400 Ericsson
Overlag Chemie, CmbH, D-6940 Weinheim, 1978
The influence of sodium chloride on the atmospheric corrosion of steel
Einflufl von N a t r i u m c hl o r i d auf die at mo s p h ar i s ch e Korrosion von Stahl
R . Ericrson *)
Department of Inorganic Chemistry, Chalmers University of Technology and the University of Goteborg, P.O. Box,
S-402 20 Goteborg 5 , Sweden
Summary
The influence of sodium chloride on the atmospheric
corrosion o f mild steel has been studied. Samples covered
with sodium chloride crystals ( 8 pg NaCl ' c m - 2 ) have been
exposed to an SO,-free atmosphere a t different relative
humidities. The combined effect of sodium chloride crys-
tals on a steel surface and sulphur dioxide in the atmosphere
has been investigated at an SO,-supply of
1 pg SO2 . c m - , . h-' ( 1 ppm SO,). The corrosion attack
was followed by periodic weighings. The corrosion products
formed have been studied using the SEM-technique. On ex-
posure of samples with sodium chloride in an SO2-free at-
mosphere the extent o f corrosion increased with increasing
relative humidity f ro m 58% to 9095, interrupted by a sharp
minimum at about 87% relative humidity. The SEM-studies
showed that tower shaped corrosion products were formed
at a high relative humidity while filiform corrosion appeared
when the relative humidity was lowered. A t 90% relative
humidity more corrosion was observed with clean steelsamp-
les and an S02-supply o f I p g . cm-, . h-' than with sodium
chloride crystals on the surfaces ( 8 pg NaCl. c m - 2 ) in the
absence of SO2. In the combined influence of sodium chlo-
ride on the steel surfaces and sulphur dioxide in the atmos-
phere a synergic effect was noticed at 90% relative humidity.
A t 70% relative humidity no influence of an S02-supply of
1 ~ . ( gSO, ' c.m-2 . h-' o n the corrosion of steel samples with
sodium chloride crystals on their surfaces could be observed,
Introduction
The atmospheric corrosion of steel is influenced by several
factors e.g. t h e S02-concentration, t h e temperature, the re-
lative humidity and t h e amount of precipitation. I n coastal
regions t h e amount of saline particles in t h e air i s of great
importance t o t h e extent of the atmospheric corrosion. In-
vestigations performed by Ambler and Bain ( 1 ) in Nigeria
showed that atmospheric corrosion decreased with decreas-
ing salt content in the air. In an investigation in Great Britain
no correlation between t h e amount of deposited chloride
particles and the extent of corrosion could be found ( 2 ) .
Sereda (3) considers that a threshold value of the chloride
content must be exceeded for t h e corrosion caused b y de-
posited chloride particles t o increase appreciably.
The initial corrosion of steel is often higher in industrial
environments than in coastal regions but t h e rate of corrosion
decreases more rapidly in industrial atmospheres. After an
exposure of two years t o a marine environment carbon steel
corrodes more rapidly than if exposed t o a n industrial en-
vironment (4, 5 ) . Coburn et a]. ( 6 ) also made a comparison
between different types of corrosion environments and
*) Present address: Avestl: Jernverk, P.O. Box, S-77401 Avesta 1 ,
Sweden.
Werkstoffe und Korrosion 29,400-403 (1978)
Kur zfassung
Untersucht wurde der Einjluj3 von Natriumchlorid auf
die atmospharische Korrosion von Stahl. Dabei wurden mit
Natriumchloridkristallen (8 pg NaCl . c m - 2 ) bedeckte Pro-
ben einer schwefeldioxidfreien A tmosphare verschiedener
relativer Feuchtigkeit ausgesetzt. Die kombinierte Wirkung
von Natriumchloridkristallen auf der Stahloberflache und
Schwefeldioxid in der Atmosphare wurde bei einer Sch wefel-
dioxidzufuhr von 1 pg SO, . c m -2* Std-' ( 1 p p m S O 2 ) uriter-
sucht. Der Angriff wurde durch kontinuierliches Wiegen ver-
folgt, und die Korrosionsprodukte wurden mittels REM-
Technik untersucht. Bei der Bewitterung der mit Natrium-
chlorid bedeckten Proben in schwefeldioxidfreier Atmospha-
re steigt die Korrosion mit der relativen Feuchtigkeit (58 bis
90x1, wobei die Kurve durch ein scharfes Minimum bei 87%
rel. Feuchtigkeit unterbrochen ist. Die REM-Bilder zeigen,
daj3 bei hoher rel. Feuchtigkeit saulenformige Korrosions-
produkte entstehen, wahrend bei niedriger rel. Feuchtigkeit
fadenformige Korrosion auftritt. Bei 90% Feuchtigkeit ist
die Korrosion von sauberen Stahlproben bei einer Schwefel-
dioxidzufuhr von 1 pg- c m - 2 . Std-' starker als die von
mit Natriumchlorid bedeckten Proben in schwefeldioxid-
freier Atmosphare. Bei Anwesenheit beider angreifender
Mittel tritt bei 90% rel. Feuchtigkeit ein synergistischer
Ejyekt auf, wahrend bei 70% rel. Feuchtigkeit die Korrosion
der mit Natriumchlorid bedeckten Proben durch Schwefel-
dioxid nicht beeinflu$t wird.
found that during extended exposures, steel corrodes more
in a marine environment than in an industrial region.
Chlorides deposited on metal surfaces can originate from
industrial plant. This fact can explain the high chloride con-
tent in the corrosion products of steel exposed far from the
sea ( 3 ) . A combined influence of chlorides deposited o n the
surface and sulphur dioxide in the atmosphere is considered
to cause extensive corrosion (7).
Laboratory investigations with different salt particles
showed that the rate of corrosion increased considerably
above a certain critical value of the relative humidity ( 8 , 9).
This critical value is equivalent t o the relative humidity above
a saturated solution of ;he actual salt. For sodium chloride
this value is 7 8 % a t 20 C ( 8 ) . However, steel with sodium
chloride particles o n its surface corrodes at below 7 8 %
relative humidity. This is probably due to the fact that mois-
ture is bound to the yteel surface a t the salt particles by ca-
pillary action (10). Thus, in t h e presence of sodium chloride
particles on a steel surface, corrosion can take place at relative
humidities that would not have caused corrosion o n clean
surfaces ( 1 1). The extent of corrosion is dependent on the
amount of moisture that has been adsorbed by the salt (1 I ,
12). Sea salt particles cause corrosion a t lower relative hu-
midities than sodium chloride particles do, which results
Werkstoffe und Korrosion 29, 400-403 (1978) The influence of sodium chloride on the atmospheric corrosion 401

from the fact that sea salt contains very hygroscopic mag- W
nesium salts ( 13). irng cm?
In a laboratory investigation with accurately determined 90 7 0
amounts of sodium chloride on a steel surface the rate of
corrosion increased with increasing relative humidity up to 04 85%
90%, interrupted b y a sharp minimum at 87.5% relative 96 %
humidity ( 14). a 1%
Studies concerning t h e initiation of the sodium chloride 86%
87%
induced corrosion of steel showed that the attacks started
where concentrated salt drops had accumulated ( 1 5 , 16). 03 75%
78%

Experimental

The mild steel had the following chemical analysis: 02 58%

%C % Si % Mn % 1' %S
0.088 0.1 6 0.50 0.035 0.01 8
% Cr % Ni % Mo %cu %N
01
0.02 <O.OI <0.01 0.1 0.005

The samples for the weight increase measurements had

the dimensions 40 x 2 5 x 2.0 mm and were polished on
600 emery paper with water and ethanol. They were finally
cleaned in ethanol using ultrasonic agitation and dried with 100 500 time l h l
hot air. The samples for the SEM-study were round with a
Fig. 2. The wcight increase of steel samples with sodium chloride
diameter of 1 1mm. They were treated as described above (8 pg NaCl . exposed to an SOz-free atmosphere at different
but were electrolytically polished and etched in addition. relative humidities
A solution of 1.5 wjo sodium chloride was sprayed on
to the sampie surfaces and the samples were immediately Abb. 2. Gcwichtszunahrne von rnit Natriumchlorid (8 p g NaCl ' cm-')
bedeckten Proben in schwefeldioxidfreier Atmosphkre bci verschie-
dried at 80 C. The sodium chloride crystals o n t h e surfaces denen relativen Feuchtigkeiten
had a size of about 0.01 m m and the amount of salt was
*
8 1 pg NaCl per cm2 surface area.
The S02-exposures were performed with equipment
shown in Fig. 1. Purified air was saturated with water vapour
w
irng c d
:lean air
SO? "2
I lrom cylindei

Or,
Flow meters

"

Thermosbled exposure chamber

Fig. 1. The measuring device Abb. 1. Mefhorrichtung

03
at a temperature that gave a desired relative humidity at
22.0 O C in the exposure chamber. The humid air was mixed
with a gas mixture of 100 ppm SOz inonitrogen. The final
gas mixture was thermostated at 22.0 C and entered the
exposure chamber. The SO2-supply to the samples was
1 pg SO2 per cm2 surface area and hour, which was equiva-
lent to 1 ppm S O 2 . Th: exposures t o humid S02-free air 02
were performed at 22 C above saturated salt solutions in
closed chambers. The corrosion attacks o n the samples were
followed by periodic weighings. The SEM-technique was
used to study the appearance of t h e corrosion products
formed. 50 60 70 80 90 130 R H 1%1

Fig. 3. The weight increase of steel samples with sodium chloride

Results
The increases in weight of the samples with sodium
chloride crystals have been measured at different relative
humidities. The results are shown in Fig. 2. Each curve in
(8 fig NaCl . cm-') exposed to humid S02-free atmospheres for
670 hours
Abb. 3. Gewichtszunahrnc von mit Natriumchlorid (8 pg NaCl . ern-*)
bedeckten Proben in feuchtcr schwcfeldioxidfreier Atmospharc
(Versuchsdaucr 670 Stundcn)
402 Ericsson Werkstoffe und Korrosion 29, 400-403 (1978)

Fig. 2 is based o n weight increase measurements of at least

three samples. At all relative humidities except at 58%, the
lowest relative humidity that was used in this investigation,
the rate of corrosion was highest at t h e beginning of the
exposure and decreased as corrosion products were formed
o n the st eel surface.
In Fig. 3 the weight increase after an exposure of 670
hours has been plotted as a function of t h e relative humidity
a t th e exposure. The rate of corrosion increased with in-
creasing relative humidity up t o 85%. After that there was
a sharp minimum at about 87% relative humidity. The
highest rate of corrosion was measured at 90% relative
humidity and above that value the corrosion decreased.
Fig. 5 shows a steel surface with sodium chlorid par-
ticles exposed t o a n S02-free atmosphere at 96% relative
humidity for 4 hours. The surface shows extensive corrosion
attacks. At certain places characteristic tower shaped cor-
rosion products rising perpendicularly t o the surface were
formed. At lower relative humidities this type of attack was
not present but filiform corrosion was observed instead;
see Fig. 6.
The results from the exposures with sulphur dioxide are
shown in Fig. 4. At 90% relative humidity the SOz-induced
corrosion was considerably higher than that induced by
Fig. 5. A steel surface with sodium chloride crystals exposed to
sodium chloride alone. The combined effect of sulphur di- an SO2-free atmosphere at 96% relative humidity for 4 hours
oxide and sodium chloride caused the greatest corrosion.
At 70%)relative humidity t h e SO,-supply had a negligible Abb. 5. Mit Natriumchloridkristallen bedeckte Stahloberflache ndch
effect o n the corrosion of samples with sodium chloride. 4 Std. in schwefeldioxidfrcicrAtrnosphire bei 96% rcl. Feuchtigkeit
Samples without sodium chloride did not corrode measur-
ably o n SOz -exposure at 70% relative humidity.

W
Img cni*~

50

Fig. 6. A steel surface with sodium chloride crystals exposed to

an SOz-free atmosphere at 78% relative humidity for 8 hours
1- 100 500 time Ihl Abb. 6. Mit NatriumchloridkriFtalIen bcdcckte Stahlobertliche
Fig. 4. The weight increase of steel samples exposed at different nach 8 Std. in schwcfeldioxidfreier Atmosphkc bei 78% rel. I:euch
corrosive conditions tigkeit
A n: 8 pg NaCl . cm;2 and 1 pg SO, ' cm-2 * h-' (1 ppm SO?)
: 1 pg SO2 . cm-L . h-' (1 ppm SO,)
0 0: 8 pg NdCl. Cm-'
Filled symbols: 90% relative humidity Discussion
Open symbols: 70% relative humidity
The decrease of the rate of corrosion that was found at
Abb. 4. Cewichtszunahme von Stahlproben bei verschiedenen Kor-
about 87% relative humidity according to Fig. 3 has been
rosionsbedingungen
A A: 8 pg NaCl . und 1 pg SO2 ' cm-' . h-' (1 pprn SO2) reported earlier ( 1 4). Then it was stated that filiforrn cor-
1 pg SO2 ' cn-2 . h-'
0: (1 ppm SOz) rosion in its most extreme form appears a t about 87% relative
0 0:8 pg NKI . cm-' humidity.
gefiillte Symbolc: 90% rel. E'cuchtigkeit A t 96% relative humidity, characteristic tower shaped
offene Symbole: 70%rel. Feuchtigkeit corrosion products, shown in Fig. 5 , were formed after a
Werkstoffe und Korrosion 29, 403 -408 (1978)
Overlag Chemie, GrnbH, D4940 Weinheim, 1978
short time of exposure. These products have an appearance
similar t o t h e corniform attacks that Hill ( 1 7) reported. The
sodium chloride induced corrosion at 78% relative humidity
totally lacked these tower shaped products and caused fili-
form corrosion products instead.
At 90% relative humidity a synergic effect of the com-
bined influence of sulphur dioxide and sodium chloride was
obtained. At 70% relative humidity 1 pg SO, . crn-2 . h-'
did not cause a measurable corrosion o n t h e samples without
sodium chloride crystals. It is notable that at 70% relative
humidity with a combined influence of sulphur dioxide and
sodium chloride, the corrosion did not increase compared
to the influence of sodium chloride alone.
This investigation has shown the great importance of
sodium chloride for t h e atmospheric corrosion of steel. Thus
sodium chloride particles o n a steel surface can cause corro-
sion at relative humidities which have been considered too
low to start SO,-induced corrosion.
Summary
On exposure of steel samples with their surfaces covered
with sodium chloride crystals (8 p g NaCl cm-') in an SO,-
free atmosphere the amount of corrosion increased with in-
creasing relative humidity from 58% to 90% with the excep-
tion of a decreased rate of corrosion at about 87% relative
humidity.
At 96% relative humidity the sodium chloride induced
corrosion resulted in tower shaped products that rose at
right angles from the steel surface. An exposure at 78% re-
lative humidity caused filiform corrosion.
Korrosionsschutz in Rohren durch Betonauskleidungen")
Corrosion protection in pipelines by concrete linings
Von G. Nagel**)
Kurzf assung
Guj3eiserne Rohre verschiedener Durchmesser sowie Stahl-
rohre konnen durch Auskleidungen aus Zementmortel wir-
kungsvoll gegen Korrosion geschutzt werden. Dabei ergibt
sich gleichzeitig eine Verbesserung der mechanischen Eigen-
schaften des Verbundes. Der Mortel erreicht innerhalb von
28 Tagen Druckfestigkeiten von 700 bis 1000 k g lcm 2 , Bie-
gezugfesrigkeit uber 100 kglcm' und vtrmittelt auch bei
Schaden am Metallrohr hohe Druckfestigkeit. Die Haftung
zwischen Ze me n t und Metal1 verbessert sich mit der Zeit:
der frische Zement ist ziemlich poros, so daf3 Wasser bis z u m
Metall durchdringen kann; das Calciumhydroxid aus d em
Zement und die Ferrihydrate auf der Metalloberfkiche wer-
den dann in den Beton eingebaut, so daf3 sich eine gute Ver-
zahnung ergibt. Sulfatbestandige Spezialzemente halten in
A bwasserteitungen Sulfatkonzentrafionen bis 3000 mgll
und kurzfristig auch pH-Werte bis 4,s aus.
*) Vortrag, gehalten auf
der 'Pagung ,,Korrosion durch Meerwasser"
an der Fachhochschule Wilhelmshaven, 28. bis 30. April 1977.
Korrosionsschutz in Rohren durch Bctonauskleidungen 403
A synergic effect of the combined influence of
1 pg SOz . crn-' . h-' ( 1 ppm S O z ) and sodium chloride
crystals o n the sample surface could be noticed at an ex-
posure at 90% relative humidity, However, at 70% relative
humidity sulphur dioxide did not affect t h e corrosion of
steel with sodium chloride particles o n its surface.
(Eingegangen: 4. 1 1. 1977)
References
1. H. R. Ambler and A . A. Bain: J. Appl. Chem. 5 (1955) 437.
2. H. R. Ambler: J. Appl. Chem. I0 (1960) 213.
3. P. J. Sereda: ASTM STP 558 (1974) 7.
4. C. P. Larrabee: Corrosion 9 (1953) 259.
5. C. P. Lurrabee: Trans. electrochem. SOC.87(1945) 161.
6. S. K. Coburn et al.: ASTM STP 435 (1968) 360.
7. H. R. Copson: ASTM Roc. 45 (1945) 554.
8. A. Bukowiecki: Schw. Arch. 23 (1957) 97.
9. A . Bukowieckiand B. G. foshi: Schw. Arch. 32 (1966) 42.
10. B. Sanyal and D. V. Bhadwar: J. Sci. ind. Res. 2 1 0 (1962) 243.
11. K. A . Chandler: Brit. Corros. J. I (1966) 264.
12. S. f. Duly: J. SOC.Chem. Ind. 6Y (1950) 304.
13. L! R. Evans and C. A . J. Taylor: Brit. Corros. J. Y (1974) 26.
14. T.Berglund: TVF 31 (1960) 325.
15. W. S. Patterson and f. H. Wilkinson: J . SOC.Chem. Ind. 5 7
(1938) 445.
16. J. F. Henriksen: Corros. Sci. 9 (1969) 573.
17. G. V, G. Hill: Brit. Corros. J. 8 (1973) 128.
W 1970
Summary
Cast iron and steel pipes of various diameters may be pro-
tected against corrosion b y internal linings o f cement mor-
tar. This lining results in addition in an improvement o f the
mechanical properties of the composit. Within 28 days the
mortar cures to compressive strength between 700 to
1000 kg/cmz and tensile fle x strength exceeding 100 kgjcm',
even when the metal tube is damaged a high pressure resis-
tance is maintained. Adhesion between lining and metal im-
proves with time: green cement is rather porous, so that
water may pmetrate to the metal; calcium hydroxide from
the cement and ferric hydrates formed on the metal surface
are incorporated in the concrete so that good adhesion re-
sults. Sulfate-resistant special grades have shown to be resis-
tant in waste water containing up t o 3000 mgfl sulfate and
having pH-values down to 4,5 for short times.
**) Ing. grad. Gerhard Nagel, Stadtoberbaurat, Tulpenstr. 1,
75 15 Linkcnheim-FIochstetten.