Web http://www.jcse.org/

Volume 2 Paper 12

Submitted 13th September 1999


Degradation of Acrylic Coatings on Galvanized Steel

A. Forsgren and N. Steihed

Swedish Corrosion Institute Roslagv. 101, hus 25 SE-10405 Stockholm Sweden

Abstract

Trials have been conducted of four types of waterborne paints on various ferrous substrates. Samples were subjected to either a Scab test for two years, or ten days of salt spray testing.

Interestingly, the acrylic paints all performed worse on galvanized steel than on carbon steel. This contradicts previous experience of waterborne acrylics at the Swedish Corrosion Institute and elsewhere. It was not possible, using grazing-angle FTIR, to establish what had happened; infrared results seem to indicate that an alkali environment is necessary to the degradation of the coatings; but the role which the zinc plays is not clearly defined.

The conclusion drawn is not that waterborne acrylic coatings are unsuitable for galvanized substrates, but that further work is necessary to understand the interactions of zinc and waterborne acrylic polymers. Theories for why the zinc caused this include:

Either the initial adhesion to the zinc was poor (a polymer composition problem) or the zinc substrate was not wetted (a paint formulation problem). In either case, water and oxygen can reach the zinc surface and initiate corrosion of the zinc. This causes an alkali environment at the zinc-paint interface, causing saponification of the polymer.

The zinc is causing further crosslinking of the binder, depending on which monomers are used. Crosslinking goes too far and the coating becomes brittle. Weathering stresses such as temperature change and wetting/drying cycles break down the cured coating.

Zn++ ions destroyed the coulombic stability of the dispersion, leading to flocculation. The paint is in effect destroyed after being applied but before a coherent film can form.

Zinc ions, together with organic acids in the paint (from additives, rather than the binder), form a zinc soap which in turn breaks down the cured paint.

Keywords: acrylic waterbornes, zinc, galvanized steel, corrosion

Introduction

Waterborne paints are not simply solventborne paints in which the organic solvent has been replaced with water; the paint chemist must design an entirely new system from the ground up. Additives to keep the pigment from clumping, for example, may be completely different for a dispersion in a polar liquid such as water than in a nonpolar organic solvent. The same can be said for the chemicals added to make the pigments integrate with the binder, so that pigment particles are not simply occupying voids in the binder. The binder itself most frequently exists as a dispersion in waterborne paints. To keep the solid polymer in a stable and even dispersion, rather than in clumps at the bottom of the paint can, requires some creative use of additives. And of course, additives may be used to prevent flash rusting of the steel before enough water has evaporated for the polymer to form a coherent film which can provide corrosion protection.

The implications of all this paint chemistry for coating-substrate interactions, and hence coating-substrate compatibility, is not well known. The subject is of more than academic interest: experience from the field shows that waterborne acrylic paints on galvanized surfaces sometimes perform extremely well – but sometimes degrade quickly, for unknown reasons.

The aim of this project was to examine the compatibility of various ferrous substrates with acrylic waterborne resins. Four different paints were studied.

Experimental

Paints:

Four acrylic paint systems (one from each of four manufacturers, to obtain a representative collection) were used in this study. Each manufacturer also provided a waterborne epoxy paint. The ranking of the four substrates were fairly well known for the epoxies; they were therefore used as a control of overall surface preparation.

All the coatings are lead-free and chromate-free. The 36 substrate-coating combinations studied are summarized in Table 1, wherein K, L, M and N represent the four coatings manufacturers.

Table 1. Substrate-coating combinations. K, L, M and N are the four coatings manufacturers. Numbers in body of table are the dry film thickness (microns), either as primer + topcoat, or primer + intermediate coat + topcoat. 'Extra-thick paint' indicates that thickness of both primer and topcoat was twice that of the other samples.

Paint Type

Hot-rolled Steel

Cold-rolled steel, phosphated

Hot-rolled steel, galvanized

Cold-rolled steel, aluminum-zinc coated

 

extra-thick paint

regular thickness

reg. thickness

reg. thickness

reg. thickness

epoxy:

K

L

M

N

 

 

 

123+128

 

45+82+51

62+72

62+62

62+32

 

43+80+44

58+72

59+59

58+31

 

43+74+48

61+74

59+57

63+30

 

48+83+40

61+75

62+57

64+31

acrylic:

K

L

M

N

 

133+136

115+131

118+102

 

 

60+76

57+65

59+54

55+67

 

53+68

61+60

64+60

57+54

 

51+65

67+57

59+53

63+54

 

56+67

68+60

59+56

56+58

Substrates

Four substrates were included in this study: hot-rolled steel, blast-cleaned to Sa 2�, surface profile medium; hot-dipped galvanized steel, lightly blasted; zinc-phosphated cold-rolled steel; and Aluzink-coated. Zinc thickness' on the hot-dipped galvanized samples ranged from 68 to 77 microns; the average zinc thickness was 72 microns.

Initial Adhesion

The initial adhesion of the newly-cured coatings to the substrates were tested using the method ISO-2409, "Paints and varnishes - cross-cut test" [1]. In this test, six parallel cuts spaced 1 mm apart are scribed through the paint to bare metal. Then six more parallel cuts, also spaced 1 mm apart, are made perpendicular to the first set of cuts; a 5-by-5 grid is thus made. The panel is brushed along both diagonals of the enscribed squares. The amount of paint removed by this procedure is classified according to Table 2 below.

Table 2. Classifications used for adhesion in the standard ISO-2409, "Paints and varnishes - Cross-cut test".

Classification

Description

0

The edges of the cuts are completely smooth; none of the squares of the lattice is detached.

1

Detachment of small flakes of the coating at the intersections of the cuts. A cross-cut area not distinctly greater than 5% is affected.

2

The coating has flaked along the edges and/or at the intersections of the cuts. A cross-cut area distinctly greater than 5%, but not distinctly greater than 15% is affected.

3

The coating has flaked along the edges of the cuts partly or wholly in large ribbons, and/or it has flaked partly or wholly on different parts of the squares. A cross-cut area distinctly greater than 15%, but not distinctly greater than 35% is affected.

4

The coating has flaked along the edges of the cuts in large ribbons and/or some squares have detached partly or wholly. A cross-cut area distinctly greater than 35%, but not distinctly greater than 65% is affected.

5

Any degree of flaking that cannot be classified by classification 4.

Scab Test:

A weakly-accelerated field test known as the ‘SCAB-test’, Swedish Standard SS 117211, was performed on the scribed samples. Prior to testing, samples were scribed down to the steel by cutting through the paint with a hand-held 1-mm wide stainless steel blade held against a steel ruler. The resultant scribe is 1 mm in width, oriented horizontally and symmetrically located, on each sample. In the SCAB-test, samples are set out at an outdoor exposure station at SCI’s facilities in Stockholm, and sprayed twice-weekly with a 3% NaCl solution. Samples were placed 45 from the vertical, facing west, with the scribe horizontal. Testing was begun in July 1995 and ended in July 1997. After the field exposure was completed, samples were evaluated for maximum one-way creep from scribe, blistering, cracking, flaking, and rust intensity.

Salt Spray:

The salt spray test, ASTM B-117, was not performed in order to rank the coatings or provide any sort of comparison of the coating-substrate combinations used in this study. Rather, it was performed in order to see if exposure to a different corrosion environment - with constant wetness instead of cycling - would provide any additional clues interactions between substrates and polymers.

Samples scribed as described previously were run in the salt spray test for ten days. The test was carried out in accordance with ASTM B-117; scribes were placed parallel to the ground. At the end of ten days, samples were evaluated for maximum one-way creep from scribe, blistering, cracking, flaking, and rust intensity.

Grazing-Angle Ftir:

FT-IR studies were performed to look for evidence of saponification. Two methods were used, the KBr (Potassium Bromide)-pellet method and the ATR (Attenuated Total Reflectance) method. The KBr-pellet method was used to analyse corrosion products and the ATR method was used to analyse the paint surface. For the KBr-pellet studies, the apparatus was a Biorad FTS-175C and for the ATR studies the apparatus was an UMA500 FT-IR microscope with an ATR objective.

Results And Discussion

Initial adhesion:

Results of adhesion testing after cure but before exposure are given in Table 3.

Table 3. Initial adhesion for each substrate-paint combination. Numbers in body of table are the adhesion classifications as determined by the standard ISO 2409, "Paints and Varnishes - Cross-cut Test"

Paint Type

Manf.

Hot-rolled Steel

Cold-rolled steel, phos.

Hot-rolled steel, galv.

Cold-rolled steel, alum.-zinc coated

Extra-thick paint

regular thickness

regular thickness

regular thickness

regular thickness

acrylic

K

0

0

0

0

0

L

0

1

1

1

1

M

1

0

0

0

0

N

 

0

0

0

0

epoxy

K

 

1

1

1

1

L

 

2

2

2

2

M

1

0

0

0

0

N

 

0

0

0

0

An analysis of variance (general linear model) was performed on this data, with adhesion as response and binder type (epoxy, acrylic, etc.), manufacturer (K-N), substrate (hot-rolled steel, phosphated etc.), and coating thickness as factors. The results are summarized in Table 4 below. ‘DF’, ‘Seq. SS’, and ‘Adj. SS’ refer respectively to degrees of freedom, sequential sums of squares, and adjusted sums of squares. It can be seen, from the large F-values and low p-values (where a p-value below 0.05 indicates that the factor may be considered significant), that binder type and paint composition are both significant in determining initial adhesion; substrate is not important for the four ferrous substrates used in this study, and neither is coating thickness for the number of datapoints included here.

 

Table 4. GLM-analysis of variance for cross-cut adhesion.

Source

Degrees of Freedom

Sequential Sums of Squares

Adj. Sums of Squares

F

p

binder type

1

6.5427

7.9303

13.33

0.000

paint composition

3

9.5603

9.5380

16.03

0.000

substrate

3

0.0071

0.0000

0.00

1.000

coating thickness

1

0.0278

0.0278

0.14

0.710

error

26

6.1477

6.1477

   

total

35

22.2857

     

Accelerated Corrosion Testing:

Results of SCAB-testing are given in Tables 5-6.

Table 5. Acrylic binders. Results after 2 years’ SCAB-testing in Stockholm, Sweden. "Ri" is rust-through, in accordance with ISO standard 4628; "UC" is maximum one-way undercutting (creep) from scribe. Blistering, flaking and cracking did not occurr except on the samples for which this data is noted, and were determined in accordance with ISO 4628.

Paint

Hot-rolled Steel

 

Cold-rolled steel, phosphated

Hot-rolled steel, galvanized

Cold-rolled steel, aluminum-zinc coated

extra-thick paint

regular paint thickness

regular paint thickness

regular paint thickness

regular paint thickness

K

Ri: 0
UC: 19 mm

Ri: 1
UC: 10 mm

Ri: 0
UC: 10 mm

Ri: 1
UC: 40 mm
Blistering: 5(S3).

Ri: 2
UC: 8 mm
Blistering: 2(S3).

L

Ri: 0
UC: 21 mm

Ri: 1
UC: 15 mm

Ri: 0
UC: 6 mm

Ri: 5
UC: 34 mm
Flaking: 3(S4)b.

Ri: 0
UC: 6 mm
Cracking: 1(S3)

M

Ri: 0
UC: 20 mm

Ri: 0
UC: 19 mm

Ri: 0
UC: 12 mm

Ri: 0
UC: 29 mm
Flaking 3(S4)b.

Ri: 0
UC: 15 mm

N

 

 

Ri: 0
UC: 16 mm

Ri: 0
UC: 11 mm

Ri 4.
UC: 17 mm
Blistering: 2(S5)

Ri: 0
UC: 9 mm
Blistering: 2(S2)

 

Table 6. Epoxy binders. Results after 2 years’ SCAB-testing in Stockholm, Sweden. "Ri" is rust-through, determined in accordance with ISO 4628; "UC" is maximum one-way undercutting (creep) from scribe. Cracking did not occurr except on the samples for which this data is noted, and was determined in accordance with ISO 4628. Blistering and flaking did not occur.

System.

Hot-rolled Steel

 

Cold-rolled steel, phosphated

Hot-rolled steel, galvanized

Cold-rolled steel, aluminum-zinc coated

 

extra-thick paint

regular paint thickness

regular paint thickness

regular paint thickness

regular paint thickness

K  

Ri: 3
UC: 7 mm
Cracking: 5(S2)

Ri: 1
UC: 11 mm
Cracking: 5(S2)

Ri: 0
UC: 2 mm
Cracking: 5(S2)

Ri: 0
UC: 11 mm
Cracking: 5(S2)

L  

Ri: 0
UC: 14 mm

Ri: 0
UC: 8 mm

Ri: 0
UC: 1 mm

Ri: 0
UC: 9 mm

M

Ri: 0
UC: 21 mm

Ri: 0
UC: 13 mm

Ri: 0
UC: 4 mm

Ri: 0
UC: 1 mm

Ri: 0
UC: 4 mm

N  

Ri: 0
UC: 18 mm

Ri: 0
UC: 6 mm

Ri: 0
UC: 3 mm

Ri: 0
UC: 7 mm

Salt spray testing:

Results of salt spray testing are given in Tables 7-8.

Table 7. Acrylic binders. Results after 10 days’ salt spray (ASTM B117) testing. A '*' indicates that it is impossible to separate blisters from rust intensity. "Ri" is rust-through, determined in accordance with ISO 4628; "UC" is maximum one-way undercutting (creep) from scribe. Blistering did not occur unless noted, and was determined in accordance with ISO standard 4628. No samples showed flaking or cracking.

Manf.

Hot-rolled Steel

 

Cold-rolled steel, phosphated

Hot-rolled steel, galvanized

Cold-rolled steel, aluminum-zinc coated

extra-thick paint

reg. paint thickness

reg. paint thickness

reg. paint thickness

reg. paint thickness

K

Ri: 0
UC: 0

Ri 2.
UC: 0
Blisters 3(S2)

Ri: 0
UC: 0

Ri: *
UC: 4 mm
Blisters 5(S3)

Ri: 0
UC: 3 mm
Blisters 5(S2), only under scribe.

L

 

Ri: 1
Blisters 3(S3)

Ri: 0
UC: 0

Ri: 0
UC: 0
Blisters 5(S3) under scribe and at edges.

Ri: 0
UC: 3 mm
Blisters 5(S4), only under scribe.

M

Ri: 0
UC: 0
Blisters 2(S2)

Ri: 1
UC: 0
Blisters 3(S4)

Ri: 0
UC: 0

Ri: 0
UC: 0
Blisters 6(S2); larger blisters under scribe.

Ri: 0
UC: 4 mm
Blisters 5(S3), only under scribe.

N

 

 

Ri: 0
UC: 0
Blisters 3(S3)

   

Ri: 0
UC: 5 mm
Blisters 5(S3), only under scribe.

Table 8. Epoxy binders. Results after 10 days’ salt spray (ASTM B117) testing. "Ri" is rust-through, determined in accordance with ISO 4628; "UC" is maximum one-way undercutting (creep) from scribe. Blistering did not occur unless noted, and was determined in accordance with ISO standard 4628. No samples showed flaking or cracking.

Manf.

Hot-rolled Steel

 

Cold-rolled steel, phosphated

Hot-rolled steel, galvanized

Cold-rolled steel, aluminum-zinc coated

extra thick paint

reg. paint thickness

reg. paint thickness

reg. paint thickness

reg. paint thickness

K

 

No defects

 

Ri: 0
UC: 0
Blisters 3(S3)

Ri: 0
UC: 0
Blisters 2(S4).

 

L

 

No defects

No defects

No defects

Ri: 0
UC: 5 mm

M

No defects

No defects

No defects

No defects

No defects

N

 

 

No defects

No defects

Ri: 0
UC: 4 mm

Ri: 0
UC: 0
1 blister, 1(S3).

Behavior of the acrylics: The results for acrylics were quite surprising. Whereas the acrylics as a group experience creep from scribe and edge corrosion on blast-cleaned hot-rolled steel, the same paints degrade completely and flake off of the hot-dipped galvanized substrates. This is a complete reverse of what had been expected, and indeed does not agree with results from other trials [2].

Are acrylics vulnerable to saponification? A possible clue is found in the Swedish pulp and paper industry standard SSG 1009, "Systems for initial painting of hot dip zinc coated steel" [3], which advises that zinc-coated surfaces should be painted with paints based on non-saponifiable binders. What is of interest here is that zinc-coated surfaces, either initially or as they undergo corrosion, can be expected to provide an alkali environment. The question then becomes, are acrylics vulnerable to saponification?

Acrylics can be divided into two groups, acrylates (or polyacrylates) and methacrylates, depending upon the original monomer from which the polymer was built. (The difference lies in a methyl group attached to the backbone of the polymer molecule of a methacrylate in place of the hydrogen atom found in the acrylate.)

Billmeyer, in his classic text on polymer science, states that methyl methacrylates are quite resistant to saponification -–but that polyacrylates are not [4]. If an acrylic paint, therefore, is composed entirely of methyl methacrylate, it should not saponify. If, however, it is a polyacrylate or mixture of the two types, it may be vulnerable to saponification.

In practice, it is not feasible to formulate a coating entirely upon methyl methacrylate. Certain mechanical properties of poly(methyl methacrylate) and polyacrylates, given in Table 9 below, indicate that methyl methacrylate needs some modification from a copolymer in order to form a satisfactory paint: while the high tensile strength has good implications for impact resistance, for example, the elongation of pure methyl methacrylate is undesirably low. High amounts of external plastisizers or coalescing solvents would be required to form a film at room temperature. The excellent mechanical properties of methyl methacrylate mean that a waterborne paint based solely on this polymer would have a minimum film formation temperature of over 100�C ! [8]

Table 9. Mechanical Properties of Methyl methacrylate and Polyacrylates [5]

 

Methyl methacrylate

Polyacrylates

Tensile strength, psi

9000

3-1000

Elongation at break, %

4

750-2000

A ‘softer’ acrylate used as a copolymer could improve the ability of the binder to flex and bend to some extent, which is desirable in a coating for cold crack resistance [6] or dimensionally active substrates such as wood [4].

It seems reasonable to postulate that the binder in the acrylic paints is not composed entirely of methyl methacrylate, but of at least a copolymer; and that the particular binders used in this study saponified in the presence of zinc.

However, this should not lead to the conclusion that all acrylic emulsion polymers on the market saponify, unless they are pure methacrylates. Other compositional variations, like the type of acrylate monomer used for co-polymerization, the polymer morphology, the molecular weight and the methods used to stabilize the emulsion in water are of major importance. And in fact all the paint ingredients, not merely the binder, must be carefully selected in order to achieve good performance on zinc.

What role does the zinc play? The four waterborne acrylic paint systems showed markedly worse performance on galvanized steel than on carbon steel; the trend was just the opposite for the four epoxies (as expected). Therefore, it seems reasonable to conclude that these results are not due to some mistake in preparing the galvanized samples which resulted in an entirely unpaintable surface. It also seems reasonable to conclude that no drastic mistakes in application of the acrylics were made, since these paints showed a normal behavior on the other ferrous substrates. The degradation of the acrylics on the galvanized steel seems, therefore, to be due to an interaction between the zinc and the acrylic coatings.

As discussed above, it is quite possible that the copolymers used in this study underwent saponification and failed. But what role, precisely, did the zinc play in this? There are several possiblities:

The initial adhesion to the zinc was poor. The polymer compositions used for these coatings were not optimized for zinc, but rather for carbon steel. The reactive groups on the polymers which are intended to adhere to the substrate, did not react well enough with zinc. The zinc does not receive sufficient corrosion protection from the coating; water and oxygen can reach the zinc surface and initiate corrosion of the zinc. This corrosion causes an alkali environment at the zinc-paint interface, causing saponification of the polymer.

The zinc substrate was not wetted, due to paint formulation rather than polymer design. The reactive groups on the polymer may be ideally suited to adhere to zinc; but if the paint does not flow – e.g., due to high viscosity – enough across zinc to wet the substrate surface, the polymer never has a chance to adhere tightly. As in the previous case, water and oxygen can reach the zinc surface and initiate corrosion of the zinc. This causes an alkali environment at the zinc-paint interface, causing saponification of the polymer.

The zinc causes further crosslinking of the binder during cure, depending on which monomers are used. Crosslinking goes too far and the coating becomes brittle. Weathering stresses such as temperature change and wetting/drying cycles break down the cured coating. Saponification is not the major mechanism of degradation in this theory.

Zn++ ions destroyed the coulombic stability of the dispersion, leading to flocculation. The paint is in effect destroyed after being applied but before a coherent film can form.

Zinc ions, together with organic acids in the paint (from additives, rather than the binder), form a zinc soap which in turn breaks down the cured paint. The formation of metal soaps by the reaction of metal compounds with organic acids is a well known phenomenon. It is known, for example, that this is the mechanism by which red-lead paint protects steel. The acids may be present in these paints already at its production, due to additives used rather than the binder. Different metal soaps have, however, different properties. The lead soaps formed in red lead paints based on linseed oil are known to be elastic, adhering and very resistant to water. Zinc soaps, in contrast to lead soaps, are known to be brittle and to be water soluble [7]; the latter can be expected to increase the driving force for the water transport through the paint film towards the metal surface.

Has evidence of saponification been observed? Grazing-angle FTIR studies were performed on the samples which had been exposed at the outdoor station for two years, and on the samples which had been in the salt spray test for ten days. The paint surface which had been immediately adjacent to the metal was examined for evidence of zinc soaps. The problem with this technique was that to examine the paint at this interface, it was necessary to look at areas where delamination had occurred; and since delamination was always associated with corrosion of the underlying substrate in these samples, these were areas heavily contaminated with corrosion products. The only identifiable spectra observed, were those attributable to the corrosion products Zn5(OH)8Cl2 H2O and ZnO. An interesting peak was seen on all the samples at 1385cm-1. Paints all have a peak here, but for these samples it was sharper than normal.

Samples with only the primer of the system which had shown the most degradation, Acrylic K, were studied with the FTIR before and after exposure to solutions containing ZnCl, and NaOH. A solution of ZnCl in distilled water (pH 5 after addition of ZnCl) was used; the other solution was sodium hydroxide in water, with pH 11. A solution of zinc ions in an alkali solution would have been of interest, of course, but was not possible; at alkali conditions the zinc precipitates out of solution and the pH drops sharply. After 1340 hours of exposure to the ZnCl solution, no change was seen in the FTIR spectra of the Acrylic K primer. The sample of Acrylic K primer which was exposed to NaOH at pH 11, on the other hand, showed a change in peaks at 1380cm-1 and 1029cm-1. After 113 hours' exposure to the alkali solution, these peaks had had increased height in comparison to the other peaks. After 1340 hours' exposure to the solution, both the height and width were acccentuated. The conclusion seems to be that the mere presence of zinc is not enough to affect the cured coating; an alkali environment is needed.

The salt spray results hint that zinc, or corrosion products, or both, are responsible for the acrylics' degradation. The pattern of blistering seen on the acrylics after 10 days’ salt spray testing is quite different from that seen after two years’ SCAB exposure. After the salt spray testing, blisters are dramatically concentrated under the scribe, and seem to follow the lines of the liquid running down from the scribe. Because of the constant wetness in the salt spray chamber, the zinc in galvanized or galvalume coatings undergoes a different corrosion mechanism from that which would occur in a field exposure; the zinc rapidly dissolves without forming corrosion products in the salt spray chamber. Thus the runoff liquid under the scribe may be expected to contain zinc, and alkaline corrosion products, from the inorganic coating exposed at the scribe. It is not known whether the blistering seen in the runoff lines under the scribe is due to a reaction in these areas between the paint and the zinc in the runoff liquid, or between the paint and the alkaline zinc corrosion products.

Conclusions

Waterborne acrylic coatings are frequently used on galvanized substrates with good results; however, the acrylic coatings used here performed much worse on galvanized steel than on carbon steel. A possible explanation is that the copolymers used for the binder were vulnerable to saponification.

More work should be done in this area to identify more clearly the processes by which these acrylics degrade on the zinc surfaces. Evidence confirming or eliminating saponification as the cause of acrylic degradation, the actual mechanism of degradation (zinc surface producing an alkali environment leading to saponification, or zinc plus binder creating zinc soaps, et cetera), and which acrylic polymer chemistries and pretreatments before painting lead to a tendency to this, are areas which could benefit from future research.

References

1. ISO-2409, "Paints and varnishes - cross-cut test". International Organisation for Standardization, Case postale 55, CH-1211 Geneva 20, Switzerland.

2. Rendahl, B. and A. Forsgren. "Field Testing of Anticorrosion paints at Sulphate and Sulphite Mills". KI Rapport 1997:6E. Stockholm: Korrosionsinstitutet (Swedish Corrosion Insitute), 1997. ISSN: 0348-7199.

3. SSG-standard 1009 "Systems for initial paint of hot dip zinc coated steel", 1994-06-01, Edition 5, Desig. TKY, page 1(5). Available from SSG, Skogsindustriernas Teknik AB, Box 140, S-851 03 Sundsvall, Sweden.

4. Billmeyer, F.W. Textbook of Polymer Science, third ed. John Wiley & Sons, New York. ISBN 0-471-03196-8. p 388. 1984.

5. Brendley, W.H. Paint and Varnish Production, July 1973.

6. Bentley, J. "Organic Film Formers" , chapter in Paint and Surface Coatings Theory and Practice, ed. R. Lambourne. Ellis Horwood Limited, Publ. Chichester (Great Britain). ISBN 0-85312-692-5. 1987.

7. Paul, S. (ed.) Surface Coatings: Science and Technology. John Wiley & Sons, New York. p365. 1996.

8. Riemann, S., (Rohm & Haas, France). Memorandum to J. Ilomaki (Rohm & Haas, Finland) titled "Comments to the Study on 'Substrate-Polymer Compatibility for Waterborne Paint Resins'" dated October 16, 1998.

Acknowledgements

The authors would like to thank Sten Palmgren of SCI for painting and exposing the samples, and performing the adhesion tests. The assistance of Akzo-Nobel, International, Teknos Tranemo and Dickursby paint companies in supplying coating samples is gratefully acknowledged. The authors are also greatly indebted to Siegfried Riemann, Alain Garzon, Kris Peeters and Robert Krasnansky (all of Rohm & Haas) for discussions of acrylic film formation and degradation.