Long-Term Performance of Epoxy Adhesive Anchor Systems (2013)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

3 Introduction The objectives of this research project were to: • Investigate the influence of various parameters (e.g., type of adhesive, installation conditions, and in-service conditions) on the sustained-load performance of adhesive anchors and • Develop recommended test methods, material specifications, design guidelines, design specifications, quality assurance guidelines, and construction specifications for AASHTO for the use of adhesive anchors in transportation structures. This chapter begins with a brief overview of the background of the behavior and design of adhesive anchors in concrete. This chapter concludes with a review of the literature on adhe- sive anchors and is organized as follows: • Parameters influencing bond strength, • Test methods and material specifications related to adhesive anchor systems, • Design guidelines and design specifications related to adhesive anchor systems, and • Quality assurance guidelines and construction specifica- tions related to adhesive anchor systems. Background on Behavior/ Design of Anchors While various design standards and design methodology will be discussed in detail later, a general review of the cur- rent behavior/design for anchoring to concrete is provided for background. This document adopts the definition of adhesive as found in ACI 355.4-11, which is as follows: Adhesive – Any adhesive comprised of chemical components that cure when blended together. Adhesives are formulated from organic polymers, or a combination of organic polymers and inorganic materials. Organic polymers used in adhesives can include, but are not limited to, epoxies, polyurethanes, polyesters, methyl methacryl­ ates and vinyl esters. Behavioral Model The behavioral model and resulting design procedures for adhesive anchors contained in most standards have been under development for the past 20 years. Detailed information on single adhesive anchor behavior is presented in Cook et al. (1998). Information on group and edge effects is presented in Eligehausen et al. (2006a). The following presents a general overview of the behavior/design model for single adhesive anchors. Figure 1 shows typical failure modes exhibited by bonded anchors. Figure 2 shows the mechanism for load transfer in bonded anchors. In the elastic range, adhesive anchors have been shown in Cook et al. (1993) to exhibit a hyperbolic tangent stress distri- bution along the bonded anchor as shown in Figure 3. Research by McVay et al. (1996) used an elasto-plastic Sandler-DiMaggio constitutive model to show how the bond stress is distributed along the length of anchor under various stress levels (Figure 4 through Figure 7). Figure 4 through Figure 7 have been modified from their original in that the percent stress level has been identified for each curve. At low load levels, the stress distribution generally follows the elastic hyperbolic tangent stress distribution in which the adhesive close to the surface is higher stressed than the adhesive deeper in the hole. As the load level is increased above approximately 30% of the peak stress, the upper portions of the adhesive become plastic and redistribute the load further into the hole. As the load is further increased, deeper and deeper portions of the adhesive become plastic. As the stress level reaches approximately 70% of the peak stress, the stress dis- tribution approaches a relatively uniform bond stress distri- bution along the entire length of the anchor. Any additional increase in load causes the adhesive to dilate providing for an increased capacity until failure. C H A P T E R 1 Background

4concrete breakout failure adhesive/concrete interface steel/adhesive interface adhesive/concrete and steel/adhesive interface Source: Cook et al. (1998) Figure 1. Potential embedment failure modes of bonded anchors. Figure 3. Hyperbolic tangent stress distribution (d = anchor diameter, tmax = maximum bond stress). Figure 2. Mechanism of load transfer of a bonded anchor. Figure 6. Stress distribution along length of adhesive anchor for hef /do = 6.67. Bond Stress (MPa) N or m al iz ed D ep th Source: McVay et al. (1996) Figure 5. Stress distribution along length of adhesive anchor for hef /do = 5.33. Bond Stress (MPa) N or m al iz ed D ep th Source: McVay et al. (1996) Figure 7. Stress distribution along length of adhesive anchor for hef /do = 8.00. Bond Stress (MPa) N or m al iz ed D ep th Source: McVay et al. (1996) Figure 4. Stress distribution along length of adhesive anchor for hef /do = 4.00 (do = hole diameter). Bond Stress (MPa) N or m al iz ed D ep th Source: McVay et al. (1996) For adhesive-bonded anchors where the hole diameter does not exceed 1.5 times the anchor diameter and with an embedment depth to anchor diameter ratio not exceeding 20, the uniform bond stress model shown in Figure 8 and given by Equation 1 (Eq. 1) has been shown to be a valid behavioral model both experimentally and numerically [Cook et al. (1998)]. In Eq. 1, the mean failure load (N−t) is a function of the product’s mean bond strength (–t) multiplied by the bond area

5 calculated at the anchor diameter (d). As noted in Cook et al. (1998), test samples in a worldwide database indicated that the hole size is less than 1.5 times the anchor diameter for adhesive anchor applications. Anchors in holes larger than 1.5 times the anchor diameter typically use cementitious or polymer grout. For these typical adhesive anchor applica- tions with hole sizes less than 1.5 times the anchor diameter, it is not practical to establish two separate interface bond strengths as shown in Figure 2 and, in fact, test data shows that the uniform bond stress model works quite well if the bond stress is determined from a series of product qualifica- tion tests by simply dividing the failure load by the bonded area calculated at the diameter of the anchor. Details of this are provided in Cook et al. (1998). .N dh Eq 1ef= τpiτ where N−t = mean failure load, lb, –t = mean bond strength, psi, d = anchor diameter, in., and hef = embedment depth, in. For design, the nominal bond strength of adhesive-bonded anchors is dependent on the mean bond strength of anchors installed in accordance with manufacturer’s printed installa- tion instructions (MPII), adjusted for scatter of the product’s test results, and for the product’s sensitivity to installation and in-service conditions. As discussed in Cook and Konz (2001) the bond strength of properly installed bonded anchor products varies considerably. Based on tests of 20 adhesive anchor products, the mean bond strength at the adhesive/ anchor interface for individual products ranged from 330 psi to 2,830 psi (2.3 MPa to 19.5 MPa). Short-Term Sensitivity The short-term load sensitivity of an adhesive to a specific variable can be determined from two series of short-term tests. A series of five baseline tests are conducted to determine the adhesive’s short-term strength under standard conditions (N−baseline). Another series of five tests are conducted with a specific variable introduced (N−variable). The alpha-reduction factor (a) is determined by dividing the average load of the variable test by the average load of the baseline test. This is illustrated in Figure 9. Eq. 2 provides the basic design relationship using load and resistance factor design (LRFD) for a single adhesive anchor. As shown by Eq. 2, the factored tension load (Nu) would need to be less than the design strength determined as a capacity reduction factor (f) multiplied by the nominal bond capacity. ≤ φ .N N Eq 2u bond where Nu = factored tension load, lb, f = capacity reduction factor, Nbond = t′ p d hef, t′ = nominal bond stress, psi, d = anchor diameter, in., and hef = embedment depth, in. The nominal bond strength (t′) is the 5% lower fractile of the mean bond strength (tk) adjusted by a series of reduc- tion factors (a) for installation and in-service conditions as shown in Eq. 3. ′τ = τ α α α .1 2 3 Eq 3k where tk = 5% lower fractile of mean bond strength and a1, a2, a3 = reduction factors determined from comparing the bond strength under different installation and in-service conditions to the baseline bond strength Figure 8. Uniform bond stress model for adhesive anchors. Figure 9. Calculation of reduction factor (a).

6• Mixing Effort: how well are the constituent parts mixed prior to installation. • Adhesive Curing Time When First Loaded: 24 hours, 7 days, 28 days, or longer. • Bond Line Thickness: how much space is there between the anchor and the sides of the hole. • Fiber Content of Adhesive: type and proportion of fillers in the adhesive. • Chemical Resistance: alkalinity, sulfur dioxide, and other compounds. Installation Factors: • Hole Orientation: downward, horizontal, upward. • Hole Drilling: rotary hammer, core drill, or drilled in accordance with manufacturer’s instructions. • Hole Cleaning: uncleaned, partially cleaned, or cleaned in accordance with the manufacturer’s instructions. • Moisture in Installation: dry, damp, submerged, or installed in holes with moisture limitation conditions in accordance with manufacturer’s instructions. • Installation Temperature: concrete below freezing, adhe- sive below freezing, or preheated. • Depth of Hole (Embedment Depth): the depth of the anchor can affect not only the bond strength but the type of failure. • Anchor Diameter: anchor diameter can affect bond strength. • Type of Concrete: Portland cement only, Portland cement with blast furnace slag, fly ash, or other additives. • Concrete Strength: low compressive strength, high com- pressive strength. • Type of Coarse Aggregate: mineralogy, absorption, and hardness (affects hole roughness). • Cracked or Uncracked Concrete: the presence of cracks can reduce the bond strength significantly. • Concrete Age: installed and/or loaded at early age. In-Service Factors Elevated Temperature. According to Messler (2004), “the greatest shortcoming of many structural adhesives is their limited tolerance of elevated temperature.” However, adhe- sives with open-ring structures (polyimidazoles and substi- tuted imidazoles) that close under high temperatures become stronger. He further adds that it is important to measure an adhesive’s resistance to creep under sustained loading condi- tions especially if exposed to high temperature. According to Adams and Wake (1984), an adhesive anchor system with sustained loads at a temperature 18°F (10°C) above its heat deflection temperature will exhibit significant creep. Experimental tests by CALTRANS in Dusel and Mir (1991) confirm this and explain that the adhesive will “soften The 5% lower fractile, or characteristic value, (tk) is deter- mined from Eq. 4: )(τ = τ −1 .Kv Eq 4k where –t = mean bond stress, psi; K = tolerance factor corresponding to a 5% probability of non-exceedence with a 90% confidence using ACI 355.4. Note, other definitions of “characteristic value” exist. For example, ASTM D7729 uses an 80% confi- dence interval; and n = coefficient of variation. Sustained-Load Sensitivity The single anchor design model is provided for refer- ence. Recommendations on how to incorporate the effects of sustained-load performance under various installation and in-service conditions are addressed in this project. For parameters that are shown to have a more aggravated effect under sustained load than under short-term load, a reduc- tion factor (a) would be dependent on stress level and dura- tion of load. This relationship is determined from the “stress versus time-to-failure” test series discussed later. Parameters Influencing Bond Strength As noted in Cook et al. (1994), Cook et al. (1996), and Cook and Konz (2001) there are many variables that affect the per- formance of adhesive anchors. Below is a list of many of the common factors with brief comments. A more in-depth dis- cussion of each follows. Most of the items in the list are incor- porated into ICC-ES AC308 (2008), ACI 355.4 (2011b), and EOTA ETAG 001 Part 5 (2002) discussed later in this chapter. In-Service Factors: • Elevated Temperature: temperature variations during the life of the structure, and effects of sustained elevated temperature. • Reduced Temperature: brittleness associated with reduced temperature. • Moisture-in-Service: adhesive anchor subjected to dry, damp, or immersed conditions during the life of the anchor. • Freeze–Thaw: magnitude and frequency of freeze–thaw cycles. Factors Related to the Adhesive: • Type of Adhesive: for example: epoxy-mercaptan, epoxy- amine, vinylester, polyester, or hybrid.

7 based adhesives have higher bond strengths than ester-based adhesives. ASTM C881/C881M classifies seven types of epoxy-resin bonding systems, specifying Type IV as those that are for use in load-bearing applications for bonding hardened concrete to other materials, but Type IV is not specifically identified for epoxies used in adhesive anchor systems. Fourier transform infrared spectroscopy (FTIR) is a test method to chemically characterize an adhesive as shown in the National Transportation Safety Board (NTSB) (2007b) report on the adhesives from the Boston Tunnel collapse. The results of an FTIR test can be used to investigate correlations in the chemical make-up of an adhesive and its bond strength. ACI 355.4 includes several fingerprinting tests (discussed later) to identify the material and compare it against the man- ufacturer’s standard. Mixing Effort. Bond strength is dependent on the proper composition of the adhesive. Adhesive anchor systems come in components that need to be mixed thoroughly and to the proper proportion prior to installation. Some systems are designed to guarantee proper proportions and thorough mixing, and some are solely dependent on the installer. Com- mon systems include: • Glass and Foil Capsule Systems, which contain spe- cific amounts of polymer resin, accelerator, and a min- eral aggregate. The capsules are placed in the hole and an anchor (with a chiseled end) is set with a hammer drill that bores through the capsule, thereby mixing the adhesive. See Figure 10 for a typical capsule anchor system. • Injection Systems typically include plastic tubes of resin and hardener. The components are commonly mixed in a special nozzle as they are dispensed. The adhesive is injected into the hole and the anchor is installed after- wards. The anchor is usually rotated slowly during instal- lation to prevent the formation of air bubbles which cause voids in the adhesive. See Figure 11 for a typical injection anchor system. • Other Systems include pouches that contain the compo- nents, which are mixed manually and then dispensed into the hole. It is also possible to purchase the components separately and mix them manually. and become rubbery” above its glass transition temperature (comparable to heat deflection temperature) and its bond strength will decrease. Reduced Temperature. Reduced in-service temperatures can make adhesives more brittle as mentioned in Cognard (2005). Currently ICC-ES AC308 (2008) has a reduced tem- perature test only during installation. The commentary for ACI 355.4 (2011b) mentions that reduced temperature dur- ing installation increases viscosity and retards the cure time of adhesives. Moisture-in-Service. While it has been widely known that the presence of moisture during the installation of the adhe- sive affects bond strength, a recent study [Chin et al. (2007)] indicates that the presence of moisture after curing can also affect the creep resistance of an anchor. Chin et al. (2007) of the National Institute of Standards and Technology (NIST) conducted thermo-viscoelastic analysis on ambient cure epoxy adhesives used in construction. This research showed that the presence of absorbed moisture after curing can create the same creep type behavior commonly seen in high tem- perature conditions. Cognard (2005) mentions that water can degrade adhe- sives in three ways: (1) Penetrate into the adhesive and soften it, (2) Penetrate between the adhesive and the substrate thereby destroying the adhesion, and (3) Penetrate into porous substrates causing swelling and det- rimental movements. Additionally, Cognard (2005) recommends that water resistance tests be performed if the adhesive will be subject to moisture during the life of the product. Freeze–Thaw. The expansion and contraction of materials due to temperature changes and the expansion of water when it freezes tend to be detrimental to structural systems. Factors Related to the Adhesive Type of Adhesive. According to Cook and Konz (2001) adhesives can vary significantly between chemical groups and even within chemical groups. For example, on average, epoxy- Source: Cook et al. (1998) Figure 10. Typical capsule anchor system.

8adhesive and the anchor and/or the concrete and a smaller bond area reduces bond strength. Section 1.6 of the Florida Department of Transportation’s (FDOT) (2009) Structures Design Guidelines and section 937 of the FDOT (2007) Standard Specifications for Road and Bridge Construction prohibit adhesive anchors to be installed in over- head or upwardly-inclined holes for the above mentioned reason. The New York and Pennsylvania departments of trans- portation have similar restrictions. From the Boston Tunnel collapse investigation report, NTSB (2007a), departments of transportation are prohibited from using adhesive anchors in sustained tensile-load overhead highway applications until the development of testing and protocols to ensure safety. Due to the sensitivity of horizontal or vertically upward installed anchors to improper installation, ACI 355.4 requires that products be specifically approved for use in these condi- tions and be installed by certified personnel. The ACI-Concrete Reinforcing Steel Institute (CRSI) Adhesive Anchor Installation Certification Program entails both a written and performance evaluation that includes installation in vertically upward holes. Hole Drilling. The two common methods of hole drilling involve diamond core drill bits, which produce a very smooth sided hole or carbide-tipped hammer-drill bits, which pro- duce a rough sided hole. Since one of the ways the adhesive bonds with the concrete is by mechanical interlock, it was thought that a rough sided hole should provide better bond. This research project showed that for the three adhesives tested, an anchor installed in a core drilled hole had an aver- age short-term strength of 74% that of an anchor installed in a rotary impact hammer drilled hole. Hole Cleaning. According to Cook and Konz (2001) the cleanliness of the hole has a significant impact on bond strength, as dust created during the drilling operation can interfere with the adhesive/concrete bond surface. Tests were performed in which some holes were cleaned with com- pressed air and a non-metallic brush. In holes that were not cleaned, the average bond stress was 71% that of the cleaned holes (with a range from approximately 20% to 150%) and had an average coefficient of variation of 20%. The type of brush is also significant. Section 416 of the FDOT (2007) Standard Specifications for Road and Bridge Whatever system is used, it is important that the components are mixed thoroughly and to the proper proportions. Manufac- turers typically recommend mixing until a certain consistency and color is reached. The adhesive must completely fill voids between the anchor and the sides of the holes as any voids will reduce the effective area and subsequently the bond stress. Adhesive Curing Time When First Loaded. According to Cook and Konz (2001), the duration of adhesive cur- ing affects bond strength. Adhesives were tested at 24 hours and 7 days of cure time. Most anchors showed a decrease in bond strength over a shorter adhesive cure time; the average bond strength for a 24 hour cure was 88% of those with a seven day cure. Bond Line Thickness. According to Çolak (2007), the smaller the dimension between the anchor and the side of the hole, the lower the potential for creep. Çolak (2007) con- ducted tests on anchors with a ratio of the hole diameter to the anchor diameter (do/d) range of 1.2 to 1.8. In these tests, it was noticed that creep resistance was increased when the bond line thickness of the adhesive was decreased. This relationship is supported by Section 2.3.7 of ACI 503.5R-92 (1997). However, according to analytical studies by Krishnamurthy (1996), anchors with a much larger ratio of the hole diameter to the anchor diameter (do/d) range of 1.2 to 4.1, the bond line thickness does not significantly affect the capacity of the anchor. Therefore, current data is not conclusive. Fiber Content of Adhesive. Section 2.3.7 of ACI 503.5R-92 (1997) and Çolak (2001) mention that creep resistance can be increased by increasing the fiber content of the adhesive. Chemical Resistance. Cognard (2005) confirms that chemicals, oils, greases, and other compounds can penetrate the adhesive and degrade the adhesion with the anchor or the concrete causing a bond failure. Installation Factors Hole Orientation. The orientation of the hole has the potential to significantly affect the performance of adhesive anchors. Vertical or upwardly inclined holes prove difficult to fill with adhesive, as the adhesive will tend to run out of the hole. The subsequent voids reduce the bond area between the Source: Cook et al. (1998) Figure 11. Typical injection anchor system.

9 Type of Concrete. The concrete mix design can affect the bond strength of the adhesive anchor. This includes but is not limited to the type of cement, mix proportions, and the types of additives (air entrainment, plasticizers, fly ash, blast furnace slag). Tests conducted at the University of Florida by Anderson (1999) showed a reduction in bond stress in anchors installed in concrete with fly ash and blast furnace slag as compared to anchors installed in regular concrete without additives. Concrete Strength. According to Cook and Konz (2001) there was no consistent correlation between bond strength and concrete strength among the adhesives tested (specimens A–T). As concrete strength was increased, some adhesives showed an increase in bond strength, and others displayed a local maximum or minimum at midrange strengths (Fig- ure 12). This reveals that no broad rules can be applied, but must be determined for each adhesive. In the extreme cases, as the concrete strength was increased 100%, the largest increase in bond strength was 120% and the largest decrease was 35%. Type of Coarse Aggregate. Cook and Konz (2001) deter- mined through lab testing in concrete specimens with lime- stone and river gravel that the type of coarse aggregate plays a factor in bond strength. Based on tests conducted by Caldwell (2001), the mineralogy of the aggregate also affects the bond strength. Of all the samples tested, concretes that used calcium-rich aggregates such as lime- stone failed at the lowest anchor loads. Additionally, concretes that used aggregates with high silicon content failed at relatively higher loads, although the findings were not conclusive. Cook and Jain (2005) conducted tests on adhesive anchors in concrete with different coarse aggregate types. It was observed that adhesive anchors installed in concrete with harder coarse aggregates produced higher bond strengths. It Construction requires cleaning with a non-metallic brush, as metallic brushes tend to polish the sides of the holes, thereby reducing the ability of the adhesive to create a mechanical interlock with the sides of the hole. Moisture in Installation. According to Cook and Konz (2001) the dampness of the hole significantly affects bond strength in two ways. It can restrict the entrance of adhesive into the pores of the concrete thereby reducing mechanical interlock, and moisture can interfere with the chemical reac- tion between the hardener and the resin. It was demonstrated that anchors installed in damp holes (wet surface) produced an average bond strength for 20 prod- ucts of 77% (with a range of approximately 20% to 150%) compared to a dry installation. Anchors installed in wet holes (standing water) produced an average bond strength of 43% (with a range of approximately 10% to 160%) compared to the dry installation. Installation Temperature. For anchors installed at low temperatures, the final degree of hardening is smaller com- pared to installation at normal temperature. This might result in a reduction of the sustained-load bond strength. Depth of Hole (Embedment Depth). Increasing the depth of the hole does have a slight impact on bond strength up to a point. According to tests by Krishnamurthy (1996), the load increases proportionally up to a limit of hef/d of 25 and then drops due to the bond stress not redistributing uniformly at depths over 25hef. Anchor Diameter. For most bonded anchor systems the bond strength measured in short-term tests decreases some- what with increasing anchor diameter according to Eligehausen et al. (2006b). In general, it is assumed that bond strength is independent of the anchor diameter if within the manufac- turer’s recommendations for hole diameter. (a) linear trend (b) local maximum trend (c) local minimum trendSource: Cook and Konz (2001) Figure 12. Various relationships of bond strength as a function of concrete strength.

10 fied effects. According to Messler (2004), the combination of several climatic factors (heat, moisture, temperature cycling, moisture cycling, ultraviolet radiation, oxidation) can be par- ticularly severe. Adhesive anchors historically have not been tested for moisture and temperature combinations. ASTM D1151-00 provides a standard for testing adhesives under different tem- perature and humidity exposures. Test Methods and Material Specifications Related to Adhesive Anchor Systems The review of test methods and material specifications related to adhesive anchors included national standards, state DOT standards, and international standards. Other test methods are also presented. National Test Methods and Material Specifications ASTM E488 Standard Test Methods for Strength of Anchors in Concrete and Masonry Elements ASTM E488 provides the fundamental test procedures to determine the static, seismic, fatigue, and shock, tensile, and shear strengths of concrete and masonry anchors. These procedures serve as the basic building blocks for anchor testing and are either adopted in full or slightly modified by governing agencies. In all tests, the anchors are installed and conditioned at standard temperature [73°F (23°C)] and 50% relative humidity. The various tests methods contained within this test standard are briefly described below. Static Tests. This standard discusses a series of tension and shear tests on five anchors for each variation of anchor size, type, embedment depth, and location. The tension test sub- jects an anchor to a tensile load and the shear test subjects the anchor to a shear load. In both tests the load is applied at a continuous load rate that will produce failure in 2 ± 1 minute. Load and displacement readings are monitored. The tension test can either have a confined or an unconfined test setup. The confined test setup isolates the failure to the adhesive bond surface in order to determine the bond strength. The unconfined test setup allows for bond failure with a shallow concrete cone or complete concrete breakout failure. Seismic Tests. This standard discusses a series of seismic tests on five anchors for each variation of anchor size and type. Procedures are specified for both a seismic tension and shear test. In both tests the load is applied in cycles according to a specified program that simulates a seismic event. Load, dis- placement, and acceleration readings are monitored. The seis- mic shear test can either be conducted with a direct-loading or was concluded that the harder aggregates created rougher sur- faces when the holes were drilled for the anchor. The rougher surface (as mentioned earlier) provided for more mechanical interlock and thus an increase in the bond strength. Cracked or Uncracked Concrete. Based on research by Eligehausen and Balough (1995) cracked concrete can have a significant impact on adhesive bond strength. The research- ers state that anchors in concrete, or the holes in the con- crete created for adhesive anchors, will attract or even induce cracks at the anchor/hole location. Cracks in the concrete at an anchor will then tend to break down the bond between the concrete and the adhesive. Based on the research findings of Eligehausen and Balough (1995) and Fuchs et al. (1995), bond strengths in cracked concrete can vary from 33% to 70% of the bond strength in uncracked concrete. Similarly, Meszaros (1999) estimates from his research that bond strengths in cracked concrete are approximately 50% of the bond strength on uncracked concrete. See Figure 13 for a crack in a typical adhesive anchor application. Concrete Age. Following casting, the concrete can remain damp for several days while it hydrates. ICC-ES AC308 requires that the anchors be installed in concrete after 21 days of curing. Part of the study for this project will be to determine if adhesive anchors installed in early-age con- crete will have lower short-term bond strengths than those installed in concrete beyond 21 days. If lower in strength, this may be due to a synergistic effect of the very low con- crete strengths and the high moisture content present in early-age concrete. Synergistic Effects. The above-mentioned factors are typically considered independently; however, their combinations can have ampli- Figure 13. Typical crack location of bonded anchor.

11 Radiation Test. This test evaluates the radiation resistance of an adhesive anchor system. The anchors are exposed to a minimum gamma radiation level of 2 × 107 rads. Static tension tests (confined or unconfined) are conducted and the irradi- ated samples are compared to baseline (confined or uncon- fined) samples. Tests on Effect of Freezing and Thawing Conditions. This test evaluates the freeze–thaw resistance of an adhesive anchor system. A minimum of three confined or unconfined tests are conducted. Freeze-resistant concrete is used and the surface of the concrete is covered with ½” of water for a minimum of 3” around the anchor. A constant tension load is applied equal to 40% of the ultimate capacity. Fifty com- plete freeze–thaw cycles are conducted by lowering the tem- perature to −10°F (−23°C), holding for 3 hours, then raising to 104°F (40°C) and holding for 3 hours. Static tension tests are conducted following the fifty freeze–thaw cycles and the residual strength is compared to the baseline strength. Test on Effects of Damp Environment. This test evaluates the sensitivity of an anchor system installed in damp or water- filled holes. A minimum of three confined or unconfined tests are conducted. Prior to anchor installation the holes are filled with tap water and kept full for seven days. The freestanding water is removed immediately before anchor installation. Fol- lowing the required curing time, static tension tests are con- ducted to failure and the results compared to the baseline test. This test can also be conducted on water-filled holes in which the freestanding water is not removed prior to anchor installation. Test on Effect of Elevated Temperature on Cured Samples. This test determines an adhesive anchor’s sensitivity to elevated temperature under short-term loads. A minimum of three confined or unconfined tests are conducted per temperature. Tests are conducted at 70°F (21°C) and at a minimum of four higher temperatures, one of which is at least 180°F (82°C). The anchors are installed and cured at 75°F ± 10°F (24°C ± 5°C) and following the cure time the specimens are heated to their test temperature. Following 24 hours at the stabilized test temperature, the specimens are removed and static tension tests are conducted. The static strengths for each test are normalized by the 70°F (21°C) test strength and presented in a chart showing the trend of normalized strength versus temperature. See Figure 14 for a sample bond strength versus temperature chart for three hypothetical adhesives. Test on Effect of Reduced Temperature on Curing. This test determines an adhesive anchor’s sensitivity to curing at reduced temperature. A minimum of three confined or unconfined tests are conducted. The test member and anchor rod are conditioned at the test temperature for 24 hours prior to installation. The anchor is then installed and cured and once curing is completed, a static tension test is conducted and the result compared to a baseline test at 70°F (21°C). an indirect-loading procedure. The indirect-loading procedure attaches a weight to the structural member via the anchor and shakes the structural member thereby applying a seismic force to the anchor. At the end of the seismic shear tests, a static shear test is conducted to determine its residual strength. Fatigue Tests. This standard discusses a series of tension and shear fatigue tests using any of the previously demon- strated test setups. In both tests the load is applied according to a fatigue program that specifies the loading method, load levels, frequency, and number of cycles. A static tension test is conducted at the conclusion of the fatigue loading program to determine the residual strength and failure mode. Shock Tests. This standard discusses a series of tension and shear shock tests to determine either (1) if an anchor system will withstand a certain shock load or (2) the maximum shock load an anchor system can withstand without failure. The shock load is applied in a ramp loading rate over a 30 ms dura- tion per shock. A static tension test is conducted at the conclu- sion of the shock test to determine the residual strength. ASTM E1512 Standard Test Methods for Testing Bond Performance of Bonded Anchors ASTM E1512 builds upon the test program established in ASTM E488 and while ASTM E488 is for all concrete anchor systems, ASTM E1512 is solely for bonded anchors. As with ASTM E488, ASTM E1512 is adopted by many governing agencies for the testing and evaluation of adhesive anchor systems. ASTM E1512 requires that static, fatigue, and seis- mic tests be conducted per the procedures set forth in ASTM E488, and specifies additional environmental test procedures. The requirements for the environmental tests are as follows: • Concrete of the same mix design in all series with the com- pressive strength between 2,500 psi and 3,500 psi at the time of testing, • Concrete cured for 28 days, • Anchors installed at 75°F ± 10°F (24°C ± 5°C), • ½”-13 UNC threaded rods embedded 4½” in concrete, and • Either confined or unconfined test, but all test series shall be the same. The following environmental tests are briefly described below. Test on Short-Term Effect of Fire. This test evaluates the performance of an anchor in a fire. This is an unconfined test on a minimum of three anchors in concrete. The slab is conditioned and the fire is applied as set forth in ASTM E119. A constant tension load is applied and temperature and dis- placement readings are recorded at 1 minute intervals until failure.

12 Concrete temperature readings are conducted during the test and if the concrete temperature falls below the mini- mum temperature for more than 24 hours, the test duration is extended to account for the total time below the minimum temperature. The test is continued for 42 days (1,000 hours). A logarithmic trendline of the displacement versus time is projected out to 600 days using a least squares fit through the data points using the equation: ∆ = ( )+a t b• ln where D = projected displacement, t = time, and a & b = constants evaluated by regression analysis. This trendline is constructed from not less than the last 20 days (minimum of 20 data points). The projected displace- ment at 600 days is compared to the displacement from the static tension test series at elevated temperature. See Figure 15 for a graphical presentation of this projection. ICC-ES AC58 Acceptance Criteria for Adhesive Anchors in Concrete and Masonry Elements ICC-ES AC58 is an acceptance criteria based on allowable stress design (ASD) developed by the ICC-ES and first approved in January 1995. The purpose of these acceptance criteria was to provide a standard method and report for manufacturers to qualify their adhesive anchor products for use in concrete and masonry elements. Beginning in 2008, ICC-ES AC58 was no longer accepted by the International Building Code for anchor- ages in concrete and was replaced by ICC-ES AC308 (2008) (discussed later) and the current version of ICC-ES AC58 (2007) only addresses anchorages in masonry elements. A brief If the adhesive anchor is to be used below 50°F (10°C) an additional test is conducted. The conditioning and installa- tion procedure is the same as described above. However, prior to removal of the specimen from the environmental cham- ber, a preload of 25% of the ultimate load is applied to the anchor. Once removed from the chamber, the specimen is heated uniformly to 75°F ± 10°F (24°C ± 5°C) over a period of 72 to 96 hours. Temperature and displacement readings are taken during this heating period. A static tension test is con- ducted to failure once the specimen has reached the desired temperature. Creep Test. A minimum of three confined or unconfined tests are conducted per creep test series. The creep test is com- prised of three separate individual tests as described below: Static Tension Test Series at 75°F ± 10°F (24°C ± 5°C). This test series conducts a static tension test in order to determine the average ultimate tension load. Static Tension Test Series at Elevated Temperature. This test series conducts a static tension test at a minimum concrete temperature of 110°F (43°C) to determine the average dis- placement at the ultimate tension load. Creep Test Series at Elevated Temperature. Upon comple- tion of the adhesive curing period, the concrete temperature is raised to a minimum temperature of 110°F ± 3°F (43°C ± 2°C) and stabilized for at least 24 hours. Next a preload of no more than 5% of the sustained creep load [40% of the ultimate tension load determined from the static tension test series at 75°F ± 10°F (24°C ± 5°C)] is applied to set the anchor and testing equipment before zeroing the test read- ings. Once the test equipment is zeroed, the remainder of the load is applied. The initial elastic displacement is recorded within the first 3 minutes of the test and subsequent displace- ment readings are taken every hour for the first 6 hours, and then daily for the remainder of the test. Source: Cook et al. (1998) Figure 14. Sample bond strength versus temperature curve for three hypothetical adhesives.

13 Test series 13 through 14 evaluated the critical and mini- mum edge distances for shear loading. Test series 15 was a combined tension and shear static test in which the direction of loading was at a 45° angle from the concrete. Suitability Requirement Tests. Fire Resistance Test (optional). This test referenced the test on short-term effect of fire found in ASTM E1512-01. The test results were used to determine loads for hourly fire ratings. Creep Test (optional). ICC-ES AC58 referred to ASTM E488 and ASTM E1512 for the general creep test procedure, with the following differences: • Anchors were installed and cured at 70°F ± 5°F (21°C ± 3°C), • The static tension test series was conducted at 70°F ± 5°F (21°C ± 3°C). • Provided an allowable temperature tolerance of ± 3°F (± 1.7°C) during the static tension test series at elevated temperature and the creep test series at elevated tempera- ture, and • The average displacement at the mean static load must have satisfied the displacement limitations presented in tables in ICC-ES AC58. discussion of ICC-ES AC58 (2005) is presented to provide a historical basis of adhesive anchor testing in concrete. Twenty-one test series were identified by ICC-ES AC58 and many were based on ASTM E488 and ASTM E1512. There were 15 service-condition tests to determine design values. Of these 15 service-condition tests, 11 were tension tests, three were shear tests, and one was an oblique tension test. There were also six suitability requirement tests to evalu- ate the adhesive system’s suitability for various conditions. It is important to note that of the 21 tests, only five were man- datory. If the anchor was not tested for the various optional tests, then it could not be qualified for that use. Service-Condition Tests. Test series 1 through 3 were static tension tests on single anchors and reference the static tension test procedure set forth in ASTM E488. These three test series were conducted at three different concrete strengths. Test series 4 through 7 evaluated the critical and minimum edge distances for tension loading. The different test series were all for single anchors and varied the concrete strength. Test series 8 through 11 evaluated the critical and mini- mum spacings for anchor groups of two and four anchors. Test series 12 was the static shear test of a single anchor and referenced the static shear test procedure set forth in ASTM E488. Figure 15. Extrapolation of sustained load displacements per ASTM E1512. Duration of load t [hours] D is pl ac em en t [m m] 2000 600 days 1000 4000 1000 Duration of load t [hours] 2 1 2 D is pl ac em en t [m m] Detail A Detail A data points used for extrapolation 600 log function extrapolation Source: Eligehausen and Silva (2008)

14 greater than 80% of the average tension load of the control specimens. Freezing and Thawing Test (optional). This test refer- enced the test on the effects of freezing and thawing condi- tions found in ASTM E1512. Seismic Test (optional). ICC-ES AC58 provided two methods for seismic testing. Seismic Method 1 referred to the Structural Engineers Association of Southern California (SEAOSC) (1997) standard method for the test procedure and acceptance criteria. Seismic Method 2 subjected five ½” diameter anchors to a simulated alternating sinusoidal load- ing cycle in both tension and shear tests. For the seismic tension tests, the maximum tension load (Ns) was 1.5 times the desired qualified tension load. The anchor was subjected to a series of sinusoidal loads of vary- ing magnitudes and frequencies as listed below: • 10 cycles at Ns, • 30 cycles at Ni = 0.625Ns, and • 100 cycles at Nm = 0.25Ns. Following the cyclic loading, a static tension test was con- ducted to determine residual capacity. The anchor was accepted if it withstood the cyclic loading, the residual capacity was at least 80% of the ultimate static tension load, and the maximum displacement satisfied the following equation: ∆ ≤ ∆N T ICC-ES AC58 Eq. 3ns s ref ult where Dns = maximum displacement during seismic test, Tref = average ultimate tension load, and Dult = displacement limitation for ultimate tension load. For the seismic shear tests, the maximum shear load (Vs) was 1.5 times the desired qualified shear load. The anchor was subjected to a series of sinusoidal loads of varying mag- nitudes and frequencies as listed below: • 10 cycles at Vs, • 30 cycles at Vi = 0.625Vs, and • 100 cycles at Vm = 0.25Vs. Following the cyclic loading, a static shear test was con- ducted to determine residual capacity. The anchor was accepted if it withstood the cyclic loading, the residual capacity was at least 80% of the ultimate static shear load, and the maximum displacement satisfied the following equation: ∆ ≤ ∆V V 2 ICC-ES AC58 Eq. 4ns s ref ult The data was projected as discussed in the creep test pro- cedure in ASTM E1512 (Figure 15). The anchor was accepted for creep if the average projected displacement at 600 days was less than (a) the average displacement at mean static load determined from static tension test series at elevated temper- ature (see Figure 16) and (b) 0.12 inches. The rationale behind the acceptance criteria for the creep test procedure for adhesive anchors in ICC-ES AC58 is described in detail in NCHRP Report 639 (NCHRP 2009) but is summarized below. The test temperature was chosen as 110°F (43°C) as it was determined that this was an acceptable peak temperature for an anchor installed in a concrete bridge located in the Califor- nia desert. The sustained load of 40% was based on a conver- sion from ASD with a factor of safety of 4 and a 1.6 multiplier for maximum anticipated sustained load. The test duration was determined from a database of tests in which tests that failed within a 120-day testing period did not pullout after 21 days. To be conservative, that duration was doubled to arrive at a 42-day testing period. The 600 day projection was chosen as it was determined that there would be approxi- mately 600 days in which an anchor could be expected to be above 110°F (43°C) over a given lifetime of 50 years. In-Service Temperature Test (required). This test refer- enced the test on the effect of service temperature found in ASTM E1512. The test results were used to establish adjust- ment factors for service loads. Dampness Test (optional). This test referenced the test on the effects of damp environment found in ASTM E1512. Control specimens, which had all the same properties as the damp specimens except they were maintained dry, were also tested. The average tension load of the damp specimens must have been at least 80% of the average tension load of the con- trol specimens. Each damp specimen result must not have varied from the average by 15% or all results must have been Lo ad Displacement 0 1 1 0creep u D is pl ac em en t 0 creep Time Lo ad Nu75 u110 Static Tension Test @ 75o F Static Tension Test @ 110o F Creep Test Series @ 40% Nu75 and 110o F u75 Nu110 600 days Displacement Figure 16. Basic pass/fail criteria per ICC-ES AC58.

15 • Permitted drilling methods. Evaluates installations in holes created with rotary hammer drill with carbide tip, core drill, and rock drill. The default drilling method is rotary hammer drill with carbide tip. • Hole orientation. Tests anchors oriented in the down, horizontal, and overhead orientation. The default orienta- tion is down. • Installation temperature. The default installation tem- perature range of the concrete is 50°F to 80°F (10°C to 27°C). Some test procedures allow installation at lower temperatures. • Embedment depth and anchor diameter. The embedment depth and anchor diameters tested are specified by the man- ufacturer and within the ranges established by ACI 355.4. • Type of anchor. Tests various materials (carbon, stainless); strengths; and geometries (threaded rod, deformed rebar, internally threaded inserts). • Environmental conditions of use. Testing conditions are dry and wet environment with a service temperature range of 32°F to 104°F (0°C to 40°C). Optional conditions are elevated temperature and freezing-thawing conditions. • Chemical exposure. Default condition is a high alkaline wet condition. The optional condition is sulfur dioxide. • Concrete condition. Either uncracked or both cracked and uncracked. • Loading. The default loading conditions are static and sus- tained loading. Seismic loading is optional. • Member thickness. Determines the minimum thickness of a member to avoid spalling on the backside. ACI 355.4 has four basic types of tests (identification tests, reference tests, reliability tests, and service-condition tests). Additional supplemental service-condition tests and assess- ment tests are also included. The testing schedule is presented in three tables divided between cracked and uncracked con- crete applications. Optional tests are identified in the tables. The tables are listed below and are included in Appendix A: • Table 3.1: Tests for adhesive anchors in uncracked concrete, • Table 3.2: Tests for adhesive anchors in cracked and uncracked concrete, and • Table 3.3: Reduced test program for adhesive anchors in cracked and uncracked concrete. The reduced testing program mentioned by ACI 355.4 Table 3.3 uses predefined ratios of the characteristic limiting bond stress for use in cracked and in uncracked concrete. The characteristic bond stress is based on the 5% fractile as dis- cussed earlier. All tests are referenced in ACI 355.4 Tables 3.1 through 3.3 by their ACI 355.4 section number. For concise- ness, the test descriptions in this report are referred to by their ACI 355.4 section number. where Dns = maximum displacement during seismic test, Vref = average ultimate shear load, and Dult = displacement limitation for ultimate shear load. Torque Tests. This test conducted five torque tests per anchor diameter. The manufacturer’s specified torque moment was applied to the adhesive anchor and the resulting prestressing force was recorded. The 95% fractile of the prestressing force must have been less than 60% of the 5% fractile of the ultimate load of the confined reference tests. ICC-ES AC308 Acceptance Criteria for Post-Installed Adhesive Anchors in Concrete Elements ICC-ES AC308 (2008) was an acceptance criteria for adhe- sive anchors in concrete elements based on ultimate strength design—LRFD—developed by the International Code Council Evaluation Service (ICC-ES). The purpose of these acceptance criteria was to provide a standard method and report for manufacturers to qualify their post-installed adhe- sive anchor products. Beginning in 2008, ICC-ES AC308 replaced the previous acceptance criteria ICC-ES AC58 for installations in concrete. ICC-ES AC308 was the source document for ACI 355.4 (2011b) Qualification of Post­Installed Adhesive Anchors in Concrete. Therefore the tests methods and specifications pre- scribed by ICC-ES AC308 are not discussed, rather a focus is made on the test procedures and specifications found in ACI 355.4. ACI 355.4 Qualification of Post-Installed Adhesive Anchors in Concrete ACI 355.4 (2011b) presents the testing and evaluation program of post-installed adhesive anchors in concrete. ICC-ES AC308 served as the basis for ACI 355.4 which was published by the American Concrete Institute (ACI) in 2011. Due to the tremendous research and development invested into ICC-ES AC308, and the consensus review pro- cess conducted by ACI, it is suggested that ACI 355.4 serve as the basis for the testing program and specifications for AASHTO. The testing program specified by ACI 355.4 evaluates the following variables and installation and use conditions: • Hole cleaning procedures. Typical manufacturer instruc- tions can include vacuuming, blowing with compressed air, and brushing. Instructions indicate the number of brushes, duration, and cycles and can vary due to mois- ture condition of the concrete at installation. The default installation condition is dry concrete.

16 Assessment Approach. Section 10.4.4 addresses the requirements on load-displacement behavior. The purpose of the procedure presented is to locate the point on the load- displacement curve that represents an uncontrolled slip under tension. This point is identified as Nadh, or the loss of adhesion. Loss of adhesion occurs when the anchor and adhesive are extracted from the hole as a unit which is dependent primarily upon the roughness of the hole and is seen as a drastic loss in stiffness on a load-displacement curve (Figure 17). The ACI 355.4 procedure to locate Nadh is as follows: • Determine a tangent stiffness at 30% of the peak static load (Nu), which is typically approximated as the secant stiff- ness from the origin to the point on the load-displacement curve at 0.30Nu; • Multiply the tangent stiffness by 2/3 and project this line until it intersects with the load-displacement curve; • Nadh is taken at the point of a sudden change in stiffness (Figure 17); • If there is not a very sudden change in stiffness, and the 2/3 secant line intersects the load-displacement curve before the peak, Nadh is taken at the intersection (Figure 18); • If there is not a very sudden change in stiffness, and the 2/3 secant line intersects the load-displacement curve after the peak, Nadh is taken at the peak (Figure 19); and • If the displacement at 0.30Nu is less than 0.002”, the origin is shifted to the point on the load-displacement curve at 0.30Nu and Nadh is taken at the 2/3-secant line and the load- displacement curve intersection (Figure 20). Source: ACI 355.4 (2011b) Figure 18. Evaluation of load at Nadh [k = tangent stiffness, d0.3 = displacement at 30% of the peak tension load (Nu)]. Source: ACI 355.4 (2011b) Figure 17. Evaluation of load at Nadh [Dlim = displacement corresponding to a loss of adhesion load (Nadh)]. Source: ACI 355.4 (2011b) Figure 19. Evaluation of load at Nadh. Most of the tests discussed (except for the identification tests and test series: §6.0, §7.7, §7.8, §7.13, §7.19, §8.8, §8.9, §8.10, §8.11, §8.13, §9.1, §9.2) have a requirement on the coefficient of variation for load and displacement which is addressed in §10.4.2 and establishes a reduction factor if the coefficient of variation from the tests exceeds a certain threshold (30% for ultimate loads in reliability tests and 20% for other tests).

17 where –tu,i = mean bond stress from reliability test series in con- crete batch or test member i, –to,i = mean bond stress from reference test series in con- crete batch or test member i, tk,i = characteristic bond stress from reliability test series in concrete batch or test member i calculated in accordance with §10.3, tk,o,i = characteristic bond stress from reference test series in concrete batch or test member i calculated in accor- dance with §10.3, and areq = controlling value for reliability tests and service- condition tests where calculation of a is required. The reference value (areq) is specific to each test and is either given in Tables 3.1 to 3.3 or determined from §10.4.6 based on the anchor category. Sensitivity to Hole Cleaning, Dry Concrete (ACI 355.4 §7.5). This test evaluates the sensitivity of an adhesive anchor to the degree of hole cleaning prior to installation in dry con- crete. The hole is cleaned with 50% of the manufacturer’s cleaning instructions. If the manufacturer does not specify the cleaning operation, no cleaning is conducted. A static ten- sion test is conducted as specified in ASTM E488, continu- ously monitoring load and displacement to determine the ratio a per ACI 355.4 Eq. 10-7. This test is not required if the manufacturer requires that the holes be flushed with water. Sensitivity to Hole Cleaning, Saturated Concrete (ACI 355.4 §7.6). This test evaluates the sensitivity of an adhesive anchor to the degree of hole cleaning prior to installation in saturated concrete. A pilot hole about one-half the diameter of the intended hole is drilled and kept filled with water for 8 days or until the concrete is saturated over a diameter of 1.5 times the hole diameter. Prior to installation, the water is removed with a vacuum and the hole is drilled to the required diameter. The hole is cleaned with the 50% cleaning effort as mentioned in §7.5 and the anchor is installed. Flushing the hole with water is allowed if specified by the manufacturer. A static tension test is conducted as specified in ASTM E488, continuously monitoring load and displacement to deter- mine the ratio a per ACI 355.4 Eq. 10-7. Sensitivity to Hole Cleaning, Water-Filled Hole (ACI 355.4 §7.7, optional). This test evaluates the sensitivity of an adhesive anchor to the degree of hole cleaning prior to installation in a water-filled hole. The test is identical to the test described in §7.6, except that the hole is filled with water after the reduced cleaning procedure. A static tension test is conducted as specified in ASTM E488, continuously moni- toring load and displacement to determine the ratio a per ACI 355.4 Eq. 10-7. Sensitivity to Hole Cleaning, Submerged Concrete (ACI 355.4 §7.8, optional). This test evaluates the sensitivity of an Identification Tests. In order to positively identify the adhesive being tested and compare it against the manufac- turer’s standard, ACI 355.4 §5.3 requires that at least three of the following tests be conducted. • Infrared absorption spectroscopy per ASTM E1252; • Bond strength per ASTM C882 or equivalent; • Specific gravity per ASTM D1875; • Gel time per ASTM C881; • Viscosity per ASTM D2556, ASTM F1080, or equivalent; and • Other appropriate tests to positively identify the material. Reference Tests. For each batch of concrete, reference static tests are performed to establish baseline values to later calculate a ratio (a) to compare a specific test’s results to the reference test results for the subsequent reliability and service-condition tests. These tests follow the ASTM E488 static test procedure and are conducted in dry concrete at standard temperature. These tests are referred to as 1a to 1d in ACI 355.4 Tables 3.1 to 3.3. Reliability Tests. Reliability tests are conducted to deter- mine an adhesive anchor’s performance under adverse instal- lation conditions and sustained load. ACI 355.4 Tables 3.1 to 3.3 refer to the tests by their section number. In the listing of the tests below, the ACI 355.4 section number is included for reference. The baseline strength determined in the reference tests is used to evaluate the results from the reliability test. This eval- uation creates a ratio, a, as calculated per ACI 355.4 Equa- tion. 10-7, which is compared to a limit referred to as areq. ; ACI 355.4 Eq.10-7 , , , , , min u i o i k i k o i reqα = τ τ τ τ     ≤ α Source: ACI 355.4 (2011b) Figure 20. Evaluation of load at Nadh.

18 Sensitivity to Crack Width, High-Strength Concrete (ACI 355.4 §7.14). This test is similar to the test specified in §7.13 except that the concrete specimen is of high-strength concrete. Sensitivity to Crack-Width Cycling (ACI 355.4 §7.15). This test evaluates an adhesive anchor’s performance in cracked concrete whose crack width is cycled. An anchor is installed so that a crack runs through the middle of the hole and a tension load of about 30% of its characteristic resis- tance is applied. While the load is maintained on the anchor, the test member is cyclically loaded so that the crack width is cycled between two set limits at a frequency of 0.2 Hz for 1,000 cycles. Load and displacement are measured during the test and following the 1,000 cycles the anchor is unloaded and the resulting displacement and crack width is measured. A static tension test as specified in ASTM E488 is conducted, continuously monitoring load and displacement to deter- mine the ratio a per ACI 355.4 Eq. 10-7. Additionally, the cumulative anchor displacement after 20 cycles must be less than 0.080” and the cumulative anchor displacement after the 1,000 cycles must be less than 0.120”. Sensitivity to Freezing and Thawing (ACI 355.4 §7.16). This test determines the performance of an adhesive anchor under freezing and thawing conditions. An anchor is installed in concrete and the top surface of the concrete is covered with ½” of water for a distance of 3” around the anchor. The anchor is loaded with a sustained load of about 55% of the average ultimate tension load of reference tests. Within two hours the temperature is lowered to −4°F ± 5°F (−20°C ± 2°C) and maintained for 14 hours. The temperature is then raised to +68°F ± 5°F (+20°C ± 2°C) within 1 hour and maintained for 14 hours. Fifty such cycles are conducted measuring load, displacement, and temperature. Following the 50 cycles, a static tension test as specified in ASTM E488 is conducted, continuously monitoring load and displacement to deter- mine the ratio a per ACI 355.4 Eq. 10-7. Additionally, the rate of displacement increase shall decrease to zero as the number of freeze–thaw cycles increase. Sensitivity to Sustained Loading at Standard and Maxi- mum Long-Term Temperature (ACI 355.4 §7.17). The sus- tained loading test is similar to the procedure from ICC-ES AC58 (based on ASTM E1512) with the following changes: • The sustained load is increased to about 55% of the aver- age tension capacity of the reference tests; • Sustained load tests are conducted at both standard tem- perature and the long-term elevated temperature; and • Following the 42 day (1,000 hr) sustained load tests, the anchors are loaded until failure to determine the residual capacity. The acceptance criteria as presented in ICC-ES AC58 were modified in the development of ICC-ES AC308 and are reflected in ACI 355.4. The displacement data is projected adhesive anchor to the degree of hole cleaning prior to instal- lation in submerged concrete. The concrete member is cov- ered with at least ½” of water during drilling, and is then subjected to the reduced cleaning effort (as described in §7.5), installation, and testing. A static tension test is conducted as specified in ASTM E488, continuously monitoring load and displacement to determine the ratio a per ACI 355.4 Eq. 10-7. Sensitivity to Mixing Effort (ACI 355.4 §7.9). This test evaluates the sensitivity of the adhesive to a reduced mixing effort. This test is only for adhesive anchor systems in which the mixing of the adhesive components is controlled by the installer such as systems that require mixing until a color change occurs, or mixing for a specific duration or number of mixing repetitions. This test is not required for systems that use a cartridge system with static mixing nozzles or capsule anchor systems. A reduced mixing effort is defined as mixing the adhesive for only 75% of the required mixing time speci- fied by the manufacturer. A static tension test is conducted as specified in ASTM E488, continuously monitoring load and displacement to determine the ratio a per ACI 355.4 Eq. 10-7. Sensitivity to Installation in Water-Saturated Concrete (ACI 355.4 §7.10, optional). This test evaluates the sensitivity of an adhesive anchor to installation in saturated concrete. This test is similar to the test specified in §7.6 except that it requires a full cleaning effort as prescribed by the manufac- turer. A static tension test is conducted as specified in ASTM E488, continuously monitoring load and displacement to determine the ratio a per ACI 355.4 Eq. 10-7. Sensitivity to Installation in a Water-Filled Hole, Satu- rated Concrete (ACI 355.4 §7.11). This test evaluates the sensitivity of an adhesive anchor installation in a water-filled hole. This test is similar to the test specified in §7.7 except that it requires a full cleaning effort as prescribed by the manufac- turer. A static tension test is conducted as specified in ASTM E488, continuously monitoring load and displacement to determine the ratio a per ACI 355.4 Eq. 10-7. Sensitivity to Installation in Submerged Concrete (ACI 355.4 §7.12, optional). This test evaluates the sensitivity of an adhesive anchor installation in submerged concrete. This test is similar to the test specified in §7.8 except that it requires a full cleaning effort as prescribed by the manufacturer. A static tension test is conducted as specified in ASTM E488, contin- uously monitoring load and displacement to determine the ratio a per ACI 355.4 Eq. 10-7. Sensitivity to Crack Width, Low-Strength Concrete (ACI 355.4 §7.13). This test evaluates the sensitivity of an adhe- sive anchor installed in low-strength concrete with a wide crack passing through the anchor location. Following anchor installation and adhesive curing, the crack is widened and a static tension test as specified in ASTM E488 is conducted, continuously monitoring load, displacement, and crack width to determine the ratio a per ACI 355.4 Eq. 10-7.

19 Tension Tests in Uncracked and Cracked Concrete (ACI 355.4 §8.4). These tests are conducted to determine the adhesive anchor’s unconfined tension strength per ASTM E488 and are used as a baseline for unconfined tests. Tests are conducted in both low- and high-strength concrete for cracked and uncracked conditions. Tension Tests at Elevated Temperature (ACI 355.4 §8.5). These tests are conducted to determine an adhesive anchor’s sensitivity to elevated temperature. Static tension tests are conducted at various temperatures per Table 8.1 in ACI 355.4 (shown as Table 1). Anchors are installed and cured at stan- dard temperature for both categories. For category A, tests are conducted at the long-term and the short-term temper- ature. For category B, tests are conducted at standard tem- perature, at the long-term temperature, at the short-term temperature, and at least two temperatures in between the long-term and the short-term temperature with a maximum increment of 35°F (20°C). Following the cure time, the anchors are heated to the test temperature and tested per ASTM E488 with continuous measurements of load and displacement. Tests must be com- pleted before the test member temperature falls below the test from the last 20 days (minimum of 20 data points) from the creep test using the Findley power law (instead of the loga- rithmic model) shown in ACI 355.4 Eq. 10-24. ACI 355.4 Eq.10-240t att b( )∆ = ∆ += where D(t) = total displacement at time t, Dt = 0 = initial displacement under sustained load, t = time corresponding to the recorded displacement, and a, b = constants evaluated from a regression analysis. Using ACI 355.4 Eq. 10-24, the displacement is then esti- mated at the service lives of 10 years and 50 years. The adhe- sive anchor is accepted for sustained load if: • The projected displacement at 10 years is less than the mean displacement at loss of adhesion for the reference test at elevated temperature, • The projected displacement at 50 years is less than the mean displacement at loss of adhesion for the reference test at standard temperature, and • The residual capacity is greater than 90% of the reference test’s capacity. Sensitivity to Installation Direction (ACI 355.4 §7.18, optional). This test evaluates the sensitivity of an adhesive anchor to hole orientation (horizontal or upward). This test installs anchors in holes that are oriented horizontally and vertically overhead. The anchors are installed with the most unfavorable installation temperature of the concrete and the adhesive. Static tension tests are conducted as specified in ASTM E488, continuously monitoring load and displace- ment to determine the ratio a per ACI 355.4 Eq. 10-7. Addi- tionally, the anchor must not displace more than 0.05 times the anchor diameter during curing. There are additional sub- jective assessments on the adequacy of the manufacturer’s procedures for overhead and horizontal installations. The effectiveness of the overhead installation procedure can be verified by the procedure shown in Figure 21. Torque Tests (ACI 355.4 §7.19). This test evaluates the maximum torque that can be applied to an adhesive anchor without damaging the adhesive bond or yielding the anchor. Torque is applied to the anchor and measurements of torque and the resulting induced tension in the anchor are recorded. The torque reached in the test must be greater than 130% of the tightening torque specified by the manufacturer. Service-Condition Tests. These tests are conducted to determine an adhesive anchor’s performance under service conditions. In the listing of the tests below, the ACI 355.4 sec- tion number is included for reference. Source: ACI 355.4 (2011b) Figure 21. Procedure for verifying the effectiveness of overhead adhesive injection. Temperature Category Long-Term Test Temperature1, Tlt Short-Term Test Temperature1, Tst ºF ºC ºF ºC A 110 43 176 80 B ≥110 ≥≥ ≥43 Tlt + 20 Tlt + 11 1All test temperatures have a minus tolerance of 0°. Table 1. ACI 355.4 Table 8.1 - Minimum test temperatures.

20 the anchor under the sustained preload portion shall stabilize prior to static tension testing. Establishment of Cure Time at Standard Temperature (ACI 355.4 §8.7). These tests are conducted to determine an adhesive anchor’s sensitivity to reduced cure time. Compari- son tests are conducted on anchors allowed to cure for the minimum curing time and on anchors that were cured for 24 hours longer than the minimum curing time. Confined static tension tests are conducted in uncracked concrete as specified in ASTM E488 while continuously monitoring load and displacement. The acceptance criterion for these tests is shown as ACI 355.4 Eq. 10-28: min N N N N cure cure k cure k cure+ +     ≥ 24 24h h ; , , 0 9. ACI 355.4 Eq. -10 28 where N−cure = mean tension capacity corresponding to the manufacturer’s published minimum cure time, N−cure+24h = mean tension capacity corresponding to the manufacturer’s published minimum cure time + 24 hours, Nk,cure = characteristic tension capacity corresponding to the manufacturer’s published minimum cure time, and Nk,cure+24h = characteristic tension capacity corresponding to the manufacturer’s published minimum cure time + 24 hours. Durability Assessment (ACI 355.4 §8.8, sulfur test is optional). These tests determine an adhesive anchor’s sen- sitivity to harsh environments. Mandatory alkalinity tests are conducted and optional sulfur dioxide tests can be con- ducted. Specimens are made by installing adhesive anchors in 6” diameter concrete cylinders cast in PVC or steel pipe. After installation and curing, the cylinders are sliced into 13/16” ± 1/8” thick slices. A minimum of 10 slices are to be made for each environmental condition tested plus 10 for reference tests. The reference slices are stored at standard tempera- ture and 50% relative humidity for 2,000 hours. The slices for the high alkalinity environment tests are stored for 2,000 hours in an alkaline solution with a pH = 13.2. The slices for the optional sulfur dioxide tests are tested accord- ing to EN ISO 6988 (Kesternich Test) with a concentration of 0.67% for at least 80 cycles. Following storage, the anchors are punched out of the slices with the concrete restrained in a device similar to that shown in Figure 22. The bond stress for each slice is the peak load divided by the circumferential area of the anchor. A reduction factor adur is calculated for each durability test. Verification of Full Concrete Capacity in a Corner (ACI 355.4 §8.9). These tests determine the critical edge distance (cac) temperature. The ratios alt and ast are calculated from the sustained load and short-term tests respectfully as shown in ACI 355.4 Eq. 10-26 and Eq. 10-27 below. ; 1.0 ACI 355.4 Eq.10-26 , , , , min N N N N lt lt o i k lt k o i α =     ≤ 0.8 ; 0.8 1.0 ACI 355.4 Eq.10-27 , , min N N N N st st lt k st k lt α =     ≤ where N−lt = mean tension capacity at long-term elevated temperature, N−st = mean tension capacity at short-term elevated temperature, N−o,i = mean tension capacity of an anchor in reference test series i, Nk,lt = characteristic tension capacity at long-term ele- vated temperature, Nk,st = characteristic tension capacity at short-term ele- vated temperature, and Nk,o,i = characteristic tension capacity of an anchor in ref- erence test series i. Tension Tests with Decreased Installation Temperature (ACI 355.4 §8.6, optional). These tests determine an adhesive anchor’s sensitivity to installation at reduced temperature. A minimum of five confined tests in uncracked concrete are conducted. The test member and anchor rod are conditioned at a test temperature below 50°F (10°C) for 24 hours prior to installation. The anchor is then installed and cured at the desired temperature. Once curing is completed, a static ten- sion test is conducted. If the test temperature is below 40°F (5°C) an additional test is conducted. The conditioning and installation pro- cedure is the same as described above. However, prior to removal of the specimen from the environmental chamber, a preload of about 55% of the ultimate load is applied to the anchor. The specimen is then removed from the chamber and is heated uniformly to standard temperature over a period of 72 to 96 hours. Temperature and displacement readings are taken during this heating period. Once the specimen has reached the desired temperature, a static tension test is con- ducted to failure. The mean and the 5% fractile of these tests shall be sta- tistically equivalent to those of the reference tests. ACI 355.4 defines statistically equivalent as follows, if “. . . there are no significant differences between the means and between the standard deviations of the two groups. Such statistical equiv- alence shall be demonstrated using a one-sided Student’s t-Test at a confidence level of 90%.” Additionally, for anchors installed in concrete below 50°F (10°C), the displacement of

21 is opened to the maximum crack width during the seismic test and a static tension test is conducted in accordance with ASTM E488 until failure. For acceptance, the anchors must complete the seismic loading cycle without failure. Upon completion, the resid- ual strength of the anchor must be at least 160% of Neq. If the anchor does not complete the seismic loading cycle, a reduced value for Neq (Neq,reduced) is used until the anchors pass the criteria. If a reduced loading cycle is performed, a reduction factor aN,seis is determined by dividing Neq,reduced by Neq. Simulated Seismic Shear Tests (ACI 355.4 §8.13, optional). The purpose of these tests is to evaluate adhesive anchors sub- jected to a simulated seismic shear load in cracked concrete. Anchors are installed in a crack which is opened by 0.020″ prior to loading. A sinusoidal shear load is applied to the anchor parallel to the crack with a frequency between 0.1 and 2 Hz. The peak shear load is initially at Veq for 10 cycles, then reduced to Vi for 30 cycles, and finally to Vm for 100 cycles, where: Veq = about 50% of the mean shear capacity of reference tests, Vi = 75% Veq, and Vm = 50% Veq. During the test, crack width, shear load, and displacement are recorded. Following the seismic loading, the crack is opened to the maximum crack width during the seismic test and a static shear test is conducted in accordance with ASTM E488 until failure. For acceptance, the anchors must complete the seismic loading cycle without failure. Upon completion, the residual strength of the anchor must be at least 160% of Veq. If the anchor does not complete the seismic loading cycle, a reduced value for Veq (Veq,reduced) is used until the anchors pass the crite- ria. If a reduced loading cycle is performed, a reduction factor aV,seis is determined by dividing Veq,reduced by Veq. in test members with the minimum thickness as specified by the manufacturer. Static tension tests per ASTM E488 are per- formed in low-strength uncracked concrete on anchors located in a corner with equal edge distances of cac. The tension capacity from these tests should be statistically equivalent to the tension capacity of reference tests performed away from a corner. Determination of Minimum Spacing and Edge Distance to Preclude Splitting (ACI 355.4 §8.10, optional). The pur- pose of these tests is to evaluate the shear capacity of adhesive anchors. Static shear tests away from edges are performed per ASTM E488. The concrete should not crack during the test and the mean failure load must be greater than 90% of the expected failure load. Test to Determine Shear Capacity of Anchor Elements with Non-Uniform Cross Section (ACI 355.4 §8.11). The purpose of these tests is to determine the shear capacity of anchors in which the shear capacity cannot be reliability cal- culated due to a non-uniform cross section. Static shear tests away from edges are performed per ASTM E488 with a few requirements on edge spacing and embedment depth. Simulated Seismic Tension Tests (ACI 355.4 §8.12, optional). The purpose of these tests is to evaluate adhe- sive anchors subjected to a simulated seismic tension load in cracked concrete. Anchors are installed in a crack which is opened by 0.020″ prior to loading. A sinusoidal tension load is applied to the anchor with a frequency between 0.1 and 2 Hz. The peak tension load is initially at Neq for 10 cycles, then reduced to Ni for 30 cycles, and finally to Nm for 100 cycles, where: Neq = about 50% of the mean tension capacity of reference tests, Ni = 75% Neq, and Nm = 50% Neq. During the test, crack width, tension load, and displace- ment are recorded. Following the seismic loading, the crack Source: ACI 355.4 (2011b) Figure 22. ACI 355.4 Punch test apparatus (hsl = slice thickness as measured immediately prior to punch test).

22 The nominal characteristic bond stress for each service- condition test (tk,nom(cr,uncr)) is calculated as per Eq. 3 shown earlier. The limiting characteristic bond stress for each service- condition test (tk(cr,uncr)) is adjusted for many reduction fac- tors as shown in ACI 355.4 Eq. 10-12: τ = τ βα α α α α α α( ) ( ) ACI 355.4 Eq.10-12 k cr uncr k nom cr uncr lt st dur p conc COV cat, , , 3 where: b = min min ;α α α req adhmin     the reliability and service- condition tests listed in ACI 355.4 Table 10.2 and Table 10.3, a = ratio of reliability test result to reference test result evaluated for all reliability tests listed in ACI 355.4 Table 10.2, alt = reduction factor for maximum long-term temperature, ast = reduction factor for maximum short-term temperature, adur = reduction factor for durability, ap = min. reduction factor for reduced sustained load in reliability tests, aconc = adjustment factor for regional concrete variation, aCOV = reduction factor associated with the coefficient of variation of ultimate loads, and acat3 = reduction factor for anchor category 3. Anchor Categories. Based on the alpha-reduction factor results of the reliability tests, anchors are classified into catego- ries depending on the required level of inspection. Tables 10.5 and 10.6 in ACI 355.4 are used to compare the alpha-reduction factors from the different reliability tests against certain thresh- old values in order to assign a strength reduction (resistance) factor for design. AASHTO Standard Specifications for Transportation Materials and Methods of Sampling and Testing AASHTO (2008) Standard Specifications for Transportation Materials and Methods of Sampling and Testing was reviewed for a framework of specifications within which to incorpo- rate specifications for adhesive anchors. The only test method dealing with adhesive systems was T 333-07 (2007c) Linear Coefficient for Shrinkage on Cure of Adhesive Systems which measures the change in length of a cured adhesive material. TP 84-10 (2010c) Evaluation of Adhesive Anchors in Concrete under Sustained Loading Conditions was recently created to evaluate adhesive anchors using the stress versus time-to- failure approach discussed in detail later in the chapter. Additional Supplemental Tests. ACI 355.4 specifies a few additional supplemental tests: Round Robin Tests (ACI 355.4 §9.1). These tests examine the effects of regional variations of concrete on the behavior of adhesive anchor systems. Tests are conducted at labora- tories located in each time zone of the United States using aggregates representative of that region. Five confined and five unconfined static tension tests per ASTM E488 are con- ducted and compared with the original laboratory results to generate an adjustment factor aconc for each laboratory and the minimum value is used. Tests to Determine Minimum Member Thickness (ACI 355.4 §9.2). These tests verify the minimum member thick- ness as specified by the manufacturer. Ten anchors are installed at the maximum embedment depth in a concrete member and the member is checked for cracking or spalling. Additional Assessment Tests. A few additional assess- ment tests are included if pertinent. Multiple Anchor Type Supplementary Tests (ACI 355.4 §3.4). These tests investigate the effects of using anchors of different metal composition within an anchor group. The entire test program is conducted with one anchor type, and the other anchor types are subjected to a series of additional tests specified in ACI 355.4 Table 3.4. Alternate Drilling Methods Supplementary Tests (ACI 355.4 §3.5). If the manufacturer permits drilling methods other than with rotary hammer drill and carbide bit, supplementary tests are conducted using the alternate drilling method. ACI 355.4 Table 3.5 lists the tests to conduct on the alter- nate drilling method. If the results of these tests are not sta- tistically equivalent to the results from their respective tests using the rotary hammer and carbide bit drilling method, all tests need to be conducted except for the shear capacity tests for an element having a non-uniform cross section (§8.11). Resulting Design Values. The previously described test- ing program provides design values to be used by ACI 318-11 Appendix D. The bond stress for each service-condition test (ti) is calculated from ACI 355.4 Eq. 10-11: ACI 355.4 Eq.10-11 , ,N dh i setup u i fc ef τ = α pi where asetup = 1.0 for unconfined test; = 0.75 for confined test; = 0.70 for confined test in cracked concrete; Nu,i,fc = peak tension load in test series i normalized to concrete strength of fc = 2,500 psi, lbs; d = anchor diameter, in.; and hef = embedment depth, in.

23 CALTRANS Standard Specifications Section 75 “Miscellaneous Metal” of the CALTRANS (2006b) Standard Specifications lists the requirements for resin capsule anchors tested under CALTRANS (2001) CTM681. A resin capsule anchor must withstand a sustained tensile load for at least 48 hours with a displacement less than 0.035”. The applied sustained load shall be in accordance with Table 2. Anchors must be made of steel or stainless steel and hot- dip or mechanically galvanized. CALTRANS has a test method for creep performance of adhesive anchor systems, but no comprehensive material specifications for adhesive anchors could be found in their standard specifications. While Section 95 of the CALTRANS (2006b) Standard Specifications deals with epoxy, there is no specific mention of an epoxy used for adhesive anchor applica- tions. However, in Section 83 (Buildings and Barriers), there is a comment that anchor bolts that are set with epoxy shall use a two-component epoxy mixture as specified in Section 95-2.01 “Binder (Adhesive), Epoxy Resin Base.” Texas Department of Transportation DMS-6100— Epoxies and Adhesives Texas Department of Transportation (TxDOT) (2007a) DMS-6100 Epoxies and Adhesives classify epoxies and adhesives into nine types and specifies Type III to be used for dowel and tie bar adhesives. Type III adhesives are further classified into three classes (A–C). Class A is a bulk material for horizontal applications, Class B is for vertical applications, and Class C is either a bulk material or cartridge dispensed material for machine applications and can be applied horizontally or vertically. Table 3 specifies the performance requirements for Type III adhesives tested according to TxDOT (2007b) Tex-614-J. TxDOT (2007b) Tex-614-J requires that each component be distinctly colored and result in a third color when thoroughly mixed. Also the filler in the components must not damage the dispensing equipment and the extruder must meter the M 235M/M 235-03 (2007b) Standard Specification for Epoxy Resin Adhesives is the AASHTO version of ASTM C881-99 that provides specifications for seven types of adhesives and subsequent tests that refer to many ASTM tests. While Type IV adhesives are for bonding hardened concrete to other materials in load-bearing applications, it is not specifically for epoxy-adhesive anchor systems. State DOT Test Methods and Material Specifications A review was made of test methods and material specifi- cations from various state departments of transportation (DOTs). California Department of Transportation Acceptance Criteria for Adhesive Anchors in Concrete and Masonry Elements Adhesive anchors for use in California Department of Transportation (CALTRANS) contracts must meet the requirements of International Conference of Building Offi- cials (ICBO)-AC58 (ICC-ES AC58) as well as additional clar- ifications and amendments as found in CALTRANS (2010). Four fingerprint tests are identified: • Qualitative infrared analysis per ASTM E1252. • Bond strength–slant shear per ASTM C882. • Density per ASTM D1875. • Gel time per ASTM C881. Six additional tests not discussed in ICBO-AC58 are also required: • Viscosity of adhesives per ASTM D2556. • Deflection temperature per ASTM D648. • Filler content per ASTM C881. • Rheological properties per CALTRANS CTM438. • Glass transition temperature per CALTRANS CTM438. • Sag test to evaluate the tendency of the adhesive to flow out of overhead hole. Four optional tests per ICBO-AC58 are required by CALTRANS: • Creep test conducted per ASTM E1512. • Dampness test. • Freezing and thawing test. • Seismic test. A fire resistance test is required if there is a concern for a particular job or application. Stud Diameter (inches) Sustained Tension Test Load (pounds) 1 ¼ 31,000 1 17,900 14,400 ¾ 5,000 4,100 ½ 3,200 2,100 ¼ 1,000 Table 2. CTM681 sustained load values.

24 New York State Department of Transportation Standard Specifications New York State Department of Transportation (NYSDOT) (2008b) Standard Specifications Section 701-07 “Anchor- ing Materials—Chemically Curing” specifies the testing and material requirements for polymer anchoring materials for anchor bolts in concrete. The material must be non-metallic, non-shrink polymer resin in prepackaged or premeasured containers. It cannot contain corrosion promoting agents and must be insensitive to moisture. The material must last at least 6 months when stored between 40°F and 90°F (4°C and 32°C). The container must include the mixing instruc- tions, setting time, and expiration date. Section 701-07 specifies certain chemical resistances as tested per ASTM D471 at 70°F (21°C) for 24 hours as noted in Table 4. Two series of tension pullout tests are specified for accep- tance by the state. Test series 1 conducts three tests using 1” diameter threaded rods embedded 10” in concrete. The pullout load must be greater than the values found in Table 5. Test series 2 conducts two sets of three tests using 5⁄8” diameter threaded rods embedded 4” in concrete. The pullout load for each set must be greater than the values found in Table 5. Section 654-3.03 “Anchorages” of NYSDOT (2008c) per- mits drilling by rotary impact drills only, and specifically does not permit core drills. proportioning and mixing of the components and handle the viscosity range of the components. TxDOT Tex-641-J Testing Epoxy Materials TxDOT (2007b) Tex­641­J, Testing Epoxy Materials is a col- lection of many material tests for adhesives. Five are required for Type III adhesives. Material specifications are covered in TxDOT (2007a) DMS-6100 Epoxies and Adhesives discussed later. Gel Time. This test measures the gel time by mixing a sample at 77°F ± 2°F (25°C ± 1°C) and probing it with a toothpick until a ball of cured material forms at the center. Viscosity. This test measures the viscosity of the adhesive using a Brookfield viscometer at 77°F ± 2°F (25°C ± 1°C). Tensile Bond. This test measures the bond strength of the adhesive between two mortar briquettes. Two sets of three specimens are prepared for Type III adhesive. For each speci- men, two mortar briquettes are joined with adhesive. One set of specimens is cured for 6 hours at 77°F ± 2°F (25°C ± 1°C) and the second set is cured for 48 hours at 120°F ± 2°F (49°C ± 1°C). Once cured, the specimens are placed into a tensile machine and loaded in tension until failure. Thixotropy Bond @ 120°F (49°C). This test forms a 2” by 4” by 0.05” thick sample of adhesive on a metal plate conditioned at 120°F ± 2°F (49°C ± 1°C). The plates are then placed in an oven at 120°F ± 2°F (49°C ± 1°C) until the adhesive has hard- ened. The thickness retained is measured and the thixotropy bond is calculated as the average of eight thickness readings. Wet Pullout Strength. In this test, a #3 (3⁄8”) grade 60-ksi rebar is installed in a 5⁄8” diameter by 3.5” deep hole in a 6” diameter by 8” long concrete cylinder. The adhesive and anchor are installed and cured at 77°F ± 3°F (25°C ± 1°C). Following a 24-hour curing time, the block is submerged upright in a 77°F ± 3°F (25°C ± 1°C) water bath for 6 days. The anchor is then loaded in tension until failure. Physical Property Requirements Class A Class B Class C Gel time, min. 25 min 25 min 6 min Viscosity of mixed components, poise (Pa-s) 1,200 (120) max 20 (2) min 150 (50) max - Tensile bond @ 6 hr., psi (Mpa) 200 (1.40) min 200 (1.40) min 200 (1.40) min Tensile bond @ 120°F (49°C), psi (Mpa) 400 (2.8) min 400 (2.8) min 400 (2.8) min Thixotropy bond @ 120°F (49°C), mils (mm) 30 (0.75) min - 30 (0.75) min Wet pullout1 strength, lbf. (kN) 4,500 (20) min 4,500 (20) min 4,500 (20) min 1The wet pullout test determines the strength of the adhesive bond between a steel anchor and the surface of a hole in concrete or masonry units. Table 3. Performance requirements for Type III adhesives tested with Tex-614-J. Chemical Resistance Gasoline Slight swell Hydraulic brake fluid No effect Motor Oil No effect Sodium chloride (5%) No effect Calcium chloride (5%) No effect Table 4. NYSDOT chemical resistance requirements.

25 of 5⁄8” (16 mm), and an embedment of 4” (102 mm). The test load must be applied within 24 hours after installation. Long-Term Load (Creep). This test conducts the creep test series listed in ASTM E1512 with the following specifications: • References Table 2 of ASTM E488 for requirements on the distance between the reaction force and the anchor; • The minimum sustained tension load of 40% of the aver- age tension failure load is established by an unconfined tension test; • The minimum testing temperature of the concrete and anchor specimens is 110°F (43°C); • A load duration of 42 days; and • Following the 42 day loading period, the temperature of the specimens is cooled to 70°F ± 5°F (21°C ± 3°C) and an unconfined tension test is performed. Unconfined Static Tension Test. This test method specifies unconfined test setups with anchor diameters and embed- ments as follows: • An anchor diameter of 5⁄8” (16 mm) and embedment of 4” (102 mm), • An anchor diameter of 5⁄8” (16 mm) and embedment of 6” (152 mm), and • An anchor diameter of ¾” (19 mm) and embedment of 6” (152 mm). FDOT Standard Specifications for Road and Bridge Construction The material specifications for adhesive anchor systems for FDOT are found in FDOT (2007) Standard Specifications for Road and Bridge Construction, Section 937 “Adhesive Bond- ing Material Systems for Structural Applications.” Only systems that are specifically intended for bonding anchors and dowels into concrete in structural applications are allowed. FDOT restricts the use of adhesives that are man- ually combined from bulk supplies and only allows systems that are prepackaged in which the two components are in separate chambers and are automatically proportioned and mixed when discharged. Only undamaged full packages can be used (i.e., packages that were previously opened cannot be used). Adhesive anchors can only be installed in positions ranging from horizontal to vertically downward. Two types of adhesive systems (HV and HSHV) are defined as follows: • Type HV Adhesives: Used in bonding materials for all horizontal installations and vertical installations other than constructing doweled pile splices, except when Type HSHV is required. Type HV adhesives may not be substi- tuted for Type HSHV adhesives. NYSDOT Engineering Instruction EI 08-012 NYSDOT (2008b) Engineering Instruction EI 08-012 was published in March of 2008 to limit the use of NYSDOT (2008c) Section 701-07 “Anchoring Materials – Chemical Curing.” This was due to recommendations from the National Transporta- tion Safety Board (NTSB) to limit the use of adhesive anchors in overhead installations or in situations that could pose a risk to public safety. In such situations, NYSDOT recommends using alternative anchoring systems such as cementitious grout or mechanical anchor systems. FDOT FM 5-568 Florida Method of Test for Anchor Systems for Adhesive-Bonded Anchors and Dowels. FDOT (2000) FM 5-568 is FDOT’s test method for anchor systems with adhesive-bonded anchors and dowels. Its pur- pose is to determine the bond strength and performance characteristics of adhesive anchors in uncracked concrete. The material specifications for this test method are contained in Section 937 of FDOT (2007) Standard Specifications for Road and Bridge Construction. The tests contained in FDOT FM 5-568 reference the test procedures specified in ASTM E488 and ASTM E1512 with a few modifications/specifications as explained below. Confined Tension. This test method specifies a confined test setup, an anchor diameter of 5⁄8” (16 mm), and an embedment of 4” (102 mm). Damp-Hole Installation. This test method specifies a con- fined test setup, an anchor diameter of 5⁄8” (16 mm), and an embedment of 4” (102 mm). Elevated Temperature. This test method specifies a con- fined test setup, an anchor diameter of 5⁄8” (16 mm), an embedment of 4” (102 mm), and a minimum temperature of 108°F (42°C). Horizontal Orientation. This test is a static tension test on an anchor installed and cured in a horizontal orientation. This test method specifies a confined test setup, an anchor diameter of 5⁄8” (16 mm), and an embedment of 4” (102 mm). Short-Term Cure. This test is a static tension test on an anchor installed and cured in a horizontal orientation. This test method specifies a confined test setup, an anchor diameter Concrete Strength (psi) Minimum Pullout Load (lbf) Test series 1 1” dia. - 10” embedment Test series 2 ” dia. - 4” embedment 4,000 51,120 8,593 4,500 54,225 9,113 5,000 57,150 9,630 5,500 59,940 10,080 Table 5. NYSDOT anchor tests minimum pullout loads.

26 systems use #5 epoxy coated 60 ksi rebar. The following tests are conducted: Dry Conditioning. A static tension test per ASTM E488 is conducted within 1 hour of installation and stopped when the load reaches 16 kips or the displacement reaches 0.1”. The anchor system is accepted if it withstood a minimum load of 13.55 kips with less than 0.1” displacement. Wet Conditioning. This test is similar to the “Dry Condi- tioning” test except the hole is filled with water for 12 hours and then removed prior to installation. A static tension test per ASTM E488 is conducted within 1 hour of installation and stopped when the load reaches 16 kips or the displacement reaches 0.1”. The anchor system is accepted if it withstood a minimum load of 13.55 kips with less than 0.1” displacement. Cold Temperature Conditioning. This test is similar to the “dry conditioning” test except that the adhesive and threaded rod are conditioned to 32°F ± 4°F (0°C ± 2°C) prior to instal- lation. The anchor is cured for 24 hours at the above tempera- ture. A static tension test per ASTM E488 is conducted at the end of the 24 hour curing period and loaded until failure. The anchor system is accepted if the displacement at failure was less than 0.1”. Compressive Strength. This test is not for glass capsule sys- tems. This test tests two 1” diameter by 2” cylinder specimens at 73°F ± 4°F (23°C ± 2°C). One specimen is tested at 1 hour and the other at 24 hours after casting. The adhesive is accepted if the 1 hour compressive strength is greater than 3,000 psi and the 24 hour compressive strength is greater than 4,000 psi. Horizontal Installation Stability. This test is not for glass capsule systems. This test installs a 1¼” diameter by 14” long smooth steel dowel bar into a horizontal 9” long by 13⁄8” diam- eter clear plastic tube at 73°F ± 4°F (23°C ± 2°C). The anchor system is accepted if the anchor could be installed by hand without “appreciable drain down” from the top of the tube. Infrared Spectrophotometer “Fingerprint.” A fingerprint record is made of the cured adhesive for future reference. IDOT Standard Specifications for Road and Bridge Construction Section 1027.01 “Chemical Adhesive” of IDOT (2007b) Standard Specifications for Road and Bridge Construction ref- erences IDOT (2007a) Laboratory Test Procedure for Chemical Adhesives for the testing and acceptance of chemical adhe- sives. Section 1027.01 states that the adhesive must consist of a two-part fast-setting resin and filler/hardener. Washington State Department of Transportation Standard Specifications Section 9-26 “Epoxy Resins” of Washington State Department of Transportation (WSDOT) (2008b) Standard • Type HSHV Adhesives: Use higher strength Type HSHV adhesive bonding materials for installation of traffic rail- ing barrier reinforcement and anchor bolts into existing concrete bridge decks. HV and HSHV systems must be packaged to be automati- cally proportioned during installation. Section 937 also specifies the minimum performance requirements for tests conducted under FM 5-568 as indi- cated in Table 6. The coefficient of variation of the uniform bond stress is limited to 20%. Three criteria are specified for the creep test and are listed as follows: • The displacement rate shall decrease during the 42 day test period. • The total displacement at 42 days (with load still applied) shall be less than 0.03” and the total displacement due to creep during the last 14 days must be less than 0.003”. • After the 42 day test, the uniform bond stress from the con- fined tension test shall not be less than 1,800 psi. Finally, a qualified products list (QPL) is maintained by FDOT in which manufacturers can apply for their products to be included once they have met the requirements of Sec- tion 937. Illinois Department of Transportation Laboratory Test Procedure for Chemical Adhesives Illinois Department of Transportation (IDOT) (2007a) Laboratory Test Procedure for Chemical Adhesives tests chemi- cal adhesives for dowels and tie bars. The test procedure is for both gun grade adhesives and glass capsule adhesive systems. The glass capsule systems are installed using threaded rods in a ¾” diameter hole and embedded 5” into 4,000 psi dry concrete at 73°F ± 4°F (23°C ± 2°C). The gun grade adhesive Test or Property Uniform Bond Stress Type HV Adhesive (psi) Type HSHV Adhesive (psi) Confined tension 2,290 3,060 Damp-hole installation 1,680 1,830 Elevated temperature 2,290 3,060 Horizontal orientation 2,060 2,060 Short-term cure 1,710 1,710 Specified bond strength 1,080 1,830 Table 6. FDOT Minimum performance requirements for adhesive systems.

27 rather Section 1.3 references other previously discussed standards such as: • EOTA ETAG 001, • ICC-ES AC308, and • ACI 355.2. Other Test Methods The following section presents various alternate test meth- ods that can potentially evaluate sustained load performance of adhesive anchor systems. Short-Term Incremental Loading Test for Adhesive Anchors ASTM E488 provides for two load rates in the static ten- sion test: a continuous load rate that will produce failure at around 2 minutes and an incremental load rate that loads at 15% intervals and holds each step for 2 minutes. Several static tension tests were conducted at the University of Florida under NCHRP 20-07/Task 255 using a modified incremental load rate. Figure 23 shows a sample anchor test loaded with the incremental load rate. Under the incremental load rate, it was noticed that at the lower stress levels the anchor would initially displace when the load was held constant but would eventually stabilize over the 2 minute interval. However, at the higher stress levels, some anchors would continue to displace over the 2 minute interval. Stress versus Time-to-Failure Test NCHRP (2009) Project 20-07/Task 255 investigated sus- tained load testing for adhesive anchors and recommended a “stress versus time-to-failure” test method for AASHTO that has been adopted as AASHTO TP 84-10. The following is a summary of that test method; more detailed information is presented in NCHRP Report 639 (2009). The test method begins by placing five specimens under con- fined static tension tests to determine the mean static load at an elevated temperature of 110°F (43°C). Subsequent sustained load test series are conducted on five specimens at two lower stress levels at an elevated temperature of 110°F (43°C). It was recommended that these lower stress levels be within the speci- fied ranges of 70% to 80% and 60% to 70% of the mean static load. Ideally, the stress levels chosen would create data points in separate log cycles. The sustained load tests are conducted until failure, which is defined as the initiation of tertiary creep. The data is plotted on a stress versus time-to-failure graph (semi-log plot). A least squares trendline is drawn through each data point and projected linearly (on the log scale). Specifications lists the various types of epoxy bonding agents per the classification found in ASTM C881. Section 6-02.3(18) “Placing Anchor Bolts” discusses the requirements for plac- ing grouted anchor bolts and does not specifically mention adhesive anchors. Michigan Department of Transportation Material Source Guide Specification 712.03J “Adhesive Systems for Structural Anchor & Lane Ties” of the Qualified Products List (QPL) in the Michigan Department of Transportation (MDOT) (2009) Material Source Guide states that anchors should be installed per the manufacturer’s instructions with a mini- mum embedment depth of 9 diameters for threaded rod. Virginia Department of Transportation Road and Bridge Specifications Section 214 “Epoxy-Resin Systems” of Virginia Depart- ment of Transportation (VDOT) (2007a) Road and Bridge Specifications lists various types of epoxy-resin systems for various uses, but does not include adhesive anchors. Section 519 “Sound Barrier Walls” specifically prohibits the use of epoxy or adhesive anchors. International Test Methods and Material Specifications The following summarizes the review of international test standards and material specifications. EOTA ETAG 001 Part 5 – Bonded Anchors Part 5 of EOTA (2002) ETAG 001 Guideline for Euro­ pean Technical Approval of Metal Anchors for Use in Concrete addresses bonded anchors. This technical approval docu- ment was created in 2002 and has undergone several amend- ments. EOTA (2002) ETAG 001 Part 5 served as the basis for ICC-ES AC308 which subsequently served as the basis for ACI 355.4 discussed above. A review of EOTA (2002) ETAG 001 Part 5 did not provide any new information than what was already discussed with ACI 355.4. Federation Internationale du Beton Design of Anchorages in Concrete The Federation Internationale du Beton (fib) (2011) Design of Anchorages in Concrete does not provide adhe- sive anchor qualification and quality control requirements,

28 cheaper, and quicker than tests that involve the entire adhe- sive anchor system installed in concrete. It is understood that the interaction of the adhesive with the concrete is an important variable to creep resistance and is essential to be included in the testing. Therefore, it was not reasonable to only test the adhesive alone for the evaluation of short-term and sustained load performance of adhesive anchors in concrete, but such tests were included in the proj- ect since they could possibly serve as: • qualifying or prescreening tests prior to further more expensive/timely testing, • fingerprinting tests to confirm the identity of an adhesive on site, and • comparison tests between adhesives. Time–Temperature Superposition and Master Curves. Time–temperature superposition is the idea that a change in temperature produces the same effect as a change in measure- ment time for a viscoelastic material. This proposal allows the researcher to conduct tests on a sample over a range of temperatures and shift the results along the time axis until According to Klompen et al. (2005), most polymers show a linear relationship between the logarithm of increasing time to failure and decreasing stress; however, some polymers do exhibit a lower bound stress level. While a linear projection would be sufficient and possibly conservative, a manufacturer can perform longer term tests at lower stress levels in order to better define the curve. See Figure 24 for a sample stress versus time-to-failure graph. The test data can also be summarized in a table of estimated failure loads at specified structure lifetimes. NCHRP Report 639 (2009) indicates that the “stress ver- sus time-to-failure” test method provided a viable means for evaluating the sustained load performance of adhesive anchors. This method was adopted as the primary method of assessing a parameter’s influence on sustained load per- formance for this project. A detailed discussion of how this method was implemented is described later. Adhesive-Alone Tests Adhesive-alone tests involve testing the adhesive without the concrete and anchor. This approach could be simpler, - 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000 0.000 0.010 0.020 0.030 0.040 0.050 0.080 0.090 0.100 Displacement (in) Lo ad (lb f) 0 100 200 300 400 500 600 Ti m e (se c) Load Time 0.060 0.070 Figure 23. Load versus displacement and time versus displacement graph for a sample anchor with incremental loading showing time versus displacement response.

29 Time–temperature superposition works well for poly- mers within the linear viscoelastic region where compliance is independent of stress. For materials whose compliance increases as stress increases, time–temperature superposition is not appropriate. Figure 25 is a sample master curve created from stress relaxation data. The left side of the figure shows the stress relaxation data for various temperatures. These curves were then shifted until they lined up and formed the master curve as shown on the right side of the figure. Time–Stress Superposition. Time–stress superposi- tion is another method to create a master curve from several short-term tests at a constant temperature at various stress levels. This approach is more practical for materials not within the linear viscoelastic region where the compliance changes as stress changes. Tests are conducted on a sample at a constant temperature over a range of stress levels. Similar to time–temperature superposition, the curves can be shifted along the time axis to create a master curve at a particular stress level (Figure 26 and Figure 27). For strains in the linear viscoelastic range, the time– temperature superposition principle works well. How- ever, the time–temperature superposition principle only relates the temperature to time and if the strain or stress is large enough to change the speed of the underlying molecular motion mechanism or even alter the mechanism, the predicted time response from only using the time– temperature superposition principle will not be accurate. A few theories tried to address this issue by assuming that there were no changes in the underlying mechanism and only the stress or strain altered the speed. Using such an they superimpose, creating what is called a master curve, thereby providing predictions of the material’s behavior over a broader range of time. Crawford (1998) explains that the glass transition tem- perature (Tg) is usually taken as the reference temperature. If the properties of an adhesive are known at Tg, then the prop- erties at any temperature can be determined. Per Hunston et al. (1980) this relationship is valid for materials with more simple chemistries, but may not be valid for more complex materials. Various ASTM test methods exist for determining the glass transition temperature. ASTM E1356 uses differential scanning calorimetry (DSC) to monitor the heat flow of a specimen as it is heated or cooled through the glass transi- tion region. ASTM E1545 uses thermomechanical analysis to measure the movement of a probe in contact with a specimen as it changes from a vitreous solid to an amorphous liquid while it is heated through its glass transition temperature. ASTM E1640 uses dynamic mechanical analysis to oscillate a specimen at a fixed or resonant frequency and monitors the change in the viscoelastic response as it is heated. The glass transition region is marked by a decrease in storage modulus and an increase in the loss modulus and tan-d. Master curves are a common method of simplifying and presenting data dealing with time-temperature equivalence and can be used to extend the data beyond the testing range. Vuoristo and Kuokkala (2002) conducted creep tests at dif- ferent temperatures and used master curves to predict the behavior of an epoxy used on rolls in the paper making industry by expanding the data by two orders of magnitude. Master curves are also used in ASTM D2990 as an accepted method to predict sustained load properties of plastics. 0% 20% 40% 60% 80% 100% 120% 0.01 0.1 1 10 100 1000 10000 100000 1000000 10000000 TIME (hr) PE R CE NT S TR ES S (% ) Figure 24. Sample stress versus time-to-failure graph.

30 E″, and tan delta curves (respectively) generated by the researchers. Creep Compliance Curves. Creep compliance is defined as the strain due to creep divided by stress. Creep compliance curves are plotted versus time and since the strain is normal- ized by stress, these curves provide an indication of displace- ment versus time and can be used to show a material’s creep deformation properties over time. In the National Institute of Standards and Technology (NIST) study by Chin et al. (2007), creep compliance curves were generated that displayed the predicted creep behavior of two adhesives over time. Fig- ure 30 clearly illustrates that the two adhesives tested are pre- dicted to have different creep properties. Chin et al. (2007) warn that these estimated creep compliance curves are “not approach, a stress or strain shift factor can also be intro- duced. The time–stress superposition shows satisfactory results for a few polymer systems but its validity needs to be verified for each material. Dynamic Mechanical Thermal Analysis Tests. Dynamic mechanical thermal analysis (DMTA) tests take thin samples of an adhesive and subject them to many cycles of a tensile load. Chin et al. (2007) conducted DMTA tests on two adhesives in which tensile strain sweeps were conducted at different temperatures and the test data was used to perform a time–temperature superposition. The storage modulus (E′), loss modulus (E″), and tan delta (E″/E′) were calculated and master curves were generated for both adhesives. Figure 28 and Figure 29 present the E′, Source: Hunston and Chin (2008) Figure 25. Sample master curve using time-temperature superposition. Source: Jazouli et al. (2005) Figure 26. Individual compliance curves used in time–stress superposition. Source: Jazouli et al. (2005) Figure 27. Sample master curve using time– stress superposition.

31 Source: Chin et al. (2007) Figure 28. E’ and E” master curves for an epoxy. Source: Chin et al. (2007) Figure 29. Tan delta master curve for an epoxy. Source: Chin et al. (2007) Figure 30. Creep compliance curve for two epoxies. a substitute for the direct measurement of creep behavior” because they are limited to the linear viscoelastic region and adhesive anchors under sustained loading may function in the nonlinear region, especially as failure is approached. However, they can be valuable as fingerprinting tests or pre- screening tests by which to indicate which adhesives warrant further/more exact testing by manufacturers. CALTRANS Test Method 438. CALTRANS (2006a) Test Method 438 determines rheological properties of adhesives using a dynamic shear rheometer (DSR). This test method, also confined to the linear viscoelastic range as discussed above, cannot be used as a direct measurement of creep per- formance, but might be able to be used as a prescreening test. Tensile Creep Tests. ASTM D2990 (2001) provides the testing procedure for a tensile creep test. Tensile creep tests load small “dogbone” specimens of adhesive using the dimen- sions for Type I or Type II dogbones as specified in ASTM D638 (Figure 31). Two specimens are required for each stress level tested or three specimens if fewer than four stress levels are used. A minimum of three stress levels is recommended for materials that show linear viscoelasticity and at least five stress levels for materials that are significantly affected by stress. ASTM D2990 specifies that the tensile creep specimens are loaded to the given stress level within 5 seconds. Measure- ments of extension, temperature, and humidity are recorded at progressively longer time intervals. ASTM D2990 suggests the following approximate time schedule: 1, 6, 12, and 30 min; 1, 2, 5, 20, 50, 100, 200, 500, 700, and 1,000 hours; and monthly beyond 1,000 hours. The tests are continued until failure. Test series can be conducted under different testing conditions (temperature, humidity, cure time) to evaluate the effect of a parameter on the adhesive’s creep performance. To determine the 100% stress level (mean static strength), static load tests on five specimens are conducted per the procedure specified in ASTM D638. ASTM D638 specifies a constant strain loading rate that produces failure between 30 seconds and 5 minutes. NCHRP Report 639 (2009) recommends that both the static load tests and the sustained load tests must be loaded with the same load transfer duration. It is recommended that the load transfer duration of both the ASTM D638 static load tests and the ASTM D2990 tensile creep tests be set at 2 ± 1 min- utes, as specified for the static load tests and sustained load tests for anchor pullout tests. Design Guidelines and Specifications Related to Adhesive Anchor Systems The review of design guidelines and specifications related to adhesive anchors included national standards, state DOT standards, and international standards. The test methods

32 (c) Pullout strength of cast-in, post-installed expansion or undercut anchor in tension; (d) Concrete side-face blowout strength of a headed anchor in tension; and (e) Bond strength of adhesive anchor in tension. Steel strength of anchor in tension. The nominal strength of the anchor in tension as governed by the steel (Nsa) is determined as: =N A fsa se N uta ACI 318 Eq. D-2, where Ase,N = effective cross section of a single anchor, in.2 and futa = specified tensile strength of anchor steel, psi. and specifications described above generate design values (e.g., bond stress) that are used in the design calculations described below. National Design Guidelines and Specifications ICC-ES AC308 Acceptance Criteria for Post-Installed Adhesive Anchors in Concrete Elements ICC-ES AC308 provides both ASD and LRFD design provi- sions. Only the LRFD method will be addressed in this report. The ICC-ES AC308 LRFD (strength design) method presented in Section 3.3 provided the basis for development of the adhe- sive anchor provisions in ACI 318-11 Appendix D. ACI 355.4 will not include design provisions. The design methodology provided in ICC-ES AC308 will be discussed under ACI 318-11. ACI 318-11 Building Code Requirements for Structural Concrete ACI 318-11 Appendix D addresses anchorage to concrete and recently incorporated the design provisions for adhesive anchors developed by ICC-ES AC308. A general overview of adhesive anchor provisions is presented below. ACI 318-11 Appendix D specifies various strength reduc- tion factors (φ) depending on type of failure, steel element (brittle or ductile), presence of supplementary reinforce- ment, and category as defined by ACI 355.2. The strength reduction factors range from 0.45 to 0.75. Tension. ACI 318 Appendix D considers the following design strengths (failure modes) for anchors in tension that are illustrated in Figure 32: (a) Steel strength of anchor in tension; (b) Concrete breakout strength of anchor in tension; Source: ASTM D638 Figure 31. ASTM D638 Type I and Type II specimens. (a) Steel strength failure (b) Concrete breakout failure (c) Pullout failure (d) Concrete side-face blowout failure (e) Adhesive bond strength failure Source: ACI 318-11 (2011a) Figure 32. ACI 318-11 tension failure modes.

33 Bond strength of adhesive anchor in tension. The nomi- nal bond strength of a single adhesive anchor (Na) or a group of adhesive anchors (Nag) shall not exceed: Single anchor: = ψ ψN A A Na Na Nao ed Na cp Na ba ACI 318 Eq. D-18, , Group of anchors: = ψ ψ ψN A A Nag Na Nao ec Na ed Na cp Na ba ACI 318 Eq. D-19, , , where ANa = projected influence area of a single adhesive anchor or a group of adhesive anchors that can be approx- imated as the base of a rectangle that is resulted by projecting the failure surface out cNa from the centerlines of the anchors in a group of adhesive anchors, in.2; ANao = projected influence area of a single anchor with edge distance equal or greater than cNa, in.2; ( )=A cNao Na2 ACI 318 Eq. D-202 cNa = edge distance required to develop the full bond strength of a single adhesive anchor, in.; = τ C dNa a uncr 10 1,100 ACI 318 Eq. D-21 da = nominal diameter of adhesive anchor, in.; tuncr = characteristic limiting bond stress of adhesive anchor in uncracked concrete, psi; ψec,Na = modification factor for eccentricity of applied loads; ψed,Na = modification factor for edge effects; ψcp,Na = modification factor for anchor in uncracked con- crete without supplementary reinforcement; Nba = the basic bond strength of a single adhesive anchor in tension in cracked concrete; = λ τ piN d hba a cr a ef ACI 318 Eq. D-22 la = modification factor for lightweight concrete; tcr = characteristic limiting bond stress of adhesive anchor in cracked concrete, psi; and hef = effective embedment depth of anchor, in. If test results for tcr from ACI 355.4 are not available, then the values in Table 7 for tcr and tuncr can be used. Note that the values in Table 7 are multiplied by 0.4 if the anchor is subject to sustained tension loading. The value (futa) used in Eq. D-2 must not exceed 1.9fya or 125,000 psi where fya is the specified yield strength of the anchor steel in psi. The 1.9fya limit is to ensure that yielding does not occur under service loads. Concrete breakout strength of anchor in tension. The nominal concrete breakout strength of a single anchor (Ncb) or a group of anchors (Ncbg) shall not exceed: Single anchor: N A A Ncb Nc Nco ed N c N cp N b= ψ ψ ψ, , , ACI 318 Eq. D-3 Group of anchors: = ψ ψ ψ ψN A A N ACI 318 Eq. D-4cbg Nc Nco ec N ed N c N cp N b, , , , where ANc = projected concrete failure area of a single anchor or a group of anchors that can be approximated as the base of a rectangle that is resulted by projecting the failure surface out 1.5hef from the centerlines of the anchor or from the centerlines of the anchors in a group of adhesive anchors, in.2; ANco = projected concrete failure area of a single anchor with edge distance equal or greater than 1.5hef, in.2; =A h9 ACI 318 Eq. D-5Nco ef 2 hef = effective embedment depth of anchor, in.; ψec,N = modification factor for eccentricity of applied loads; ψed,N = modification factor for edge effects; ψc,N = modification factor based on presence or absence of cracking; ψcp,N = modification factor for anchor in uncracked concrete without supplementary reinforcement; Nb = the basic concrete breakout strength of a single anchor in tension in cracked concrete; = λ ′N k f hb c a c ef ACI 318 Eq. D-61.5 kc = 17 for post-installed anchors; = 24 for cast-in-place anchors; la = modification factor for lightweight concrete; and f ′c = specified compressive strength, psi. Pullout strength of cast-in, post-installed expansion or undercut anchor in tension. This failure mode does not apply to adhesive anchors. Concrete side-face blowout strength of a headed anchor in tension. This failure mode does not apply to adhesive anchors.

34 The value (futa) used in Eq. D-29 must not exceed 1.9fya or 125,000 psi where fya is the specified yield strength of the anchor steel in psi. The 1.9fya limit is to ensure that yielding does not occur under service loads. Concrete breakout strength of anchor in shear. The nom- inal concrete breakout strength of a single anchor (Vcb) or a group of anchors (Vcbg) in shear shall not exceed: Single anchor: = ψ ψ ψ ACI 318 Eq. D-30, , ,V A A Vcb Vc Vco ed V c V h V b Group of anchors: = ψ ψ ψ ψ ACI 318 Eq. D-31, , , ,V A A Vcbg Vc Vco ec V ed V c V h V b where AVc = projected area of the failure surface on the side of the concrete member at its edge for a single anchor or group of anchors, in.2; AVco = projected area for a single anchor in a deep member with a distance from the edge greater than 1.5ca1 in the direction of the shear force, in.2; ( )= 4.5 ACI 318 Eq. D-321 2A cVco a ca1 = distance from the center of the anchor to edge of the member in the direction of the shear load, in.; ψec,V = modification factor for eccentricity of applied loads; ψed,V = modification factor for edge effects; ψc,V = modification factor based on presence or absence of cracking; ψh,V = modification factor for anchor in concrete where the member thickness is less than 1.5ca1; Vb = the basic concrete breakout strength of a single anchor in shear in cracked concrete is the lesser of Eq. D-33 and D-34; ( )=       λ ′7 ACI 318 Eq. D-33 0.2 1 1.5V l d d f cb e a a a c a The nominal tension strength (Nn) is then the lesser of Nsa, Ncb, and Na for single adhesive anchors or Nsa, Ncbg, and Nag for a group of adhesive anchors. In addition, per ACI 318-11 D.4.1.2, a 55% limitation is placed on the anchor in a connec- tion that resists the highest sustained load. Shear. ACI 318 Appendix D considers the following design strengths (failure modes) for anchors in shear that are illustrated in Figure 33: (a) Steel strength of anchor in shear, (b) Concrete breakout strength of anchor in shear, and (c) Concrete pry-out strength anchor in shear. Steel strength of anchor in shear. The nominal strength of an anchor in shear as governed by the steel (Vsa) is determined as: = 0.60 ACI 318 Eq. D-29,V A fsa se V uta where Ase,V = effective cross section of a single anchor in shear, in.2 and futa = specified tensile strength of anchor steel, psi. Installation and Service Conditions Moisture Content of Concrete at Time of Anchor Installation Peak In-Service Temperature of Concrete (°F) cr (psi) uncr (psi) Outdoor Dry to fully saturated 175 200 650 Indoor Dry 110 300 1,000 Notes: Where anchor design includes sustained tension loading, multiply values of cr and uncr by 0.4. Where anchor design includes earthquake loads for structures assigned to seismic design category C, D, E, or F, multiply values of cr by 0.8 and uncr by 0.4. Table 7. Characteristic bond stress to use in absence of test results. (a) Steel strength failure (c) Concrete pry-out failure (b) Concrete breakout failures Source: ACI 318-11 (2011a) Figure 33. ACI 318-11 shear failure modes.

35 where Nua = factored tension force, lbs; Nn = nominal strength in tension, lbs; Vua = factored shear force, lbs; Vn = nominal strength in shear, lbs; and j = strength reduction factor. AASHTO LRFD Bridge Design Specifications The AASHTO (2010b) LRFD Bridge Design Specifications was reviewed for a framework of design specifications related to general anchor bolt design within which the epoxy adhe- sive design standards could be incorporated. Article 14.8.3 “Anchorages and Anchor Bolts” presents the design require- ments for anchor bolts. It refers to article 6.13.2.10.2 for the tensile resistance, article 6.13.2.12 for the shear resistance, and article 6.13.2.11 for combined tension–shear resistance. The commentary in C14.8.3.1 refers the designer to ACI 318 Appendix D for global design of anchorages. These three ref- erences to Section 6 only evaluate the resistance of the bolt and do not consider concrete failure. Article 5.7.5 addresses the bearing resistance of the concrete, but there is no provi- sion for concrete breakout and side-face blowout in tension and concrete breakout and pry-out failure in shear. Comparison Between ACI 318-11 and AASHTO LRFD Bridge Design Specifications As AASHTO does not currently have design provisions for ad hesive anchors, one possible solution is to reference ACI 318-11 Appendix D. A comparison of the nominal and factored resistances for similar design situations found in ACI 318-11 and AASHTO (2010b) LRFD Bridge Design Specifications was conducted to determine if any changes to the resistance and/or load factors in ACI for reference by AASHTO. The results are summarized below and more detail can be found in Appendix B. Resistance Factors. For concrete sections, Table 8 lists the resistance factors found in the AASHTO (2010b) LRFD Bridge Design Specifications and ACI 318-11. ( )= λ ′9 ACI 318 Eq. D-341 1.5V f cb a c a le = load-bearing length of anchor, in.; da = nominal diameter of adhesive anchor, in.; la = modification factor for lightweight concrete; and fc′ = specified compressive strength, psi. Concrete pry-out strength of anchor in shear. The nomi- nal pry-out strength of a single anchor (Vcp) or a group of anchors (Vcpg) in shear shall not exceed: Single anchor: = ACI 318 Eq. D-40V k Ncp cp cp Group of anchors: = ACI 318 Eq. D-41V k Ncpg cp cpg where kcp = 1.0 for hef < 2.5 in.; kcp = 2.0 for hef ≥ 2.5 in.; Ncp = basic concrete pry-out strength of a single anchor, lbs; and Ncpg = basic concrete pry-out strength of a group of anchors, lbs. The nominal shear strength (Vn) is then the lesser of Vsa, Vcb, Vcp for single adhesive anchors or Vsa, Vcbg, Vcpg for a group of adhesive anchors. Tension and Shear Interaction. ACI 318-11 Appen- dix D uses a tri-linear approach to tension–shear interaction expressed in the following equation with two conditions: ϕ + ϕ ≤ 1.2 ACI 318 Eq. D-32 N N V V ua n ua n but if Vua ≤ 0.2jVn, then full strength in tension can be used; if Nua ≤ 0.2jNn, then full strength in shear can be used; Factor ACI 318-11 AASHTO Tension-controlled section 0.90 9.3.2.1 0.90 5.5.4.2.1 Compression-controlled sections (Spiral reinforcement) 0.75 9.3.2.2 0.75 5.5.4.2.1 Compression-controlled sections (Tie reinforcement) 0.65 9.3.2.2 0.75 5.5.4.2.1 Shear (normal weight) 0.75 9.3.2.3 0.90 5.5.4.2.1 Shear (lightweight) 0.601 9.3.2.3 0.70 5.5.4.2.1 Bearing 0.65 9.3.2.4 0.70 5.5.4.2.1 Note: Assuming an average value of = 0.80 from ACI 8.6.1, and for comparison with AASHTO, the ACI phi factor reported in this table is = = (0.75)(0.80) = 0.60. Table 8. Comparison of ACI strength reduction factors and AASHTO resistance factors.

36 the researchers’ opinion that if ACI is conservative for the cases examined, then it should be conservative for AASHTO to use ACI design provisions for adhesive anchor design. Load Factors. Load factors found in AASHTO (2010b) of 1.25 for dead loads and 1.75 for live loads are higher than the load factors used in ACI Equation 9.2 of 1.2 for dead loads and 1.6 for live loads. The difference is associated with the design life and importance of the structures. Summary. In every comparable case of factored resistance as determined by ACI and AASHTO, ACI is either the same or more conservative than AASHTO. Additionally, for the load cases considered, AASHTO’s factored loads are higher. There- fore, it seems reasonable that using factored resistances for anchor design from ACI 318-11 with AASHTO load combi- nations will result in a conservative design. As a result, it is both appropriate and conservative for AASHTO to use the factored resistances (design strengths) in ACI 318-11 Appendix D in conjunction with the load fac- tors in AASHTO for adhesive anchor design. There does not appear to be any justification for AASHTO using higher fac- tored resistances than ACI for adhesive anchors. Perhaps a future reliability study could be performed to determine if higher factored resistances could be permitted for adhesive anchors. AASHTO Standard Specifications for Structural Supports for Highway Signs, Luminaires, and Traffic Signals Article 5.17 of AASHTO (2009) Standard Specifications for Structural Supports for Highway Signs, Luminaires, and Traf­ fic Signals was also investigated for a possible anchor bolt design framework in which to incorporate adhesive anchors. AASHTO (2009) provides ASD design guidelines for cast- in-place anchor bolts. No provision is made for adhesive anchors. To ensure a ductile failure, anchor bolts must be designed so that they reach their minimum tensile strength prior to concrete failure. It specifies that the following failure modes be addressed: • Bolt failure, • Load transfer from anchor to concrete, • Concrete tensile strength, • Lateral bursting of concrete, and • Base plate failure. The ASD design provisions in Article 5.17 were adapted from NCHRP Report 469 (2002). AASHTO (2009) mainly addresses the design of the anchor bolt itself and provides ASD equations for allowable tension, compression, and shear stresses on the bolt as well as interaction equations The design provisions per ACI 318-11 and AASHTO (2010b) LRFD Bridge Design Specifications produce identical nominal resistances for flexure, shear, axial compression, and bearing of concrete. As the nominal resistances are identical, the factored resistances (design strengths) will only vary by their resistance factors (strength reduction factors). Table 9 presents the ratio (ACI/AASHTO) of the above factored resistances. For all the cases evaluated, ACI is either identical to or more conservative than AASHTO in determining the nominal and factored resistances. ACI 318-11 determines the factored resistance for anchor bolt steel failure in Appendix H and Table 10 lists the ratio (ACI/AASHTO) of the factored resistance for anchor bolt steel failure in tension. In all cases, ACI is more conservative than AASHTO. The factored resistances for adhesive anchor design cannot be compared between ACI and AASHTO as no design provi- sions for adhesive anchors exist in AASHTO. Therefore, it is Factored Resistance ACI/AASHTO Tension controlled section 1.00 Compression-controlled sections (spiral reinforcement) 1.00 Compression-controlled sections (tie reinforcement) 0.87 Shear (normal weight) 0.83 Shear (lightweight) 0.86 Bearing 0.93 Table 9. Ratio of factored resistance determined by ACI (2011a) to AASHTO (2010b). Bolt Diameter (in.) Nsa(ACI)/ Tn(AASHTO) 0.91 ¾ 0.93 0.95 1 0.95 1 0.95 1 ¼ 0.97 1 0.96 1 ½ 0.98 1 ¾ 0.97 2 0.98 Notes: The resistance factors used are = 0.75 for ACI design equation and = 0.80 for T n = nominal resistance of a bolt in tension. Fub = specified minimum tensile strength of a bolt. Fub = futa. AASHTO design equation. Table 10. Ratio of factored resistance for steel strength in tension as computed by ACI (2011a) and AASHTO (2010b) for ductile steel elements.

37 adhesive anchors. FDOT does not allow adhesive anchors for overhead or upwardly inclined holes. Furthermore, adhesive anchors are not allowed for loading conditions with a pre- dominately sustained load. A predominately sustained load is defined as a load where the permanent portion of the fac- tored tension load exceeds 30%. The reduction factors (f) specified by FDOT are as follows: fc = 0.85 for adhesive anchors controlled by concrete embedment, fc = 1.00 for extreme events, and fs = 0.90 for adhesive anchors controlled by anchor steel. The following design requirements are specified by FDOT: Tension. FDOT considers the following design strengths for anchors in tension: Tensile strength controlled by anchor steel. The design tension strength controlled by anchor steel (jNs) is defined as: ϕ = ϕN A f FDOT Eq.1-2s s e y where js = strength reduction factor; Ae = effective tensile stress area of steel anchor (may be 75% of gross area for threaded anchors), in.2; and fy = minimum specified yield strength of steel, ksi. Tensile strength controlled by adhesive bond. The design tension strength controlled by adhesive bond (jNc) is defined as: ϕ = ϕ Ψ Ψ FDOT Eq.1-3N Nc c e gn o where jc = strength reduction factor; Ye = modification factor to account for edge distances; Ygn = modification factor for groups; No = nominal tensile strength of adhesive bond, kips; = ′ pi FDOT Eq.1-4N T d ho e T′ = nominal bond strength of adhesive product, ksi: T′ = 1.08 ksi for type V and HV adhesive product on FDOT QPL; T′ = 1.83 ksi for type HSHV adhesive product on FDOT QPL; d = nominal diameter of adhesive anchor, in.; and he = anchor embedment depth, in. The design tension strength (jNn) is the smaller of jNs and jNc. for combined tension and shear and combined compression and shear. Bending stresses are considered for double-nut anchor bolt connections if the clearance between the bottom of the leveling nuts and the top of the concrete exceeds one bolt diameter. While outside of the scope of AASHTO (2009), it recom- mends other design considerations such as: • Block shear rupture, • Shear lag, • Prying action, and • Base plate stiffness. NCHRP Report 469: Fatigue-Resistant Design of Cantilevered Signal, Sign, and Light Supports NCHRP Report 469: Fatigue­Resistant Design of Cantilev­ ered Signal, Sign, and Light Supports (2002) includes a “Rec- ommended Anchor Rod Specification and Commentary” as Appendix A of the report. The specification is for the design, installation, and inspection of cast-in-place anchor rods and does not cover post-installed anchors, but allows them as “alternative design anchors.” NCHRP (2002) uses an LRFD approach and designs anchor bolts for tensile strength, compressive strength, shear strength, combined tension and shear, bearing at anchor rod shear holes, and tensile fatigue. The American Concrete Institute (ACI) is referred to for concrete design. State DOT Design Guidelines and Specifications A review was made of design guidelines and specifications from the various state departments of transportation listed earlier. NYSDOT Bridge Design Manual NYSDOT (2008a) Bridge Manual refers to Section 14.8.3 of AASHTO (2010b) LRFD Bridge Design Specifications for the design of anchor bolts. Section 6.8.5.4 of NYSDOT (2008a) allows for post-installed grouted anchors that are allowed for rehabilitation projects and recommends that proof-load tests be conducted. It further recommends an embedment depth of 12″ for 1″ diameter bolts. FDOT Structures Manual Section 1.6 of volume 1 of the FDOT (2009) Structures Design Guidelines provides FDOT’s design guidelines for

38 In this design approach, the anchor bolts are only designed for shear. The number of anchor bolts is determined by dividing the base shear at the bearing by the allowable shear force per anchor. When soil conditions are poor or it is not possible to enlarge seat lengths, the anchor bolts must be designed to remain elas- tic during the seismic event. In this situation, the anchor bolts must be designed for combined shear and tension. The only specific reference to epoxy anchor bolts is made in Section 3.7.4, which address the conversion of an existing abutment into a semi-integral abutment. Pennsylvania Department of Transportation Design Manual Part 4—Structures. Pennsylvania Department of Transportation (PENNDOT) (2007) Design Manual Part 4— Structures Section 3.6.4.9 refers to sections 5.17 and 5.12 of AASHTO (2009) Standard Specifications for Structural Sup­ ports for Highway Signs, Luminaries, and Traffic Signals for the design of anchor bolts. PENNDOT (2007) has restrictions on the use of adhe- sive anchors. Section 3.6.8 addresses adhesive anchor design in general and states that adhesive anchors are not allowed in tension applications for permanent installations. Sec- tion 3.6.4.9 addresses the anchor bolt design for sound bar- rier walls and specifically does not permit adhesive anchors. Section 5.5 pertains to bridge rehabilitation strategies and provides a detail in Section 5.5.2.4 (Figure 5.5.2.4-4) for a repair of expansion dams using adhesive anchors for cases in which the bolts were sheared off. WSDOT Bridge Design Manual Section 10.1.2 “Bridge Mounted Signs” of WSDOT (2008a) Bridge Design Manual specifically mentions using resin bonded anchors in new and existing structures. The anchors must be installed per the manufacturer’s specifications in dry concrete and the nuts must be torqued to the proof load. MDOT Bridge Design Manual Section 7.06.02 of the MDOT (2005) Bridge Design Manual provides design guidelines for bonded anchors. The embed- ment depth for A307 bolts is nine times the nominal anchor diameter. Bonded anchors are designed for tension and shear, but only steel failure is addressed. The allowable tension and shear loads are defined as follows: Allowable tensile load = 1.25 f A FS y T Allowable shear load = 0.30 fy AT where: AT = tensile stress area (net section through threads), fy = yield strength, and FS = factor of safety = 4. Shear. FDOT considers the following design strengths for anchors in shear: Shear strength controlled by anchor steel. The design shear strength controlled by anchor steel (jVs) is defined as: ϕ = ϕ 0.7 FDOT Eq.1-7V A fs s e y where js = strength reduction factor; Ae = effective tensile stress area of steel anchor (may be 75% of gross area for threaded anchors), in.2; and fy = minimum specified yield strength of steel, ksi. Shear strength controlled by concrete breakout. The design shear strength controlled by concrete breakout (jVC) is defined as: ϕ = ϕ Ψ ′0.4534 FDOT Eq.1-81.5V c fc c gv c where jc = strength reduction factor; Ygv = modification factor for groups; c = anchor edge distance from center of anchor to free edge, in.; and f ′c = minimum specified compressive strength of concrete, ksi. The design shear strength (jVn) is the smaller of jVS and jVC. Tension and Shear Interaction. FDOT uses a linear approach to tension–shear interaction expressed in the fol- lowing equation: ϕ + ϕ ≤ 1.0 FDOT Eq.1-10 N N V V u n u n where: Nu = factored tension load, kips; Nn = design tension strength, kips; Vu = factored shear load, kips; Vn = design shear strength, kips; and j = strength reduction factor. IDOT Bridge Manual The only reference to anchor bolt design in IDOT (2008) Bridge Manual was found in Section 3.7.3, which addresses seismic design of bridge bearings. The design approach taken by IDOT for bridge bearing during seismic events is to pre- vent the loss of span. Loss of span is prevented by adequately detailing seat widths and span lengths. Connection elements or anchor bolts are designed to fail at a certain level of acceleration. When the connection elements fail, the bearing seat width or span length must be large enough to prevent loss of span.

39 shear loading. The design methodology is only applicable for anchors with a predominate static load. Part 3 presents both an elastic and a plastic design procedure summarized in two flowcharts. Most of the calculations are similar to those pre- sented by ACI 318-11. Tension. The design procedure for tensile resistance evalu- ates the following resistances. Steel resistance. The equation to calculate steel resistance is similar to ACI 318 Eq. D-3. Concrete pullout resistance. The characteristic resistance of a combined pullout and cone failure (NRk,p) is as follows: = Ψ Ψ Ψ Ψ Ψ fib Eq.16.2-1, ,0 , , , , ,N NRk p Rk p A Np s Np g Np ec Np re Np where: N0Rk,p = characteristic bond resistance similar to ACI 318- 11 Eq. D-22; YA,Np = modification factor due to geometric effects, comparable to the A A Na Nao factor in ACI 318-11 Appendix D; Ys,Np = modification factor due to edge effects, compa- rable to Yed,N in ACI 318-11 Appendix D; Yg,Np = modification factor accounting for the failure surface of groups: Often this is neglected for simplification; Yec,Np = modification factor due to eccentricity effects in groups, comparable to Yec,Na in ACI 318-11 Appendix D; and Yre,Np = modification factor due to shell spalling in cases of low embedment depth and closely spaced reinforcement. Concrete cone resistance. The characteristic resistance of an anchor or group of anchors due to cone failure (NRk,c) is as follows: = Ψ Ψ Ψ Ψ fib Eq.10.1-2, ,0 , , , ,N NRk c Rk c A N s N ec N re N where: N0Rk,c = characteristic resistance similar to ACI 318-11 Eq. D-6; YA,N = modification factor due to geometric effects, comparable to the A A Nc Nco factor in ACI 318-11 Appendix D; Ys,N = modification factor due to edge effects, comparable to Yed,N in ACI 318-11 Appendix D; Yec,N = modification factor due to eccentricity effects in groups, comparable to Yec,N in ACI 318-11 Appen- dix D; and Yre,N = modification factor due to shell spalling in cases of low embedment depth and closely spaced reinforcement. Adhesive anchors are specifically prohibited in overhead applications with a sustained tension load. MDOT Moratorium on the Use of Adhesive Anchors in Sustained Tensile-Load-Only Overhead Applications MDOT (2008) Bureau of Highway Instructional Memoran- dum 2008-07 “Moratorium on the Use of Adhesive Anchors in Sustained Tensile-Load-Only Overhead Applications” imposed a moratorium on overhead sustained tension loading applications. VDOT IIM-S&B-40.2 Sound Barrier Wall Attachments VDOT (2007b) memorandum IIM-S&B-40.2 “Sound Bar- rier Wall Attachments” prohibits using adhesive anchors in attaching structure mounted walls. VDOT IIM-S&B-76.2 Adhesive Anchors for Structural Applications VDOT (2008a) memorandum IIM-S&B-76.2 “Adhesive Anchors for Structural Applications” limits the use of adhe- sive anchors to shear loading only. It specifically prohibits using adhesive anchors in applications of sustained, cyclical, and fatigue tension loadings. International Design Guidelines and Specifications Two international design guidelines were reviewed for anchor design provisions. EOTA ETAG 001 Annex C – Design Methods for Anchorages Annex C of EOTA (1997b) ETAG 001 Guideline for Euro­ pean Technical Approval of Metal Anchors for Use in Concrete presents a design methodology for bonded anchors. This tech- nical approval document was created in 1997 and has under- gone several amendments. EOTA (1997b) ETAG 001 Annex C served as the basis for the adhesive anchor provisions in ACI 318-11 Appendix D. A review of EOTA (1997b) ETAG 001 Annex C did not produce any new information than what was already discussed with ACI 318-11 Appendix D. fib Design of Anchorages in Concrete The fib (2011) Design of Anchorages in Concrete has design methodology for adhesive anchors subject to tension and

40 (MPII) were also reviewed for an understanding of what is typically required in adhesive anchor installations. National Quality Assurance and Construction Specifications ICC-ES AC308 Acceptance Criteria for Post-Installed Adhesive Anchors in Concrete Elements Section 14 of ICC-ES AC308 includes quality assurance guidelines for the inspector of adhesive anchor installa- tions. Since ACI 355.4 was developed from ICC-ES AC308, the provisions set forth in ICC-ES AC308 will not be discussed, but will be addressed under the discussion of ACI 355.4. ACI 355.4 Qualification of Post-Installed Adhesive Anchors in Concrete ACI 355.4 presents a quality assurance program for the inspector of post-installed adhesive anchors. Section 13 of ACI 355.4 specifies the quality assurance requirements. Man- ufacturers must have an approved quality assurance program with a quality control manual for each product. Manufac- turers must undergo unannounced inspections according to the requirements of ISO/IEC 17011 by an inspection agency under ISO/IEC 17020. Manufacturers must supply inspec- tion manuals for each product and anchors must be installed with special inspection in accordance with the building code and ACI 355.4. When required, continuous special inspection shall be conducted in which all aspects of the installation must be inspected by an inspector. However, holes can be drilled without an inspector present as long as the inspector inspects the drill bit and verifies the hole sizes. The following must be verified: • Hole drilling method in accordance with the manufac- turer’s specifications; • Hole location, diameter, and depth; • Hole cleaning per the manufacturer’s specifications; • Anchor type, material, diameter, and length; • Adhesive identification and expiration date; and • Installation in accordance with the manufacturer’s speci- fications. When required, periodic special inspections shall be con- ducted in which the inspector inspects all aspects listed above for each anchor type for the same construction personnel. Only the initial installation needs to be inspected and the rest can be installed without the inspector as long as the same Concrete splitting. The characteristic resistance of an anchor or group of anchors due to splitting failure is cal- culated using fib Eq. 10.1-2 with an additional modifica- tion factor (Yh,sp) to account for the influence of member thickness. Shear. The design procedure for shear resistance evalu- ates the following resistances: Steel resistance. The equation to calculate steel resistance is similar to ACI 318 Eq. D-29 except it specifies a constant of 0.5 instead of 0.6. Concrete pry-out resistance. The equation to calculate concrete pry-out resistance is similar to ACI 318 Eq. D-40 and Eq. D-41. Concrete edge resistance. The characteristic resistance of an anchor or group of anchors close to an edge (VRk,c) is as follows: = Ψ Ψ Ψ Ψ Ψ Ψα fib Eq.10.2-5, ,0 , , , , , ,V VRk c Rk c A V h V s V ec V V re V where V0Rk,c = characteristic resistance similar to ACI 318-11 Eq. D-33; YA,V = modification factor due to geometric effects, comparable to the A A Vc Vco factor in ACI 318-11 Appen- dix D; Yh,V = modification factor due to edge effects, comparable to Yh,V in ACI 318-11 Appendix D; Ys,V = modification factor due to edge effects, comparable to Yed,V in ACI 318-11 Appendix D; Yec,V = modification factor due to eccentricity effects in groups, comparable to Yec,V in ACI 318-11 Appen- dix D; Ya,V = modification factor to take into account the angle of the applied load; and Yre,V = modification factor due to type of edge reinforce- ment used. Tension and Shear Interaction. The fib (2011) Design of Anchorages in Concrete uses a tri-linear approach to tension–shear interaction similar to that found in ACI 318-11 Appendix D. Quality Assurance Guidelines and Construction Specifications Related to Adhesive Anchor Systems The review of quality assurance guidelines and construc- tion specifications related to adhesive anchors included national standards, state DOT standards, and international standards. Manufacturer’s printed installation instructions

41 provides guidance for installation and construction inspec- tion. While this report only addresses cast-in-place anchors and not adhesive anchors, most of the information covers the casting of anchor bolts in concrete and tightening of the anchor bolt following concrete curing. The specification provides guidance for straightening a misaligned bolt. The maximum misalignment allowed is 1:40 from vertical. If an anchor bolt does not exceed a misalign- ment of 1:20 from vertical it can be straightened by hitting it with a hammer or bending it with a jack or pipe. State DOT Quality Assurance and Construction Specifications A review was made of quality assurance and construction specifications from various state departments of transportation. CALTRANS Standard Specifications Section 75 of CALTRANS (2006b) Standard Specifica­ tions refers to the manufacturer’s specifications for installa- tion requirements. Anchors must be installed such that the equipment attached to it will bear firmly against the con- crete. If there is no mention in the manufacturer’s instruc- tions regarding the installation torque, the anchors should be torqued to the values listed in Table 11. It should be noted that using the torque values proposed by CALTRANS gener- ates different stress levels in the resin adhesive, and therefore the commensurate loads associated with these thread stresses vary substantially and do not reflect uniform conditions in the fastener or adhesive. TxDOT Standard Specifications for Construction and Maintenance of Highways, Streets, and Bridges TxDOT addresses the installation of anchor bolts in Section 420.4 of TxDOT (2004) Standard Specifications product is installed by the same personnel. For long construc- tion projects, the inspector should regularly verify that the adhesive product is being installed correctly. When required, a proof loading program should be con- ducted which includes the following: • frequency of proof loading based on anchor type, diam- eter, and embedment depth; • proof loads by anchor type, diameter, and location; • acceptable displacement at proof load; and • action taken to remediate a case of excessive displacement or the failure to achieve the proof load. Proof-load tests should be confined tension tests with the proof load not exceeding 50% of the expected ultimate load based on adhesive bond strength nor 80% of the anchor yield strength. AASHTO LRFD Bridge Construction Specifications Section 18.9 of AASHTO (2010a) LRFD Bridge Construc­ tion Specifications addresses anchor bolts for bearing devices and only references cast-in or grouted anchor bolts. Section 29 specifically mentions adhesive anchors and requires that they be prequalified by universal test standards. The user is encouraged to follow the MPII for drilling and only allows core drilling if it is specified by manufacturer or anchors have been tested in core drilled holes. Core drills are allowed to cut rebar. The user is referred to the MPII for proper cleaning procedures. This section provides for sacrificial and proof-load testing and guidance on torquing of the anchor bolts. AASHTO Standard Specifications for Structural Supports for Highway Signs, Luminaires, and Traffic Signals Section 5.17.5 of AASHTO (2009) Standard Specifications for Structural Supports for Highway Signs, Luminaires, and Traffic Signals specifies that anchor bolts must be installed with sufficient length, cover, and anchorage in concrete to ensure a ductile failure. Additionally, AASHTO (2009) places a limit on misalignment of 1:40 from vertical for anchor bolt installation. NCHRP Report 469 Fatigue-Resistant Design of Cantilevered Signal, Sign, and Light Supports NCHRP Report 469: Fatigue­Resistant Design of Canti­ levered Signal, Sign, and Light Supports (2002) includes a “Recommended Anchor Rod Specification and Com- mentary” as Appendix A of the report. The specification Stud Diameter (inches) Resin Capsule Anchors (foot-pounds) 1 ¼ 400 1 230 175 ¾ 150 75 ½ 30 18 ¼ - Table 11. CALTRANS installation torque values.

42 (2007) Standard Specifications for Road and Bridge Construc­ tion that the installation of adhesive anchors and the equip- ment used for the installation must be in accordance with the manufacturer’s specifications. FDOT only allows an adhesive anchor product that meets Section 937 of FDOT (2007) Stan­ dard Specifications for Road and Bridge Construction and is included in the qualified products list (QPL) maintained by the state. The following requirements pertain to the installation of adhesive anchors: • Install in structurally sound concrete member free of cracks in the area of the anchor; • Use a rotary hammer drill and carbide bit unless otherwise specified by the manufacturer; • The hole diameter must be greater than 105% and less than 150% of the anchor diameter; • Clean the hole according to the manufacturer’s require- ments, but at a minimum blow with compressed air, then brush, and blow again with compressed air; • Use only a non-metallic brush to prevent polishing the hole; • Follow the manufacturer’s requirements regarding limits on anchor position, dampness, ambient temperature, and curing time; and • Fill the hole with the adhesive such that it is within ¼″ of concrete surface after placement of anchor. IDOT Standard Specifications for Road and Bridge Construction Section 509.06 “Setting Anchor Rods” of IDOT (2007b) Standard Specifications for Road and Bridge Construction requires that the holes be drilled to the diameter and depth specified by the manufacturer. The rods should be set with capsule or cartridge systems previously approved by the state and installed per the manufacturer’s instructions. Section 521.06 “Anchor Bolts, Rods, and Side Retainers” of IDOT (2007b) Standard Specifications for Road and Bridge Construction requires verification of the holes for depth and diameter prior to installation. Holes are required to be kept dry and to be blown clean prior to installation. Following installation, the top of the bolt shall be measured in order to determine proper embedment. The anchor bolts should allow for ½″ to 2″ above the top of the nut. WSDOT Construction Manual Section 6-3.2C “Use of Epoxy Resins” of the WSDOT (2009) Construction Manual warns the user against viewing epoxy resins as a cure-all for bonding applications due to their inherent limitations. Specific caution is mentioned regarding for Construction and Maintenance of Highways, Streets, and Bridges. For epoxy installations it specifies a hole diameter of 1/16” to 1/4” greater than the anchor diameter. For prepackaged systems, it requires that the manufacturer’s cleaning instruc- tions be followed exactly. A procedure must be established for the cleaning and preparation of the holes, which includes cleaning the holes of loose material, grease, oil, and other substances. The holes should be blown with filtered com- pressed air and be in a dry condition prior to installation. The space between the anchor and the sides of the hole must be completely filled with adhesive. Section 420.4 specifies a Type III adhesive per TxDOT (2007a) DMS-6100 for neat epoxies and Type VIII for epoxy grout. NYSDOT Standard Specifications Section 586-2.01 “Drilling and Grouting Bolts” of NYSDOT (2008c) Standard Specifications restricts the use of adhesive anchors in overhead installations or for applications with a sustained tensile load. Section 586-3.01 “Drilling and Grouting Bolts” of NYSDOT (2008c) specifies the installation requirements for adhesive anchors. A rotary impact drill should be used but if reinforce- ment is encountered during drilling, a core drill can be used only to cut the rebar, and the rotary impact drill used for the remainder of the drilling. Lubricants cannot be used during drilling and drilling should not cause damage to concrete. Prior to installation, the holes must be dry and clean of loose material. The bolts should be inserted the full depth of the hole and jiggled to ensure complete coverage by the adhesive. Excess adhesive should be struck-off flush with the surface. Horizon- tal installations are allowed and care should be taken to ensure that the adhesive does not run out of the hole. Section 586-3.02 “Pullout Testing” of NYSDOT (2008c) specifies the requirements for pullout testing. A table is pro- vided in the specification to determine the number of anchors to test depending on lot size. The load applied should not exceed 90% of the ASTM proof load (ASTM A568 for anchor bolts and ASTM A615 for reinforcing bars) or 90% of the anchor yield strength if the ASTM proof load is not given. Once the test load is reached, the test is stopped. Anchors pass the test if they can attain the load without permanently displacing. Section 586 of NYSDOT (2008c) includes the changes addressed by NYSDOT (2008b) Engineering Instruction EI 08-012 discussed earlier. FDOT Standard Specifications for Road and Bridge Construction FDOT specifies in Section 416 “Installing Adhesive-Bonded Anchors and Dowels for Structural Applications” of FDOT

43 and the products chosen were not necessarily what were used in the testing program of this research project. Manufacturer’s specifications contain information regard- ing storage conditions (temperature and humidity ranges) and warnings to check that the expiration date has not passed prior to installation. There are many similarities amongst manufacturer’s instal- lation instructions. Most include instructions on cleaning the hole which can include blowing with compressed air and brushing. There is a procedure to confirm that the adhesive is thoroughly mixed, by number of squeezes of the applicator or by visually inspecting the color of the adhesive. Addition- ally, there is a process for injecting the adhesive in the hole to avoid air voids. And finally there are instructions for inserting the anchor. Manufacturer X This product is an epoxy resin with quartz and titanium dioxide. The hole is prepared by drilling to the proper depth with a drill bit 1/8” larger than the anchor diameter. The hole is then blown out using a nozzle and 80 psi (minimum) oil-free compressed air for four seconds. The hole is then brushed up and down four times with a nylon brush. And finally the hole is blown for another four seconds with compressed air. The adhesive is discharged using a cartridge and self- mixing nozzle. Initially the adhesive is discharged to the side until the discharge has a uniform color signifying complete mixing. The hole is filled by inserting the nozzle to the bottom and discharging the adhesive. The nozzle is extracted as the hole fills in order to avoid the formation of air voids. The hole is filled to ½ to 2/3 full in dry and damp holes and completely full in water-filled holes in order to remove all the water. The clean oil-free anchor is installed in the hole while slowly turn- ing it until it contacts the bottom of the hole. The anchor should not be disturbed until the adhesive has fully cured. Horizontal and overhead installations are allowed and the installation is the same, except that a retaining cap is placed over the hole to keep the adhesive within the hole. Manufacturer Y This product is an epoxy resin with an amine hardener. The hole is drilled with a rotary hammer drill and a drill bit confirming to American National Standards Institute (ANSI) B212.15.1994. The drill bit diameter is equal to the rod diam- eter plus 1/16” for anchor diameters of 3/8” and ½” or the rod diameter plus 1/8” for anchor diameters of 5/8” and above. The manufacturer limits anchors of 5/8” and above to horizontal and downward installations only. using epoxy resins below 50°F (10°C). Several guidelines are provided for the inspector of epoxy-resin systems: • Epoxy resin must be completely mixed, • Verify the temperature and/or moisture limitations of the epoxy resin, • Area should be cleaned and prepared according to the manufacturer’s specifications prior to installation, and • The epoxy should completely fill the space around the anchor. The material portion of WSDOT (2009) includes Section 9-4.60, which addresses “Epoxy Systems” and Section 9-4.61 for “Resin Bonded Anchors.” Section 9-4.61 refers to a quali- fied products list (QPL) maintained by the state for material approval. If a resin bonded anchor system is not on the QPL, test results from ASTM E488 and manufacturer’s certificate of compliance can be submitted for approval. MDOT Standard Specifications for Construction Section 712 “Bridge Rehabilitation—Concrete” of MDOT (2003) Standard Specifications for Construction allows the installation of adhesive anchoring of bars in vertical and horizontal applications. International Quality Assurance and Construction Specifications The following summarizes the review of international standards related to quality assurance guidelines and con- struction specifications. EOTA ETAG 001 Part 1 of EOTA (1997a) ETAG 001 does not provide much information regarding quality assurance or construction specifications. Reference is made to the manufacturer’s instal- lation requirements, but limits the installation to a tempera- ture range of 23°F to 104°F (−5°C to 40°C). Part 5 of EOTA (2002) ETAG 001 allows for installation in dry, wet, and flooded holes. It also specifies that the holes are to be drilled as specified by the manufacturer. Manufacturer’s Installation Recommendations Due to the fact that many specifications refer to the manu- facturer’s printed installation instructions (MPII), the instal- lation requirements from three different products (names withheld) have been included to serve as a reference for what is typically specified by manufacturers. The manufacturers

44 water, brushed, and then flushed with water again. The stand- ing water must be removed prior to inserting the adhesive. Depending on the size of the cartridge, the adhesive from the first two or three trigger pulls are discarded [four trigger pulls are discarded if the temperature is below 41°F (5°C)]. The adhesive is inserted in the hole without forming air pock- ets to ½ to 2/3 full. The anchor rod is inserted while twisting and can be adjusted during the specified gel time. The anchor should not be disturbed between the gel time and the cure time. Summary This chapter summarized the findings from the literature review, which investigated the behavior of adhesive anchor systems as well as test methods and material specifications, design guidelines and specifications, and quality assurance guidelines and construction specifications related to adhe- sive anchors in concrete. Extensive investigation and research was involved in the development of the documents reviewed in this chapter, which provided a solid base upon which to develop specifications for AASHTO. One of the significant limitations of the state-of-the-art in adhesive anchors is the effect of various installation and in-service parameters on the sustained load performance of adhesive anchors. Most state DOTs rely on tests that emphasize short-term tests that consider conditions (wetness, time of cure, application of lubricating oils and salt water, etc.); placement and pullout to failure; or a specified acceptable deformation. However, creep rates are not specifically addressed in state DOT tests, although they are very important with respect to tempera- ture, longevity, and the ability to carry loads. The next chapter presents the test program developed to investigate those effects for inclusion of adhesive anchors into the AASHTO specifications. The hole is cleaned with 50 psi to 100 psi compressed air starting at the bottom using a nozzle and oscillating the noz- zle in and out of the hole four times for a total of four sec- onds. If the hole is filled with water or sludge, the hole can be cleaned with pressurized water. The hole is then cleaned with a brush by inserting the brush into the hole in a clockwise fashion. The brush is turned one complete revolution for each ½” of depth. Once the brush has reached the bottom, the brush is turned four complete times. The brush is then removed from the hole by rotating it one complete revolution for every ½” of depth. Alternatively, the brush can be attached to a drill. The hole is then blown with compressed air or flushed with pressurized water as before. The adhesive is discharged and discarded from the car- tridge tool until the adhesive is of a uniform color. The nozzle is inserted to the bottom of the hole and slowly pulled out while discharging in a circular motion maintaining the tip of the nozzle under the level of the adhesive. The hole is filled to 60% full. For holes underwater, the hole is filled entirely with adhesive thereby displacing all the water. The concrete must be between 50°F (10°C) and 110°F (43°C) during installation. The anchor is inserted in a counterclockwise motion and jiggled to remove air pockets. The anchor must not be dis- turbed during working time until the cure time has elapsed. Manufacturer Z This product is a hybrid with methacrylate hardener, cementitious material, and quartz filler. The hole is drilled with a carbide bit to the proper depth and diameter. The hole is then cleaned with 80 psi compressed air using a nozzle inserted to the bottom of the hole. The hole is cleaned three times with a wire brush that is twisted while inserting. The hole is then blown with compressed air again. For holes with standing water, the hole must be flushed with