The impact of exploitation and environmental factors on the degradation of steel road safety equipment

The work is aimed at determining the corrosivity of atmosphere in the vicinity of roads, taking into account the characteristics of local emission sources, including traffic intensity of vehicles along with climatic and exploitation factors. Determination of the corrosivity of atmosphere was carried out according to the procedures described in PN EN ISO standards. Samples for testing were made of low carbon steel DC05, zinc and hot dip galvanized steel. Samples were assembled at 19 sites in the close vicinity of roads and highways near the measurement points of vehicle traffic intensity. The mass loss of exposed samples was the basis for determination the atmosphere corrosivity at each of 19 test sites. Regarding steel, the corrosivity category of C4 was observed at 8/19 sites. Corrosion losses outside roads are 2-4 times lower and ranged within the categories of C2 and C3. Zinc corrosion losses classified to category C4 occurred at 2/19 stations. In the remaining ones they corresponded to category C3. In areas outside of roads, zinc corrosion losses are about 20-100% lower (C2). The first attempts to model the dependence of operating and environmental parameters on zinc and zinc coating corrosion losses indicate significant correlation between zinc and zinc coating corrosion losses as a function


Introduction
The environment created around the road area differs significantly in its qualitative and quantitative composition from areas even slightly distant from sources of road emissions [1]. Vehicles emit a gas mixture, which within a fraction of a second, creates various substances with strong corrosive properties: nitrogen oxides, nitric acid, sulfur dioxide, ozone, reactive radicals, soot, organic acids and others [2]. There is no available information on the relationship between the rate of corrosion of metals assembled in the construction of road infrastructure, and environmental parameters including vehicle traffic and/or other exploitation conditions. The work is aimed at the determining of corrosivity of atmosphere in the vicinity of selected roads, taking into account the characteristics of local emission sources, including traffic intensity of vehicles along with climatic and exploitation factors. The data obtained will be used to develop methods for predicting the rate of corrosion of

Method
Determination of the corrosivity of atmosphere was carried out according to procedures described in PN EN ISO standards [3][4][5][6]. Samples for testing with dimensions of 150 x 100 x 1 mm were made of low carbon steel DC05, zinc and hot dip galvanized steel. Samples were assembled at 19 sites in the close vicinity of roads and highways and monitored within the period of 07.2017-10.2018. These places are located near the measurement points/stations of vehicle traffic intensity carried out by GDDKiA. For stations with automatic measurement of pollution (Katowice, Kraków, Warszawa/Niepodległości Avenue), data came from the network of Regional Inspectorates for Environmental Protection (www.wios.katowice.gov.pl, www.wios.krakow.gov.pl, www.wios.warszawa.gow.pl).
In other cases the following parameters were monitored directly at corrosion sites: concentration of NO2 and SO2 by means of passive samples [7], deposition of chlorides and sulphates according to [5]. An average annual traffic volume divided by a type of vehicle is shown in Fig. 1. The average annual traffic intensity at corrosion sites ranged from 19599 vehicles/day to 127822 vehicles/day. Passenger cars and microbuses accounted for 90% (Mysiadło) to 54% (Stryków) of the total number of vehicles.

Results and discussion
The first set of results contain an evaluation of corrosive agents on site. An average concentration of NO2 related to traffic intensity emitted by vehicles at corrosion sites is illustrated in Fig. 2. Chloride ions also belong to the list of common corrosion agents. In the case of roads and highways direct corrosion risk results from the application of de-icing salts during a winter period. An average Clions deposition within an exposition period is shown in Fig. 3.   The next set of results is related to an evaluation of corrosion losses of steel, zinc and hotdip galvanized steel (table 1). The mass loss of exposed samples is the basis for a determination of atmosphere corrosivity at each of 19 test sites. Regarding steel, the corrosivity category of C4 was observed at 8 sites, while at the remaining ones the corrosivity corresponds to the C3 category. Corrosion losses outside roads calculated on the basis of environmental parameters are 2-4 times lower and ranged within the categories of C2 and C3. The latter is determined for Silesia and Myślenice. Zinc corrosion losses classified to category C4 occurred at two stations. In the remaining ones they corresponded to corrosivity category C3. In areas outside of roads, zinc corrosion losses are about 20-100% lower and correspond to the corrosivity category of C2. Corrosion losses of zinc coating on roadside stands are higher than zinc corrosion losses by a factor 1.35. Corrosion losses at the off-road areas are approximately 20-100% lower than on roads. Linear correlation coefficients R between the parameters describing the environment and corrosion losses of steel, zinc and zinc coatings for the sample size N = 21 are given in Table  2. Correlation at the good level between corrosion losses of zinc and zinc coating and traffic as well as corrosion losses of zinc and chloride deposition is observed. On the other side, there is no good correlation between environmental parameters and corrosion losses of steel. It should be noted that the small sample size (N = 21) does not guarantee of an obtainment of precise relation between analyzed parameters. The estimation error for this sample is 21% with the assumption of an infinite population.
In preliminary steps to model the dependence of operating and environmental parameters on zinc and zinc coating corrosion losses a trial of constructing certain dose -response functions was attempted taking into consideration the most significant parameter -traffic intensity(NR). Examples are shown in Table 3.