2.1 Methodology utilized

Most of the information contained in the present study was obtained from DNIT's Special Engineering Structures Management System (SGO) database, and from detailed analyses of several bridges and viaducts located on federal highways crossing the State of Pernambuco, whose denominations are shown in Table 1 below.

A survey was also carried out to identify the construction periods and the typical mobile loads of each structure. Failure to supply most of the original designs meant that much of this information was obtained by comparing the evolution of the bridge cross-sections of federal highways over time with the mobile loads that were used on bridges built in the respective periods, as shown in Figure 1.

Figure 1 Evolution of the transverse cross-section of federal road bridges (Measurements in cm) 

For the analyses of the structural and functional conditions, DNIT Standards 010/2004 - PRO were used, which specify the stability condition of the structure by assigning a technical grade that varies from 1 (critical structure) to 5 (structure without problems). The new version of NBR 9452 (^{ABNT, 2016}) was also used, which introduced structural, functional, and durability parameters to guide the classification of bridge structures. The procedures foreseen in these two norms are usually adopted in Brazil to identify and diagnose the main anomalies in bridge structures, including pathological manifestations and the most frequent types of damage.

Finally, the results obtained in the research were interpreted and evaluated, in order to determine the structural and functional conditions of the SES studied and to propose solutions for the correction of existing problems, applying a more comprehensive view that can be applied not only to the bridges and viaducts of Pernambuco, but to other structure that make up the Brazilian road system as a whole.

2.2 Bridges analyzed by construction period, typical load, and construction system

Table 1 shows the federal highways within the jurisdiction of the State of Pernambuco, totaling 546 bridges and viaducts, and the amount inspected in this study, corresponding to 332 structures. This sample represents approximately 60% of all of the existing bridges in the state road network.

The study also classified the Special Engineering Structures by period of construction (Table 2) and by the project model load, which is defined as the set of mobile loads to be applied to the structure in a position that produces the most unfavorable forces for each section and combination of loads calculated. As previously explained, the information was acquired from the SGO/DNIT and through the analysis of the evolution of the transversal cross-section of the structure, associated with the changes that occurred in the Brazilian standards specifying mobile load values over time for those projects where data from the construction time period was not available. The Special Engineering Structures for which this information could not be obtained were considered unidentified.

Regarding the project model load, Figure 2 shows that 50.60% of the structures were qualified with 360 kN loading. The mobile loads of 450 kN, currently used, represent 28.31% and those of 240 kN represent 7.83% of the total sample. There were also 13.25% of structures whose project model loads were not possible to identify.

Figure 2 Project model load of bridges evaluated 

Regarding the constructive system, it was verified that reinforced concrete poured on-site was predominant, used in 76.20% of the bridges and viaducts. The second most prominent system was prestressed concrete, adopted in 23.19% of the structures. Pre-cast reinforced concrete accounted for 0.60%. These data are shown in Figure 3.

Figure 3 Constructive systems used on evaluated bridges 

2.3 Pathological manifestations identified in the studied sample

In order to identify pathological manifestations, observations were made at each site during the inspections and the SGO survey reports were analyzed. The major pathological problems identified are illustrated in Figures 4 to 7.

Figure 4 Presence of efflorescence, displacement of concrete coating, and reddish-brown spots due to the corrosion of reinforcement 

Figure 5 Exposed and rusty reinforcement on beams, some with broken strips 

Figure 6 Erosion of the foundation and misalignment of the pillar in relation to the pipe 

Figure 7 Metal bearing on a Gerber tooth with visible damage and infiltration 

Figure 8 synthesizes the highest incidences of pathological manifestations for the evaluated set, showing the predominance of infiltrations and efflorescence, present in 293 SES. The following anomalies are also worth mentioning: displacement of the concrete (of the superstructure) in 175 structures, exposed and oxidized reinforcement (of the superstructure) in 172 structures, and damaged and/or nonexistent expansion joints in 157 SES.

Figure 8 Occurrences of pathological manifestations in the analyzed structures 

2.4 Structural damage identified

Similar to the pathological manifestations, some of the main types of damage observed in the bridges and viaducts of this study are illustrated in Figures 9 to 12.

Figure 9 Absence of lateral guard-rail 

Figure 10 Damaged drain causing infiltration and moisture spots between the beams 

Figure 11 Deformed elastomeric bearing 

Figure 12 Extreme erosion around the foundation 

Figure 13 quantifies the principal damage types found in the analyzed set of structures. It was verified that 92 bridges had damaged and/or nonexistent drains. Next most prevalent were damaged guardrails (50 SES), followed by breakdowns in stone/concrete masonry (46 SES), as well as the lack or deficiency of bearings (25 SES), and holes in the concrete (20 SES). Other types of damage occurred less frequently.

Figure 13 Principal types of structural damage found in the set of structures analyzed 

3.1 Evaluation by DNIT Standard 010/2004 - PRO

The criteria for evaluating the structural stability conditions considered by DNIT standard 010/2004-PRO are shown in Table 3 below.

The application of these criteria to the sample studied is shown in Figure 14, which indicates that 156 SES can be classified as potentially problematic (grade 3), or 46.99%. It is important to note that 174 structures were evaluated as having grades between 1 and 3 (critical to potentially problematic), which is equivalent to 52.41% of the sample studied. Only 59 structures (17.77%) received a grade of 5.

Figure 14 Stability conditions according to DNIT 010/2004-PRO 

Table 4 shows the correlation between the construction period and the stability conditions. It also shows that 34% of the structures studied were built between 1960 and 1975, meaning that they are over 40 years old, while 8% are almost 70 years old.

3.2 Evaluation by NBR 9452/2016

The structural evaluation according to NBR 9452/2016 followed the criteria shown in Table 5. In Table 6 it can be seen that, according to these criteria, a grade of 4 was most common, attributed to 116 bridges. A grade of 5 was obtained by 107 SES, followed by grade 3 (87), grade 2 (20) and grade 1 (2). By these criteria, the 113 structures built between 1960 and 1975 had scores different from those obtained by the criteria of the DNIT standard.

3.3 Considerations on the analyses made by the DNIT and ABNT standards

Both standards used in this study assign a classification grade from 1 to 5 for stability conditions. The DNIT 010/2004-PRO Standard also includes concepts, inspection types, the necessary qualifications for inspectors, and the minimum frequency of surveys that allow some evaluation criteria to be defined. However, the recent revision of NBR 9452 (^{ABNT, 2016}), in addition to adopting the concepts and guidelines contained in the DNIT standard, introduced greater flexibility by including structural, functional, and durability parameters. These parameters made inspections and classification criteria of bridges and viaducts in Brazil more realistic.

It is worth highlighting that the two standards made the diagnosis of the main pathological manifestations possible without significant differences in the evaluation of damages and other anomalies in the sample bridges.

On the other hand, when analyzing stability conditions, a considerable difference was observed in the results obtained by each of the two standards, as can be seen in Table 7.

Even considering that inspections were carried out visually, depending essentially on the experience and judgment of the inspector, the information in Table 7 shows that, with regard to grades 1 and 2, referring to structures in the worst conditions, there were no differences in the results obtained through the two standards.

However, results related to the other grades assigned (3, 4 and 5) show that analysis by NBR 9452 leads to values representing significantly better structural conditions, when compared to those obtained by the DNIT Standard. This means that several bridges and viaducts in this study that may have poor durability conditions and functionality assessments but have satisfactory structural conditions. It also means that these structures can be used without reaching the ultimate state limit. The opposite situation was also verified: bridges and viaducts classified without durability and functionality problems (grades 4 and 5) that had a low structural condition classification.

In general terms, it is possible to affirm that the new version of the ABNT standard that introduces greater variability in the evaluation parameters and applies them to a larger quantity of the elements that make up a bridge, guarantees a more consistent and more realistic evaluation than the DNIT standard, which scores only five elements and is therefore a more conservative evaluation. This is evident in the case of a Special Engineering Structure without major breakdowns that can have one of the elements evaluated with a low rating, and therefore will have the final technical grade correspond to the lowest value among those received all of the elements. This produces a conservative final assessment that does not represent the actual situation of the bridge.

In this sense, NBR 9452 (^{ABNT, 2016}), even though it also bases its final evaluation on the lowest individual grade, introduces a greater amount of parameters for judgment and therefore is less conservative, because it determines the grades through a more comprehensive analysis. This standard also allows for a more comprehensive and realistic diagnosis of the structure, which is a breakthrough in Brazil, which lacks methods and literature to assess the safety of existing bridges, unlike other countries where sophisticated analyses are used, including probabilistic methods.