Ensuring Durability with Stone Restoration Techniques— Critical Decisions for Common Repairs Matthew Farmer, PE Wiss, Janney, Elstner Associates | Tampa, FL mfarmer@wje.com IIBEC International Convention & Trade Show | SeptembeBEr 15-20, 2021 Farmer | 177 Matthew Farmer, PE Wiss, Janney, Elstner Associates | Tampa, FL Matthew Farmer, PE, is a principal investigator on evaluations of buildings and monuments, concentrating his practice in the areas of engineering, design, investigation, analysis, and repair of masonry building enclosure systems. His experience includes institutional and commercial projects, as well as work on numerous historic landmarks. Farmer received a bachelor of science in architectural engineering and a bachelor of environmental design from the University of Colorado, and a master of civil engineering from Cornell University. He is a registered professional engineer in the District of Columbia, Maryland, and Virginia, and an active member of ASTM C18 (Dimension Stone), ASTM C27 (Cast Stone), and the Masonry Society. 178 | Farmer II BEC International Convention & Trade Show | September 15-20, 2021 ABSTRACT SPEAKER Architects, engineers, and preservationists have been specifying restoration repairs of stone masonry and cladding for decades. Execution of these repairs has fallen to craftspeople with a range of stone masonry experience, who often receive inadequate or no expert guidance and instead rely on general industry practices. Although most professionals will agree on many of the basic design principles for commonly applied repairs, each project has unique parameters that can affect the final repair design and its ultimate service life. This presentation will be of interest to design professionals, as well as building owners, managers, and engineers tasked with maintaining buildings constructed of stone masonry or stone cladding. It will summarize best practices for designing and implementing repairs for stone wall systems, and will identify the key decision points that influence the type of materials and repair procedures that we use to optimize durability. The presenter will also challenge some commonly held beliefs regarding popular material choices and repair techniques, and highlight how, in some instances, minor changes in approach can make significant differences in repair performance. This presentation will be pitched at an intermediate level; some prior knowledge of exterior wall systems and stone masonry is recommended. Natural stone is perhaps the oldest and one of the most durable materials used in our built environment. While the stone type can range in hardness and durability, stone conveys a sense of permanence—literally and figuratively—that is not replicated by any other material. Throughout history, stone has been used to construct civilization’s most prized and important structures in an effort to stand the test of time. Whether to build structures that symbolize our democracy, house our governments, pay tribute to our leaders, allow us to worship, or perform as critical infrastructure, stone is sought after as the material of choice (Fig. 1 and 2). If stone is such a durable material, why then should we have to consider repairs at all? First, the inherent durability of stone extends the maintenance cycle, but does not eliminate it. The time between repair interventions can be substantially greater than that of other building materials (decades rather than years), but repairs will eventually be necessary. Often it is not the stone itself that is to blame for deterioration, but the manner in which it is used or attached to the structure that initiates the need for repair. Sometimes it is circumstances completely beyond our anticipation, such as a natural disaster or in-service damage from vandalism, that require us to repair even the most durable of our building materials. Second, all stone types are not created equal for the environment in which we place them; for example, a sedimentary stone (such as brownstone) or softer metamorphic stone (limestone) will not stand up as well to harsher weather as an igneous stone, such as granite (Fig. 3), but may perform extremely well in a milder climate. Alternatively, limestone may perform well at a higher elevation of a structure, but not at grade, where increased moisture and deicing chemicals are present (Fig. 4). Using a “weaker,” less durable stone for some applications does not automatically lead to repairs, provided the stone used is well suited, well detailed, and of good quality for the particular application. Stone masonry mass walls can serve both as a primary structure and as a primary enclosure material. Repair of stone can be essential to maintain established, as-designed load paths; excessive deterioration can compromise structural integrity and risk partial or full collapse. II ii B E C International Convention & Trade Show | September 15-20, 2021 Farmer | 179 Ensuring Durability with Stone Restoration Techniques— Critical Decisions for Common Repairs Figure 1. The Washington Monument, Washington, DC. Construction began in 1848. Figure 2. The Cathedral of Mary Our Queen, Baltimore, Maryland. Construction began in 1954. Figure 3. Granite is used as the base course of many limestone buildings. Figure 4. Limestone at grade is prone to freeze/thaw-related deterioration and cryptoflorescence. While it may not be a structural concern, deteriorated or unstable stone can be a safety hazard if present above an area where pedestrians or vehicles pass by (Fig. 5). As an enclosure material, stone masonry mass wall assemblies often serve as the primary barrier between our interior and exterior environment. The walls prevent infiltration of environmental moisture and excessive air movement, plus they serve as thermal mass that buffers the interior environment against rapid heat loss or gain. Stone repair is necessary to maintain weather barrier integrity and prevent reduction in the quality of our interior environments (such as excessive moisture, humidity, or air infiltration). If left unrepaired, stone deterioration can also accelerate as more of the previously protected and concealed stone surfaces are exposed to greater amounts of moisture. TYPES OF STONE DISTRESS The subject of this paper is the critical decisions that are necessary to design durable stone repairs; therefore, it is important to first understand some of the common forms of stone damage that can compromise its performance, accelerate its deterioration, or lead to more significant problems such as water leakage or structural failure. The more common forms of stone damage that are repaired generally fit into the following broad categories. Cracking/Spalling A spall is a location where a piece of stone has broken off from the parent material, leaving a void. An incipient spall is a piece of stone that has cracked and largely separated from the parent material but remains in place and at risk of complete separation. Physical damage to the stone compromises its ability to carry load or facilitates excessive moisture or air infiltration (Fig. 6). Cracking and spalling are related in that cracks, depending on their geometry and proximity to stone unit joints or edges, can create a spall (Fig. 7). Delamination A delamination is the separation of a stone surface, typically along bedding planes, seams, or rifts inherent with the geologic formation of the stone. The binders are weakest along these planes; exposure to thermal cycling, moisture infiltration, and other environmental effects causes the outer layers of stone to progressively disengage from the inner layers (Fig. 8). In rare instances, delaminations can result from nonenvironmental causes, such as joint material incompatibility or loading. 180 | Farmer II BEC International Convention & Trade Show | September 15-20, 2021 Stone repair is necessary to maintain weather barrier integrity and prevent reduction in the quality of our interior environments (such as excessive moisture, humidity, or air infiltration). Figure 5. Large spall failure at a limestone cornice. Figure 8. Failure of “face bedded” stone facade units along the bedding planes. Figure 7. Incipient spalls located along a stone joint. Figure 6. Cracks that will allow water to enter the stone. Displacement Displacement is movement of a stone unit or portion thereof along a joint or crack boundary from its original position (Fig. 9 and 10). If the displacements are excessive, the load path can be compromised or the stone may be unstable. Displacements can be caused by sudden changes in load, corrosion of embedded ferrous metals, or impacts. Joint Material Deterioration Joint materials (typically mortar or sealant) serve several purposes in a stone assembly, including limiting air and water infiltration. Mortar is also used to transfer and evenly distribute load between stone units as well as bond the stones together to prevent them from moving. Joint material deterioration can lead to excessive water infiltration, stress concentrations as joint material between stone units breaks down, changes in load path, or stone instability (Fig. 11 and 12). Strength Loss Some marble varieties (and, to a lesser degree, other calcitic stones) lose strength over time. This is a result of hysteresis, a deterioration process whose primary symptoms are cladding panel bowing and disaggregation of the stone surface, commonly referred to as “sugaring” (Fig. 13). Thermal hysteresis is the permanent distortion of stone masonry units from differential expansion of one stone surface relative to another. The crystalline structure of the stone disassociates when heated and then does not return to its original position when cooled, resulting in asymmetric volume change and reduced strength. If the strength loss is excessive in mass masonry, stone damage, such as erosion cracking, anchorage failure, and instability, can result. Erosion/Scaling Largely an aesthetic concern, erosion results in the loss of detail and material at the surface of stone, most often from constant exposure to water (Fig. 14) or airborne abrasives such as sand. II ii B E C International Convention & Trade Show | September 15-20, 2021 Farmer | 181 Figure 13. Weakening of a marble cornice due to hysteresis. Figure 14. Surface erosion and scaling of granite. Figure 9. Displaced stone coping unit. Figure 10. Displaced stone facade unit. Figure 11. Failed joint mortar has led to water infiltration and efflorescence. Figure 12. Failed sealant allows water to reach stone anchors and ultimately the building interior. PARAMETERS FOR EVALUATING REPAIR TECHNIQUES First and foremost, stone repairs should be performed if the structural integrity of the stone assembly is compromised or a risk to the public. If the stone assembly is unstable or load paths have changed, it is imperative that the condition be addressed quickly to avoid further instability or collapse. Temporary repairs, such as shoring or limiting access to portions of the structure, may be warranted until more-durable repairs can be designed and implemented. Any long-term repair strategy should include a sound method for re-establishing the load-carrying capacity of the stone assembly, whether it be the original load path or a modified path that meets the structural requirements. Perhaps equally important to structural integrity when selecting a repair approach is to confirm that it will slow deterioration of the parent stone. Otherwise, what is the purpose of performing the repair at all? The purpose of the repair should be not only to replace or re-establish the original building fabric, but also to prevent (or at least substantially reduce) further deterioration of the fabric itself. For example, repairing a stone fragment caused by corrosion of embedded anchors without addressing the anchors will not stop the deterioration process, nor prevent failure of the repair and the surrounding stone (Fig. 15). One must also consider the location and continuity of the existing load paths. If the damaged stone is loaded in bearing, the load path is likely already redirected away from the damaged portion and into the sound stone that remains. Adequacy of the surrounding stone should be verified for the applied load. If stone is damaged at a lateral anchor, it is recommend that the connection be abandoned and a new connection installed to avoid loading the repair, which is usually not capable of accommodating load. A repair should be evaluated for its own durability and any effect on durability of the parent stone assembly. A repair that is not durable or that requires frequent maintenance may be worse than no repair at all, particularly if access to the repair location is difficult or expensive. A less durable repair may be inexpensive in the near term but can become costly once reduced service life and increased maintenance are factored in. Funding for subsequent repairs and maintenance are also often not known; therefore, a less durable repair may need to remain in place past its useful service life, risking repair failure and increased damage to the parent stone. The most durable repair appropriate for a given application should always be considered when evaluating repair options. When evaluating repair approaches, one must also consider the potential risk of damaging the stone at or adjacent to the repair site as part of the process (Fig. 16). If the risk is too high or the repair too invasive, then perhaps the repair is not in the best interest of the structure. An alternative approach that may not directly repair the damage but that minimizes the potential for additional damage or slow deterioration may be a better option in the long run. Whenever a decision is made about whether to perform a stone repair, its importance to the building fabric and its historical sensitivity should be evaluated. If the element is not consequential to the performance of the structure or historically important, and it presents an ongoing maintenance concern, then removal or redesign of the element might be considered as a repair option. However, if the condition is critical to protect other elements or is historically sensitive, then clearly restoration and perhaps preventive maintenance would be more appropriate. Cornices and water table features are examples of stone elements whose presence may be critical not only from a historical perspective, but also as a protecting element for the stone assemblies below (Fig. 17). Two of the most common drivers of stone repair selection are cost and aesthetics. Unfortunately, cost is a reality that any entity responsible for the care of a structure must contend with; funds are never unlimited. When evaluating repairs, it is critical to consider all the costs associated with them. These include the initial cost; required maintenance; life-cycle replacement (service life); access; and the relative cost of doing nothing, which will result in future (perhaps more extensive or invasive) repairs. It is also important to consider the overall cost of not only the repair itself, but also the access to perform the repair. If access to repair areas is difficult or costly, then it may be more cost-effective overall to perform a more durable repair that may carry a higher unit cost but will require less maintenance and replacement over the building service life. The aesthetics of a repair not only can significantly affect the overall appearance of a structure, but can also influence the motivation for the repairs in the first place. One of the most common misconceptions in stone repair is that the repair will look as good as or better than the original stone assembly, or that it will not be noticeable. This is rarely the case. Often the best aesthetic repair is to do nothing, as the “cure can be worse than the disease.” When trying to market for repair funds, the aesthetics are often overpromised, and expectations that something old will look completely new again 182 | Farmer IIiiBEC International Convention & Trade Show | September 15-20, 2021 Figure 15. Spall caused by corrosion of a stone anchor. Figure 16. Minor chipping that would require a complex dutchman profile to repair. Figure 17. A limestone cornice is vulnerable to weathering, but partially shields the limestone facade below. are unmet. This disappointment can lead to reduced demand and subsequently reduced funds to support the repairs, which results in greater levels of deterioration due to deferred maintenance. Ironically, a successful repair program is often one where the repairs and modifications are not apparent, especially for historic structures (as further discussed in the following section). So it is critical to keep expectations regarding repair aesthetics realistic by communicating throughout the repair design process and encouraging early repair mock-ups. TREATMENT OF HISTORIC STRUCTURES In the United States, many stone structures are historic in nature and have been granted some level of landmark status that is accompanied by specific requirements for restoration and preservation. Work on landmark structures is largely influenced by local authorities having jurisdiction and the Secretary of the Interior’s Standards for the Treatment of Historic Properties or similar local jurisdiction ordinances. While these standards go into great detail regarding the classification of treatment, limits on intervention, extent of replacement, and integration of new construction into existing historic assemblies, one can distill several simple tenets that are generally applicable to stone repair decisions and include the following: • Retain as much original building fabric as possible. • Design repairs to be reversible whenever possible. • Minimize the level of intervention to only what is necessary to achieve the repair objectives. • Differentiate any new construction from the original structure fabric, contrasting rather than blending the two types of assemblies. • Use the gentlest repair or maintenance methods possible that achieve the minimally acceptable result. Avoid the temptation to overreact to the observed distress or damage. COMMON REPAIRS AND CRITICAL DECISION POINTS To address the types of distress identified previously, there are several common repair types that are used for stone masonry mass wall systems and older or historic structures. Many of these repairs also can be applied to more modern stone assemblies, such as stone cladding systems, though that is not the focus of this document. Each of these repairs seems relatively straightforward, but requires several critical decisions to be made to confirm the repairs are appropriate for the application and to optimize their implementation. No Repair One “repair” option that should always be considered is to do nothing. This is the benchmark by which all other repair options should be evaluated. Doing nothing has the advantage of no risk of damage to the remaining structural fabric, no initial cost, no maintenance, and no replacement cost (Fig. 18). For small-scale, isolated damage that does not draw attention, there is no reason to execute a repair. If structural stability is a concern, if a lack of repair will accelerate deterioration of the remaining structure fabric, or if the appearance of the damage is unacceptable (Fig. 19), then doing nothing may not be a wise option. Dutchman A dutchman is a piece of salvaged or new stone that is carefully cut and fitted precisely into a carved opening within an existing stone unit to replace a damaged section. It is important to understand that by the nature of excavating the damaged portion of the stone, the load path no longer travels through the damaged section and will not load the dutchman. Before proceeding with dutchman repairs, it is important to assess the potential sources of stone for use as the dutchman. New stone of the same variety as is being repaired will frequently appear significantly different from the original stone due to variations in weathering, quarrying, and material source (Fig. 20). Often, newly quarried stone of varieties with more color variation tends to blend better with the original stone. It is also possible to improve II ii B E C International Convention & Trade Show | September 15-20, 2021 Farmer | 183 Figure 18. Small spall on a stone tread; no consequential damage is likely if left unrepaired. Figure 20. New limestone used as a dutchman. Figure 19. Small incipient spall near joints that will contribute to water infiltration if left unrepaired. the blending of new stone with the old by artificially weathering the surface to better match the texture of the original weathered stone (Fig. 21). One of the best options for dutchman stock is salvaged stone from the structure; salvaged stone comes from the same source, has typically been in service for the same period of time, and has experienced similar weathering patterns. Often it is better to remove stone for use as a dutchman from locations on the structure that are less visible (for example, backs of parapets, roof structures, and interior sides of columns) to use at more prominently located repair locations. The areas where the salvaged stone is removed can be repaired with newer and perhaps less-well-matched stone. It is important to determine whether the repair appearance is important enough to warrant the increased expense of removing and the salvaged stone and repairing the source location. Dutchman repairs rely on initial bond of the joint material and shear friction facilitated by completely filling the perimeter joints. To maximize shear friction, the excavation made to fit the dutchman should be rectilinear and have sharp edges perpendicular to the stone surfaces. If the sides are sloped inward, the perimeter shear friction is reduced, so stability relies on the bond of the joining material. To address the potential variabilities with installation that can affect stability, it is recommended to include mechanical engagement of the dutchman to the substrate. Using dowels or postinstalled anchors across the dutchman/substrate bond interface reduces reliance solely on the joint material as a binder and joint filler. The size of a dutchman repair is determined primarily by the size of the stone area that is damaged, but practical limits to the size of the dutchman must also be evaluated. A dutchman excavation that is significantly larger than the damaged stone area is economically wasteful and removes more of the original stone fabric than is necessary. Alternatively, a dutchman excavation that is too small may not allow enough room for the tools necessary to precisely trim the original damaged stone to receive the new dutchman. The dutchman itself must also be large enough to accommodate mechanical anchors to secure it into place. Sometimes postinstalling anchors is a better option than increasing the size of the dutchman sufficiently to accommodate concealed anchors (Fig. 22). When determining the size of the dutchman and how much of the original stone to excavate, one should also consider the joints and how best to conceal the seams between the dutchman and the original stone. It is preferable for the dutchman edges to coincide with existing joints, thereby reducing the visible edges of the dutchman. It is also critical to ensure any existing joints are maintained through a repair area and are not interrupted by the dutchman. Cracking can occur in the dutchman if the existing joints are not allowed to continue in their original configuration, which may require two dutchmen to be installed instead of a single dutchman (Fig. 23). There are options and preferences regarding dutchman perimeter joint treatment. Traditionally, stone epoxy is used to bond the dutchman to the substrate. Using epoxy enables the joints to be narrow at the surface, helping to conceal the repair. However, the joint must be wide enough to facilitate full contact of the epoxy with the substrate and the dutchman. If the joints are too narrow, the adhesive may not spread, causing poor bond and voids that can retain water. If the joints are too wide, shrinkage during cure can reduce bond. Selecting the correct epoxy for the application is also critical, as some products are sensitive to ultraviolet (UV) light, which can affect the surface color and integrity. Stability of the dutchman relies on the epoxy for adhesion and also to facilitate shear friction with the substrate. Over time, epoxies undergo shrinkage and lose adhesion, but a joint properly filled with epoxy will maintain shear friction between the dutchman and the substrate. Some practitioners choose to use mortar at the dutchman perimeter joints instead of epoxy. Setting the dutchman with mortar like a masonry unit requires a more visible joint that must be packed from the exterior, but enables maintenance in the future by repointing. The stability of a dutchman set with mortar relies almost entirely on shear friction because the initial bond is minimal and it is subject to early drying shrinkage. Fragment Reattachment If a stone unit cracks and the stone fragment is recoverable, often the least invasive repair is to reattach it. It eliminates the need to match the stone, and often the fracture plane is sufficiently tight that the repair is hardly visible. Reattaching a fragment is an option if the condition that caused the damage can be removed, the fragment is large enough to be removed and reattached intact, and the fragment surface is relatively undamaged along the fracture plane. Otherwise, a dutchman may be the better choice. Because the fracture plane of a stone fragment is often oriented toward the outer surface 184 | Farmer IIiiBEC International Convention & Trade Show | September 15-20, 2021 Figure 21. Salvaged granite used as a dutchman, then microabraded to simulate weathering. Figure 23. A dutchman repair that straddles an existing joint. Figure 22. Schematic design for dutchman repair. of the stone or a joint (Fig. 24), shear friction cannot be relied upon to hold the fragment in place, and stone adhesive will lose bond over time, mechanical anchorage is critical to long-term repair durability. Larger fragments can be removed and reinstalled with anchors positioned across the crack plane. If the fragment is too oddly shaped or too small to preinstall anchors, the fragment can be first bonded into position and then small-diameter through-face anchors used to provide redundant attachment back to the parent stone unit (Fig. 25). If the fragment cracked away cleanly and remained intact, then a thin layer of stone adhesive is all that is needed at the stone mating surfaces. The “crack” at the stone surface will often not require any additional repairs for concealment, but if so, then it can be routed and pointed with a repair mortar that matches the stone. Crack Grouting/Injection Not all cracks require repair. There are really only two reasons to repair cracks: structural instability or excessive water penetration into or through the stone masonry. If neither of these conditions is of concern, then the risk of damaging the original stone during repair, the increased maintenance, or the diminished appearance may not be worth the marginal reduction in water permeance achieved by repairing the crack. One way to think of a crack is as another joint in the stone assembly, and to determine what, if any, impact an additional few feet of joint will have on the performance of the stone masonry structure. The primary task when evaluating cracks is to determine their cause. Cracks can be considered either structural or nonstructural and each type needs to be evaluated differently. If cracks result from a failure of a load path, compromise the load-carrying capacity of the stone masonry, or enable fragments or whole stone units to become disengaged from the structure, they should be considered structural in nature and should command an immediate repair response. Most often the best repair approach is to grout the cracks solid. By grouting the separation, the ability to transfer compression through bearing and shear loads through aggregate interlock is re-established (Fig. 26 and 27). It is always better to first consider re-establishing an existing load path than to attempt to rationalize an alternative load path because the as-built material characteristics of older stone masonry assemblies cannot always be well defined (Fig. 28 and 29). Using II ii B E C International Convention & Trade Show | September 15-20, 2021 Farmer | 185 Figure 24. Reattached stone fragment. Figure 28. A critically damaged load-carrying stone element. Figure 29. Grout has re-established load transfer across a previously open crack. Figure 26. Stone crack prior to grout injection. Figure 27. Stone crack after grout injection. Figure 25. Optional anchorage placement for reattaching a stone fragment. cementitious and resin grouts with a high modulus of elasticity and compressive strength in excess of the stone masonry assembly being repaired is not recommended because the stiffer grout can concentrate localized stresses in the substrate stone. Unless the applied loads are increased and the original assembly was structurally inadequate, supplemental reinforcement is not necessary to increase system capacity and can be detrimental by increasing stiffness, concentrating load, and inducing localized cracking. Gravity-feed and injection grouts are highly susceptible to mixing errors and our experience suggests that their actual physical properties often do not match the published values for strength, air content, and bond strength. If grouting is being relied upon for structural reasons or is being used in a challenging environment, then mock-ups and field sampling for quality assurance testing are strongly encouraged to ensure the selected materials are appropriate for use. Also challenging with grout installations are monitoring and verification of flow. Depending on the complexity of the masonry system, the crack width, and the grout viscosity, the grout may not reach the desired depth in the assemblies or may flow excessively and wind up where it is not desired (Fig. 30 and 31). Monitoring the volume of the grout being used is critical and establishing volume expectations through trial repairs or mock-ups is essential to confirm volume estimates. Cracks in stone masonry that are caused by restraint, minor cyclic movements, or a lack of accommodation of volume change, but that have no significant impact on the load path or masonry stability, are considered nonstructural. Nonstructural cracks need not be repaired unless they are thought to be contributing to excessive water penetration. Crack Pointing/Sealing If the appearance of a crack (whether structural or nonstructural) at the surface of the stone is objectionable, the crack surface can be routed and filled with a patching mortar colored to match the stone. Joint mortar should be avoided because it rarely is close to the original stone color and may not bond well to the routed edges. The width of the routed crack surface can be as narrow as feasible to minimize its appearance but still be wide enough to enable compaction of the patching mortar, preferably in lifts, to enhance mortar-stone bond (Fig. 32). If increased resistance to water penetration at the crack surface is desired, elastomeric sealant and an appropriate backer can be used instead of patching mortar (Fig. 33). Patching mortar is usually less noticeable and more durable, but will typically crack along the bondline with the stone substrate due to shrinkage and differential movements caused by thermal changes. Sealant is more forgiving of substrate movement but has a more pronounced appearance and is often less durable due to loss of adhesion with the substrate over time. Where cracks are expected to move significantly, sealant should be used to accommodate the crack width changes. Joint Repointing Repointing is the process of removing deteriorated mortar at the surface of masonry joints and replacing it with compacted lifts of new mortar. Similar to cracks, there are two main reasons to repoint stone masonry joints: loss of stone unit stability or excessive water penetration through the stone assembly. Repointing of stone masonry joints carries great risk of damage to the substrate stone from tools used by the masons to remove the existing mortar. The principal purpose of joint mortar is to enable stone masonry units to be set plumb and level. Because the mating 186 | Farmer IIiiBEC International Convention & Trade Show | September 15-20, 2021 Figure 30. Craftsmen gravity feeding cementitious grout into a cracked joint. Figure 31. Craftsman injecting resin grout into a crack. Figure 33. Routed crack filled with sealant. Figure 32. Routed crack pointed with repair mortar. surfaces of two stacked stone units will never be completely smooth, surface irregularities must be accommodated and the voids filled to reduce air and water penetration. Stone masonry joints are often narrower (1/8 in.) than other unit masonry systems, which can make removal of mortar without damaging the stone extremely difficult. Stone is also often set with a higher-compressive-strength mortar, then pointed with a lower-strength mortar. An overly strong pointing mortar can create stress concentrations at the stone face and lead to damage or mortar failure (Fig. 34). The pointing mortar should also be designed with an appropriate aggregate gradation for the joint widths to facilitate compression and bond with minimal shrinkage. Material property testing such as that specified in ASTM C8561 and ASTM C13422 is recommended at multiple locations to help estimate the compressive strength, as well as identify the binders and other constituents of the original mortar. When mortar joints are exposed on horizontal surfaces, they are more susceptible to water penetration and deterioration. Often, sealant or lead joint fillers can be used at skyward-facing joints to reduce water entry into the joints and improve durability (Fig. 35). It should be noted that sealant requires periodic replacement to remain watertight and that lead joint fillers are not completely watertight, so avoiding the exposed mortar joints by flashing over them can sometimes be the best solution if the design can tolerate the visual changes. Patching One of the most cost-effective repairs available for stone masonry is patching with repair mortar. Patching is the process of applying a mortar with a proprietary blend of aggregates and binders into an excavated repair area. The patch material is then formed to the original surface profile, where it cures and hardens. Key to performance of the patches is the compatibility between the patching mortar and the stone substrate with respect to absorption and compressive strength, the ability to achieve and maintain bond in spite of inherent shrinkage, and substrate preparation to promote bond and shear friction for stability. There is general consensus in the industry that patches should be of a minimum thickness, have rectilinear perimeters, and have sides that are perpendicular to the stone surfaces for developing shear friction (Fig. 36). Because performance of patches is tied closely to substrate preparation, the minimum size of patches should provide enough space to achieve the preferred excavation parameters. However, if the patches are too large, material shrinkage can have a significant impact on bond, cracking, and ultimately stability of the patch. To reduce the risk of patch bond failure over time, adding mechanical anchorage across the bond interface is strongly encouraged so that if the bond fails, the patch can remain secured to the substrate until it can be replaced. Avoiding patches over areas where failure would pose a high risk or life safety concern is also recommended (Fig. 37). Although patching can be cost-effective when compared with other repair methods, even the best patches may not meet owner expectations. Patches can change in appearance due to UV light exposure and variations in moisture content, experience plastic drying shrinkage (cracking), debond from the substrate, and ultimately require replacement sooner than other repair options. II ii B E C International Convention & Trade Show | September 15-20, 2021 Farmer | 187 Figure 34. Failed joint repointing and widened joint. Figure 35. Sealant installed at a skyward-facing joint. Note that the mortar is installed in the joint below. Figure 37. Poorly installed and placed patch. Figure 36. Properly installed stone patches. CONCLUSION There are a number of critical decisions regarding appropriate selection and design of stone repair options that may not be readily apparent to design practitioners or craftsmen who execute stone repairs. All too often, stone repairs are performed without appropriate context, resulting in increased life-cycle cost for stone masonry assemblies. Carefully considering various stone repair options and the project-specific conditions affecting their implementation will result in improved repair service life, lower risk of future damage, and reduced life-cycle cost. REFERENCES 1. ASTM Subcommittee C09.65, Standard Practice for Petrographic Examination of Hardened Concrete, ASTM C856. West Conshohocken, PA: ASTM International, 2020. 2. ASTM Subcommittee C12.02, Standard Test Method for Examination and Analysis of Hardened Masonry Mortar, ASTM C1324. West Conshohocken, PA: ASTM International, 2020. 3. ASTM Subcommittee C18.06, Standard Guide for Selection, Design, and Installation of Dimension Stone Attachment Systems, ASTM C1242. West Conshohocken, PA: ASTM International, 2021. 4. ASTM Subcommittee C18.07, Standard Guide for Assessment and Maintenance of Exterior Dimension Stone Masonry Walls and Facades, ASTM C1496. West Conshohocken, PA: ASTM International, 2018. 5. ASTM Subcommittee C18.07, Standard Guide for Cleaning of Exterior Dimension Stone, Vertical and Horizontal Surfaces, New or Existing, ASTM C1515. West Conshohocken, PA: ASTM International, 2020. 6. ASTM Subcommittee C18.07, Standard Guide for Repair and Restoration of Dimension Stone, ASTM C1722. West Conshohocken, PA: ASTM International, 2018. 7. Alexander Newman, Structural Renovation of Buildings: Methods, Details, and Design Examples. New York, NY: McGraw-Hill, 2001. 188 | Farmer IIiiBEC International Convention & Trade Show | September 15-20, 2021
