Astm stp 841 1985

NEW CONCEPTS FOR COATING PROTECTION OF STEEL STRUCTURES A symposium sponsored by ASTM Committee D-1 on Paint and Related Coatings and IVIaterials and Steel Structures Painting Council

NEW CONCEPTS FOR

COATING PROTECTION

OF STEEL STRUCTURES

A symposium sponsored by ASTM Committee D-1 on Paint and Related Coatings and IVIaterials and Steel Structures Painting Council Lake Buena Vista, Fla., 26 January 1983

ASTM SPECIAL TECHNICAL PUBLICATION 841

D M Berger, Gilbert/Commonwealth, and

R F Wint, Hercules Incorporated, editors

ASTM Publication Code Number (PCN) 04-841000-14

181b 1916 Race Street Ptiiladelpho, Pa 19103

Library of Congress Catalog Card Number: 83-82647

ISBN 0-8031-0236-4

NOTE The Society is not responsible, as a body, for the statements and opinions advanced in this publication

Printed in Baltimore, Md (b) First Printing, July 19*4 Second Printing, October 1985

Foreword

The Symposium on New Concepts for Coating Protection of Steel

Struc-tures was held in Lake Buena Vista, Florida, on 26 January 1983 Sponsors

were ASTM Committee D-1 on Paint and Related Coatings and Materials

and the Steel Structures Painting Council D M Berger,

Gilbert/Common-wealth, and R F Wint, Hercules Incorporated, served as symposium

chair-men and have edited this publication

ASTM Publications Permanence of Organic Coatings, STP 781 (1982), 04-781000-14

Selection and Use of Wear Tests for Coatings, STP 769 (1982), 04-769000-29

A Note of Appreciation

to Reviewers

The quality of the papers that appear in this publication reflects not only

the obvious efforts of the authors but also the unheralded, though essential,

work of the reviewers On behalf of ASTM we acknowledge with appreciation

their dedication to high professional standards and their sacrifice of time and

effort

ASTM Committee on Publications

Janet R Schroeder Kathleen A Greene Rosemary Horstman Helen M Hoersch Helen P Mahy Allan S Kleinberg Susan L Gebremedhin

Contents

Introduction 1

KEYNOTE ADDRESS

Evolution of Steel Protection; A Personal View—s L LOPATA 5

SURFACE TECHNOLOGY CONCEPTS

Quantitative Evaluation of Blistering and Corrosion in Organic

Coating Systems—M E MCKNIGHT AND J W MARTIN 13

Blast Cleaning with Zinc-Coated Abrasives—K W LOWREY 21

Detrimental Materials at the Steel/Paint Interface—

W C JOHNSON 2 8

Effects of Rotary Peening Surface-Conditioning Products on

APPLIED COATING SYSTEMS AND SAFETY PRACTICES

Zinc-Plus-Paint System for Corrosion Protection of a Steel Bridge—

M M L W I N 5 3

Spray-Applied Fluoroelastomers for Protection of Carbon Steel

Structures in Flue Gas Desulfurization Service—

C A MCCLAIN AND T DOLAN 6 3

Perspectives on 100% Solid Spray-Applied Polyurethane

Minimum Film Thickness for Protection of Hot-RoUed Steel: Results

after 23 Years of Exposure at Kure Beach, North Carolina—

M MORCILLO 95

ASTM Safety Alert System—s IOHN OECHSLE 118

SUMMARY Summary 127

Index 133

STP841-EB/JUI 1984

Introduction

The Symposium on New Concepts for Coating Protection of Steel Structures

was sponsored by ASTM Committee D-1 on Paint and Related Coatings and

Materials and the Steel Structures Painting Council (SSPC) ASTM

Subcom-mittee D-01.46 on Industrial Protective Painting directly relates to the work of

SSPC This meeting, held in Lake Buena Vista, Florida, represented the first

time SSPC met anywhere other than in Pittsburgh, Pennsylvania The meeting

also represented the first time Committee D-1 met in joint session with SSPC

The success of the meeting was attributed to the fact that one could attend

Committee D-1 meetings in the beginning of the week, the joint symposium on

Wednesday, and the SSPC meeting later in the week Over 180 members

at-tended the ASTM meetings and over 220 atat-tended the SSPC meetings This

joint venture is of particular value when air travel and other expenses are

con-sidered, because it allowed the individual members to attend both meetings

under one travel expense It is expected that future symposia and joint

meetings will be held by these two organizations Owing to the presence of

SSPC, representatives of 15 other organizations were present

This symposium was the first sponsored by SSPC The selection of Stan

Lopata as keynote speaker set the tone of the subject Protection of Steel

Struc-tures Mr Lopata, Chairman of the Board, Carboline Company, St Louis,

Missouri, inventor of alkyl silicate inorganic zinc-rich primers, has

con-tributed significantly to the technology of protection of steel structures

Sidney B Levinson, Chairman of ASTM Committee D-1, and John D

Keane, Director of the Steel Structures Painting Council, welcomed the

at-tendees The speakers were introduced by Dean Berger and Rufus Wint, who

served as symposium chairmen and who have edited this publication

The Summary at the end of this volume reviews the presented papers and

highlights the important issues raised by the speakers

sym-Stanley L Lopata^

Evolution of Steel Protection:

A Personal View

REFERENCE: Lopata, S L., "Evohitlon of Steel Protection: A Persona] View," New

Con-cepts for Coating Protection of Steel Structures, ASTMSTP841, D M Bergerand R F

Wint, Eds., American Society for Testing and Materials, 1984, pp 5-9

ABSTRACT: This paper covers 45 years of personal experience in the development of

corrosion-resistant protective coatings for steel and concrete and experience in industry

practices in corrosion prevention since World War II The performance of traditional

coatings for the protection of steel is outlined, followed by a comparison with various

high-performance coatings based on vinyl, epoxy, and ethyl silicate binders The effect of

sur-face preparation on coating performance is described, together with an indication of the

formulation of coatings of superior performance in chemical environments on rusty steel

when abrasive blasting is impossible under plant conditions The paper concludes with a

plea for thorough, statistically designed long-term testing of experimental formulations, so

as to ensure adequate correlations between laboratory results and plant exposures

KEY WORDS: steel protection history, protective coatings, inorganic zinc silicate, ethyl

silicate, postcured primers, self-curing primers, long-term testing

My first involvement with corrosion was in 1936, when the tarnishing of

brass eraser holders for wooden pencils was brought to my attention The

tar-nish was caused by hydrogen sulfide in the atmosphere Laboratory testing

consisted of dipping the brass eraser holders in cellulose acetate lacquer, then

putting the pencil top into jars of hydrogen sulfide to see how long it took for

tarnish to develop The test was crude, but satisfactory Needless to say, I did

not develop a satisfactory coating from the starting raw materials

My next involvement with corrosion developed from efforts to sell

acid-resistant pumps and valves At hundreds of chemical plants I observed

firsthand the corrosion and erosion of pumps, concrete foundations, cast iron

bases, malleable u-on packing gland holders, etc At that time I was not

con-cerned about the corrosion of the structural steel or the concrete floors

'Chairman of the Board/Chief Executive Officer, Carboline Co., St Louis, MO 63144

In 1945 I visited a titanium dioxide plant and again observed severe

corro-sion of structural steel and concrete The resident corrocorro-sion engineer, noting

the need for a primer for use on rusty metal, suggested that there would be a

good market for it and, of course, an opportunity to make a lot of money This

idea was most appealing and so I began investigating the coating practices of

the day Other than the "wash primer" developed by Union Carbide just

be-fore World War II, an adhesion-promoting coating that needed a chemically

clean or sandblasted surface, there did not seem to have been any major

improvements in the previous 25 years Red lead primers based on raw and

boiled linseed oil, alkyd primers using iron oxide pigmentation, and short oil

and long oil primers, none of which had very much corrosion resistance, were

commonly accepted In severely corrosive environments such as the sulfuric

acid leaching areas of the titanium dioxide process, paint would last only one

to two years and show as much as 40% failure Sandblasting was seldom used

in the field and in this plant would have contaminated the product that was

be-ing manufactured It seemed that what was required by industry was an

acid-resistant primer that would perform fairly well on rusty steel This contrasted

with existing primers which were good for rusty metal but had only mediocre

acid resistance and were useful only on rust with a pH above 7

The best approach to this problem appeared to be the development of a

prod-uct that had good adhesion Consequently a literature search was condprod-ucted to

find acid-resistant vehicles with this property An attempt was then made to

pigment the better formulations in order to evaluate corrosion inhibition This

work led to a line of primers for rusty metal useful in both acid and alkali

envi-ronments Some of these are still used, naturally with modifications to improve

certain properties

At the same time work was being carried out by Union Carbide to develop a

more acid- and alkali-resistant primer for use on sandblasted surfaces This

work produced vinyl primers that could be applied in coats 25 to 50 fim (1 to 2

mils) thick instead of the 7.5 to 12.5 ^m (0.3 to 0.5 mils) typical of earlier wash

primers

Simultaneously, Kenneth A Tator developed a method for accelerated

testing of paint systems in corrosive environments His first work, done on

coating systems on sandblasted steel, proved rather conclusively that the

mini-mum system should be 125 fim (5 mils) thick and consist of three coats The

125 fitn (5 mils) was thick enough on plane surfaces to give about 50 to 75 /tm (2

to 3 mils) film thickness on sharp edges This was certainly better than the 25

to 37 fiTa (1 to 1.5 mils) of edge thickness obtained from standard two-coat

in-dustrial systems yielding 50 to 100 /*m (2 to 4 mils) thick films Mr Tator's

work also produced accelerated results; thus failure of paint systems could be

determined in one third to one sixth of the time previously required His

method of testing, with some modifications, is still being used today

By 1949 to 1951, two new developments began to spur recognition of the

need for the protection of steel:

LOPATA ON EVOLUTION OF STEEL PROTECTION 7

1 Sodium silicate-zinc-rich primers for steel, developed in Australia, were

being talked about The first of these, after heat cure, produced a glass-like

zinc silicate surface on the steel A pipeline coated with 75 fim (3 mils) of this

type of coating had been in service for some nine or ten years with excellent

re-sults Soon the heat cure was replaced by an acid wash, originally dilute

phos-phoric acid, to insolubilize the sodium silicate This system consisted of (1) a

sandblast with a minimum 37 /xm (1.5 mils) profile, (2) a 75 to 100 ^m (3 to 4

mils) film of the inorganic zinc rich coating, and (3) an acid wash (sometimes

two washes if the film was too thick) This coating system provided industry

with its first true substitute for galvanized steel

The idea of substituting an inorganic zinc silicate for galvanizing had

suffi-cient merit to cause investigators to look at other sources of silicates that would

permit the formation of an inorganic zinc silicate film without infringing on

existing Australian-American patents One of the more useful sources of

sili-con was ethyl silicate, which could be hydrolyzed to form a tough binder One

of the original uses for partially hydrolyzed ethyl silicate was to promote

ad-hesion of coatings to glass surfaces Experimental work using different

pig-ments and fillers in this partially hydrolyzed ethyl silicate vehicle yielded an

in-organic zinc silicate coating deposited from a solvent system During and after

drying, the moisture from the environment completed the hydrolysis and the

formation of a basic zinc silicate coating This coating had both advantages

and disadvantages over the sodium silicate postcured primer Some of the

ad-vantages were (1) less sensitivity to surface preparation, (2) insensitivity to rain

within 20 min after application, and (3) because the surface was more

per-meable, better adhesion of intermediate coats and topcoats The

disadvan-tages of the self-curing versus the postcured product were (1) a softer film until

it built up a layer of dense basic zinc carbonates, and (2) many different types

of topcoats tended to bubble because of void spaces These two types of

inor-ganic zinc primers still dominate the field for steel protection in areas where

high relative humidity or corrosive fumes are involved Of course, there have

since been many modifications of these two basic products

2 At the same time (between 1949 and 1951) epoxy resins developed in

Switzerland and the United Kingdom became available in the United States

These resins could be applied in heavy coatings, 75 to 500 /tm (3 to 20 mils) per

coat, that could be catalyzed so they would harden in 4 to 8 h without skinning

The use of these catalyzed coatings eliminated the requirement for the slow

oxidation process that takes place in heavy films of coatings when drying oil

ve-hicles were used The adhesion of these films to steel was phenomenal For

ex-ample, by using combinations of various epoxy resins and semiplasticizers in a

heavy (125 to 375 nm [5 to 15 mils]) properly pigmented coating primer on

cold-rolled thin steel 375 to 500 nm (15 to 20 mils) thick, and then bending the

steel, it is possible to remove the thin blue mill scale from steel sheet The mill

scale will remain bonded to the epoxy surface as the coating cracks and peels

Because of the various molecular weights of the epoxy resins avaSable, it

soon became possible to apply films 125 to 750 /xm (5 to 30 mils) thick on the

surface of steel in one coat These heavy epoxy primers had certain

disad-vantages: (1) the steel had to be sandblasted for satisfactory adhesion, (2) poor

resistance to oxidizing conditions, and (3) water permeability Formulators

and polymer chemists developed means of improving the resistance of the

resins to oxidizing acid conditions and by catalyst variation developed

materials that had excellent resistance to water penetration Many of the

prod-ucts developed between 1952 and 1962 are still being used, and new modified

epoxy resins and coatings formulated from these resins have flooded the

market (for example, epoxy-phenolic combinations, epoxy-polyester, and

epoxy-acrylic) Various mechanisms for increased levels of cross-linking were

developed to meet particular requirements, such as raising the heat distortion

point of the coating During recent years epoxy urethane materials have been

developed These are used more for finish coats than prime coats, but in areas

where drying at lower temperatures is required epoxy urethane formulations

permit application and hardening of primers at temperatures as low as — 18°C

(0°F)

By the end of the 1950s, epoxy coating usage had developed in volume, and

various formulators were trying to correct some of the disadvantages of the

original primers One major disadvantage was that the use of epoxy coatings

called for commercial or white blast surfaces It was soon realized that if

in-dustry could develop a coating for application over hand-cleaned rusty metal,

the market for epoxy primers would substantially increase Formulators began

modifying epoxy primer formulations so they would penetrate and stabilize

rust It was apparent that the primer had to possess retained flexibility when

exposed to outdoor environments If it became brittle it would pull off the rust

and delaminate from the steel within a year or two, since the coating and the

steel expanded and contracted at substantially different rates After seven or

eight years of experimental work the correct combination of resins and

inhibit-ing pigments was developed This coatinhibit-ing had to be relatively thick to make

sure that the uneven surface was covered by a minimum 125 to 150 /im (5 to 6

mUs) dry film The performance of this modified epoxy primer has been so

good that it is now being sold in over 40 countries around the world where

sandblasting cannot be done or is too expensive

As one might suspect, there are probably as many different formulations as

there are paint technologists, each person thinking that his product will do

something that some other product doesn't do In certain cases this is true, but

what is most important in the development of paint primers is for thorough

long-term testing to be done to ensure consistently satisfactory performance

Fortunately, the development of the aforementioned primers has raised the art

of steel protection to a science It is still not possible to obtain 100%

reproduci-LOPATA ON EVOLUTION OF STEEL PROTECTION 9

ble results, but with proper care and formulation, manufacturing, and

appli-cation, 90% correlation can be obtained

Over the years, various insufficiently tested coating systems have been

mar-keted One of these was the "Harvey System", which consisted of an alkyd

primer, a chlorinated rubber intermediate coat, and a vinyl chloride topcoat

Delamination between coats developed on exterior aging

The use of tannic acid in the primer reduced surface rust on steel to

magne-tite or iron (Fe) If the rust was only 2.5 pim (0.1 mil) thick, the procedure

worked, but where the rust layer was thicker it did not work

A German development of an iron oxide pigment with minimal aluminum

content dispersed in an alkyd resin vehicle which caused the conversion of rust

(f«rric oxide) to magnetite worked fairly well over a 12.5 jum (0.5 mil) rust

layer, but the topcoats had to be impervious for long periods of time to permit

the reaction to take place slowly

Lead suboxide pigment in an alkyd vehicle, developed in Switzerland about

50 years ago, has worked well on lightly rusted galvanized steel and on neutral

pH rust, that is, rust formed in nonacid environments Unfortunately in many

areas the use of lead-containing pigments has been prohibited

What does the future hold for the protection of steel? A protector that will

do a better job for less money—and there are probably 10 000 paint

for-mulators working towards that end today Much of this research will involve

minimizing surface preparation while still providing reliable performance

Much work will be done on water-based coatings possessing equal or superior

performance to solvent-based coatings Further development of silicone,

sili-cate, aliphatic urethane, and fully cross-linked polymers to provide longer

ser-vice, less deterioration on aging, and better resistance to heat, abrasion, and

sunlight will occur This is the challenge for the balance of the 20th century

Each formulator who is working on improving paint primers and topcoats

should do his own long-term testing and not blindly accept the results of

others Short-term testing alone will not suffice In addition, there should be a

larger number of series of duplicate tests With this increased research the

steel protection industry will not only proudly maintain its present credibility

but also improve it

MaryE McKnight^ and Jonathan W Martin^

Quantitative Evaluation of Blistering

and Corrosion in Organic Coating

Systems

REFERENCE: McKnight, M E and Martin, J W., "Quantitative Evaluation of

Blister-ing and Corrosion in Organic CoatBlister-ing Systems," New Concepts for CoatBlister-ing Protection of

Steel Structures, ASTMSTP841, D M Berger and R F Wint, Eds., American Society

for Testing and Materials, 1984, pp 13-20

ABSTRACT; A nondestructive laboratory procedure using infrared thermography for

de-tecting air- and water-filled blisters and localized corrosion at the coating/metallic

sub-strate interface is described Deteriorated areas are observed in real time as varying gray

levels on the cathode ray tube of an infrared thermographic camera or after digitization of

the signal on a TV monitor Digitization of the analog signal permits (1) image

enhance-ment through signal averaging techniques, (2) association of gray levels with degraded

areas, (3) quantitative analysis of the panel for amount, location, and type of

degrada-tion, (4) computerized storage of the digitized signal for dynamic analysis of the degraded

coating, and (5) graphic display of thermographic images In addition to providing a

non-destructive test for assessing degradation at the coating/substrate interface, these analyses

will contribute to further understanding of corrosion mechanisms and the kinetics of

deg-radation This knowledge will be useful in developing improved accelerated tests for

pre-dicting coating durability

KEY WORDS: infrared thermography, corrosion, organic coatings, steel, aluminum,

de-tection, accelerated testing, degradation

Suppliers and users of corrosion control coatings are showing increased

in-terest in rapid quantitative measurement techniques for detecting and

assess-ing early coatassess-ing failure and in usassess-ing accelerated agassess-ing tests for predictassess-ing the

service life of coating systems The demand for these developments is fueled

by (1) the large economic losses resulting from corrosion, (2) the high costs

resulting from coating failures, (3) the mandatory displacement of

tradi-tional coatings by new, less familiar coating systems, (4) the inadequacy of

'Research Chemist and Materials Research Engineer, respectively, Building Materials

Divi-sion, Center for Building Technology, National Bureau of Standards, Washington, DC 20234

13

currently available accelerated aging procedures and degradation

measure-ment techniques for predicting service life [/,2], and (5) lack of knowledge of

the effects of changes in surface preparation, solvents, and inhibitors on the

durability of coating systems

Our research in developing improved accelerated aging tests and

measure-ment techniques uses reliability analysis techniques for analyzing accelerated

aging test results, guiding the design of experiments, and verifying the

mathe-matical models derived in analyzing the accelerated aging data As a

prereq-uisite to this analysis, a quantitative measurement tool for assessing early

deg-radation is necessary One such tool in infrared thermography The objective

of this paper is to report on this technique Results of the reliability analysis

will be reported later

Experimental Procedures and Results

The experimental procedure using infrared thermography to detect changes

at the coating/substrate interface has been described previously [3] Briefly,

test specimens are heated uniformly from the back to 10 to 20°C above the

ambient temperature and the thermal energy emitted from the painted

sur-face is measured with an infrared thermal line scanning system Deteriorated

areas appear as varying gray levels on the cathode ray tube (CRT) of the

infra-red system The infrainfra-red camera used in this research detects radiation in the

2 to 5.6 pim range The temperatu.re resolution was 0.2°C for an object

tem-perature of 30°C; the spatial resolution was 1 mrad with the 8-deg

field-of-view lens The area of coated panel examined at any one time is about 70 by

70 mm

Spatial resolution and the types of coating defects that can be detected were

determined empirically by examining coated steel panels (both smooth and

sandblasted) that had been fabricated with known defects Experiments

showed that corrosion under intact, nonblistered films can be detected, as can

air- and water-filled blisters A typical thermograph of a coated smooth steel

panel with localized corrosion under a pigmented alkyd film is shown in Fig 1

The corroded regions correspond to the lighter areas On smooth substrates,

corroded areas greater than 1 mm in diameter can easily be detected On

sandblasted substrates, the diameter of a detectable corroded area has to be at

least 1.2 to 1.5 mm Air-filled blisters appear dark ("cold"), as illustrated in

Fig 2, and water-filled blisters appear hot

The ability to spatially resolve defects is strongly dependent on the camera

and lens selected For studying mechanisms of degradation, observing early

stages of deterioration, relating surface preparation techniques to

perfor-mance, or comparing performance properties of different inhibitive

pig-ments, a close-up lens with high magnification should be selected However,

for observing large areas of a coating system, such as in field inspections, a

wide angle lens would be more appropriate

MCKNIGHT AND MARTIN ON QUANTITATIVE EVALUATION 15

FIG 1—Thermograph of corroded sandblasted steel panel with pigmented atkyd coating

Cor-rosion spots are 5 mm in diameter

FIG 2—Thermograph of air blisters under clear coating Small blisters are 1 mm in diameter

Quantification of Analog Infraied Thermographic Signals

To obtain quantitative information from the tliermographic image,

re-search was conducted with the following objectives: (1) to digitize the analog

signal transmitted from the thermographic camera according to its gray level,

(2) to enhance the digitized signal by averaging techniques, (3) to associate

the digitized gray level with deteriorated areas of the coating or substrate,

(4) to analyze the digitized picture with respect to deteriorated areas, (5) to

store digitized pictures in a computer for future retrieval, and (6) to

graphi-cally display pictures and results of analyses

The system used in quantifying the thermographic output includes the

in-frared camera, a microprocessor, and a minicomputer (A video camera can

be used to take a picture of the thermographic image if the infrared camera is

not video-compatible.) The function of each of these is described below

The scanning infrared camera detects radiation emitted from a coated

specimen Inside the camera, the signal appears as a matrix of dots (pixels);

each pixel is characterized by its location and luminance (somewhere between

black and white) In creating a picture, the camera receives a video signal

in-put for each pkel At the same time the camera is receiving inin-put, it is

trans-mitting a frequency-modulated signal voltage to the microprocessor The

magnitude of this voltage is equivalent to the luminance or gray level of the

pixel

The microprocessor receives and digitizes the analog signal voltage to one

of up to 256 gray scale levels This digital signal is stored in the

microproces-sor memory line by line until the picture is complete A new picture is

pro-cessed every one thirtieth of a second The digital image can be propro-cessed in

many ways [4], To reduce the effects of noise the signal can be enhanced by

ei-ther spatial or temporal averaging In spatial averaging, noise in a picture

frame is reduced by averaging the signal for each pixel with those of the

sur-rounding pixels In temporal averaging, successive picture frames are

cap-tured and then averaged to create an enhanced picture Other techniques,

such as background subtraction, contrast enhancement, and edge sharpening

can also be used to improve the image After processing, the image can be

analyzed in several ways The fraction of area each gray level occupies can be

determined The varying gray levels are most likely related to the degree of

de-terioration and can be displayed by using a three-dimensional plot Often,

only the fraction of area that has deteriorated is of interest To pinpoint this

area, a high-contrast function can be programmed into the microprocessor

such that all gray levels above some chosen level are made black and those

be-low that level become white The level is chosen so that only the area of interest

is highlighted For example, it may be chosen so that all corroded areas

ap-pear white and everything else apap-pears black Since the fraction of area that is

black or white now corresponds to the fraction that has deteriorated, it can

easily be computed Other functions programmed into the microprocessor

in-MCKNIGHT AND MARTIN ON QUANTITATIVE EVALUATION 17

elude the ability to invert gray levels so that black goes to white and white to

black, and the ability to store thermographs on disk or send them to a

mini-computer

The minicomputer is used to store and retrieve large numbers of

thermo-graphic images as data These data are used for analyzing dynamic changes in

the degradation state of the coating The results of these analyses can then be

displayed on graphics terminals or output as desired The steps that would

typically be used to obtain quantitative information from an infrared

thermo-graph are described below

The typical thermographic image of a deteriorated panel, as seen on the

in-frared camera CRT, is shown in Fig 3 With an inin-frared camera and

micro-processor, as described above, a digitized image of the analog output

dis-played on the CRT is obtained on a TV monitor A picture of the digitized

image, as seen on the TV monitor, is shown in Fig 4 As can be seen, there is

little or no difference in the two pictures Now that a digitized image is

avail-able, averaging techniques are used to enhance the image The temporal

aver-age of eight successive frames of the analog thermographic imaver-age is shown in

Fig 5 This averaging technique improves the spatial resolution by reducing

the noise The enhanced digitial image can then be analyzed as desired If the

percentage of the area that is degraded is needed, then the high-contrast

func-tion can be used to transform the gray levels to either white (corroded area) or

black (intact area) The percentage of area corroded then corresponds to the

FIG 3—Thermograph of corroded steel panel with clear acrylic coating, as displayed on

camera CRT

FIG 4—Pictures of digitized thermograph (Fig 3) as displayed on TV monitor

white level and is easily determined The result of using the high-contrast

function on the thermograph illustrated in Fig 5 is shown in Fig 6 The

frac-tion of white or corroded area for the example shown in Fig 4 is 11.4%

Analysis of the digital thermographs can take many directions Anticipated

procedures include determining rates of corrosion as a function of such

vari-ables as exposure conditions, coating type, substrate preparation, thickness

of coating, and dynamic pattern analysis

Summai;

A nondestructive evaluation procedure using a scanning infrared

thermo-graphic camera has been described which will enable researchers to determine

the extent of deterioration of an organic coating/metallic substrate system It

has been demonstrated that:

1 Thermographic techniques may be used to observe corrosion under

in-tact, pigmented films on metallic substrates

2 The analog thermographic image as displayed on the CRT can be

digi-tized and the signal enhanced by using a video camera and microprocessor

MCKNIGHT AND MARTIN ON QUANTITATIVE EVALUATION 19

FIG 5—Thermograph shown in Fig 4 after temporal averaging

3 Quantitative information on the extent and location of deterioration can

be obtained Thermographs can be transferred to a minicomputer for storage

and further analysis

4 The method offers both a quantitative evaluation of coated metals and

earlier detection of deterioration than visual comparison methods

It is felt that this quantitative evaluation procedure will contribute to

fur-ther understanding of corrosion mechanisms and to the development of

accel-erated tests which will allow better predictions of coating durability

Acknowledgments

The authors gratefully acknowledge the support they received for this

re-search from the Federal Highway Administration (FHWA) Particular thanks

in this regard go to Dr Lloyd Smith of FHWA and Dr Bernard Appleman,

formerly with FHWA The authors are also greatly indebted to Dr Geoffrey J

Fronsdorff, Mr Larry Masters, and their other colleagues at the National

Bu-reau of Standards for the many fine interchanges during the course of this

work

FIG 6—Result of using high-contrast function on thermograph shown in Fig 5

References

[/] Campbell, P G., Martin, J W., and McKnight, M E., "Short-Tenn Evaluation for

Coat-ings for Structural Steel," TNH49, National Bureau of Standards, Washington, DC, Sept

1981

[2] Leidheiser, H., Jr., Corrosion, Vol 38, No 7, July 1982, pp 374-383

[3] McKnight, M E and Martin, J W., "Nondestructive Corrosion Detection under Organic

Films Using Infrared Thermography," 14th National SAMPE Technical Conference, Society

for Advancement of Materials and Process Engineering, Azusa, CA, 1982

[4] Stucki, Peter, Ed., Advances in Digital Image Processing, Plenum Press, New York, 1979

Kenneth W Lowrey^

Blast Cleaning with

Zinc-Coated Abrasives

REFERENCE: Lowrey, K W., "Blast Cleaning with Zinc-Coated Abrashes," New

Con-cepts/or Coating Protection of Steel Structures ASTM STP 841, D M Berger and R F

Wint, Eds., American Society for Testing and Materials, 1984, pp 21-27

ABSTRACT: In the early 1960s the concept of cleaning/protecting steel surfaces was

pio-neered by the Steel Structures Painting Council (SSPC) The discovery was made that zinc

could be transferred from the steel shot to the steel substrate during the cleaning operation

in such a way that the onset of re-rusting was delayed significantly This discovery prompted

others to look at the potential which was offered by this particular technique

Almost 20 years later Colebrand Ltd decided to make use of zinc-coated abrasives as

part of a program into the repair and protection of offshore structures Colebrand research

chemists were not satisfied with the results obtained with the then commercially available

zinc-coated abrasives As a result, work was undertaken to develop a new product which

would be deemed acceptable for cleaning in the splash zone area and also for submerged

steel structures This paper examines the parameters involved in the selection of a suitable

type of grit and an acceptable particle size range from those currently available for

indus-trial use Mention is also made of the selection of the binder used to bond the zinc to the

surface of the grit Details are given on laboratory and field trials and on aspects of health

and safety when using zinc-coated abrasives Finally, a case is made for the use of such

ma-terials in certain applications in industry

KEY WORDS: steel, cleaning, protection, painting, zinc-coated adhesives, splash zone,

salt spray adhesion

The concept of using an abrasive to fulfill a dual function is not new, and

many experiments have been perfonned in this particular area of

steel-ing techniques Experiments ussteel-ing zinc-coated shot for abrasive blast

clean-ing were carried out by the Steel Structures Paintclean-ing Council/International

Lead Zinc Research Organization (SSPC/ILZRO) in 1963 [/] The results

showed that sufficient zinc was deposited on the steel substrate during the

cleaning operation to exhibit a definite improvement in the length of time

elapsed before re-rusting occurred This important discovery led many other

technicians to look closely at this method for the cleaning of steel Almost a

'Technical Director, Colebrand Ltd., London, England

21

decade later patents were granted for steel cleaning using zinc-coated

abra-sives [2] Some 20 years after the original work Colebrand Ltd became

partic-ularly interested in this method of cleaning/protecting a steel surface and

in-cluded this method in a technical development investigating the repair and

protection of offshore structures

Experimental Procedure

Selection of Abrasive Grade

The requirements of a material suitable for the preparation of a metal

sur-face by impact cleaning can be summarized as follows:

1 A granular material of regular structure of controlled size range and

shape (grit being sharp and angular, shot being spherical,) having extreme

hardness and roughness The material must have high density in order to

im-part maximum energy to the blasting process

2 An abrasive material devoid of any deleterious component likely to

ad-here to the cleaned metal surface and thus likely to create possible sites for

corrosion attack of the adhesion of the surface protection

3 A material able to produce the desired surface cleanliness and profile

with maximum efficiency in terms of time and effort Within the confines of

this study the use of expendable abrasive was the sole abrasive considered

4 An inert material of minimal health and environmental hazard

As the final product would eventually be used under site conditions where

reclamation would not be a viable or economic process, the investigation was

narrowed to three grades of expendable metallic slags The sieve analyses are

shown in Table 1 It can be seen from the sieve analyses that Grades B and C

contain a large proportion of fines It was thought that this could cause

prob-lems during manufacture of the coated grit because weight for weight they

would have a larger surface area than Grade A Further, there could be a

ten-dency for these small particles to form agglomerates which could give rise to

problems on storage and subsequently in use

As all three grades have an equal hardness and shatter index, it was decided

TABLE 1—Sieve analyses

Grade A Grade B Grade C

% thttjugh B.S 18 mesh

% through B.S 25 mesh

% through B.S 35 mesh

^B B Christensen, British Patent No 1,377,484, 18 Dec 1974; B B Christensen, U.S

Pa-tent No 3,765,923, 16 Oct 1974

2.5 1.1 0.6

57.1 7.7 0.8

61.6 32.2 0.6

LOWERY ON BLAST CLEANING WITH ZINC-COATED ABRASIVES 2 3

that the work be performed using Grade A with a particle size 90% of which is

between 0.3 and 1.5 mm

Selection of Binder

A series of initial trials in which quantities of zinc dust were simply mixed in

with the grit and subsequently used to clean steel surfaces produced rather

in-conclusive results as well as an increase in dust levels during the cleaning

operation Therefore it was decided to bond the zinc to the grit and thereby

give a much more usable product At first sight there is obviously a large range

of materials, both organic and inorganic, that lend themselves to this

particu-lar use, namely drying oils, vinyl, acrylic, silicate, etc When looked at in

depth, however, many have to be discarded since they tend to give rise to

cak-ing, are prone to softening at high ambient temperatures, or give rise to

exces-sive dusting

After a series of trials the best practical binder was found to be a solution of

a comparatively high molecular weight epoxy The choice of solvent as carrier

for the resin was also important in that its evaporation rate had to be slow

enough to allow thorough wetting of the grit but fast enough to allow rapid

drying once the zinc had been added so as not to have a protracted production

rate

The ratio of grit/zinc-dust/resin was determined with regard to

perfor-mance and cost effectiveness

Laboratory Trials

In these tests 4.7-mm (•'/i6-in.)-thick flat steel plates, 152 by 305 mm (6 by

12 in.), with a tight mill scale were used as the substrate The plates were

di-vided into two groups: those to be blast cleaned with standard abrasive and

those to be cleaned with the zinc-coated abrasive to a standard of cleanliness of

SA 2V2 Once blasted, the panels were placed in a salt spray cabinet and a pair

were removed at various times intervals to be subsequently coated The paints

were allowed to cure for seven days at room temperature before being subjected

to a pull-off test using the Elcometer Adhesion Tester This technique,

involv-ing a direct tensile pull, is described in Ref 2

Four paints were selected for adhesion testing:

1 Red lead pigmented liquid epoxy/aromatic amine system

2 Inorganic zinc silicate

3 Red oxide pigmented solid epoxy/polyamide adduct system

4 Zinc phosphate/titanium dioxide pigmented solid epoxy/isolated amine

adduct system

In each case, the panels were removed from the salt spray cabinet after V2,

1, 2, 3, 4, 6, and 8 h and then coated Panels that had not been exposed to salt

spray were used as controls

Results

During the first time interval, from Vi to 2 h, there was a fluctuation in

re-sults After this period, however, a trend was clearly shown in all four cases

that the adhesion levels obtained using the coated abrasive were greater than

those obtained using the standard abrasive (Fig 1) Why the initial

^ 1

g«APH 2

2 ? h 5 6 7 «

•fiME ELAPsti/'noofts) IN S^ALT speAy

EPoxV'fifb oxibt — PaVAMiit AiSbucT BASED <i>ATiN&

LOK<'£b AfcRASlKE

1 a i * s fr T

•fine IfLAPSOi [»<OoAi) IN SALT i P t t A V

FIG 1—Adhesion levels

LOWERY ON BLAST CLEANING WITH ZINC-COATED ABRASIVES 2 5

tion in results occurred is not readily understood; in some small part, at least,

it may be attributed to the fact that the modes of failure which occurred

dur-ing the pull-off tests varied between adhesive, cohesive, dolly, and a

combina-tion thereof

A further factor may have influenced the adhesion of paint coatings If the

substrate which has been cleaned using the zinc-coated abrasive is viewed

un-der a scanning electron microscope, it can be seen that approximately 25% of

the steel surface is randomly covered by single grains of zinc This variation in

substrate may influence adhesion in certain cases

Field Trials

A number of visits were made to Ardyne Point, a disused construction site

on the Clyde Estuary in Scotland This was an ideal site where the preparation

of test panels could be carried out under similar conditions to those

encoun-tered in an offshore environment

The first visit was made in November 1979 Comparative tests were run

us-ing coated and standard grit in conjunction with a red lead epoxy primer and

a pitch epoxy top coat Areas of steel work were blasted within the tidal zone

at low tide and then coated with the paint system mentioned previously The

areas blasted with standard grit were coated immediately after blasting; those

with coated grit were left exposed for 2 h before coating Pull-off adhesion

tests were performed every six months as part of an on-going test program,

and the substrate beneath the specimens was examined for signs of rusting

Some of the test results obtained by January 1982 are listed in Table 2 It

can be seen that as yet there is little difference between the respective levels of

adhesion The major observation at this stage is that the use of coated abrasive

does not detract from long-term paint adhesion Further testing revealed the

degree of protection offered by coated abrasive when exposed to different

en-vironments (Table 3)

Health and Safety

Concern had been expressed by various potential users as to the dust levels

and the ill effects caused by the respiration of zinc dust In Great Britain the

recommended safe levels for toxic dusts to which workers can be exposed are

given in Guidance Note EH15/79, Threshold Limit Values for 1979 This

guidance note is essentially a reprint of the Threshold Limit Values adopted

by the American Conference of Government Industrial Hygienists (ACGIH)

The level recommended for inert or nuisance dusts is 10 mg/m-' for total

air-borne dusts There is in fact no level quoted for zinc dusts, but zinc oxide dust

is classed as an inert or nuisance dust Tests were performed by the National

Occupational Hygiene Service Ltd., who monitored dust levels during a

clean-ing operation The results obtained durclean-ing this survey showed that high

air-TABLE 2—Test results obtained by January' 1982

35 cohesive failure

25 cohesive failure

30 cohesive failure

25 cohesive/adhesive failure

TABLE 3—Protection offered by coated abrasive when exposed to different environments

Environment Time to Onset of Flash Rusting North Sea splash zone 4 h

Clyde tidal zone 4 h

Constant immersion in highly oxygenated water 3 h

Above water, semi-rural 1 week

borne dust levels could be expected within 5 m of the grit-blasting operation

These high dust levels were found using both types of grit; however, in neither

case did the level exceed 10 mg/m^

Economics

Because of the significant difference in cost between conventional and

zinc-coated abrasives, £40 ($68) and £172.5 ($293) per tonne respectively, the

se-lection of the zinc-coated abrasive for a particular contract will require careful

consideration In instances where climatic conditions are not of the best, the

coated abrasive could be a worthwhile investment This is particularly true if a

major part of the working day is likely to be lost owing to unfavorable

condi-tions for blast cleaning using the conventional abrasive It is worth bearing in

mind that with the increasing number of "all-weather" coatings this

combina-tion of cleaning process and coating applicacombina-tion could effectively mean that

downtime reductions are a real possibility

DiscusHon

Comment has been made that the presence of zinc on the steel surface will

inevitably lead to the formation of zinc salts which could be detrimental to

long-term paint adhesion and hence overall performance From the results to

LOWERY ON BLAST CLEANING WITH ZINC-COATED ABRASIVES 2 7

date this does not appear to be the case, probably because the surfaces have

not been left exposed for a long enough period for salt formation to occur

As a cautionary note, it should not be thought that a coated abrasive

per-forms the function of a cleaning agent plus a prefabrication primer It has

been mentioned previously that the coating of zinc on the metal is not

continu-ous and hence cannot be classified as a true priming coat

Whilst the use of coated grit will extend the time between blasting and

painting, it should not be abused Coated abrasive will obviously have a time

limit in its ability to resist the onset of flash rusting and thus a safety margin

must be allowed This safety margin will depend upon the location and

preva-lent weather conditions during the cleaning operation

Although only results obtained using epoxy-based coatings have been

re-ferred to in this paper, media such as vinyl, acrylic, and chlorinated rubber

have also been evaluated In almost all cases an increase in adhesion and

sub-sequent performance have been reported when these media are used in

con-junction with coated abrasive

Conclusion

Tests have shown that the use of coated abrasive provides a dual function,

firstly by cleaning the surface and secondly by protecting that surface for a

limited time period before the application of a paint or coating system The

ability of a coated abrasive to perform these functions has been shown in

sub-merged, splash zone, and above-water environments

Whether coated abrasive should be used as the sole cleaning media or

whether it should be used for sweep blasting after the major cleaning has been

carried out using standard abrasive is a question of cost effectiveness

Never-theless, coated abrasive does offer a new outlook on surface preparation

References

[/] Keane, J D., "Zinc Shot Blasting of Structural Steel," SSPC/ILZRO Report ZE-66, Steel

Structures Painting Council/International Lead Zinc Research Organization, April 1964

[2] Paint Testing Manual, ASTM STP500, 13th ed., G G Sward, Ed., American Society for

Testing and Materials, 1972, p 458

Detrimental Materials at the

Steel/Paint interface

REFERENCE: Johnson, W C , "Detrimental Materials at the Steel/Paint Interface,"

New Concepts for Coating Protection of Steel Structures, ASTMSTP841, D M Berger

and R F Wint, Eds., American Society for Testing and Materials, 1984, pp 28-43

ABSTRACT: Materials harmful to paint life can remain on the surface of steel not only

af-ter hand cleaning but also afaf-ter the more thorough white metal blast cleaning Harmful

materials that can interfere with protection by paint include grease and oil, moisture, iron

oxides, and chemical contaminants Chemical contaminants come from acid rain, snow

removal salts, seawater, nearby chemical plants, or contaminated blasting abrasives

These materials are discussed with emphasis on significant and novel concepts and with

respect to occurrence, identification, removal, and the role of the material in causing

cor-rosion Experimental evidence and handbook data are used to support conclusions

KEY WORDS: cell, chlorides, concentration, conductivity, contamination, corrosion,

electrolyte, grease, inorganic salts, interface, ions, mill scale, osmosis, osmotic pressure,

paints, primer (coatings), scale (corrosion), steel surface, sulfates, vapor pressure

In the past, when environments were milder and coatings were simpler, steel

surfaces were prepared for painting by hand wire-brushing Today, however,

when environments are more corrosive and coating systems are less forgiving

of poorly prepared surfaces, greater rigor is required in surface preparation

In contaminated chemical areas, detrimental materials in sufficiently high

quantities to cause trouble can escape removal, even with more thorough

methods of cleaning such as blasting to white metal Thus, interest is growing

in methods of removing these contaminants and in paints that can tolerate

their presence on steel surfaces

Interest is also growing in finding alternatives to current blasting

proce-dures When highway bridges over rivers are blasted, toxic dusts are generated

from red lead and chromate paints that have been applied in the past To

pro-tect wildlife from these toxic materials, regulations require the daily

place-'Staff, KTA-TATOR Inc., Pittsburgh, PA 15275

28

JOHNSON ON DETRIMENTAL MATERIALS 29

ment and removal of straw dams across rivers, to catch some of the toxic dusts

This costly procedure and others arising from regulations generate interest in

alternative forms of surface preparation and in coatings that do not require

blasting

Cleaning surfaces and formulating primers for metal are more complex

tasks than has generally been recognized by the industry Harmful materials

that remain on the steel surface after cleaning are now receiving more attention

As a basis for study and discussion of cleaning methods and (non-zinc-rich)

metal primer compositions, materials that remain on the surface are reviewed

in this report with respect to (1) source or occurrence, (2) identification,

(3) roles in corrosion of the steel and in paint performance, and (4) methods

of removal or of neutralizing their effects

Materials that can interfere with paint protection are divided into the

In addition, oxygen and gaseous water permeate films readily and are

avail-able to join with the foregoing material

Graases and Oils

Occurrence

During handling and fabrication, greases from lubricants and coolants can

accumulate on steel, generally in patches or spots as the result of (1) contact

with lubricated tools, skids, or machinery, (2) accidental drips from cranes,

or (3) purposeful application during drilling or machining operations

Steel structures can be splattered by grease or oil from motorized traffic,

tools, and lubricated machinery or equipment, and by fumes

Such contamination commonly occurs in patches but can also be uniformly

distributed by contaminated blasting abrasive Careless solvent cleaning—for

example, leaving dirty solvent on the steel to evaporate instead of wiping to

dryness—merely results in the redistribution of the oils and greases

Identification

The only satisfactory field test for oil or grease contamination is visual

in-spection of the surface before blasting Grease (or oil) patches often have

char-acteristic shapes, such as circular spots from globs of grease, or charchar-acteristic

locations, such as the areas around drill holes A water test of blasted surfaces

to check for dewetting can be misleading when water does not draw back from

grease- or oil-lined valleys in the profile

A method of evaluating grease on the abrasive is to add water, level to the

top of an abrasive sample, and observe for nonwetting Experience is

benefi-cial in judging the degree of wetness

Ultraviolet light causes some greases to fluoresce in the visible range The

use of ultraviolet light, however, does not cause fluorescence in some common

oils and greases; also, the test is usually not sufficiently sensitive

Removal of Grease and Oil

Grease and oil at the interface between steel and a coating can interfere with

the formation or bond of the coating to the steel

Oils and greases must be removed before blasting or other surface

prepara-tion, not only to prevent contamination of abrasive and to avoid spreading the

grease over a larger area, but also to avoid leaving harmful but invisible

amounts Methods are outlined in Steel Structures Painting Council (SSPC)

Specification SP-1 on Solvent Cleaning Briefly, this method consists of

clean-ing the surface with solvents, aided by rags, brushes, or sprays, and usclean-ing

clean solvent and rags for the final wiping It includes the alternatives of vapor

degreasing, immersion, emulsion, or steam cleaning

Moisture

Condensation and Hydration

Moisture condenses when the surface temperature is at or below the dew

point Conversely, water evaporates when the surface temperature is above the

dew point

Moisture also accumulates through hydration and solution of salts, such as

ferric chloride or ferric sulfate, that escape removal during cleaning Water is

absorbed because a saturated solution of a salt has a lower vapor pressure

than pure water

Identification

Measuring the temperature of the surface and determining the dew point

indirectly identify the presence of water on a clean surface that is caused by

condensation The presence of water resulting from absorption by salt

con-tamination may be estimated from the results of an analysis for salts

Role of Water

Moisture can prevent paints from wetting the surface, with the result that

the coating does not bond to the steel Alone or in conjunction with oxygen,

moisture will corrode the steel surface by formation of an electrochemical cell

JOHNSON ON DETRIMENTAL MATERIALS 3 1

Because moisture is an essential ingredient of corrosion cells, electrolytic

cor-rosion does not occur in the absence of water

Evaporation and Dehydration

Painting specifications normally require surface temperatures 3°C (5°F)

above the dew point thrpughout surface preparation and painting, in order

that the surface will be sufficiently dry The dew point is determined from wet

and dry bulb temperature measurements in conjunction with a humidity chart

or table

When deliquescent salts or hygroscopic solutions'of salts are present in

detectable quantities, they must be removed Data on surface temperatures

necessary to avoid water absorption are given in Appendix A The data are

in-structive in showing the futility of heating to remove absorbed water from

salt-contaminated surfaces The salt's demand for water remains, and it will begin

absorbing water upon cooling A better approach is to remove the salts

Chemical Contaminants

Occurrence

Chemical contaminants on rusted steel come from the atmosphere, snow

removal salts, splashes, fumes, or mists in salt marine or chemical plant

loca-tions Salts settle from the atmosphere during fogs, dews, inversions, and acid

rains Upon evaporation of the dews, fogs, and rains, the water-soluble salt

contaminants are concentrated at localized areas on the steel (substrate)

sur-face The dews and fogs concentrate low-lying contaminants On the other

hand, long heavy rains can cleanse salts from surfaces exposed to the

atmosphere

Acid rain, now common particularly throughout the northeastern United

States and Canada, results largely from the presence of gaseous oxides of

sul-fur and nitrogen in the air Snow removal chloride salts are used on highways

and bridges In marine locations, chloride sea salts are present Chemical and

industrial plants have severe concentrations of a variety of chemicals

Other sources of chemical contaminants include (1) blasting abrasive, (2)

rinse water used in wet cleaning methods, and (3) ingredients of the coatings

Contaminants remain even after thorough blasting of rusted steel Traces

of chemical contaminants such as soluble salts and acid can be present in

harmful although invisible amounts

Identification

Information about the specific chemicals and concentrations in the bulk

corrosion product can generally be obtained by analysis of a small sample of

representative corrosion product for pH, anions, and concentration The

presence of anions have been identified in surface corrosion pits with a

scan-ning electron microscope (SEM) [/]

When KTA-Tator encountered abrasives that differed in rust back of

blasted steel, tests for salts were conducted The abrasive that resulted in

cor-rosion did indeed contain a substantially higher concentration of soluble salt

Because a method was needed to assess the amount of salt on the abrasive

un-til a consensus group acted on methods of testing and acceptable limits for

salts, we devised a field method for measuring soluble salts on abrasives

Sam-ples of abrasives from around the country, both by-product and natural, have

been evaluated by measuring the resistivity of filtrate from a slurry of

deion-ized water and abrasives

Three somewhat distinct ranges have been encountered: (1) a high

concen-tration (resistance reading of 2000 to 3500 Q-m), (2) an intermediate

con-centration (resistance reading of 7100 to 8300 Q-m), and (3) a low

concentra-tion (resistance reading of 15 000 to 39 000 Q-m) The resistance readings of

samples of abrasives from around the country and the corresponding ranges

of salt concentration are given in Table 1

A procedure for measuring soluble salts on abrasives has been written into a

tentative method and has been given to the SSPC Surface Preparation

Com-mittee

Role of Chemical Contaminants

The corrosive action of chemical contaminants can supplant the inhibitive

effects of anti-corrosion pigments Chemicals such as salts on the steel surface

act (1) to absorb and hold moisture from the air (Appendix I); (2) to draw

moisture through a coating by means of the process of osmosis (Appendbc II);

(3) to reduce the corrosion cell resistance (Appendix III); (4) to establish

concentration potentials (Appendix IV); and (5) to lower the pH (Appendix V)

Any or all of these actions, singly or in combination, are detrimental to the

protection provided by paints These corrosive actions are destructive forces

beneath paint films

Removal of Chemical Contaminants

Minimizing the amount of detrimental soluble salts is a matter of removing

them, excluding them, or chemically inactivating them Included are salts on

corroded steel surfaces that are found in the bulk of the corrosion product and

anions that are held tenaciously in the corrosion pits, as well as salts that are

inadvertently introduced on blasting abrasive, in the rinse water of wet

clean-ing methods, or with pigments Controllclean-ing these salts will take new methods,

practices, or specifications for painting system design, field procedures,

abra-sive supply, and water source, along with paint and pigment manufacture

JOHNSON ON DETRIMENTAL MATERIALS 3 3

TABLE 1 —Soluble salts on blasting abrasives

"Range of salt concentration corresponding to resistance readings is given in parentheses

Ranges of low, intermediate (medium), and high are used

* Part per million on abrasive

Suggested test methods, practices, and specifications are listed or outlined in

Table 2 The table has a design section, a section on high-performance

coat-ing systems, and a section on oil/alkyd penetratcoat-ing primer systems

With increasing awareness of and technical dialog on the presence of salts,

new standards based on these proposals may be forthcoming Design

engi-neers can then incorporate the new standards into paint system specifications

Paint System Design

Because corrosion products vary widely in salts, they introduce complex

technical problems to the design of a painting system and to the preparation

of the surface Where the steel is not rinsed by rains or otherwise washed,

re-sults of a laboratory analysis of the contamination in the bulk of the corrosion

product is representative of the general contamination level and also of the

contaminants that are present at the anode areas On the other hand, where

the steel has been rinsed by rains, the bulk analysis may be helpful, but only

an SEM analysis will detect a concentration of anions at the corrosion pits

An engineer designing a painting system will make sounder judgments if he

knows the pH, amounts of ionic material, and kinds of ions present The

ana-lytical results should be consistent with the knowledge of the environmental

area and can be helpful in deciding the appropriate SSPC "corrosion severity"

zone [2] and the corresponding painting system

When rust is analyzed for design purposes, a sample of rust may be scraped

TABLE 2—Removal or limitation of water-soluble salts

Standards Source or

Occurrence Operation

Available Specifications Proposed Procedures or Standards

Design

Bulk corrosion

product

design of ing system

paint-Steel surface

anodes

design of ing system

paint-Design analysis conducted in a laboratory

on filtrate from slurry of powdered rossion product:

cor-1 Determine amount of electrolyte by using conductivity meter; ASTM D

1125 M e t h o d s

2 Determine pH

3 Identify principal anions a) Specific ion meter b) Semiquantitative wet chemistry c) Test strips

scanning electron microscope

High-Pertormance Coating Systems

Bulk corrosion surface prepa- SSPC Surface limit salts on abrasives (see Abrasive

prepa-Preparation Specifications SP2, SP3, SP5, SP6,' SP7, SPIO

SSPC SP5 or SPIO, followed

by water rinse

below)

go/no-go field analysis, Merkoquant, potassium ferri-cyanide, Saltesmo, Quantabs

limit salts on abrasives (see Abrasive below) and on rinse water (see Water below)

Abrasive blasting limit salts on abrasive by method based

on ASTM D 1125 Method B Water rinse of blast none

or last water contacting surface

limit salts in rinse water with water purity meter by using variation of ASTM D

1125 Method B

Pigment paint and

pigment manufacture

none limits salts on pigments using conductivity

meter per ASTM D 2448

Oil/Alkyd Penetrating Primer System

Bulk corrosion paint and lead containing formulate coatings to contain lead or

product and pigment oil/alkyd barium compounds

steel surface manufacture primers

anodes

Pigment paint and

pigment manufacture

lead pigment containing oil/

alkyd primer

insolubilize sulfates and chlorides on auxiliary pigments

JOHNSON ON DETRIMENTAL MATERIALS 35

from existing surfaces, or a special panel can be placed in a representative

lo-cation to collect contaminants for subsequent analysis

Analysis of salts as a design tool can be performed in a laboratory by using

sensitive and accurate methods and instruments, such as a Wheatstone bridge

conductivity meter, pH meter, specific ion meter, and scanning electron

mi-croscope With background information from this laboratory analysis, a

quick go/no-go field inspection test is sufficient to indicate the presence of

anions remaining on the steel surface after the cleaning operation

Designs of two hypothetical painting systems are given in Table 3 One uses

a barrier system, while the other is based on an inactivation approach The

in-activation approach may only work in oil/alkyd systems where water and ions

move through the film more readily

A choice among the various barrier systems and the various inactivation oil/

alkyd systems can be made on the basis of the design analysis and the

environ-mental zone, along with economic and other factors

High-Performance Systems

Dry blasting to white metal, as in SSPC SP5, removes most of the

contami-nants that are associated with the corrosion product Multiple procedures,

such as blasting-rinsing-reblasting or blasting-rinsing with inhibited water,

have been tried for removal of the anions at the steel surface with partial success

A mechanical process for cleaning steel, such as blast cleaning, removes the

contaminants that are associated with the bulk corrosion products but leaves

detrimental anions on the steel surface

A go/no-go field inspection test method for the complete removal of soluble

salts from the steel surface offers a two-pronged approach when used in

con-junction with the results of a design analysis Such a field test is a useful

inspection tool in surface preparation A description of several field test

pro-TABLE 3—Salt analysis and corresponding coating system design

Contamination and Environment Analysis

Concen-Major tration Envii-onmental Coating

Anion of Salt pH Zone Class

Treatment of Salts

Inactivation Limitation Removal of Anions of Salts Chloride

Sulfate

high

low

SSPC3A chemical exposure (pH 2.0 to 5.0) SSPC IB exteriors normally dry

barrier SSPC-SP5 none

plus water wash

inactivation SSPC-SP2 red lead