TROUBLE with PAINT
Non-Osmotically-Induced
Blistering Phenomena
on Metal
L
ast month, we discussed one low and red iron oxide over the
of the most common types of same immersion period. In less than
blistering of paint films on 11 days, the latter systems delami-
metal鈥攐smotic blistering. Now we nated entirely. Because both binder
will consider other forms of blister- system and pigment volume concen-
ing, including electroendosmosis, tration (PVC) were normalized in
cathodic blistering, cathodic dis- all systems, it was believed that
bondment, and compressive stress- electroendosmotic effects and elec-
related blistering. trolytic resistances of the film were
associated with pigmentation, not
Electroendosmosis the binder.
In the 1940s, Kittleberger and Elm Although the phenomenon is pri-
noted that blisters on panels im- marily seen under electrolytic im-
mersed in sea water occurred only mersion (i.e., sea water immersion),
after areas of the panels had begun it is often noted on scribed panels in
to corrode, either because they had salt spray testing (Fig. 1).
been incompletely coated, or be- Some authors have concluded that
cause the coatings were breaking once corrosion begins, electroendos-
down at cuts and abrasions. Totally motic blistering may be a primary
protected panels showed no blister- driving or accelerating force for blis-
tering.1,2,3 This seems to be the case
Fig. 1 - Electroendosmotic blistering around a
ing. The researchers later confirmed
cut scribe and discrete corroding points.
that the blistering of these panels for the oil paints used by Kittelberg-
Figures courtesy of the author
was electroendosmotic in origin, er and Elm, but the justification of
caused by the establishment of a such conclusions is generally open
paint films on steel immersed in salt
naturally occurring electrical gradi- to some dispute. Funke, for exam-
ent across the film.1 During elec- water and the extent of associated ple, believes that this type of driving
blister formation. They found that
troendosmosis, or electro-osmosis, force is much less universal than is
the osmotic force. 4 More work is
levels were much greater when the
water is forced through the paint
panels were electrically coupled to
film by an electrical potential gradi- needed to discover the importance
freely corroding (anode) panels than
ent. The flow is toward the elec- of electroendosmosis in blistering.
when the panels were uncoupled. It
trode that has the same charge as
was shown that electroendosmotic
that acquired by the immersed film. Cathodic Blistering
influences accounted for between 95
(This may not be the same with all As discussed in the December 1997
and 99 percent of all water absorbed
films or in all liquids.) Most films ac- column, alkali is produced at the
by the coatings involved, and that
quire a negative charge in water, cathode under basic and neutral
under these conditions, osmotic ef-
and the metal substrate around the conditions as part of the corrosion
fects were minimal. Pigmentation
corroding area is cathodic and there- process. This reaction is expressed
as O2 + 2H2O + 4e- = 4(OH-).
was also found to be important to
fore rich in electrons. Thus, the
the process. Oil paints pigmented
water is naturally drawn through the Cathodic blistering is caused by
with red lead and zinc chromate
film, so that blisters will occur this electrochemical reduction of
showed much lower levels of water
around the corroding site. oxygen beneath intact (and some-
absorption and blistering than did
The 2 researchers measured the times defective) coatings. Water,
systems pigmented with chrome yel-
amounts of water absorbed by oil continued
17
JPCL 鈥? PMC / MARCH 1998
Copyright 漏1998, Technology Publishing Company
TROUBLE with PAINT
beneath the coating film, the hy-
droxyl ions are trapped by the semi-
Oxygen Water Cations permeable film at the site of adhe-
sion loss (the incipient blister).
Conductive
There, ionic accumulation equates
electrolyte
with increasing pH.
(salt water)
Although the actual mechanism of
film delamination from the coated
Paint film
cathode is only partially understood,
Substrate it seems likely that the hydroxyl ions
may attack the coating binder, the
metal oxide surface, or the bond be-
At the cathode, oxygen reduction [O2 + 2H2O + 4e- = 4(OH-)] creates an increasingly
tween the two. A loss of adhesion in-
alkaline condition.
volving the alkaline hydrolysis of
Alkali attacks metal oxide and paint film as well as any conversion coating
(e.g., zinc phosphate), producing delamination and expansion of the blister. coating films based on ester and pos-
Where film is defective and corrosion is occurring at a nearby site, the blister may be sibly amide (including urethane) link-
entirely cathodic, with electrons being supplied from the nearby corroding site.
ages is reasonable. Or the alkali may
dissolve phosphate conversion coat-
ings on pretreated steel, leading to
Fig. 2 - Cathodic blistering
undercutting of the coating system.
The saponification and solubiliza-
oxygen, and alkali metal cations tions. Anodic sites may be located at tion of oil paint films around the
(e.g., sodium cations) diffuse the center of the blister or in defec- scribed area of salt spray panels is
through the coating to cathodic sites tive films at nearby corroding sites common. Carboxylated residues
to produce strongly alkaline solu- (Fig. 2). As the reaction is localized continued
18 MARCH 1998 / JPCL 鈥? PMC Copyright 漏1998, Technology Publishing Company
TROUBLE with PAINT
from this type of attack have been area of the film immediately adja- alkali attack on the metal oxide sur-
cent to the steel. 6 This indicates face7, with pH values up to 14 being
identified beneath cathodic blisters
in melamine cross-linked epoxy breakdown of the ester to the rele- noted at cathodic delamination
ester and polyester films.5 Byrnes sites. 8 These values are basic
vant alcohol and (presumably) acid.
examined infrared spectra of vinyl As discussed in our February col- enough to solubilize ferric ions and
chloroacetate linings after exposure umn, alkaline hydrolysis can result disrupt the metal oxide surface.
to excessive cathodic protection po- in propagation of adhesion loss by The effect of alkali on the in-
tential. The data show depression of film undercutting around the periph- creased rate and magnitude of water
the carboxylic ester peaks and in- eries of simple osmotic blisters. take-up by polymers was noted in
creased hydroxyl functionality in the However, there are also reports of the September 1997 column. Thorn-
ton et al. concluded that the greater
degree of water saturation of the
film resulting from increased hy-
droxyl ion concentration may exac-
erbate de-adhesion.9 Then, mechan-
ical distortion of the film (leading to
rupture of interfacial bonds) and the
pooling of water at the interface are
likely to propagate adhesive loss.
Cathodic Disbondment
Where steel is to be buried (e.g., un-
derground pipelines and tanks) or
immersed in water, it is normal for
the first line of defense against cor-
rosion to be cathodic protection via
impressed currents. The process es-
sentially protects the steel surface by
connecting the structure electrically
to a permanent anode (or anode
array). An impressed current is di-
rected between the steel and anode
in such a direction so that no current
is discharged from the steel to the
environment (soil, backfill, or water).
Thus, the steel is maintained as the
cathode of a huge electrochemical
cell and is thereby protected. Ca-
thodic protection by impressed cur-
rent systems is detailed by several
authors.10 The technique is highly
effective, but large steel areas require
large amounts of current to maintain
flow in the right direction. To mini-
mize the consumption of electricity,
the size of the steel surface exposed
to the environment is usually dimin-
ished by coating it (thus reducing the
area requiring protection to small
areas at bare spots).
In practice, the electrical potential
of the coated steel should be slightly
more negative than -0.8 volts (with
20 MARCH 1998 / JPCL 鈥? PMC Copyright 漏1998, Technology Publishing Company
TROUBLE with PAINT
reference to a copper/copper sulfate cause of this risk, the anodes are to protect potable water tank interi-
reference electrode), usually about often shielded with special devices ors. Failures occur, however, in many
-0.85 volts. This effectively maintains to withstand the cathodic conditions instances because the systems are im-
near the anode.11
protection at remote locations in properly installed, misused, or im-
the presence of holidays, locally Impressed current systems are also properly maintained. Anodes are left
uncoated areas, and steel protru- widely used for protection under unconnected or not replaced, or sys-
sions, where the current may drain fresh water. Again, the same over- tems are not switched on. The instal-
to earth. voltage limitations apply. These sys- lation must be made and maintained
However, the applied voltage can- tems are often used along with suit- by competent professionals.
not be too high, for high over-volt- able coating systems (usually epoxies) continued
age potentials (more negative than
about -1.1 volts with reference to the
same electrode) may produce blister-
ing and delamination of the coating
in areas around the same bare spots
and defects.6,11,12 Such delamination
increases the areas of bare steel on
the structure and therefore the cur-
rent requirement and cost needed to
maintain cathodic protection.
Cathodic delamination in under-
water systems (such as ships鈥? hulls),
which employ sacrificial zinc or alu-
minum anodes with fixed potential
differences with respect to the steel,
are not as liable to produce the
same degree of over-voltage poten-
tial as are impressed current sys-
tems. However, cathodic disbond-
ment is not an unusual coating
failure under these conditions.
The voltage of the structure with
reference to a copper/copper sulfate
reference electrode must be con-
stantly monitored and adjusted auto-
matically or manually to changing
environmental conditions if the non-
corroding condition is to be safely
maintained without developing an
over-voltage. This may be particular-
ly important where the vessel may
move through waters having differ-
ent salinities, and, therefore, differ-
ent resistivities. The applied poten-
tial must be decreased as the vessel
moves into water of lower resistivity
(higher salinity). In high resistance
electrolytes, the protected area may
become quite limited unless the ap-
plied potential is increased. The in-
creased potential may severely en-
danger the coating system in the
immediate areas of the anode. Be-
21
JPCL 鈥? PMC / MARCH 1998
Copyright 漏1998, Technology Publishing Company
TROUBLE with PAINT
For example, it is important that wall effects occur most often at the
the reference electrode be properly interface of the metal and the coat-
placed in the same general area as ing. However, failure may also occur
the anode array. Dubcak reports an at a primer/finish coat interface.
instance of severe cathodic blistering Cold wall effects may also occur
in a tank where voltage readings at on non-immersed exterior coatings
the cathodic protection controller of tanks containing cold water or
box were no higher than -0.9 mV.13 fuel in humid environments.18 The
In this case, the reference electrode propensity of the phenomenon to
was electrically shielded by its instal- show up on storage tanks is proba-
lation in a well, separate from the bly related to the heat sink proper-
anodes. High current readings (4.0 ties of the large body of liquid with-
mA) tipped the analyst that some- in the tank. Cold wall blisters are
thing was wrong, and over-voltage often very large. The author has
measured at anodes well removed noted large blisters 10 in. (25 cm) in
(shielded) from the reference elec- diameter on the coated exterior
trode was found to be as high as walls of fuel storage tanks contain-
-2.7 mV. When the reference elec- ing cold fuel in summer (Fig. 3).
trode was raised into the belly of the When these large blisters were
tank, closer to the anode array, cur- punctured, copious amounts of
Fig. 3 - Blistering on tank exterior induced
rent demand dropped markedly, and water flowed out. This occurred in
by thermal gradients. (Inside fluid was cold;
potential readings for all anodes New York State on a dry day, fol-
outside environment was warm and humid.)
dropped to near -1.0 mV. When the lowing a relatively humid spell of
tank was repaired, the new align- weather. Delamination occurred at
ment of the reference electrode was seen on the coated interior surfaces the primer/finish interface but not at
standardized. There has been no fur- of cold tanks containing warm the metal itself. Obviously, this inter-
ther blistering. water. Hendry reports examples of face was less secure than was that of
As Leidheiser states, good protec- the phenomenon in the Middle East the thin film of primer to the metal.
tion against cathodic disbondment where blisters appeared on the inte- It seems likely that osmotic effects,
(in systems with cathodic protection rior walls of tanks containing warm, probably related to inhibitive pig-
distilled or deionized water.16 Ap-
and in those without) requires that ments within the primer, also played
the interfacial region be highly resis- parently, water is absorbed into the a role in this failure. It is possible
tant to alkali.14 Optimizing the alkali coating at the warm side of the paint that during periods of high humidi-
resistance of the coating film (elimi- film (away from the substrate). ty, the film takes up water vapor,
nating hydrolyzible groups such as Water condenses at the interface which condenses against the cold
esters, amides, urethanes, and urea with the metal, where the tempera- metal interface on the exterior of the
tank.19 Subsequently during the heat
linkages), perfecting wet adhesion, ture and therefore the permeability
improving dielectric strength, and of the film are low enough to pre- of day, the condensed water may
minimizing film transport properties vent the same rate of movement expand and attempt to evaporate,
seem to be avenues toward this back into the film towards the high- developing high pressures over a
goal. At a high enough potential er temperature interface. (Thermo- relatively short time on the under-
(more negative than -1.0 volts vs a dynamic activity of the water/paint side of the impermeable film, which
standard calomel electrode), hydro- film model will be greater at the may lead to the physical expansion
gen formation is also possible, i.e., warm face of the coating than at the of the blister.
2H2O + 2e- H2 + 2OH-. The cold face next to the metal.) Presum- Cold wall phenomena are intensi-
evolution of hydrogen gas may itself ably, water must condense at a point fied by the relative thermal conduc-
force the film from the steel. of weakened adhesion (perhaps a tion properties of the metal (excel-
point of contamination or stress cen- lent) and the coating (poor). Small
Thermal Gradient-Induced ter) and there accumulate. temperature differentials may pro-
Blistering (Cold Wall Effect) Tator reports studies indicating duce cold wall blistering. It is most
Blisters result from water condensa- coatings with low permeability are often eliminated by proper tank or
tion produced by thermal gradients more resistant to cold wall effects pipe insulation.
across coatings.15 This effect is often than more permeable ones.17 Cold continued
23
JPCL 鈥? PMC / MARCH 1998
Copyright 漏1998, Technology Publishing Company
TROUBLE with PAINT
common.4 However, they are possi-
ble. Discrete, dry blister formations
in a marginally adherent latex paint
system over coil-coated aluminum
stock have been noted by the author
as a response to the application of
heat alone. On cooling, these blis-
Fig. 4 - Steel exposed
ters disappeared.
by thermal stress-
Leidheiser studied the response of
induced blistering.
(Note circular patterns an epoxy film to swelling stresses
relating to stepwise produced by exposure to 0.1M
progression of failure.)
H 2 SO 4 at 60 C (140 F). 22 He ob-
served that blister formation was re-
This type of failure looks different
Discussing the same phenomenon lated to adhesion. When the sub-
than other forms of blistering. Often
on the coated interior walls of tubes, strate was abraded, the film
in thermosetting systems such as
Schwenk found that the 鈥渋ncubation expanded in a single large blister;
epoxies, urethanes, polyesters, and
time for the onset of blistering鈥? in- when the substrate was blast
older alkyds, de-adhesion generally
creased as the temperature gradient cleaned (presumably resulting in
occurs in large, irregular flakes and
between the warm water on the in- greater film adhesion), several much
sheets rather than as discrete hemi-
side of the pipe and the external en- smaller blisters were noted. The ef-
spherical blisters. The interiors of
vironment was reduced, either by fect is often seen in both acid and
these stress-induced 鈥渂listers鈥? are
reducing the water temperature solvent resistance testing.
quite dry, and the domes are less
within the pipe or (presumably) by Martin et al. propose a non-osmot-
defined. The delaminating film seg-
increasing the temperature of the ex- ic model for defect-controlled ca-
terior environment.20 The danger of ments have a typical convexity to- thodic disbondment that relies on a
wards the coating/air interface (con-
peeling could also be reduced by sort of stress corrosion cracking
cave to the substrate), which betrays
using a thicker coating in the tube. process. In the model, the simulta-
response to compressive stress. Usu-
Corrosion is not expected to im- neous effects of in-plane compres-
ally, such delaminations can be de-
mediately follow the onset of cold sive stress and alkaline attack con-
tribute to de-adhesion.23
tected before any actual flaking oc-
water blistering, and will largely de-
curs; an audible response to a
pend on the rate of oxygen trans-
tapping of the areas with a finger in-
mission through the film. Other Blistering Phenomena
dicates some intra-system cavity.
Several humidity tests (e.g., ASTM Other blistering phenomena on steel
Failure may sometimes progress out-
D 4585, Standard Practice for Testing are not as conveniently classified, al-
wards from a central point of initia-
Water Resistance of Coatings Using though one or more of the above-
tion, leaving a series of concentric
Controlled Condensation, and ASTM mentioned mechanisms (including
cracks (Fig. 4).
C 868, Standard Test Method for osmosis) may be involved. An ex-
In thermoplastic systems (especial-
Chemical Resistance of Protective ample of such is the Dia Phenome-
non discussed by Van Laar.24 In this
ly where the ambient temperature is
Linings) are based on blistering in-
above the glass transition tempera-
duced by thermal gradients. effect, steel coated with a paint sys-
ture ratio), the films are often too tem predisposed to blistering in
elastic to blister. However, if
Blistering and Delamination of fresh water was observed to bear
stretched beyond their yield point,
Films from Compressive Stress blister-free areas on coated surfaces
the films may exhibit wrinkling on
In 1964, Brundt postulated that blis- corresponding to bare (freely cor-
removal of the stress (cooling or
tering could result from the paint roding) areas on the back of the
drying out). There may be some ad-
film鈥檚 response to compressive stress same panel. Simultaneously, corro-
hesive recovery at this point, al-
in the absence of osmosis or any sion was noted beneath the non-
other driving force.21 Where the ad- though if adhesion is very poor, the blistered areas. Areas of the same
film may be removed in large, flexi-
hesive strength of the coating was panel coated on both sides exhibit-
ble sheets of paint.
unable to accommodate the swelling ed heavy, water-filled blisters. The
Blisters having discrete, localized
of the film from hygroscopic stress, effect was originally ascribed to the
formation (that may be related to
thermal stress, or other factors, the diffusion of atomic hydrogen from
continued
compressive stress alone) are un-
film would delaminate as a blister.
24 MARCH 1998 / JPCL 鈥? PMC Copyright 漏1998, Technology Publishing Company
TROUBLE with PAINT
the freely corroding back areas ic blistering, which may be confirmed while dithizone may be an indicator
through the steel to the coated side by the presence of a fluid of high for zinc cations.
of the panel, where on reaction with pH. Blister formation between coats Oils, grease, perspiration, and simi-
oxygen, it causes the establishment may contain soluble residues derived lar contaminants can lead to coating
of oxygen concentration cells. While from a primer or intermediate coat. failure. Specifically patterned or irreg-
there seem to be several problems These residues may be detected by ular delamination is more likely than
with this interpretation, the phenom- atomic absorption or wet chemical discrete, symmetrical blistering. The
enon deserves more investigation. analysis; diphenyl carbohydrazide presence of moisture and dew during
Similar front side/back side effects may be an indicator for hexavalent recoating may also cause problems;
were noted by Hare when investigat- chromium compounds, for example, continued
ing the corrosion resistance of latex
paint on steel.25 The intensity of cor-
rosion on the front face of the coated
panel was much more intense when
both sides of the panel were com-
pletely coated than when the back
face of a panel bearing the same sys-
tem was allowed to corrode freely. In
this case, the phenomenon was at-
tributed to protective galvanic effects.
These effects were provided by the
freely corroding bare anode side of
the panel to the coated side of the
panel (cathode). They were not pre-
sent on the totally coated panel,
which corroded by local cell action.
Blister Analysis
Much can be revealed about the
cause of blistering from a study of
the blister interior, often found to
contain fluid. When a blister breaks,
the fluid may be ejected with some
force, indicating the presence of a
pressure gradient. The fluid may be
extracted with a syringe and exam-
ined for pH and inorganic or organic
residues. Solvents and carboxylic acid
residues may be identified by gas
chromatography. Often, the film will
reveal entrapped solvent by a defi-
nite odor that is detectable as the
blister is opened. The appearance of
the steel surface is also telling. The
presence or absence of visible scarifi-
cation (and, therefore, maximized or
compromised adhesion) and the con-
dition of the steel give clues about
the cause of blistering. Bright steel
may indicate the presence of active
inhibitors from the primers. (Chro-
mate may give yellow fluids.) Uncor-
roded steel may also indicate cathod-
27
JPCL 鈥? PMC / MARCH 1998
Copyright 漏1998, Technology Publishing Company
TROUBLE with PAINT
these problems are more difficult to
determine on recoats than on the steel
(which may exhibit corrosion). Mois-
ture contamination is sometimes made
evident by other problems: cissing and
crawling of the finish, and bubbling
and foaming with urethanes. Many of
the more polar systems today (espe-
cially the primers) may partially dis-
place or take up water.
Blisters may contain no water. In
this case, the possibility of other
forms of distress should be investi-
gated, such as the generation of
gases by the coating. Bubbling in-
duced by entrapped gas (CO 2 in
urethane films) will result in dry
blisters, although the blister will
form within the distressed film rather
than at the interface.
Blistering and Delamination
Blister formation is preceded by loss
of adhesion, which was identified as
continued
28 MARCH 1998 / JPCL 鈥? PMC Copyright 漏1998, Technology Publishing Company
TROUBLE with PAINT
a necessary precursor to corrosion Organic Coatings, ed. H. Leid- Corrosion Induced Paint Adhe-
by Funke.26 Initiation of adhesion sion Loss,鈥? Journal of Coatings
heiser (Houston, TX: NACE Inter-
Technology (Aug. 1979), 45.
loss may be difficult to identify be- national, 1981).
fore blisters actually become visible. 6. G.B. Byrnes, 鈥淏listering of Im-
4. W. Funke, 鈥淏listering of Paint
Various techniques that seem mersed Coatings Under Cathodic
Films and Filiform Corrosion,鈥?
Protection,鈥? Materials Perfor-
promising in identifying the aspect Progress in Organic Coatings
mance (Sept. 1989), 31.
of de-adhesion have been used to (Sept. 1981), 29.
study both phenomena. These meth- 7. J.J. Ritter, 鈥淓llipsometric Studies
5. J.S. Hammond, J.W. Holubka,
ods include the crossover point on the Cathodic Delamination of
and R.A. Dickie, 鈥淪urface Analy-
continued
technique in the water transmission sis of Interfacial Chemistry in
of supported and unsupported films
developed by Funke27; and the in-
frared thermography techniques of
McKnight and Martin.28 Jin et al. in-
jected a fluid between the coating
and the substrate using a hydraulic
pump.29
Using tests of this type, it is not
only possible to quantitatively mea-
sure the adhesion of coated speci-
mens to steel as immersion (or salt
spray testing) progresses, but also to
identify recovery of adhesion as the
systems dry out after testing.
Other investigations are needed to
add to our understanding of blister-
ing and delamination.
Conclusion
This concludes our discussion of
blister formation in coating films on
metal in aqueous environments. The
next several columns will focus on
improving the understanding of the
mechanisms by which coatings im-
pede corrosion. 鉂?
References
1. W.W. Kittelberger and A.C. Elm,
鈥淲ater Immersion Testing of
Metal Protective Paints鈥擱ole of
Electroendosmosis in Water Ab-
sorption and Blistering,鈥? Industry
and Engineering Chemistry (July
1947), 876.
2. C.G. Munger, Corrosion Preven-
tion by Protective Coatings
(Houston, TX: NACE Internation-
al, 1984), p. 338.
3. D.M. Berger, R.J. Trewella, and
C.J. Wummer, 鈥淓valuation of
Linings for SO 2 Scrubber Ser-
vice,鈥? in Corrosion Control by
31
JPCL 鈥? PMC / MARCH 1998
Copyright 漏1998, Technology Publishing Company
TROUBLE with PAINT
Organic Coatings on Iron and
Steel,鈥? Journal of Coatings Tech-
nology (Dec. 1982), 51.
8. J.S. Thornton and R.W. Thomas,
鈥淎dhesive Systems鈥擠urability
Testing,鈥? Sonar Transducer Reli-
ability Improvement Program,
STRIP 4th Quarter Report, NRL
Memo Report, ed., R.W. Timme,
Oct. 1995.
9. J.S. Thornton, J.F. Cartier, and
R.W. Thomas, 鈥淐athodic Delami-
nation of Protective Coatings:
Cause and Cure,鈥? Chapter 15 in
Polymeric Materials for Corro-
sion Control, ed., R.A. Dickie
and F.L. Floyd, ACS Symposium
Series 322, (Washington, DC:
American Chemical Society,
1986), p. 169.
10. M.E. Parker, Pipe Line Corrosion
and Cathodic Protection, 2nd
Edition (Houston, TX: Gulf Pub-
lishing Co., 1962).
11. M.H. Bingham and P.W. Mann,
鈥淚mpressed Current Cathodic
Protection and its Effect on Ma-
rine Paint Systems,鈥? Journal of
Coatings Technology (March
1978), 47.
12. H. Leidheiser, W. Wang, and L.
Igetoft, 鈥淭he Mechanisms of Ca-
thodic Delamination from a
Metal Surface,鈥? Progress in Or-
ganic Coatings (Nov. 1983), 19.
13. T.O. Dubcak, 鈥淚nspecting Water
Tank Linings: The Importance of
the First Anniversary,鈥? Protective
Coatings Europe (April 1996),
26.
14. H. Leidheiser Jr., et al., 鈥淓nviron-
mental Control of Coatings Com-
position as a Means of Reducing
Cathodic Delamination of Or-
ganic Coatings,鈥? Journal of Coat-
ings Technology (Oct. 1984), 55.
15. H. Leidheiser and W. Funke,
鈥淲ater Disbondment and Wet
Adhesion of Organic Coatings
on Metal: A Review and Interpre-
tation,鈥? JOCCA (May 1987),
121.
continued
33
JPCL 鈥? PMC / MARCH 1998
Copyright 漏1998, Technology Publishing Company
TROUBLE with PAINT
16. C.M. Hendry, 鈥淒esigned Perme- heiser (Houston, TX: NACE, 25. C.H. Hare, 鈥淥bservations on the
ability of Micaceous Iron Oxide 1981), p. 103. Corrosion of Latex Painted Steel
Coatings,鈥? Journal of Coatings 21. N.A. Brunt, 鈥淏listering of Paint in Electrolytic Environments,鈥?
Technology (July 1990), 33. Journal of Paint Technology
Layers as an Effect of Swelling by
Water,鈥? JOCCA (Jan. 1964), 31.
17. K.B. Tator, 鈥淏listering of Coating (June 1975), 69.
Materials,鈥? The Coatings Consul- 22. H. Leidheiser, 鈥淢echanics of 26. W. Funke, 鈥淭he Role of Adhesion
tant Newsletter, Issue #2 (Pitts- Deadhesion of Organic Coatings in Corrosion Protection by Or-
ganic Coatings,鈥? JOCCA (Sept.
burgh, PA: KTA-Tator, Inc., from Metal Surfaces,鈥? Paper 12
in Polymeric Materials for Corro-
1987). 1985), 229.
18. C.G. Munger, Corrosion Preven- sion Control, ed., R.A. Dickie 27. W. Funke, 鈥淐orrosion Tests for
tion by Protective Coatings and F.L. Floyd, ACS Symposium Organic Coatings鈥擜 Review of
(Houston, TX: NACE Internation- Series 322 (Washington, DC: their Usefulness and Limita-
tions,鈥? JOCCA (Feb. 1979), 63.
al, 1984), p. 50. American Chemical Society,
19. M. Yaseen, 鈥淧ermeation Proper- 1986), p. 124. 28. M.E. McKnight and J.W. Martin,
ties of Organic Coatings in the 23. J.W. Martin, E. Embrose, and W. 鈥淒etection and Quantitative
Control of Metallic Corrosion,鈥? in Tsao, 鈥淣on Osmotic, Defect-Con- Characterization of Blistering
Corrosion Control by Organic trolled Cathodic Disbondment of and Corrosion of Coatings on
Coatings, ed., H. Leidheiser a Coating from a Steel Sub- Steel Using Infra-Red Thermogra-
strate,鈥? Journal of Coatings Tech- phy,鈥? Journal of Coatings Tech-
(Houston, TX: NACE Internation-
nology (Nov. 1990), 25. nology (Aug. 1989), 57.
al, 1981), p. 24.
20. W. Schwenk, 鈥淎dhesive Loss of 24. J.A.W. Van Laar, 鈥淏listering of 29. X.H. Jin, K.R. Gowers, and J.D.
Painted Steel,鈥? Paint and Varnish
Organic Coatings鈥擟auses and Scantlebury, 鈥淭he Effect of Envi-
Production (July 1961, Aug.
Consequences for Corrosion Pro- ronmental Conditions on the Ad-
tection,鈥? in Corrosion Control by 1961, and Oct. 1961), 31, 49, hesion of Paints to Metal,鈥?
Organic Coatings, ed., H. Leid- JOCCA (March 1988), 78.
and 41, respectively.
34 MARCH 1998 / JPCL 鈥? PMC Copyright 漏1998, Technology Publishing Company
|