deteriorates from 100 points (in good condition) to 60 points (boundary of class 3 and 4) in 32 years, the superstructure in 38 years, and the substructure in 56 years. The deterioration rate of the deck system is approximately 1.2 times that of the superstructure, and 1.8 times that of the substructure. Therefore, sufficient attention should be given to the deck system during bridge maintenance.
According to the exponential function, the general form of the exponential deterioration model can be obtained by the following:
where
f(
t) is the technical condition score,
t is the bridge age, and
A is the deterioration coefficient. The larger the absolute value of
A, the faster the deterioration of the structure.
2.3. Deterioration in Different Regions of Shaanxi Province
The different regions of Shaanxi Province have different natural geological conditions and climatic environments, as well as wide differences in economic development, which has resulted in the strong geographical characters of the service environments of the bridges. This section analyzes the deterioration rates of voided slab bridges in three regions: Northern Shaanxi, Central Shaanxi (the Guanzhong area), and Southern Shaanxi, as shown in
Figure 5. A comparison of the bridge deterioration rates in each region is shown in
Figure 6.
In
Figure 6, the deterioration rates of the bridges in each region show significant geographical differences, with the fastest deterioration in Guanzhong, followed by Northern Shaanxi, and the slowest in Southern Shaanxi. The deterioration coefficients (
A) of the bridges in each region are shown in
Table 2. These were consistent with the degree of regional economic development in Shaanxi Province. The economy of Guanzhong is more developed than those of Northern and Southern Shaanxi, with higher traffic volumes and bridge utilization rates. Therefore, the bridges in Guanzhong have larger deterioration coefficients.
2.4. Influencing Factors on Deterioration Coefficients (A)
Bridge deterioration is a complex process with many uncertainties, and it is mainly influenced by the materials, environment, and loads. The material is an intrinsic factor in determining the structural deterioration. Different materials have different deterioration rates and are affected by the environment differently. For example, concrete and steel have decidedly different deterioration characteristics. The temperature, humidity, and load of the service environment are the external factors of bridge deterioration, and the coupling effect of the load, temperature, and humidity exacerbates the deterioration and leads to different rates of deterioration for each part of the bridge. The deck, for example, is directly exposed to the environment and is affected by the temperature, humidity, and live loads, which result in rapid deterioration. The superstructure and substructure are protected by the deck and are therefore less affected by the service environment. In terms of the overall bridge deterioration rates, the rate in Guanzhong is higher than those in Northern and Southern Shaanxi due to the high traffic volume and complex service environment.
Therefore, a generic model for bridge deterioration should be able to reflect the effects of the materials, the service environment, and other influencing factors, and it can be simplified as follows:
where
C(
t) indicates the technical condition of the bridge in the age of
t;
C0 is the initial technical condition, which is generally taken as 100;
ξ1 is the influencing factor of the service environment, which is determined according to the natural environment (temperature, humidity, sunlight, wind, etc.) and live loads;
ξ2 is the influencing factor of the material, determined according to the difference in the deterioration characteristics of steel, concrete, wood, etc., with concrete as the benchmark with a value of 1.0 (few cases of steel and wooden bridges are found in Shaanxi, and so they are not discussed); and α is the generic annual deterioration rate related only to the structure itself.
By corresponding the deterioration models for each part of the bridge developed in
Section 2.2 and the bridge deterioration models in each region developed in
Section 2.3 to the coefficients in Equation (2), the relevant coefficients for each deterioration model were obtained, as shown in
Table 3. The component deterioration model is based on the superstructure, with
α = 13.3 × 10
−3. The values of
ξ1 are 1.39, 1, and 0.67 for the deck system, superstructure, and substructure, respectively. The bridge deterioration model is based on Southern Shaanxi, with α = 8.5 × 10
−3. The values of
ξ1 are 1.06, 1.39, and 1 for Northern Shaanxi, Guanzhong, and Southern Shaanxi, respectively. It can be seen that the generic deterioration model of Equation (2) applies to each bridge component and the bridge as a whole in each region of Shaanxi Province.
The values of the service environmental influencing factor (ξ1) vary considerably between the structures of the bridges, with the value of ξ1 for the deck system approximately twice that for the substructure. The reason for this is that the deck system is the protective layer for both the superstructure and substructure and is directly exposed to environmental effects, such as rain and snow and the impact of vehicles; therefore, the deck system is most affected by the service environment. The bridge deck system is subjected to environmental effects, which are transferred to the superstructure and substructure through the deck pavement, bearings, and expansion joints, causing feedback from the superstructure and substructure to the environmental effects, and leading to structural diseases in severe cases. In summary, the bridge is subjected to environmental effects in a top-down process. A similar process may also exist in the generation of bridge diseases. This hypothesis is tested below through the study of bridge disease development patterns.