Content uploaded by Toshihiko Aso
Author content
All content in this area was uploaded by Toshihiko Aso on Jan 04, 2017
Content may be subject to copyright.
Integrated Solutions for Infrastructure Development
Edited by Lau, H. H., Tang, F. E., Ng, C. K., and Singh, A.
Copyright © 2016 ISEC Press
ISBN: 978-0-9960437-3-1
1
THE INFLUENCE OF TEMPERATURE AND
HUMIDITY ON RUST GROWTH ON
WEATHERING STEEL
RYOYA NAGATA1, SHUSAKU TACHIBANA2 and TOSHIHIKO ASO1
1Dept of Civil and Environmental Engineering, Yamaguchi University, Ube, Japan
2Nippon Steel
and
Sumikin Engineering, Osaka, Japan
In this study, to develop a growth prediction method of rust on weathering steel, the
rust progress characteristics under constant temperature and humidity have been
conducted. In order to achieve a long life of weathering steel bridges, determination of
several problems are required. Especially, rust evaluation technology, rust prediction
method and repair technology of corroded weathering steel are significant. Different
concentration (0.3%, 1.0%, 3.0%) of salt water sprayed on steel test pieces and,- were
placed in a constant temperature and humidity test vessel, to observe the progress of
rust in in-plane direction. Four patterns of exposure environment were tested.
Exposure environment were combination of temperature (40°C, 20°C) and relative
humidity (50%, 95%). To calculate the increase of rust area, specimen surface was
confirmed. Higher salt water concentration (3%) brought rapid rust development.
Furthermore, the increase of rust area was highly dependent on temperature than
humidity. From time observations, it becomes clear that increase of rust area can be
approximated to a logarithmic curve.
Keywords: Weathering steel, Rust, Salinity, Temperature, Humidity.
1 INTRODUCTION
In recent years, cost in construction has been based on life-cycle-cost. Life-cycle-cost consists of
not only initial construction cost but also maintenance cost (Kihira 2007). Generally, a large part
of maintenance cost of steel structures is paint. Paint has been used for corrosion control of steel
structures.
Weathering steel is a steel which can reduce the corrosion rate by generating a dense protective
rust on the steel surface and exert anti-corrosion performance. Thereby, weathering steel could
be provided corrosion protection performance without painted layer. Due to these advantages in
maintenance cost, number of weathering steel bridges has been increased in Japan (Kage et al.
2007). However, in some cases, very severe surface corrosion which be able to reduce load
bearing capacity has been observed (JSSC 2006, JSSC 2009).
In order to achieve a long life of weathering steel bridges, rust rating technique, method of rust
predictive simulation and repair technique of corroded weathering steel member must be
established.
To establish a growth prediction method of rust, relationship between rust growth speed and
the corrosive environment has to be clarified. In this study, exposure tests of weathering steel
were performed under constant temperature and humidity environment. During exposure test,
Lau, H. H., Tang, F. E., Ng, C. K., and Singh, A (eds.)
2
area of rust was observed by image analysis to estimate spread speed of rust on specimen surface.
From the results of test, horizontal spread of rust could be approximate as logarithmic curve.
2 MATERIALS AND METHODS
2.1 Exposure Test
Exposure tests were started from October 20, 2015. A growth of rust is horizontal spread on steel
surface and three dimensional progress by occurrence of steel wear and corrosion product. In
exposure test, weathering steel of 70×70×6(mm) was used as specimen. Salt water of 0.02 ml
was dropped on the specimen. Salt water was made from sodium chloride and deionized water.
One specimen was supplied only one kind of salt water. Then moisture on the specimen surface
was dried. The specimen was installed in a small environment tester. The concentration of salt
water was set to 0.3%, 1% and 3%. In this study, the specimen which is dropped salt water was
exposed. Furthermore, to clarify the effect of salt to steel corrosion in this tests, non-supply
specimens ware also set. Table 1 gives conditions of the test. Temperature and humidity were
decided due to the performance of small-environmental-tester and observations of bridge sites.
Formed rust on each specimen were photographed every day. The image was trimmed to 280 ×
280 pixels around center of a salt water drop part. The area of this image was 1238.278mm2.
Shot images are divided into three components, red, green and blue(RGB). Then using analysis
software, the image was taken out only green component from the RGB components. The area
ratio of rust part was calculated by extracted image. The area of rust part was the ratio of rust and
trimmed image area.
Table 1. Test conditions.
Temperature-Humidity
Number of
specimen
Salt water
concentration
Period of
exposure
Case1
40°C-95%
4
Not dropped
30
Case2
40°C -50%
4
0.3%
29
Case3
20°C -50%
4
1%
18
Case4
20°C -95%
4
3%
20
3 TEST RESULTS
3.1 Appearance Evaluation
Figure 1 shows the appearance of the specimen which is exposed of day1 and day 30 in
temperature 40°C and relative humidity 95% (3% salt water). The color of the salt water drop
part changed to black during 30 days. Therefore, it was possible to confirm clearly that the rust
grew. Except for the salt water drop part, the occurrence of uniform corrosion was recognized.
However, this uniform corrosion was not verification target. Figure 2 indicates the banalization
of green component of Figure 1(b). This process is effective to enhance the rust region in images.
Figure 3 shows the image of particle analysis. The area ratio of all images was calculated from
this step.
Integrated Solutions for Infrastructure Development
3
(a) Day1 (b) Day30
Figure 1. Variation of area ratio at day1 and day30 (3% salt water)
Figure 2. Image of green component Figure 3. Particle analysis image
3.2 Evaluation of Area Ratio of Rust Part
Influence of corrosion environment is evaluated from the ratio of rust-region. Figure 4 (a) and
Figure 4 (b) illustrates the variation of the area ratio with duration in Case1 and Case4, Case2 and
Case3, respectively. In Case1 and Case 4, relative humidity is set for 95% in both case but
temperature is different (40°C in Case1, 20°C in Case4). In Case2 and Case 3, relative humidity
is 50%. In all cases, the area ratio of the specimen which was dropped 3% salt water was big.
Therefore, the influence of salt on the rust growth could be confirmed. However, increasing
tendency of the area ratio was different in Case1 and Case4. The area ratio which is dropped 3%
salt water in Case1 was 25% in day 20. In contrast, the area ratio in Case4 was 3%. It was found
that tendency did not depend on the salinity. Furthermore, it was observed that the influence of
temperature to growth of rust is big from Fig.4 (a) and Fig.4 (b). This tendency could be
confirmed in Case2 and Case3 that are low relative humidity. From the comparison with Case1
and Case 2, in the same temperature and same salt water concentration, high relative humidity
provides large area ratio. However, this tendency could not have observed in low temperature
(Case3 and Case4). The logarithmic approximation formula is shown to Fig.4 (a) and Fig.4 (b)
10m
m
10m
m
10m
m
10m
m
Lau, H. H., Tang, F. E., Ng, C. K., and Singh, A (eds.)
4
dashed line. But the approximation was impossible from data of Case3 and Case4. The
calculated approximate value from these approximate formula was close to the actual measured
value. In all cases, the coefficient of determination was more than 0.75.
(a) Case1 and Case4
(b) Case2 and Case3
Figure 4. Variation per day of area ratio with duration.
4 GROWTH PREDICTION METHOD OF RUST
This paper also aims to establish a growth prediction method of rust to quantitatively evaluate a
corrosion process of weathering steel. From the results of exposure tests it became clear that rust
growth is due to temperature, humidity, salinity. The relationships between horizontal growth of
rust and exposure duration can be assumed by using logarithmic function. The rust growth
prediction would be expressed by equation (1)
Y=Aln(X)+B (1)
Where, Y: rust thickness, X: exposure days, A and B: Constants.
Integrated Solutions for Infrastructure Development
5
In Case1 and Case2, rust growth could be confirmed clearly. The relationship between
concentration of salt water and coefficient or intercept of the logarithmic formula was verified.
Figure 5 shows the relationship between coefficient of the logarithmic formula and
concentration of salt water in Case1 and Case2. Also Figure 6 shows the relationship between the
intercept of the logarithmic formula and concentration of salt water. In Case1, linear relationship
was recognized between the coefficient and concentration of salt water. In addition, three values
of intercept are almost the same in the case of concentration. And on the other hand, the tendency
was not recognized in Case2. In Case2, the value of 1% salt water was protruded. The reason is
that the rust growth of the specimen which was dropped 1% salt water was big in 10 days later in
Case2. In Case1, the rust growth prediction equation was calculated. It became clear that A is
1.805C (C: concentration of salt water) and B is 6.027. The rust growth prediction equation was
expressed by equation (2) in Case1.
Y=1.805Cln(X)+6.027 (2)
Where, Y: area ratio of rust, X: exposure days, C: concentration of salt water.
Figure 7 shows measured value and calculated value from equation (2). Equation (2) shows
good agreement to measured value. Therefore, equation (2) is applicable as rust growth
prediction equation in Case1. Accuracy of approximation could be improved by increasing the
experimental results.
Figure 5. Coefficient and concentration of salt water.
Figure 6. Intercept and concentration of salt water.
Lau, H. H., Tang, F. E., Ng, C. K., and Singh, A (eds.)
6
Figure 7. Measured value and calculated value.
5 CONCLUSION
Rust growth on weathering steel under several corrosion environment was evaluated by ratio of
rust region. Results of this research could be pointed as below.
1. Rust growth is depending on temperature and salt condition strongly. An effect of relative
humidity was small in this test.
2. Ratio of rust region could be approximated by logarithmic formula. Accuracy of
approximation could be improved by increasing the experimental results.
3. Coefficient of approximated curves are related to concentration of salt water.
To apply the result of this research to actual bridges, corrosion environment (temperature,
relative humidity and amount of airborne salt) of these bridges are required. A database of
corrosion environment should be constructed for planning and design of weathering steel bridges.
References
Japan Society of Steel Construction, Potentiality of Unpainted Weathering Steel Bridges, Technical
Report No.73, Japan Society of Steel Construction, Tokyo, 2006. (in Japanese)
Japan Society of Steel Construction, Proactive Maintenance of Weathering Steel Bridges, Technical
Report No.86, Japan Society of Steel Construction, Tokyo, 2009. (in Japanese)
Kage I., Kyono K., and Matuda Y., Prediction for Corrosion Loss of Weathering Steels, JFE
Technical Report, No.18, 62-66, Nov 2007. (in Japanese)
Kihira H., Systematic approaches toward minimum maintenance risk management methods for
weathering steel infrastructures, Corrosion Science 49, 112-119, 2007