COLOR AND LONG-TERM COLOR RETENTION IN COMPOSITE PATCHING SYSTEMS FOR STONE AND MASONRY

 

Summary Statement:

 

Precise measurement and understanding of composite patch color and color change are prerequisites to the optimization of long-term repair aesthetics.

 

 

Abstract:

 

Anecdotal and photographic data have limited value in assessing color and long-term color retention in composite patching systems for stone and masonry. Observer subjectivity and process inaccuracies are inherent in these methods. Modern spectrophotometers and color management software provide objective tools for precise measurement and communication of color and color change data.

 

Objective color measurement tools were used to define color changes in brownstone repair mortars exposed to accelerated weathering in the laboratory and to natural weathering over a twelve-year period. The data derived describe the extent and nature of the changes in color, and are the basis for discussion of the role of efflorescence as a contributing cause.

 


COLOR AND LONG-TERM COLOR RETENTION IN COMPOSITE PATCHING SYSTEMS FOR STONE AND MASONRY

 

The intended aesthetic outcome of employing composite patching systems in architectural masonry restoration is a repair that is visually indistinguishable from the host substrate. Close matching of composite repair mortar color to the original stone or masonry is fundamental to the fulfillment of that objective. The long-term retention of closely matched color is a more complex challenge, and ultimately may be the more decisive determinant of aesthetic success or failure.

 

Long-term color retention of composite repair mortars has assumed increased significance as a conservation issue as a result of prolonged service life expectations for composite repairs. No longer relegated to the role of short term, temporary or minor repair, some composite patching products are now routinely providing decades of service on repairs of sizable scale.

 

The work undertaken to date in this ongoing study has been aimed at gaining a better understanding of how composite repair material color responds to extended periods of both natural and accelerated weather exposure, how to catalogue and communicate color changes, and how to apply this understanding to the objective of achieving aesthetically successful restoration work. It is part of a broader, continuing effort, aimed at making incremental improvements in various aspects of repair mortar performance. Such development work is multi-faceted and constantly evolving, rather than narrowly focused with clearly defined beginning and endpoints. The objective is not to develop “perfect materials”, which may never be realized, but to move existing, effective materials further along the path toward that ideal. Opportunities for improvement are part of the intimate knowledge that can be gained by a material’s manufacturer over a long period of both ongoing development and large-scale commercial use.

 

In the course of pursuing such advances a knowledge base is developed that is not particular to any product. Although most of the work performed in this study employed cementitious, latex-modified brownstone repair mortars manufactured by Edison Coatings, Inc., the research concepts, methods and direction have broad relevance to color-matched repair mortars in general, whether field-mixed or produced by any of a number of manufacturers. Objective definition and communication of repair mortar color and color retention represent a common interest.

 

 

The Role of Anecdotal Evidence

 

The determination as to what constitutes aesthetic success or failure is, today, largely a matter of personal judgment. Sites may be revisited periodically after an intervention, and individuals may observe and judge how well repair materials and their color are withstanding the test of time. The record may be augmented through the taking and cataloguing of photographic images.

 

There are times, particularly in extreme cases where the evaluation of relative aesthetic success or failure is not terribly difficult, when these methods may be sufficient. For example, the following cases of two restored 19th Century brownstone churches, situated just a few streets apart in Middletown, Connecticut, represent distinctly antithetical ends of the spectrum in terms of the color retention achieved.

 

Figure 1

 

The First Church of Christ, on the left side of the figure, was restored in 1996 using three standard colors of a commercial, prepackaged brownstone repair mortar. St. John’s Church, on the right, was restored in 1999 using a field-mixed repair mortar, reportedly prepared in accordance with guidelines learned by the contractor during a restoration training course. The churches were photographed in 2001, five years and two years after their respective restorations.

 

The extensive five-year-old repairs at First Church, including the depicted upper window molding (left side completely rebuilt), buttress caps (both repaired) and base molding (section projecting forward) are not visually obvious because color remains a close match to the surrounding brownstone. Two-year-old repairs to similar areas at St. John’s, however, are immediately distinguishable due to extreme color change. Cracking, which is both aesthetically and functionally objectionable, was also evident in some of the St. John’s repairs.

While anecdotal data of this type can be useful, there are limits to the value of this methodology for observing and reporting on color and color retention. If aesthetic success or failure is to be assessed on a subjective basis, then it must also be recognized that reports by individual observers are inherently disposed toward inaccuracies and distortions, however unintentional they may be. Whether a particular repair represents a “good” color match is subject to a complicated set of personal values, expectations and perceptions.

On a perceptual level, one basic obstacle to accurate anecdotal reporting is that not everyone “sees” color in the same way or with equal accuracy. At one extreme, approximately one in 10-12 men and one in 165 women exhibit some degree of colorblindness1. At the other end of the spectrum, only a small percentage of individuals exhibit sufficient visual color acuity to accurately distinguish between multiple close shades of the same color, as assessed in color perception tests. Such tests have been used by some paint manufacturers to evaluate an individual’s potential capacity for performing color-matching work. Most observers fall somewhere between the two extremes, with levels of perceptual acuity that vary accordingly.

 

The Need for Objective Color Measurement & Communication

 

Color tolerance determinations based exclusively on subjective individual observations represent undefined standards. Although the ultimate goal may be to satisfy the aesthetic sensibilities of a particular individual or group of individuals, undefined standards leave ample room for disagreement and have the potential to engender conflict. Except for the most obvious cases, terms like “aesthetic success” and “aesthetic failure” become arbitrary, due to the lack of standardized definitions, guidelines or practices.

 

Based on the illustrative cases of the two churches in Middletown, one might argue that photographic imaging offers a less subjective way of representing aesthetic success or failure, because it allows each of us to see and judge the results for ourselves. Although the black-and-white photographs in Figure 1 have obvious limitations in illustrating color change, even full-color photographic imaging is of limited value as a measure of Color or Color Retention. Photographic images are strongly affected by lighting conditions, including weather, season and time of day. Angle of observation, photographic film bias, printing processes and media, and projection or viewing devices all introduce significant potential sources of color distortion. These factors, unrelated to actual color change of the repair materials themselves, affect the colors ultimately seen in the photographs and undermine the viewer’s ability to make critical judgments with regard to color.

 

While photographs have some value in allowing us to compare patch appearance to the adjacent stone, at a particular instant, they do not provide an objective way of evaluating how or precisely how much these materials are changing in color over time.

 

OBJECTIVE COLOR MEASUREMENT & COMMUNICATION

 

There is essential knowledge and experience to be gained by studying color and color changes over time in a precise and objective manner. Such assessments can provide critical insight into the nature of color change as well as implications as to its probable causes. The data also provides a basis for further optimization of long-term color retention of composite repair systems. Prerequisite to any such studies, however, must be the definition of a common language for precisely defining and communicating color information.

 

The Munsell color system2, developed in 1905 and still in use to some extent today, was an early attempt to catalogue colors in an objective and quantitative manner by creating a series of fixed color standards. The colors of the spectrum are divided into ten color groups or “hues”, organized around a vertical axis called “value”, which runs from light at the top to dark at the bottom. “Saturation”, or how pure the colors appear, runs from grayer/”muddier” at the center to more saturated/”purer” at the outer perimeter. The problem with the Munsell system is that it still relies on human observers to compare colors to a particular Munsell color standard, and the number of Munsell standards with which colors can be compared is generally limited. The Munsell tree shown in Figure 2 contains fewer than 400 standards, and even the most extensive sets of Munsell standards typically provide no more than 1550 colors3.

 

The development of the spectrophotometer, twenty-three years after Munsell’s system, opened the door to more precise and objective color measurement. Mathematical models were developed for describing color and they have evolved over time, facilitating more precise color communication.

 

Figure 2

 

The most widely used model for measuring and communicating color today is the 1976 CIE L*a*b* sphere4, which provides a means of precisely describing every possible color in terms of geometric coordinates within its three-dimensional “color space”.

 

As in the Munsell tree model, the Vertical axis, “L”, is lightness and darkness, with White at the top and Black at the bottom. If the sphere is bisected along its equator, the two-dimensional, horizontal cross-section is described by two perpendicular axes: “a”, the red-green axis and “b”, the yellow-blue axis. Colors become grayer or “muddier” as we move toward the center and cleaner or more saturated as we move toward the edge. Alternatively, the same color space can be described in polar coordinates. This is the basis for the “L C h” model (not shown).

 

Stated simply, the spectrophotometer measures reflected light across the spectrum of visible light wavelengths. These reflected light measurements are converted into numerical values, which describe the color of an object in terms of a 3-dimensional set of coordinates in color space. Today’s color software can rapidly calculate color differences in comparison to an established standard, and can report those differences in a number of different ways. These include both numerical and graphic reports.

 

The combination of modern spectrophotometer and commercial color software provides a means of measuring and communicating millions of colors with objective precision. That capability cannot be equaled by human observers working with limited numbers of “hard” printed standards.

 

Coatings manufacturers utilize spectrophotometers and color management software to aid in their daily color matching and batch color correction work. These tools provide the means for quickly matching and correcting colors of such products as breathable masonry coatings, adhesives, sealants and repair materials. They replace an otherwise tedious, time-consuming, trial-and-error process requiring a skilled and experienced operator.

 

Laboratory Study of Composite Patch Color and Color Retention

 

The particular challenges of color control and retention in brownstone patching compounds was the basis for selecting deep brownstone colors for evaluation in Edison Coatings’ study of composite patch color and color retention. Slight color changes in deep brown-colored patches tend to be more noticeable and objectionable than in lighter patches, such as those matched to typical Indiana limestone and many lighter types of sandstone. If good color retention could be achieved in this “worst case scenario, it was reasoned, then it should prove much easier to achieve on other, less challenging color substrates.

 

Studies performed in the Edison Coatings, Inc. laboratory utilized a Dataflash 100 spectrophotometer and Colortools v.1.3 Quality Control software to measure and report changes in brownstone patching material color over time and exposure. The exposure was provided by the use of a QUV accelerated weathering apparatus.

 

The Dataflash 100 dual beam spectrophotometer utilizes a pulsed xenon light source and dual 128-element diode arrays to measure reflected light in the 400 to 700 nm wavelength range. It can measure reflectance percentages from 0 to 200%. A small area view aperture was used to allow careful selection of particular “spots” for color measurement, when required. The area measured in each reading was 5 mm in diameter. Each reading reported was the average of three measurements of the same 5 mm “spot”. The instrument is calibrated and tested at the beginning of each eight-hour session.

 

ColorTools QC software communicates directly with the spectrophotometer through a cable connected to the computer’s parallel port. It records and catalogues color measurements for items designated as “standards”, against which color samples can later be compared. Its typical use in paint manufacturing would involve measurement of color for each batch of paint manufactured. Batch color is compared with the established Standard for that color, in order to determine whether the batch is within a pre-defined color tolerance. The software includes analytical tools for calculating color tolerance in accordance with a variety of commonly used scales, and for expressing batch color vs. standard color graphically. History files, “trend” displays and statistical quality analysis are also provided.

 

For the work involving measurement of patching compound color change, sample panels that had been properly cured and aged were measured and designated as the “Standard”. Panels exposed to accelerated weathering were measured and compared to that Standard to assess color changes.

 

It is understood that color change is a consequence of weathering, and that the changes in many building products are closely associated with exposure to ultraviolet radiation and wet-dry cycling5. The QUV apparatus, which is operated in accordance with ASTM G536, produces alternating cycles of ultraviolet radiation and hot water condensation to simulate the stresses of natural weathering. A variety of ultraviolet sources may be chosen, and operating temperatures and cycle lengths can be varied as well.

 

The Edison Coatings laboratory began use of its QUV apparatus in 1983, and has since logged tens of thousands of testing hours for many types of architectural and industrial coatings, cementitious and synthetic repair materials, weatherproofing treatments, stains and sealants. Specimens tested have included the Company’s existing product formulations, potential ingredient substitutions or changes to those formulations, potential new products and a wide variety of competitive products. The objective is to observe and compare how the various materials perform when tested under the same conditions, and to predict whether proposed changes will enhance or diminish weather resistance.

 

Over the course of this testing, and through correlation with observations made under natural weathering conditions, the laboratory has defined ultraviolet sources, operating temperatures and cycle times which correlate well with natural weathering. Hundreds of laboratories have similarly developed empirical “rules of thumb” to permit rough estimation of the correlation between hours of accelerated weathering exposure and years of natural exposure5. This method has proven extremely useful in ongoing product development and evaluation work.

 

For the accelerated weather exposure of brownstone patching compound samples, UVA 340 lamps were used. The apparatus was operated in alternating 4-hour cycles of ultraviolet radiation (dry) and hot water condensation (wet) at 60 and 50 degrees Centigrade, respectively.

 

Test panels were prepared by mixing the patching compound, designated as 701R, as specified by the manufacturer, and filling into 3” (7.6 cm) x 6” (15.2 cm) x ¼” thick (6.4 mm) open aluminum frames. Hardened panels were removed from the shallow, open molds after 72 hours, and initial color was measured. They were then allowed to air cure for 28 days. Following this curing period, panels were re-measured for color, prior to the start of the 800-hour accelerated weathering test. Some of the panels were exposed to accelerated weathering, while others were stored as Controls under ambient laboratory conditions.

 

Figure 3

 

Figure 3 shows two pairs of test panels, representing two different brownstone repair products. 701R is on the left. The product on the right is a commercial cementitious mortar that is no longer being produced by its manufacturer.

The two upper panels are the unweathered Controls for each product. The two lower panels have been exposed in the QUV apparatus for 500 hours. Although there is very little visible difference between the weathered and unweathered 701R panels on the left, the product on the right exhibits severe white streaking after exposure, suggesting a deficiency in this product’s weather resistance.

QUV exposure for the 701R mortar (left) was continued for a total of 800 hours, at which time the spectrophotometer was used to measure the precise changes that occurred. Color of an aged, unweathered Control was also measured at the conclusion of the 800-hour test period. This measurement was designated as the “Standard”.

 

The unexposed Controls were measured yet again after a total of six months’ aging to aid in estimating the color changes unrelated to weather exposure that are to be expected with time.

 

Figure 4: Color and Color Difference Data for Various Controls and Weathered Panels of Composite Brownstone Patching Compound

Specimen

da

Red/Green

db

Yellow/Blue

dL

Lightness

dE

Overall

Demolded panel, 72 hrs. old

0.39

0.08

-0.56

0.69

Panel 1, 28 days ambient cure

0.06

0.32

-0.14

0.35

Panel 2, 28 days ambient cure

-0.06

0.01

-0.21

0.21

CONTROL, 800 hours lab air

0.00

-0.01

0.02

0.02

Duplicate Control, 800 hours lab air

0.23

0.21

0.01

0.31

Triplicate Control, 800 hours lab air

0.34

0.39

0.11

0.53

CONTROL, 6 months lab air

0.44

0.29

0.45

0.69

Duplicate Control, 6 months lab air

0.09

-0.02

-0.42

0.43

Panel 1, 800 hours QUV Exposure

-0.12

-0.16

3.91

3.92

Panel 2, 800 hours QUV Exposure

0.59

0.04

4.09

4.13

Note: All values shown are averages of three measurements. Larger positive or negative values represent greater change. (No change = 0). dE is overall color change and is calculated as the square root of the sum of dL2 + da2  + db2 .

The significance of positive and negative data values are as follows:

Data

Interpretation

+da

Sample redder (or less green)

-da

Sample greener (or less red)

+db

Sample yellower (or less blue)

-db

Sample bluer (or less yellow)

+dL

Sample lighter

-dL

Sample darker

 

Figure 4 lists the results obtained for each of the sample panels measured. The unexposed panels exhibited relatively small color differences over time, with dE ranging from 0.21 to 0.69. The weathered panels exhibited dE of around 4.

 
Figure 5

 

The graphic output shown in figure 5 is the color difference plot for the brownstone patch after 800 hours of accelerated weathering. The color difference plot has two segments: The circular region at the center of the plot represents Chromaticity, or the degree to which the balance of red-green and yellow-blue are the same when comparing the weathered test panel with the unweathered control. The vertical bar at the right side of the plot is Value, expressed as dL, the difference in lightness-darkness between the two measured panels. The circle in the center of the Chromaticity segment of the figure represents a tolerance range of da, db = 1.0, an arbitrary color difference limit often used in paint color matching. Color differences of 1.0 or less are generally considered acceptable as paint color matches, although the eye can often detect some color difference below this limit.

 

There are no specific standards governing acceptable levels of color change in composite repair materials. The color changes or deviations observers consider acceptable can only be ascertained through testing of statistically significant numbers of human observers. Defining the means of developing such tests and standards are beyond the scope of this study, but ultimately the aim would be to correlate particular dE values in a number of different color ranges with human responses as to the acceptability of those color differences.

 

Color control of cementitious materials is more challenging than for typical architectural coatings such as latex paints. Cementitious products incorporate much lower levels of pigment than typical architectural paints and appearance is more strongly influenced by the binding matrix. While latex paint undergoes relatively simple drying and film formation after application, the process of cement hydration is more readily influenced by cure conditions and several other factors.

Depending on the temperature of cure, cement will form different crystal shapes, affecting color7. The amount of mixing water added to cementitious materials (water-cement ratio)8 and the rate at which the mortar mixture dries can also have an influence on color shading. Accordingly, colored cement products cannot be expected to maintain the same degree of color consistency as paints, and any eventual standard established for mortar color change would necessarily be less stringent as well.

 

The change in Chromaticity (da, db) of the 701R patching compound tested over 800 hours of QUV exposure in this study is less than 1.0 (See Figure 4). The dL data, the change in Lightness/Darkness, provides nearly all of the overall color change, dE, in this case.

Figure 6

This is graphically illustrated by the Reflectance vs. Wavelength Plot in Figure 6. The curves indicate the percentage of reflected light across the visible light spectrum for both the Control and a weathered specimen. Overall, the curves have nearly the same shape, with a difference on the order of two to three percent across the entire spectrum. The upper curve, the weathered specimen, is reflecting more light across the entire spectrum than the Control, accounting for its relative lightness. This provides an important clue as to why and how the material changed, slightly, over the course of the exposure.

 

 

HOW AND WHY DOES COLOR CHANGE?

 

While it is informative to describe the extent and nature of color change, it is even more useful if we can expand upon this information to gain an understanding of how and why composite patching systems change in color over time.

 

The potential sources of color change in cementitious repair systems include fading of colorants, chalking of the matrix, yellowing associated with the aging of Portland cement, surface soiling and development of efflorescence.

 

It is commonly assumed that when patches become lighter in color, the pigments are “fading”. This is unlikely, however, when using high quality iron oxide pigments meeting the requirements of ASTM C9799,10. The nature of the measured color change as illustrated in Figure 5 further discounts the assumption of pigment fading. Had the saturation of the pigments in the patching compound been reduced by fading, we would expect to see corresponding changes in the Chromaticity data (da, db). In particular, “fading” patches would be expected to become “greyer” or lower in saturation, rather than lighter or higher in Value, due to the increased influence of the grey Type I Portland cement upon which they are based.

 

Chalking of the patching compound matrix is related to mechanical breakdown of the cementitious binder, which relates to surface erosion.

The influence of a light-colored aggregate could also predominate in a much later stage of weathering, when there is pronounced erosion of the cementitious binder. There was no indication, however, of any such erosion at this stage of exposure.

 

Color changes associated with the aging of Portland cement have been reported to trend towards a more yellow hue11, a development that is not evident in this case. Neither of the two 800-hour weathered test panels showed a significant positive db, which would have indicated increased levels of yellow. It should be noted that pigments used with Portland cement tend to mask yellowing associated with cement aging11.

 

While surface soiling can have a dramatic effect on naturally exposed surface colors, the enclosed QUV system offers little or no opportunity for soiling to develop. Whereas the system produces wetness by condensation of vapor on test panel surfaces, the water deposited on the panels is free of contaminants that may build up as surface soiling. Soiling can therefore be dismissed as a factor contributing to color change, in this case.

 

This leaves efflorescence as a final potential cause of color change. For the purpose of illustration, let us assume that instead of describing the change in color of the patching material over time and exposure, the data in Figures 4 - 6 had been describing an attempt to color-match a brownstone patch. The formula for that brownstone patch might incorporate a number of different color pigments, possibly including white, carbon black, red iron oxide, yellow iron oxide and perhaps some others. In the course of adjusting the color of that patch, if we were to obtain readings indicating the same color differences shown in Figures 4 - 6, color differences which are almost entirely in the dL (Lightness/Darkness) portion of the data, we would recognize that the need for color adjustment was in the level of the white pigment. Since in this case dL >0, indicating that the color is lighter, or “whiter” than the original brownstone standard, the data would be indicating that a little too much white pigment has been added.

 

Of course, there could have been no actual addition of white pigment to the patching compound over time in the laboratory accelerated weathering exposure tests. An alternative source must therefore be sought for the elevated whiteness level that has developed over the 800 hours of accelerated weathering.

 

Measurement of dL variations over time have previously been used to describe corresponding changes in levels of efflorescence.11 Efflorescence is known to be a common cause of short to medium term color change in cementitious compositions12 and is an obvious potential source of increased whiteness over a particular range of times and exposures. Understanding this phenomenon is essential to understanding color and color change in composite repair compounds.

 

Efflorescence is the deposition of soluble salts on masonry surfaces as moisture evaporates. The salts can originate from multiple sources in the building envelope, including the masonry itself, masonry mortar, back-up materials and ingredients in the patching compounds. To some extent, all materials containing Portland cement or lime incorporate constituents capable of being transported to the surface and deposited as efflorescence. While aesthetically objectionable, efflorescence is generally not considered harmful in and of itself, but it can be an indication of continuing moisture infiltration problems that may be damaging if left unresolved.

A common assumption is that if patches discolor or efflorescence develops on patch surfaces, that this represents a problem with the patching system. As we have seen in Figure 3, sometimes that is the case, as some patch formulations demonstrate strong tendencies to “whiten” under both QUV and natural exposures. The presumption of patch instability is not valid, however.

 

Figure 7: Effloresced Replacement Stone

 

Figure 7 shows a restored building section exhibiting some moderate to heavy white efflorescence. In this case, the efflorescence cannot be a sign of patching compound instability or sensitivity, because the depicted section is not a patch, it is new replacement limestone. The source of the salts in this case is the backup masonry, which has remained saturated for many years and has built up a high level of dissolved salts within a wet zone behind the surface. As the building dried out following restoration, the water migrated to the surface and evaporated, leaving behind the salts.

 

The mechanisms by which efflorescence forms are related to the way in which masonry assemblies which are wet go about drying. The extent to which efflorescence develops is highly variable, and will depend on such factors as the concentration of available soluble salts within the masonry, the volume and pattern of water infiltration and flow through the building envelope, and the length of time over which the process has been working. Efflorescence may be as light and unobtrusive as to be unnoticeable, may appear as a slight surface haze, or may be distinctly white and sufficiently heavy to measurably alter surface profile. The point is that many things in the building envelope can be sources of efflorescence, and repairs or coatings can alter the way in which a masonry assembly dries. In some cases, this will induce the formation of efflorescence.     

 


Studies of efflorescence and color in concrete products have suggested that most of the color change associated with efflorescence will diminish with time and weather exposure11. Some level of permanent lightening has been reported, however, and available data suggests that long term lightening on the order of dL = 3 or more is to be expected9, 11. Efflorescence is more noticeable over darker, more intense color shades. Experience with composite patching of brownstone confirms this observation, as even very slight deposits of efflorescence can create very noticeable changes in color.

 

Efflorescence and its tendency to form have also been associated with the density and porosity of cementitious materials. Generally, the lower the density and the higher the porosity, the more freely liquid water can move through a material and the greater will be the tendency for efflorescence to form11. Composite repair mortars for soft sandstones are intentionally designed for relatively high permeability in order to be compatible with the porous host stone. Unfortunately, this also makes them ideal for the passage of salt-laden water, which can lead to the deposition of efflorescence.

Change in Lightness, dL, on the order of 3 to 4, as recorded in the 800-hour QUV test for compound 701R, is unlikely to be objectionable. It is conceivable, however, that more pronounced efflorescence (higher dL) may develop in any particular case. If heavy efflorescence does occur, efforts should be made to determine whether the causes are continuing moisture infiltration, repair mortar instability, or a simple one-time event related to drying of the structure. Once the causes are understood and addressed, if required, any remaining aesthetic issues can be resolved by such means as removal, staining, or simply allowing nature to weather it away in due course.

 

Correlating Laboratory Studies with Natural Exposure Data


Laboratory methods for evaluating the various performance characteristics of repair materials are only valuable if they correlate reasonably well with the actual performance of the same materials under natural exposure conditions. Accelerated weathering tests do not provide quantitative predictions of the rates at which materials will change over time and exposure. Rather, their objective is to evaluate relative resistance to weathering, and good correlations are those that accurately predict the relative ranking of the formulations or materials being tested and their resistance to change under natural weather exposures. The mode of change to be expected, such as yellowing, chalking, swelling or flaking are also important observations. A reality check for the methods used in the laboratory’s accelerated weathering study of color change involved the evaluation of naturally weathered twelve-year-old brownstone patches on a building in Hartford, Connecticut.

 

In 1989, Christ Church Cathedral, a 19th century brownstone church listed in the National Register of Historic Places, was patched using composite repair compounds. In 2001, additional repairs were undertaken, affording an opportunity to examine the twelve-year-old work at close quarters. Patch fragments were removed to the color lab, and the spectrophotometer and color software were used to compare the color of the twelve-year-old patch to a Control chip of the same color, which had been deposited in the lab’s color library in 1989. Results are listed in Figure 8.

 

Figure 8: Color Difference Data, three sample points, 12-year-old naturally weathered brownstone patches. All data are averages of three measurements.

 

Specimen

 

da

 

db

 

dL

 

dE

 

2001-1

 

1.49

 

1.32

 

1.05

 

2.25

 

2001-2

 

1.50

 

2.37

 

2.74

 

3.92

 

2001-3

 

1.10

 

1.08

 

3.14

 

3.50

 

Overall color change, dE, was of similar magnitude to the color changes recorded in the laboratory study, typically on the order of 3 to 4. Chromaticity differences were slightly higher in the naturally exposed material than in the 800-hour QUV weathered materials. The difference in dL, Lightness/Darkness, was again a significant element of overall color difference, and as in the accelerated weathering study, colors generally became slightly lighter with time and exposure. The naturally weathered patches in this case were also slightly more yellow and redder than the control.

 

Figure 9

 


Visually, the most conspicuous difference between the laboratory-weathered specimens and the naturally weathered specimens was in the amount of mica evident in the naturally exposed twelve-year old mortar. Mica flakes were incorporated in the brownstone patch mixtures to replicate the mica found within the natural stone. Over the course of twelve years of natural exposure, slight surface erosion occurred, exposing more of the mica flakes. This may have influenced Chromaticity measurements of the weathered material, potentially obscuring actual color change in the composite patch matrix.

This observation relates to the importance of considering the correlation between natural and laboratory exposure times. Reference has been made to the common practice of employing empirically derived rules of thumb for correlating accelerated and natural weathering exposures5. The 800-hour QUV exposure would correspond with a significantly shorter period of natural exposure in Southern New England, according to the guidelines derived over time by the Edison Coatings laboratory. Twelve years of natural weathering represents a more stringent test, amounting to perhaps two or three times the length of equivalent exposure.

Additional research is required to establish more precise correlations between natural and accelerated weather exposures, and the extent of the color changes to be expected in composite repair mortars over time. Ideally, the work would include greater numbers of samples spanning a wider range of colors, and involving more frequent color measurements over the course of both natural and accelerated weathering. Generally, however, the extent of the overall color changes measured and the changes in lightness (dE, dL on the order of 3-4) in the naturally weathered materials corresponded sufficiently well with the laboratory weathered materials to confirm the merit of the research concept and direction. 


KEEPING SIGHT OF THE LARGER CONTEXT

No technical study, focused as it must be on the details of a component of a problem, can hope to resolve all of the greater issues of which it is a part. The work of measuring and optimizing color and color retention has linkages to much broader issues and implications. While the work of a materials development laboratory centers on formulation issues over which it has control, the limitations of what can be achieved through formula optimization work must also be kept in perspective.

 

In particular, proper analysis and resolution of the underlying problems that caused the damages being repaired is critical. Moisture infiltration is the most common cause of damage to masonry. Control of moisture infiltration is also critical to controlling efflorescence12. Controlling efflorescence is critical to controlling color.


The work also relates to the broader issue of composite masonry repair material selection. There are no consensus standards addressing the various properties of these materials or the priorities or process by which they should be selected. Color and Color Retention must be kept in perspective as but one aspect of what repair materials must achieve. Every material selection inherently involves a series of prioritized compromises13, as “perfect” repair materials do not exist. Engineering basics, including the importance of such properties as Tensile Bond Strength, Drying Shrinkage, and Modulus of Elasticity must be given due attention and priority in masonry repair14. Material selection should be based on the best available overall balance of performance and aesthetic properties, as required to meet the specific needs of each project.

 

TOWARD OPTIMIZATION OF COLOR AND COLOR RETENTION

The optimization of composite patch color and color retention requires the integration of design and workmanship elements that also contribute to the goal of long-term color compatibility. On a basic level, there are available techniques for improving long-term aesthetics for composite repairs.

 

The use of multiple colors of patching compound, for example, makes it less evident to the casual observer that slight color differences in the repairs represent anything other than the natural variations typically found in building stone. The deliberate selection of colors that are initially slightly darker than the host substrate can counterbalance any tendency of the material to become slightly lighter with age. The close matching of texture as well as color should not be overlooked as a critical component of perceived color. The use of compatible accessory products such as translucent breathable stains to simulate aging, and consolidants to stabilize substrates that are deficient in strength, and water repellent treatments for substrates with poor weather resistance can also profoundly improve repair color and long-term color retention.

 

Conclusion


The overall quality of composite repairs has improved greatly over the past two decades through greater development of the knowledge base, better training of repair mechanics and advances in repair materials technology. While some composite repair systems have proven their ability to withstand the test of time for at least a decade or two, aesthetic expectations will and should continue to rise. This provides the impetus for researching and developing further improvements.

 

There is a great deal more work to be done in the area of understanding, controlling and optimizing composite patch color and long-term color retention. The research is only at the beginning, because acceptance of the notion that some composite patches are capable of enduring over the long-term is a relatively recent development.

 

The development of objective, acceptable color tolerances is an important long-term goal. If conflict over subjective aesthetic judgments is to be avoided, consensus guidelines must be developed. Eventual tolerance standards for composite repair color and long-term color retention will have to balance what is technically feasible, commercially available, economically affordable and aesthetically desirable. Precise measurement and communication of color and color change in composite repair mortars are important starting points. They begin to define what is being achieved now, and that is the first step along the path toward a higher future ideal.

The work will be multi-faceted and constantly evolving, rather than narrowly focused with clearly defined beginning and endpoints. Other work in progress involves measuring how patching compound color responds to moisture. There is opportunity for further aesthetic improvement if patch color can be controlled to match substrate color not only when dry, but also when wet.

 

 

Michael P. Edison, Chemical Engineer, is President and Founder of Edison Coatings, Inc., of Plainville, CT. The Company develops and manufactures customized repair mortars, repointing mortars, coatings and adhesives used in stone, masonry and concrete restoration work. Edison has been actively involved in chemical process engineering, product development and technical service work for nearly 30 years.


 

Endnotes:

 

1. Drs. Jay and Maureen Neitz, Department of Cell Biology, Neurobiology & Anatomy, Department of Ophthalmology, Medical College of Wisconsin, Color Vision Basics, June 29, 1999,  http://www.mcw.edu/cellbio/colorvision/index.html

 

2. Hunter Associates Laboratory, Reston, VA; Corporate publication, undated.

 

3. Munsell Color Communication Products, www.munsell.com

 

4. Colorimetric Fundamentals, CIE 1976 L*a*b* (CIELAB), July 29, 2001, Datacolor International, www.datacolor.com/cie.htm

 

5. "Correlation Questions and Answers- A Discussion of the most frequently asked questions about accelerated weathering, Douglas M. Grossman, Q-Panel Co., January, 1984.

 

6. Standard Practice for Operating Light- and Water-Exposure Apparatus (Fluorescent UV-Condensation Type) for Exposure of Nonmetallic Materials, ASTM G53 - 88, American Society for Testing and Materials

 

7. “A Guide to Pigmenting Concrete”, p. 10, Bayer Corporation, Industrial Chemicals Division, Publication #52-B707(5)C, undated.

 

8. Ibid, p. 8.

 

9. Bayferrox Synthetic Iron Oxide Pigments, Specification Requirements for Coloring Concrete Products, Bayer Corporation, Industrial Chemicals Division, 09/97

 

10. ASTM C979, “Standard Specification for Pigments for Integrally Colored Concrete”, American Society for Testing and Materials

 

11. “A Guide to Pigmenting Concrete”, p. 12, Bayer Corporation, Industrial Chemicals Division, Publication #52-B707(5)C

 

12. ASTM C270-96a, “Standard Specification for Mortar for Unit Masonry”, Appendix X2, Efflorescence, American Society for Testing and Materials

 

13. Guideline #03733, “Guide for Selecting and Specifying Concrete Repair Materials”, International Concrete Repair Institute

 

14. “Aesthetic Repair Materials: Properties and Selection”, Michael P. Edison, Concrete Repair Bulletin, January/February 1998, International Concrete Repair Institute


Captions:

 

Figure 1: Comparison of Composite Patch Color Retention on various elements of two restored 19th Century brownstone churches in Middletown, CT. Stone on the First Church of Christ (left) was repaired with three standard colors of a commercial pre-packaged repair mortar in 1996. Repair work on St. John ‘s Church (right) was completed in 1999, using a field-mixed repair mortar. Photographs were taken in 2001, five years and two years after the respective restorations.

 

Figure 2: The Munsell and CIE L*a*b* systems for cataloguing color

 

Figure 3: Two panels of the same brownstone patching compound, Custom System 45 #701R; Upper panel is an unweathered control, lower panel is shown after 800 hours’ accelerated weathering exposure. The section of panel exposed to accelerated weathering shows slight change.

 

Figure 4: Color and Color Difference Data for Various Controls and Weathered Panels of Custom System 45 #701R Brownstone Patching Compound

 

Figure 5: CIE Color Difference Plot, Weathered and Unweathered Test Panels, Brownstone repair mortar #701R.

 

Figure 6: Wavelength vs. Percentage Reflectance Plot, Unweathered Control/Standard and Weathered Brownstone Repair Mortar #701R (800 hours).

 

Figure 7: Efflorescence formation is evident on a recently replaced limestone element.

 

Figure 8: Color Difference Data, three sample points, 12-year-old naturally weathered brownstone patches. All data are averages of three measurements.

 

Figure 9. Color Difference Plot for naturally exposed twelve-year-old brownstone patch vs. a twelve-year-old control chip of the same color