A retrospective view of coatings adhesion testing.

A large number of short-term destructive laboratory tests have been devised to access the adhesion of coatings to various substrates. Some questions arise about the ability of such tests to predict adhesion over extended periods of time. These tests are reviewed and their advantages and shortcomings are examined.

An alternative useful test has been generally available. It is invariably correct in predicting the adhesion of coatings over extended periods of time. It has had no name but Normal Service Life Adhesion (NoSLA) is suggested.

INTRODUCTION

The term adhesion, as generally understood in the coatings industry, implies the continuous attachment of the coating to the object upon which it is applied throughout its normal service life. If the coating remains attached during this time, it is judged to adhere. If it detaches for any reason, even in minor areas, it is judged to have failed in adhesion.

How can such potential failures or successes be measured in the laboratory in a short time?

DISCUSSION

Many tests have been devised over the years to try to accomplish this desired result. More than 60 are listed. (1) Here, in random order, are the most commonly used tests:

Cross-Cutting  Mandrel
Knife-Cutting  Deceleration
Blister        Peel
Impact         Shearing
Centrifugal    Tensile
Scratch        Abrasion

Some are easy to perform, requiring minimum equipment. Some require sophisticated and complex instruments. Considerable information about the physical and fundamental aspects of coatings behaviors has been developed in some of these studies.

Of the long list, some tests are so impractical that they have not been found useful. Others yield some semblance of applicable information, provided that they correspond to stresses experienced during the coating's lifetime exposures. They are all mechanical and short-term, destroying the coating used to make the test.

Possibly the method that has yielded the most sophisticated scientific data of all the methods listed above, is the use of the instrumented sharp knife.

KNIFE-CUTTING TESTS

Knife-Cutting studies can be used to explain the complicated interrelationship between the interfacial bond strength, which is commonly referred to as adhesion, and the cohesive properties of the coatings system, often called toughness. It is quite intuitive that such a relationship should exist, (2,3) but no detailed study had been made.

Thus, the thickness of a coating possessing a given interfacial bond strength can show major changes in both practical and theoretical adhesion. This is generally regarded as the limiting component of adhesion because cohesive strength increases linearly with thickness. To make the distinction between interfacial bond strength and cohesion, the term hesion was coined. The then new instrument developed to measure this was named the Hesiometer. (4) The Hesiometer is an instrument that employs a very hard, sharp knife. The knife is pushed along the top of the substrate of the coating. The force required to remove the coating is measured. A very hard Carboloy knife is employed because it retains a sharp cutting edge over prolonged use. The sharpness of the knife at its very tip influences the measurements.

When coatings of increasing thickness are measured, the force required to remove them increases linearly. (3) Thus, by changing the profile of the sample coating to a wedge shape allows the Hesiometer to measure the force regardless of the coating's thickness. Some unexpected results were obtained. It was found that, starting at the thin end, the force required to remove the coating initially increased linearly with the thickness, as expected. (4) However, at some value of the thickness, which was quite arbitrary with any coatings system applied to any arbitrary surface, the cutting force would suddenly decrease, often to a small fraction of the maximum. Whereas a pure cutting mechanism persisted in the initial stages of the measurement, this would suddenly change to a cracking or chipping mechanism. The measured force here changed rapidly from a maximum, where the chipping was initiated, to zero while the chip was being removed.

At even greater thicknesses of the coating, another reduction in the measured force to still lower values may be encountered. Upon inspection, this is because the removal process has turned into a steady state pure peeling process. The various cutting removal processes are shown in Figures 1-4. They are enlarged frames of a high speed movie.

Figure 5 shows the cutting force required to remove the coating plotted against its thickness. The typical measured responses in the force required for removal are labeled. The initial cutting section shows that the interfacial bond strength is higher than the cohesive physical strength of the coating in shear. In other words, the coating is defined to adhere to the substrate.

Where the abrupt lowering of the cutting force occurs, a switch in the removal mechanism occurs, as described. The interfacial bond strength is lower than the gross cohesive properties of the coating; thus, the coating does not adhere as well. The consequences of these radical changes of coatings thickness will be discussed later.

Another very puzzling result found in the work of Rossman et al. (2) and Green et al. (3) is that at zero thickness, the force value is never zero. This was very difficult to explain until it was found that the measured results change with the rake angle of the knife. Although the knife is manufactured to a specific measured rake angle for the bulk of the blade, this changes radically at the very tip. It may be very close to cylindrical at very low thickness because of the graininess of the Carboloy material at microscopic values. Of the various materials for knives that were examined, none have been free of this problem. This fact is emphasized because one of the original papers on knife-cutting methods (3) interpreted the intercept on the thickness axis as a measure of the adhesion of the coating. This, perhaps unfortunately, is not true due to the reasons previously described.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

Coatings have been cleanly removed from most metals, plastics, glass, wood, and dense paper substrates with the Hesiometer. Topcoats can also be removed from primers. A prerequisite is that the substrate surfaces be reasonably flat and smooth. The measured results are remarkably repeatable, within a few percentage points. Measurements can be made either from the thin end to the thick end or the reverse.

By sophisticated examination of the vectors involved in cutting mechanisms, vector diagrams, such as shown in Figure 6, can be constructed. The analysis is adapted from metal cutting theory. (5) The actual vectors change radically with the rake angle [alpha] (alpha). Instead of cutting the coating from the substrate, rake angles can be found where the vector [F.sub.T]--the thrust force lifting the coating--can actually be reversed, so that the force is directed downward. Then, even poorly adhering coatings will seem good.

It was found that the final and important results of the measurements made with the Hesiometer could be expressed in fundamental scientific units. One is Surface Free Energy of the separation of two bodies in dynes/cm or ergs/[cm.sup.2] (lb/in. (2) in English units). Similarly, the cohesive shear or tensile properties can be derived in dynes/cm (lb/in. (2)) from measurements made with the Hesiometer.

These studies were perused in detail because it was expected that such fundamental scientific information could lead to conclusions concerning the long-term service life of the coatings studied. It should be noted that none of the other 60 or so measurement methods listed can reveal this type of information.

PEEL TEST

One other new method for coating adhesion measurement, which the author has suggested, is the Wedge Peel test. (4) This method consists of applying a wedge-shaped coating of a tough, flexible material, such as a plasticized vinyl, on a rigid substrate. Removing such a coating by a simple peeling action is often possible, starting at the thick end, as illustrated in Figure 7. As more of the coating is pulled from the substrate toward the thinner and still-adhering end, a point will usually be found where the coating can no longer be removed by simple peeling. Despite how hard one tries, the peeling action can no longer be effected at thicknesses below this critical value. Where the coating peels, it would normally be judged to fail; whereas below the critical thickness, it would be judged to adhere.

[FIGURE 6 OMITTED]

This simple test again illustrates quite clearly the close relationship between the interfacial bond strength of a coating and its cohesive strength due to such factors, among others, as increased thickness. The method can provide results in both the scientifically desired nomenclature of Surface Free Energy, by use of the split beam theory, (6) and in the practical nomenclature of Interfacial Bond Strength. Unfortunately, most coatings do not bend sufficiently without cracking, so this test cannot be universally applied.

Both the instrumented Knife test and the Peel test clearly illustrate that there are two important factors in the practical adhesion of coatings. One is the interfacial bond strength. This is often thought of as the adhesion of coatings. Second is the cohesive strength, often thought of as the toughness of the coating. This consists of tensile, shear, elasticity, and other bulk properties of coatings. To exercise either of these tests, the coating must be removed and, thus, the tests are destructive.

A simple paradigm can be constructed to illustrate this point:

Hesion = [adhesive force]/[cohesive force]

When this ratio is more than one, the coating is defined as having satisfactory adhesion. When this ratio is less than one, the coating is defined as having poor adhesion.

Next, we will examine briefly the other proposed traditional adhesion measuring methods listed.

OTHER TEST METHODS

CROSS-CUTTING: Scratches are made close together in parallel rows perpendicular to each other. A razor blade is most commonly used. An adhesive-faced tape is then pressed against this area and rapidly pulled up. If a substantial portion of the coating sticks to the tape, adhesion is judged to be poor. If very little adheres, adhesion is satisfactory. It is essentially an augmented Scratch test, which can predict little of long-term adhesion. It is also very subjective.

MANDREL: A coated substrate, usually a metal, is bent over a sharp angle, either by hand or by use of an instrument, such as the brake employed in the sheet metal business. If the coating cracks, its adhesion is judged to be poor. If it survives, it is satisfactory. In some cases, it is a practical test that may have limited application.

KNIFE-CUTTING: This test, as performed with the Hesiometer, has been discussed previously. It is unique among the tests in being able to explain the important interaction between the interfacial bond strength and cohesive properties of coatings.

DECELERATION: This is really an unusual test. The coating on a substrate is shot out of an air rifle against a target that has a small hole in it. The substrate is sized to be stopped by the target; the hole allows part of the coating to pass through. If it detaches when it hits the target hole, it has poor adhesion, but deceleration is what pulls the coating off. If it survives, the adhesion is judged to be satisfactory. Considerable interpretation is required to use this method. It is impractical.

BLISTER: The Williams Blister test (7) is not equivalent to the blister test normally encountered in a humidity cabinet. In the Williams test, a small hole is drilled through a test panel. Then, a small debonding spot is created around the hole. The most common material used for this is grease. The movement of the outer edge of the debonded coating is measured in relation to an applied gas pressure. Use of this method is restricted since the edge of the debond must be visible for the test to continue. This laboratory test is the only one, other than the original Hesiometer test and the Peel test, that allows the calculation of the interfacial bond strength in terms of the surface free energy in absolute units. It is not a practical test since this type of coating's deformation will rarely be encountered in service.

WEDGE PEEL: This method is useful only for certain limited coatings, as described before. However, it does allow the calculation of interfacial bond strength and cohesive properties in absolute scientific units.

IMPACT: This test consists of propelling gravel, stones, or sand against the face of the coating applied to a substrate. In addition, steel darts with various pseudospherical points are sometimes pushed or dropped against the surfaces. If the coating pops off, the adhesion is poor; if the coating remains, it is satisfactory. In some forms, the test may have practical applications for measuring adhesion, as will be discussed later.

[FIGURE 7 OMITTED]

SHEARING: A circular probe is cemented to the top of a coating and a torque force is applied. Force required for the coating to detach is measured to determine adhesion. Sometimes an upper limit of torque is arbitrarily set; if the coating does not detach, it is considered satisfactory adhesion. The adhesive bond must be stronger than either the interfacial bond strength or cohesive strength of the coating; otherwise, only the adhesive bond is measured.

CENTRIFUGAL: This is another unusual test. (8, 9) A small iron top measuring a quarter-inch in diameter and a quarter inch long is magnetically suspended in a vacuum chamber along its axis. A little spot of paint is applied to the surface of the top, and it is spun fast enough for the spot to detach. Centrifugal force here detaches the coating. If the speed at which it detaches, and its specific gravity and thickness are known, the interfacial bond strength can be calculated upon detachment. Speeds of up to two million rpm or more are often required to achieve the detachment. I had one of these built for my laboratory. While a logical idea, the test did not achieve usable results due to an unexpected reason. The induction current needed to spin the top heated it, like in an induction furnace, to unacceptable temperatures. Substantially slowing the rotational acceleration mitigates somewhat the heating problem, but the test then can take many hours. It is impractical.

TENSILE: Mechanically, this is an easy test. Like the Shearing test, a probe is cemented to the top of the coating and a tensile force is applied to pull the coating off the substrate. Again, the force required to detach the coating is measured to determine adhesion. A maximum value might be arbitrarily set above which adhesion is always judged to be adequate. It has the same problems as the Shearing test.

SCRATCH: This test drags a sharp object across the surface of a coating. This test is often used on coatings applied to metals, such as automobiles or appliances. The coating's ability to resist scratching is used to determine adhesion. A crude example is dragging a key across the finish of an automobile. This will be discussed later.

ABRASION: This test indirectly determines the cohesive properties of the attached coating. It does not yield useful information concerning adhesion.

RESULTS

After reviewing all these tests, an important question immediately arises. In the laboratory, which of these tests, individually or in combination, can predict the successful long-term adhesion of all formulated coatings on any chosen substrate? They are all mechanical and short-term, destroying the coating used to make the test. Their value is questionable.

A test has been generally available and recognized, but it does not have a name. It is invariably correct in not only assessing the adhesion, but, in its broader sense, also all the other properties of coatings systems' success or failure. The name Normal Service Life Adhesion (NoSLA) test is proposed. The examples that clearly explain this test follow.

Examples of the Normal Service Life Adhesion (NoSLA) Test

AUTOMOTIVE COATINGS: These coatings, everyone will agree, have come a long way in the recent past in protecting the iron and plastic, including fiberglass, parts of automobiles. It is really amazing that a very thin layer of pigmented plastic only several thousands of an inch thick, which is called a coating, can protect car bodies for 10 or 15 years. Cars are routinely exposed to heat, cold, high humidity, rain, salt water, and more. When a new car is purchased, the exterior appearance will often largely persuade the buyers which model to choose. We look at the gloss and general appearance with the full expectation that the coating will adhere throughout the normal service life of the car.

This is the NoSLA test: a coated object is exposed to its expected operating environment during its normal service life. Does the coating act as expected? We are evaluating the coating's adhesion, as well as many other properties, over the Normal Service Life of the substrate. Subconsciously, we have used this test throughout our careers in the coatings industry. We have just never given it a name. It is an infallible test that has resulted, over the years, in improvements to coating performance on the bodies of automobiles, leading to today's excellent results.

Rusting and blistering of automotive coatings are occasional coatings failures. In rusting, the coating is considered to have failed although the iron substrate is really the culprit. Many attempts have been made to mitigate this problem through coatings protection, mostly unsuccessfully, in the past. Look at the bodies of some older cars. Rarely has the whole substrate rusted out from under the coating. Not surprisingly, the coating alone could no longer hold the body together, as some buyers might have hoped. This is despite the fact that the consumer expects the coating to compensate for all manufacturing defects.

The solution has been achieved, not strictly by coatings protection of the raw substrates, but by treating the iron surface. For instance, by passivating the substrate, both outside and inside, with tin or zinc plating, the rusting of automobile bodies now seems to have become a minor problem. Coatings are now performing satisfactorily in this respect. Rusting, a type of NoSLA failure, has been studied with such relatively short-term laboratory tests as the Salt Spray Cabinet. However, this has lost much of its significance with the passivity treatment of the car body.

Blistering is another problem that has confronted the industry under NoSLA conditions. For some unlucky owners, ugly blisters have appeared on the surfaces of the cars, under normal driving conditions. Again, this problem has been quite successfully solved by a relatively short-term laboratory test using the Humidity Cabinet. However, none of the mechanical tests listed would have predicted these failures.

Typically, these two failures pose a dilemma to the coating's formulator. Rusting usually occurs in cars driven in the northern states, where salt is often used in winter months. In contrast, blistering is usually found in the southern states, where high humidity and heat create the ideal conditions for blistering.

As an extreme example, and with tongue in cheek, perhaps the coatings industry could persuade the northern states and the southern states, respectively, or even the Federal Government, to prohibit drivers from the North to drive to the South. Then the NoSLA test would show only one problem occurring in each area. There could be a North NoSLA condition and a South NoSLA test condition. The problems, of course, are each different. The proposal fails because I live in the North, and occasionally like to spend some time in the South in a place like Florida. Consequently, it is imperative that cars pass a Continental NoSLA test. Both rusting and blistering must be avoided. Different NoSLA tests are needed for cars in different locations, circumstances, and areas.

Sometimes, the NoSLA test can be accelerated by exposing coated test panels to extreme weather conditions, as will be described in the section on House Paints.

The Occasional Service Life Adhesion Test

Another problem occasionally affects the finish of automobiles. If we are unlucky enough to drive behind a truck loaded with gravel or coarse sand, some of this may fall off to make dents in the finish. With the hard, brittle coatings needed to satisfy other requirements of car coatings, little spots of the coating may be often chipped off. This occurs despite the fact that no adhesion problems had previously been encountered. This type of event is the Occasional Service Life Adhesion test, a subsidiary of the NoSLA test.

Massive failures on many cars, resulting in rusting or blistering, may invoke recalls. In contrast, occasional chipping of the paint from the front of the car due to gravel is usually dismissed as minor and unavoidable. If we are meticulous, we will touch up the damage. Otherwise, we may just let it go if the damage is not severe enough to be a serious issue. This is also true of scratches or other minor damage. Some few owners purchase a vinyl cover to protect the finish of the car from impact damage. The majority are willing to accept this damage. It occurs rarely under normal driving conditions.

Mechanical, short-term laboratory adhesion tests, like the Impact test, predict that such coatings, where the coating chips off, have poor adhesion. In fact, the adhesion on the car has been satisfactory under all conditions except impact. The fact that the paint chipped off clearly shows that the interfacial bond strength is lower than the cohesive strength of the coating, indicating that by definition it should have failed. While showing unsatisfactory adhesion according to the Impact test, the bulk of the coating has already adhered quite satisfactorily before the gravel impacts. It will undoubtedly continue to do so for a long time after the damage has been repaired.

[FIGURE 8 OMITTED]

Many cases of this type of behavior can be described. The mechanical, short-term, and destructive adhesion tests cannot predict the long-term NoSLA of coatings on a practical basis. Examples include the Cross-Cutting test, the Knife test, and the Scratch test. They may predict poor adhesion when, in fact, good adhesion exists on the body of the car, even if the interfacial bond strength is lower than the cohesive strength. These ideas carry through all adhesion problems of any coating or substrate.

This concept may be extended to direct ex-post-facto testing on cars. The results of Impact tests have already been described. There are the other relatively easily performed tests such as Cross-Cutting, Knife-Cutting, and Scratch tests, which can be performed on the body of any car.

Automobile wrecking yards contain many vehicles, particularly older ones. The coatings have adhered satisfactorily, despite the fact that some serious damage may have been inflicted on them.

A representative sample of vehicles was tested using the methods listed herein. Almost invariably, they failed these tests. All three tests are subjective to a high degree. The Cross-Cutting test is particularly subjective, dependent on spacing of the cuts and how strongly the tape adheres to the surface of the coatings. How should these failures be judged?

Photographs of the results are shown in Figure 8. The samples span the automotive years from 2000--the newest--back to 1982 for the oldest. All clearly show that although the NoSLA is satisfactory, in each case, almost all of the short-term tests predict failure. New cars were not available for this type of testing. One would assume that the finish on new cars would have also failed the short-term tests.

APPLIANCES: The exteriors of metal appliances require separate NoSLA standards than painted cars. Rusting virtually never occurs unless we throw massive quantities of salt water on the surface of a refrigerator or dishwasher. Neither blistering nor chipping of the surfaces occur under normal service use, either. Could detergent resistance be a problem? An accelerated NoSLA test could certainly detect this.

HOUSE PAINTS: For accelerated exterior testing of house paints, paint farms are commonly employed. These are weathering tests where the main object is to detect if the coatings will fail in such characteristics as color fastness, loss of gloss, or other appearance factors. The tests are performed by exposing test panels to the extreme weather conditions, particularly ultraviolet radiation that emanates from bright sunlight. This is accomplished by keeping the panels tilted directly toward the sun. These tests reputedly accelerate the time to judge performance by a factor of up to three--one year of farm exposure equating to about three years of NoSLA testing. The farms are usually in severe climates, although many are in various parts of the country for specific reasons.

The tests, besides evaluating the appearance, may also cause adhesion problems due to the deterioration of the coating. Flaws may appear as blistering, cracking, peeling, or other problems. Again, not one of the mechanical, short-term, destructive adhesion tests could have predicted the adhesion failures due to embrittlement of the house coating as a result of radiation exposure over time. A NoSLA test is required.

Two types of results may occur in the Normal Service Life Adhesion testing of coatings. In one case, no flaws develop during the many years that the coating is expected to adhere to the substrate. The other result is that the coating fails for some unexpected reason. This need not always be bad, particularly if it happens in the early part of the test period.

The reasons for short-term failure can usually be identified easily. Hopefully, the formulation can be improved. A short-term test can check to see if this problem has been mitigated, and the coating brought up to NoSLA standards.

Thus, the NoSLA test for automobiles must apparently be different from that for appliances, exterior house paints, interior house paints, or any paints for specific industrial or private use.

This has been the traditional and accepted means for improving coatings over the years. Shall we give this type of test a name? Is NoSLA the best choice?

The question remains: How can short-term mechanical and destructive laboratory tests be expected to accomplish all the different requirements for all these long-term service conditions? How can adhesion failures such as those caused, for instance, by rusting or blistering be predicted by any of these short-term laboratory tests? Can they predict failures caused by embrittlement due to long-term exterior exposure? Can they predict the possibly false assessments measured by Impact or Scratch tests? Here, for instance, automotive coatings will have adhered for extended periods until stressed by such rarely occurring accidents as gravel impacts or deliberate scratching. The Impact or Scratch tests would have originally predicted failure by a loss of adhesion, although these coatings may continue to perform for many years to come.

However, we should not ignore the fact that some of these laboratory tests have given considerable insight into the various factors contributing to both the theoretical and practical performance of adhesion. Thus, for instance, understanding the ratio between the interfacial bond strength and the cohesive strength of coatings gives valuable information about possible adhesion failure problems.

While all this is true, we must come to the very unfortunate conclusion that none of the short-term laboratory tests can predict the Normal Service Life Adhesion success or failure of any coatings system.

With all the problems encountered, it is, of course, highly desirable for the perceived needs of the coatings formulations technician to have a quick, simple test to predict the NoSLA of a newly formulated coatings system.

Although such a test does not yet exist, what can be done? Will a test combining the physical, chemical, and electrochemical attributes of coatings solve this baffling question? Whoever succeeds will be long remembered.

CONCLUSIONS

* None of the proposed short-term, mechanical laboratory tests--no matter how simple or sophisticated--will unambiguously predict the long-term adhesion of any coating to any substrate.

* However, if other relatively short-term laboratory tests, such as those described, parallel the normal service life of the coating, they may be used for very specific tests under very specific circumstances. These may be mechanical or chemical.

* A new name is suggested: The NoSLA Test. The NoSLA test unambiguously predicts the adhesion of a coating over the Normal Service Life of a coating. For high-performance coatings, this may require many years. Under extreme conditions, the test may give results in a very short time.

* The NoSLA test has been used for many years, but until now had not been given a name.

* Is Normal Service Life Adhesion, with the acronym NoSLA, acceptable, or is there a better name?

ACKNOWLEDGMENTS

Many sincere thanks are due to Dr. Harvey Alter and Dr. Kenneth L. Hoy, both formerly of the Union Carbide Corporation, now retired, for the many valuable suggestions and additions to the contents of this article. They have added significantly to its value.

ADDENDUM 

Dr. Alter also has pointed out, correctly, that the NoSLA concept applies directly to other areas as, for instance, medicine, dentistry, cosmetics, and more. In medicine, the recently developed liquid bandages, sutures, and homeostasis employing cyanoacrylates use coatings technology. In dentistry, teeth are often protected with acrylate and other coatings. Certainly lipstick is a coating. This is in addition to the subject for which this article was really intended, the paints used on automobiles, appliances, houses, etc.

There are many different types of coatings. The term Normal Service Life Adhesion, with the acronym NoSLA, applies to all of them. Will any and all coating really "stick" long enough to fulfill their required Normal Service Life in order to be useful?

FINAL NOTE

Much appreciation is due to Paul Guevin, Jr. of P.R. Guevin Associates for contacting me at Dr. John Weaver's suggestion. I had paid scant attention to the problems of the adhesion of coatings in the 30 years after I had retired from Union Carbide Corporation. Hindsight can often open our eyes to factors that we might have missed while we were immersed in the subject. This article is the result of such hindsight.

I hope it will inspire thoughtful contemplation of the factors discussed. All comments, positive or negative, are welcome and should be directed to: Editor, c/o JCT CoatingsTech, 492 Norristown Rd., Blue Bell, PA 19422-2350; [e-mail], or the author at the noted address.

References

(1) Paint Testing Manual, Sward, G.G. (Ed.), 13th Ed., 314-332, 338 (1972).

(2) Rossman, E., Weise, K., and Schubbe, A., Farben Ztg., 43, 1247 (1938).

(3) Green, H. and Lamattina, T.P., Anal. Chem., 20, 523 (1948).

(4) Asbeck, W.K., "Measuring the Adhesion of Coating Systems," JOURNAL OF PAINT TECHNOLOGY, 43, No. 556, 84 (1971).

(5) Merchant, M.E., J. Applied Phys., 16, 267, 318 (1945).

(6) Berry, J.P., J. Applied Phys., 34, 62 (1963).

(7) Williams, M.I., J. Appl. Polymer Sci., 13, 29 (1969).

(8) Malloy, A.M., Soller, W., and Roberts, A.E., POCRA, 116, 18, 14 (1953).

(9) Alter, H. and Soller, W., J. Appl. Polymer Sci., 5, S7-8 (1961).

by Walter K. Asbeck

Coatings Consultant*

*301 Knollwood Trail, Richmond Heights, OH 44143.