11.1 Definition. Organic coatings deteriorate and fail with time. Failure analysis does not concern itself with this type of deterioration. It is defined as an investigation to determine the cause or causes of premature deterioration of coatings or coating systems. It is obvious, however, that failure analyses are often also directed at obtaining additional information than that stated in the above definition. Thus, the failure analyst may also wish to determine the extent of the damage, whether all requirements of a specification of a contract or work order had been met, who might be responsible for the failure and thus be liable for repairs, or what is the best remedial action to correct the existing condition.

11.2 Documentation of Findings. Measurements, photographs, specimens, and other observations made at the job-site or later in the laboratory should be firmly documented with dates, locations, etc., because they may at a later time become legal evidence. Personnel conducting failure analyses should routinely follow the procedures necessary for such documentation to prepare for any eventuality.

11.3 Scope of Failure Analysis. Paint failure analysis can be conducted by anyone with a basic understanding of coatings. However, they are best conducted by someone specially trained for the work. This is particularly true if the investigation becomes part of a dispute, since credibility of the analyst may be a determining factor in a dispute. In some instances, an analysis need not be extensive, but care must be taken not to make important conclusions based on superficial observations. A complete paint failure analysis includes most or all of the following actions:

a) Review of specification including modifications

b) Review of supplier’s data

c) Review of inspector’s daily reports

d) Thoroughly documented on-site inspection

e) Laboratory analysis of retained and/or field


f) Analysis of data

g) Preparation of a report containing findings and conclusions

11.3.1 Review of Specification for Coating Work. The specification and the submittals required in the specification for the coating work should be thoroughly reviewed and understood. The specification states precisely the work that was to have been done and the methods and materials that were to be used, so that any deviations from them should become apparent.

The review may also point out discrepancies or lack of clarity in the document that may have contributed to the failure.

11.3.2 Review of Supplier’s Data. Supplier data sheets should be reviewed, because they describe the intended purpose of the coatings used, along with recommended surface preparation and application practices. They may also include compositional information that can be checked later by laboratory analysis to determine if the batch actually used was properly prepared.

11.3.3 Review of Inspector’s Daily Reports. The inspector’s daily reports should be reviewed, because they provide information about the conditions under which the work was accomplished and the quality of the surface preparation and coating application. Any compromises in the conditions required by the specification or recommended by the supplier may lead to early failure. These reports may also reveal field changes that were made to the original specification.

11.3.4 On-Site Inspection. Just as the inspector on the job, the person analyzing paint failures must have access to areas where failures have occurred. This may require ladders or lift equipment, lighting, or mirrors. The analyst should also have photographic equipment to document conditions and be skilled in its use. Scales should be used to show relative size, and permanent markings should be made on each photographic exposure for positive identification. Dates should also be placed on the photographs.

The analyst should have a standard kit of field test equipment including one or more thickness gages and calibration standards, a knife, a hand lens, and containers for samples. Obviously, he should be well trained in their use and use them systematically, as described elsewhere in this text. A container of methyl ethyl ketone (MEK) or other strong solvent may be useful in either determining paint solubility (e. g., verifying the general paint type or its complete cure) or to strip off a coating to examine the condition of the underlying surface or the thickness of the underlying galvanizing or other insoluble coating. Standard forms for manually recording data or equipment for voice recording are also very useful. A failure analysis checklist can ensure that no important item is overlooked. Obviously, all items on the list may not be important at all times, but to inadvertently skip an important one may be a serious oversight.

11.3.5 On-Site Inspection Techniques. An overall visual analysis should first be made to determine the areas where the deterioration is most extensive and where any apparent deviation from specification may have occurred. This should then be followed by a closer examination as to the specific types of deterioration that may be present.

a) Use of a hand lens may provide information not otherwise visually apparent. All types of failure, including color changes and chalking, should be described fully. For example, does peeling occur between coats or from the substrate? Are blisters broken or filled with water? This detailed information may be necessary for finalizing conclusions as to the type of failure. The terms defined later in this section should be used to describe failures rather than locally used terms that may not be clear to other people. Care must be taken not to come to final conclusions until all the data are analyzed. It is a good practice to state at the inspection site that the final conclusions on causes of failure cannot be made until completion of laboratory testing.

b) Dry film thicknesses should be routinely measured and recorded, as any significant deviations from recommended thicknesses can be a factor contributing to coating failure. The procedure for measurement of these thicknesses required in the specification should be followed.

c) Other measurements that may be important are coating adhesion and hardness, since they may provide important information on application or curing of the coating. Adhesion can be easily determined with a simple tape test described in Section 9 or by using more sophisticated instrumentation (refer to Sections 8 and 9). Hardness can be tested in the field with a knife or special hardness pencils.

d) It is generally important to verify the identity of the finish coatings and occasionally the identity of the entire coating system. If wet samples of the paints used have been retained, they can be submitted for laboratory analysis for conformance to specification or manufacturer’s data sheet. If these are not available, a simple solvent rub test may be useful in determining whether the exterior coatings are thermoplastic, thermosetting, or bituminous. A cotton-tipped swab stick is dipped in MEK or acetone and rubbed against the paint surface. A thermosetting coating such as a vinyl which has been deposited on the surface by simple solvent evaporation will redissolve in the solvent and be wiped onto the cotton. A bituminous (coal tar or asphalt) coating will also behave in this manner, but it will impart a characteristic brown stain to the cotton. Properly cured multiple-component thermosetting coatings such as epoxies that cure by chemical reaction will not be affected by the solvent. These solvents can also be used at the job site to remove thermoplastic coatings to examine the condition of the underlying substrate. The presence of mill scale may establish the extent of surface cleaning. If rust is found, care must be taken to determine if it was present before painting or resulted from underfilm corrosion. Samples of the finish coat can also be removed by sanding and taken to the laboratory for identification as described in par. 11.3.6.

e) Once the various types of failure that may be present have been identified, the extent of each type of deterioration should be estimated. Standard block methods that help to quantify the extent of coating deterioration are described in ASTM F 1130, Inspecting the Coating System of a Ship. Two sets of drawings are used to illustrate failures. One set is used to identify the portion of the surface on which the paint is deteriorated. The other set is used to identify the level of deterioration within the deteriorated areas. For example, a fourth of the surface could exhibit blistering and within the areas 10 percent of the surface could be blistered.

11.3.6 Laboratory Testing. A more definitive laboratory analysis of deteriorated paint is generally desired and may become critical if the problem goes into litigation. Such analyses require several representative paint samples to be collected at the job site. Peeled and blistered paint is easily sampled, but it may be necessary to obtain samples from sound paint by scraping or sanding. Each sample should be placed in a sealed container and properly identified and dated. Chain of custody procedures (ASTM D 4840, Sampling Chain of Custody Procedures) should be used if litigation is involved.

If samples of wet paint used on the job are available, they can be tested by standard laboratory tests for conformance to any SSPC, Federal, military, or State specification referenced in the contract specification. If none of these standards was

referenced in the specification, the paints can be tested for conformance to manufacturer’s listed composition or properties. Microscopic Examination. Samples of paint collected at the job site should be examined under a light microscope. An edge examination may reveal the number of coats and the thickness of each coat. An examination of the surface may reveal fine cracking or other irregularities. Examination under a scanning electron microscope (SEM) can reveal much more detailed information about the paint film. Also, the SEX often has an attachment for energy dispersive x-ray analysis which can identify the metals and other elements in the pigment portion of small areas of the coating. Spot Tests. There are several simple laboratory spot tests that can be run on samples of deteriorated paint collected at the job site. They generally provide specific information about the paint binder (ASTM D 5043, Field Identification of Coatings) or pigment. Special chemicals and training are usually required by the analyst. Infrared Spectrophotometric Analysis. The most widely used technique in laboratory analysis of paint failures is the infrared spectrophotometry. The use of new Fourier transform infrared (FTIR) spectrophotometers permits much more versatility and precision than earlier instruments. The technique can identify the resin components of paints from the shapes and locations of their characteristic spectral peaks. It is highly desirable to separate the resin from the paint pigment before analysis, because the pigment may cause spectral interference. This is easy to do with thermoplastic but not thermosetting paints. Thermoplastic resins can be dissolved in solvents that are transparent in part or all of the infrared region, filtered to remove the pigment, and the solution placed in standard liquid cells or cast as films onto sodium chloride or other plates that are transparent in the infrared region. Thermosetting coatings can be pressed into potassium bromide pellets or their spectra measured using diffuse reflectance equipment. Although the pigment is not separated in these procedures, the spectrum of the pigment can often be separated from that of the total coating by the FTIR spectrometer. Spectra of field samples are compared against published standards or authentic samples. It should be remembered that exterior weathering may cause oxidation that may appear in spectral analyses. Other Specialized Instrumentation. There are other specialized instruments that can be very helpful in failure analysis. These include emission spectroscopy, atomic absorption spectroscopy, and x-ray fluorescence instruments that can identify and quantity the metals present in a coating. Their methods of operation are beyond the scope of this document.

11.3.7 Forming Conclusions and Preparing Reports. Field and laboratory data should be analyzed logically and systematically to form conclusions as to the causes of paint failure. No data should be overlooked, and the conclusions should be consistent with the data. The report should include conclusions and recommendations requested by the activity for which the analysis was made.

The report is perhaps the most important part of the failure analysis, because it presents the findings and conclusions of the investigation. No amount of good field or laboratory work will be useful unless it is presented well in the report. There must be a clear, systematic, and logical presentation of the findings, so that the conclusions are well supported. The report should not contain errors or otherwise be subject to challenge. Where conclusions are not firm, the extent of uncertainty should be stated.

11.4 Expert System for Failure Analysis. An expert system

for failure analysis provides a systematic approach first to make a preliminary identification based on visual observations and then to verify it with supplementary information. The initial identification is based upon the answers to a series of questions designed to distinguish one type of failure from another.

Decision trees 1 and 2 are used for this, one for surface defects and one for film failures. This same approach can be used in an expert system for a computer. The importance of a systematic approach cannot be overemphasized. One should be careful not to make permanent decisions on types and causes of failure until all the evidence is considered.

The first step in the identification of a coating failure is to determine which decision tree to use. Tree l for cosmetic defects should be used if only surface damage is present, i. e., if the surface coat has not been completely penetrated to the underlying coat or structural substrate. Tree 2 for film failures should be used if coating damage has completely extended through the surface coat.

After a preliminary decision of the type of failure has been made, look at the additional comments in the verification section below to obtain further support for this selection. If this information does not support the initial decision, reexamine the evidence or reconsider answers to the decision tree, until you are satisfied that you received the best overall answer.

Remember, answers are not always easily obtained in failure analysis, and there may be multiple types and causes of failure. Thus, one may in some cases have to be content with the most probable cause or causes of coating failure.

11.4.1 Cosmetic Defects. The following paragraphs further describe the cosmetic defects chosen in the initial decision. Chalking. Chalking occurs only on exterior surfaces, since it is caused by the sun’s ultraviolet rays. The loose chalk will be the same color as the coating, and, if it is severe, an undercoat may be visible. Chalking should not be confused with loose dirt which will not be the same color as the finish coat. Mildew. Mildew may resemble dirt but generally grows in discrete colonies rather than being uniformly distributed. In addition, it can be bleached with household bleach, but dirt cannot. Also, it can also be identified microscopically by its thread-like (hyphae) structures and its groups of spherical spores. Mildew is usually black in color but some microorganisms on coatings may have a green or red coloration. Dirt. Dirt may be tightly or loosely held. It can normally be removed by washing with detergent solution. However, it may resist washing, if the dirt became embedded in the wet or soft paint. Uneven Gloss. Localized glossy spots may often be detected only if observed from a particular angle. They occur most frequently from spray application of heavy areas that do not penetrate into wood or concrete/masonry surface. Blushing. Blushing is a defect from spraying fast­evaporating coatings, particularly lacquers such as vinyls and chlorinated rubbers, on hot, humid days. Condensation of moisture on the wet film dulls the finish to cause an opalescence. Bleeding. Bleeding occurs when solvent-containing coatings are applied to a bituminous coating or pavement. The stronger the solvent and the slower its evaporation, the greater will be the tendency to dissolve the bituminous material and cause it to bleed to the surface of the finish. New asphalt pavements or toppings should be allowed to remain 4 weeks before marking with paint to allow evaporation of volatile materials in the asphalt. Fading. Fading of paint pigments occurs greatest in the sunlight. Thus, there will be less fading of coatings under eaves and other shaded areas. It also occurs more with synthetic organic pigments than with naturally-occurring mineral pigments (earth tones). Discoloration. Discoloration is caused by exposure of unstable pigments or resins to sunlight. Unstable resins like polymerized linseed oil may yellow. The only prevention is to use light-stable materials. Pigment Overload. Pigment overload frequently results in a mottled appearance or a poor quality film. It can occur when attempting to tint a white paint to a deep tone. Latex paints are particularly susceptible to this problem. By specifying colors produced by the supplier, this problem can be avoided. Checking. Early checking may be caused by improper formulation or application that causes the coating to shrink upon curing. Excessive thickness or rapid curing may be responsible. Aging will eventually cause checking of most coatings. It will often occur when existing paints are topcoated with more rigid finish coats that do not expand or contract as easily. Dry Spray. Dry spray produces an uneven, pebbly finish with holidays. The holidays can be verified on a metal substrate with a holiday detector. It occurs most frequently when applying coatings with fast evaporating solvents on warm days or when the spray gun is held too far from the surface being painted. Sagging. Sags may not permit complete curing of the body of oil-based coatings and so may be soft below the surface. Reduced film thickness in the areas where the sagging initiated may be detected using a magnetic thickness gage on steel surfaces and by using a Tooke gage on other surfaces. Orange Peel. Orange peel is a defect of spray application. It usually is widespread, when it occurs, and is easily identified by its resemblance to the skin of an orange. Wrinkling. Wrinkling occurs mostly with oil-based paints that are applied so thickly on hot days that the surface of the film cures rapidly to form a skin that does not permit oxygen to reach the interior of the film to cure it. The coating under the ridges is usually soft. Ridges generally occur in parallel rows.

11.4.2 Film Failures. The following paragraphs further describe the film defects chosen in the initial decision. Crawling. Crawling, sometimes called bug eyeing or fish eyeing, occurs during coating application, frequently on contaminated surfaces. It can usually be detected at the time of application. The smooth, oval shapes resembling eyes are characteristic of crawling. Alligatoring. The characteristic checkered pattern of cracked coating will identify alligatoring. The coating is quite inflexible and cannot expand and contract with the substrate. It is a special form of cracking or checking. Intercoat Delamination. Intercoat delamination is simply the peeling of a stressed coat from an undercoat to which it is poorly bonded. It may occur in a variety of situations, but occurs frequently when a chemically curing coating such as an epoxy or a urethane has cured too hard to permit good bonding of a topcoat. It may also occur when coating a contaminated surface. Intercoat Blistering. Intercoat blistering in a storage tank or other enclosed area is likely due to solvent entrapment. In water tanks or other areas exposed to water, the trapped solvent will cause water to be pulled into the blister.

If the blisters are large, unbroken, and filled with water, it is sometimes possible to smell the retained solvent after breaking them. Intercoat blistering may lead to intercoat delamination. Pinpoint Rusting. Pinpoint rusting is frequently caused by applying a thin coating over a high profile steel surface. A thin coating can be verified using a magnetic thickness gage. It may also occur when steel is coated with a porous latex coating system. Pinpoint rusting may initiate corrosion undercutting of the coating. Cracking. Cracking is the splitting of a stressed film in either a relatively straight or curved line to an undercoat or the structural substrate. Cracking may occur from rapidly curing coatings such as amine-cured epoxies. Mudcracking is a more severe condition caused by rapid drying, particularly by heavily pigmented coatings such as inorganic zincs. It also occurs with latex coatings applied too thickly on hot days. On wood, too thick or too inflexible a film (usually a buildup of many layers) can cause cracking perpendicular to the grain of the wood. Blistering to Substrate. The blisters that arise from the substrate may be broken or unbroken. If broken, they may lead to peeling and be hard to identify. Blistering to wood or concrete/masonry substrates may be caused by moisture in the substrate. Peeling. Peeling is the disbanding of stressed coatings from the substrate in sheets. It is generally preceded by cracking or blistering. Flaking (Scaling). Flaking or scaling is similar to peeling, except the coating is lost in smaller pieces. Flaking of aged alkyd coatings occurs commonly on wood.

11.4.3 Examples of Using Decision Trees. The decision trees 1 and 2 (Figures 24 and 25) can best be understood by using examples. Example of Surface Defect. This example is a surface defect that does not penetrate the finish coat so that use of decision tree 1 is required. We start with Question 1, "Does detergent washing remove the defect?" In our example, the answer is "Yes," so we proceed to Question 2, "Does wiping with a dry felt cloth remove defect?" This time the answer is "No," so we proceed to Question 3, "Does defect disappear when treated with household bleach?" In our example, the answer is "Yes," so we have tentatively identified the defect as "Answer 2" mildew. Example of a Film Defect. This example is a defect that penetrates the finish coat so that use of decision tree 2 is required. We start with Question 10, "Do oval voids that originate at time of coating application expose an undercoat or the structural substrate?" In our example, the answer is "No," so we proceed to Question 11, "Does the failure expose only an undercoat?" This time the answer is "Yes," so we proceed to Question 12, "Which best describes the failure?" In our example, the answer is "Peeling topcoat to expose undercoat," so we have tentatively identified the defect as "Answer 17" intercoat delamination.





Подпись: Figure 25 Decision Tree 2: Support for Analysis of Coating Defects That Penetrate the Finish Coat