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.

C-1 A P P E N D I X C Anchor Pullout Tests—University of Florida

C-2 ANCHOR PULLOUT TESTS—UNIVERSITY OF FLORIDA Test apparatus This section describes the test apparatus used for the short-term and sustained load (creep) tests. For each case, the “standard” apparatus is described that was used in the majority of the test series and variations for specific test series are discussed later. Short-Term Test Apparatus The short-term confined testing apparatus conformed to the requirements in ASTM E488. The testing apparatus for the short-term test (Figure 1) used a 6” x 6” x 0.03” thick Teflon PTFE (Polytetrafluoroethylene) confining sheet placed under an 8” x 8” x 5/8” thick steel confining plate. The confining sheet was used to correct for any surface irregularities in the concrete. A 1- 1/4” hole was drilled though the center of the confining sheet and confining plate to fit around the anchor in accordance with ASTM E488. Two 3” x 5” x 1/4” rectangular steel tubes 8” long were placed parallel to each other on either side of the anchor. A 10” x 10” x 1” thick steel plate with a 2-3/4” diameter hole in the center was placed on the rectangular steel tubes to support an Enerpac model RCH-603 Holl-O-Cylinder hydraulic ram (60 ton). A Houston Scientific Model 3500 100-kip load cell was placed on top of the ram sandwiched between four 3” x 3” x 1/4” square plates (two above and two below) with a 1-1/8” diameter hole in the center. A washer and a nut were placed above the square plates. The 5/8” diameter anchor was fed through an 11/16” diameter hole in a non-rigid coupler and secured with a nut. The oversized hole in the coupler prevented bending forces from being transferred from the coupling rod to the anchor. A 1” diameter loading rod was threaded into a hole in the top of the coupler and passed through the ram and load cell and was secured at the top with a washer and two nuts. A 2” x 16” x ¼” steel flat bar was welded to the bottom of the coupler and BEI Duncan Electronics model 9610 linear motion position sensors (linear-pots) were secured to each end of the flat bar equidistant from the center line of the anchor. The linear-pots were oriented downwards and measured displacement between the flat bar and the surface of the concrete. The linear-pots were oriented in this manner so that as the flat bar raised, the plunger extended, ensuring that the linear-pot was not damaged if the anchor failed drastically. A 2” x 2” x ¾” steel

C-3 baseplate was placed on top of the concrete surface underneath each linear-pot plunger to raise the initial bearing point of the plunger and to provide a smooth measuring surface. Figure 1: Short-term confined test apparatus. Test Series 16 (Unconfined) Short-term Test Apparatus The short-term confined testing apparatus conformed to the requirements in ASTM E488. The 6” x 6” x 0.03” thick Teflon PTFE (Polytetrafluoroethylene) confining sheet and the 8” x 8” x 5/8” thick steel confining plate were not used in this test series. The 3” x 5” x 1/4” rectangular steel tubes were placed parallel to each other on either side of the anchor no closer than two times the embedment depth. An 18” x 18” x 1” thick steel plate with a 2-3/4” diameter hole in the center was placed on the rectangular steel tubes to support an Enerpac model RCH-603 Holl- O-Cylinder hydraulic ram (60 ton). Figure 2 shows the modifications to the short-term test apparatus for test series 16.

C-4 Figure 2: Test series 16 short-term unconfined test apparatus. Standard Sustained Load (Creep) Test Apparatus The sustained load confined testing apparatus conformed to the requirements in ASTM E488 and ASTM E1512. Other testing apparatus exist (e.g. cantilevered dead load testing machines) and can be used as long as they conform to the requirements in ASTM E488 and ASTM E1512. The testing apparatus for the sustained load (creep) test (Figure 3) used the same Teflon PTFE (Polytetrafluoroethylene) confining sheet and steel confining plate as in the short- term load test apparatus. Existing steel frames from previous sustained load tests conducted at the University of Florida by Cook et al. (1996) were used to contain compression springs to apply the sustained load. Springs were chosen instead of a hydraulic ram for these sustained load tests in order to reduce the chance of loss of load caused by a hydraulic leak. The springs used were provided by the Florida Department of Transportation (FDOT) State Materials Office in Gainesville, Florida. Two sets of steel wire springs (large and small) were used individually or in parallel. The large springs were approximately 5.5” in diameter by 8” in uncompressed height and had an average approximate spring stiffness of 10.2 kips/in. and a working load range up to 16 kips. The small springs were approximately 3” in diameter by 8” in uncompressed heights and had an average spring stiffness of 3.2 kips/in. and a working load range up to 5 kips. For loads up to 15 kips, the large springs were used individually, for loads

C-5 between 16 kips and 21 kips the large and small springs were used in parallel with an average combined spring stiffness of 13.4 kips/in. The 5/8” diameter anchor was connected to the 1” diameter loading rod by means of the same non-rigid coupler as in the static load test apparatus. Linear-pots were used to measure displacement in the same configuration as in the static load test apparatus. Figure 3: Sustained load (creep) confined test apparatus. A hydraulic jack chair of four parallel Central Hydraulics Model 95979 20-kip rams with a 7/16” throw was used in order to smoothly and quickly apply the sustained loavd to the anchor. During loading, a load cell on top of the hydraulic jack chair measured the transfer of force from the spring to the anchor. Once the desired load was achieved, a nut was tightened on top of the spring below the hydraulic jack chair and the pressure in the rams was released. The use of the hydraulic jack chair allowed for one load cell to be used for all tests. A test frame with the hydraulic jack chair is shown in Figure 4.

C-6 Figure 4: Test frame with hydraulic jack chair. To prevent the test apparatus from falling over due to the dynamic load on the frame caused by an anchor pullout, the test frames were secured to the concrete blocks with loading straps. Test Series 16 (Unconfined) Sustained Load Test Apparatus The sustained load unconfined testing apparatus conformed to the requirements in ASTM E488 and ASTM E1512. The 6” x 6” x 0.03” thick Teflon PTFE (Polytetrafluoroethylene) confining sheet and the 8” x 8” x 5/8” thick steel confining plate were not used in this test series. The 3” x 5” x 1/4” rectangular steel tubes were placed parallel to each other on either side of the anchor no closer than two times the embedment depth. An 18” x 18” x 1” thick steel plate with a 2-3/4” diameter hole in the center was placed on the rectangular steel tubes to support the load frame. Figure 5 shows the modifications to the sustained load (creep) test apparatus for test series 16.

C-7 Figure 5: Test series 16 sustained load (creep) unconfined test apparatus. Specimen preparation The test specimens consisted of three parts; the concrete test member, the adhesive, and the anchor rod. Concrete Test Member The concrete test members for the short-term tests were poured in 60” x 16” x 12” forms. Minimal reinforcement of two #3 60 ksi steel reinforcing bars were placed longitudinally 5” from the bottom of the slab with approximately 1” cover. Four ½” diameter PVC pipes were placed at mid-height which allowed for ½” diameter rods to later be passed through the concrete test member in order to accommodate handling. The concrete test members for the sustained load (creep) tests were poured in 16” x 16” x 12” forms. No reinforcement was provided. Two ½” diameter PVC pipes were placed at mid- height in order to accommodate handling. All the forms used in this project were made of high density overlay plywood and were assembled with threaded rod and wing nuts, which allowed for multiple uses due to the large number of test members required. Since only twenty sustained load tests could be conducted simultaneously, the production of the concrete test specimens (Figure 6) was staggered in eleven separate pours denoted as concrete series A–J. The pour dates and number of test blocks

C-8 produced in each series are listed in Table 1. In order to provide a smooth testing surface, the blocks were cast upside down against the high density overlay plywood. After the first pour, it was decided to place a 1/8” thick sheet of Teflon PTFE (Polytetrafluoroethylene) at the bottom of the form to provide an even smoother surface and ensure against a lesser quality surface in the later pours. Figure 6: Concrete test specimens being cast. Table 1: Concrete pour details. Concrete Mix Pour Date Number of Short-term Test Members Number of Sustained load Test Members Notes A May 18, 2010 3 12 B May 26, 2010 3 12 C June 2, 2010 3 12 D June 15, 2010 3 8 E June 22, 2010 3 9 F June 29, 2010 3 9 G July 27, 2010 3 9 H August 3, 2010 3 12 20% Fly Ash I August 10, 2010 3 12 50% Blast Furnace Slag J March 25, 2011 3 12 Standard DOT Mix

C-9 The concrete for mixes A–I were batched, mixed, and placed at the Florida Department of Transportation (FDOT) State Materials office in Gainesville, FL. Concrete with round river gravel without any admixtures (except for series H and I, which included 20% fly ash and 50% blast furnace slag respectfully) was specified with a mean compressive strength between 4000 and 6000 psi during testing. All of the materials were batched by weight. Moisture samples were taken of the coarse aggregate (#7 and #89 stone) and allowed to dry in one of two Blue M large ovens (Figure 7) at 230°F (110°C) for 24 hours in order to determine the percent moisture. The sand was oven dried in the same ovens at 230°F (110°C) for 24 hours. Concrete was mixed in a Lancaster 27 CF counter current batch concrete mixer (Figure 8) and discharged into a large hopper and then placed into the forms with shovels and vibrated with an electric vibrator. Due to the size of the concrete mixer and forms, the concrete for each series was made in three batches. Plastic properties (slump, percent air, temperature, and unit weight) were evaluated and 4” x 8” cylinders were made for each batch of every series. Figure 7: Ovens. Figure 8: Mixer. The concrete for mix J was batched and mixed by Florida Rock Industries, a local ready- mix plant and placed at the University of Florida (UF) Structures Laboratory in Gainesville, FL. Mix J was designed by the NCHRP panel and included granite aggregate, water reducer, fly ash, and air entrainment. The mix designs and plastic properties for concrete series A–J are included in Appendix G. Following the pour, the concrete was covered with plastic for 24 hours. After the first 24 hours, the concrete was covered with wet tarps and plastic and maintained wet for 5 days. The

C-10 cylinders were capped with plastic lids. After 6 days the forms were removed and the cylinders demolded. The concrete test members and cylinders were maintained in the UF structures laboratory thereafter. Concrete compressive strength was determined by testing the cylinders in general accordance with ASTM C39 on a Test Mark Model CM-5000-DG compression machine (Figure 9) calibrated in August 2009 and August 2010 located at the FDOT State Materials Office in Gainesville, FL. The cylinders were ground smooth on a Hi-Kenma cylinder grinding machine (Figure 10) prior to testing. A concrete strength-age relationship was determined for each series by testing 4” x 8” cylinders at 7, 14, 28, 56, 112, 224, and 448 days or at the end of testing for that series, whichever came first. The average compressive strength for each series is presented in Table 2. Figure 9: Compression machine. Figure 10: Cylinder grinding machine. Adhesive The three adhesive products were stored in an environmentally controlled room maintained within the temperature and humidity range specified by the manufacturers prior to installation.

C-11 Table 2: Concrete series average compressive strength. Concrete Series Pour Date Average Compressive Strength (psi) 7 day 14 day 28 day 56 day 112 day 224 day 448 days Final8 No of days9 A May 18, 2010 3180 3930 4200 4350 4480 44602 38703 4210 679 B May 26, 2010 3170 3950 4260 4390 4 4 4 4100 671 C June 2, 2010 3260 3840 4410 43401 4210 4140 3830 3960 664 D June 15, 2010 3180 4080 4320 4800 4 4 4 4740 651 E June 22, 2010 3130 3790 4210 4430 4 4 4 4570 644 F June 29, 2010 2660 3670 4050 4290 4500 4 4 4310 637 G July 27, 2010 3100 3810 4260 4720 4650 4570 4 4470 609 H (FA5) August 3, 2010 2220 3010 3610 3810 3500 3760 4 3540 602 I (BFS6) August 10, 2010 1710 2740 3240 3460 3000 3020 4 2700 595 J (DOT7) March 25, 2011 4530 5490 5940 5930 5310 5540 10 4830 368 1 Test conducted at 55 days. 2 Test conducted at 231 days. 3 Test conducted at 251 days. 4 Not enough samples to conduct tests at these times, last group of three samples held until end of project. 5 FA = Fly Ash. 6 BFS = Blast Furnace Slag. 7 DOT = Department of Transportation concrete mix. 8 Samples tested at end of project on March 27, 2012. 9 Number of days since casting for final compression test. 10 Test at 448 days not conducted. Anchor Rods The anchor rods were ASTM A354 grade BD 5/8” diameter 11 threads per inch (UNC) steel threaded rod fabricated by Glaser & Associates from Martinez, CA. This grade of steel has a specified yield strength of 130 ksi and a specified tensile strength of 150 ksi. The anchor rods were cut to a length of 5.75” from 6” stock and their ends ground and chamfered with a bench grinder and steel brush to remove burrs and to clean up the threads in order to install the nuts. The anchors were stored in a sealed bucket in oil-soaked shredded paper to prevent rusting. Prior to installation, the rods were cleaned with acetone, allowed to air dry, and protected with paper until installed. Instrumentation Measurement Displacement. Direct measurement of the anchor displacement was not possible due to the location of the test apparatus. Therefore, a 16” x 2” x ¼” ASTM A36 steel flat bar was attached to the bottom of the non-rigid coupler that connected the anchor to the 1” diameter

C-12 loading rod (See Figure 1–Figure 5). Two BEI Duncan Electronics model 9610 linear motion position sensors (linear-pots) were fixed to this flat bar, one on each end, equidistant from the centerline of the anchor. The displacement was calculated as the average of the two linear-pot measurements. Load. The tension in the anchor was measured indirectly as a compressive reaction of either the hydraulic ram or the compression spring in the test apparatus. For the short-term tests, the load was measured by a Houston Scientific Model 3500 100-kip load cell excited by a 10VDC amplifier with a gain of 500. For the short-term (creep) tests, the loads were measured by the same Houston Scientific load cell during loading. Once the load was applied to the anchor, the load cell was removed and the load was monitored by the spring stiffness and displacement. Temperature. Temperature in each concrete test slab was measured by National Semiconductor LM35 Precision Centigrade Temperature Sensors. The temperature sensors were located 2 inches deep in the top of the concrete test specimen placed in 1/2” diameter holes drilled just prior to conditioning and sealed with rubber grommets to allow for reuse. Ambient air temperature in the test chamber was measured by a Cincinnati Sub-Zero EZT-560i Environmental Chamber Controller installed in the Cincinnati Sub-Zero Model WM-STH-1152- 2-H/AC Walk-In Stability Chamber. Analog cards installed in the Cincinnati Sub-Zero EZT-560i Environmental Chamber Controller provided an analog signal output allowing the ambient air temperature to be monitored by the data acquisition system. Humidity. Relative humidity in the test chamber was measured by a Cincinnati Sub-Zero EZT-560i Environmental Chamber Controller installed in the Cincinnati Sub-Zero Model WM- STH-1152-2-H/AC Walk-In Stability Chamber. Analog cards installed in the Cincinnati Sub- Zero EZT-560i Environmental Chamber Controller provided an analog signal output allowing the humidity to be monitored by the data acquisition system. Time. Time was measured using the computer’s internal clock. Instrument Calibration Displacement. The linear motion position sensors were calibrated against a Fowler digital caliper over their full range of 1” at 1/8” increments. The measurements were adjusted for variations in power supply voltage and normalized to a 10 volt power supply.

C-13 Load. The Houston Scientific Model 3500 100-kip load cell was calibrated in July 2010 at the Florida Department of Transportation State Materials Office in Gainesville, FL on a Test Mark Model CM-5000-DG compression machine. The load cell was calibrated over a range of 0 to 80 kips with nine data points. The compression springs were calibrated in June 2010 on an INSTRON System 3384 150 kN universal testing machine to determine their stiffness and working load. The large springs were calibrated individually over a range of 0 to 15 kips with about 500 data points. For loads above 15 kips, the large and small springs were calibrated in parallel over a range of 0 to 20 kips with about 630 data points. The large springs had an average stiffness of 10.2 kips/in. and a COV of 0.03. The large and small springs together had an average stiffness of 13.4 kips/in. and a COV of 0.02. The average drop in load between the end of loading and rupture was around 3%. Temperature. The National Semiconductor LM35 Precision Centigrade Temperature Sensors factory calibration was validated in June 2010 against a high quality mercury thermometer over a temperature range of 100°F to 120°F (43°C to 49°C). The temperature sensor in the test chamber was calibrated by the factory. Humidity. The humidity sensor in the test chamber was calibrated by the factory. Environmental control Standard Temperature An air conditioned space was used to store and condition the adhesive at 75ºF ±10ºF (24°C ±5°C) and 50% ±10% relative humidity. When conditions allowed, the test slabs were stored prior to installation and testing on the shop floor of the UF Structures Laboratory at 75ºF ±10ºF (24°C ±5°C) and 50% ±10% relative humidity. Elevated Temperature A 12’ by 12’ by 8’ tall Cincinnati Sub-Zero Model # WM-STH-1152-2-H/AC Walk-In Stability Chamber (Figure 11) was used to condition and test at the elevated testing temperature of 110ºF +10ºF/-0ºF (43ºC +5ºC/-0ºC) and below 40% relative humidity for the short-term and sustained load (creep) test. The chamber has a temperature range of -20°C to 60°C (-4°F to

C-14 140°F) and a relative humidity range of 10% to 95%. The chamber was equipped with a CSZ EZT-560i Touch Screen Controller to monitor and control the temperature and humidity. Figure 11: Walk-in stability chamber. The concrete test specimens were placed on furniture dollies in order to facilitate test rotation and to raise them off the ground by a few inches to promote better air flow and a uniform temperature within the concrete. The stability chamber was able to simultaneously house 20 sustained load anchor pullout test specimens and one short-term anchor pullout test specimen on the floor. Shelves were built along the walls to house the 16 adhesive-only test frames. Figure 12 and Figure 13 show the testing chamber with 20 sustained load tests running (2 not visible). A short-term testing slab is in the foreground. Figure 14 shows the layout of the anchor pullout test specimens in the stability chamber. Figure 12: Left side of testing chamber. Figure 13: Right side of testing chamber.

C-15 Figure 14: Layout of anchor pullout test frames in the stability chamber. Data management and acquisition During the testing and conditioning of the test slabs to the elevated temperature, a Microsoft compatible computer ran several National Instruments LabVIEW 8.6 software programs developed to collect, record, and display the data. Measured values included load, displacement, temperature, humidity, and time. Data acquisition was performed with a National Instruments NI cDAQ-9172 chassis with several National Instruments NI 9205 modules to interface with the instrumentation.

C-16 Due to minor fluctuations in the 10 volt power supply, the LabVIEW programs recorded the power supply voltage with each data reading and the position readings were appropriately adjusted to a normalized ten volt power supply. Data Sampling Program A LabVIEW 8.6 program (Figure 15) was developed to centrally sample data for every test. This program provided a half second time averaged record sampled at 2000 Hz. Global variables for each of the twenty sustained load test frames and the one short-term test frame were updated every half second to the computer memory to be read when needed by the separate LabVIEW programs for each test frame. Each global variable included a timestamp, and the voltage readings for the two linear-pots, power supply, load cell, concrete temperature sensor, and environmental chamber temperature and humidity. Figure 15: Data sampling LabVIEW program. Short-term test program A LabVIEW 8.6 program (Figure 16 & Figure 17) developed for this project was used for the short-term tests. Load, displacement, temperature, and humidity readings were recorded at half second intervals. A load versus displacement curve was displayed on the screen for real-time feedback. Load rate control was monitored by plotting the actual load rate from the hydraulic

C-17 hand pump against an ideal load rate to cause bond failure of the expected load in 120 seconds on a load versus time graph. This real-time plot was used to assist the pump operator in applying a constant load rate. The latest data readings were displayed on the screen and each data reading was automatically recorded in a Microsoft Excel spreadsheet. Figure 16: Short-term test LabVIEW program (main screen).

C-18 Figure 17: Short-term test LabVIEW program (chart page). Long-Term (Creep) Test Program A LabVIEW 8.6 program (Figure 18 to Figure 20) developed for this project was used for the sustained load (creep) test. Load, displacement, temperature, and humidity readings were recorded at progressively longer intervals over the course of the test as discussed previously. If it became necessary to apply additional load to the anchor during the test, the program entered a tightening phase in which data was recorded every half second. Once tightening was completed, the program began sampling every 5 seconds and proceeded through the previously discussed sampling schedule. A displacement versus time curve (Figure 19) for each anchor and a percent mean static load versus time curve (Figure 20) were displayed on the screen for real-time feedback. The latest data readings were displayed on the screen and each data reading was automatically recorded in a Microsoft Excel spreadsheet.

C-19 Figure 18: Sustained load test LabVIEW program (main screen). Figure 19: Sustained load test LabVIEW program (displacement plot).

C-20 Figure 20: Sustained load test LabVIEW program (percent load plot). Test Specimen Conditioning Program A LabVIEW 8.6 program developed for this project was used to monitor the test specimen during conditioning. Concrete specimen temperature as well as the temperature and humidity of the environmental chamber were recorded at five minute intervals. A concrete specimen temperature versus time graph was displayed on the screen for real-time feedback. The latest data readings were displayed on the screen and each data reading was automatically recorded in a Microsoft Excel spreadsheet. Since the three large concrete test specimens used for the test series 1 (baseline) short-term tests were conditioned simultaneously, one concrete test specimen was monitored during conditioning and all three were checked prior to testing. Installation procedure The standard installation procedure is described below and was followed for test series 1, 12, 14, 15, and 16. Exceptions to this standard installation procedure as used in test series 7, 9, and 13 follow.

C-21 Standard Baseline Installation Procedure All anchors were installed according to the manufacturer’s specifications. The holes were created with a 3/4” (11/16” for adhesive A) carbide tipped concrete bit as specified by the manufacturer and a Hilti model TE52 hammer drill. A drilling jig (Figure 21) with a depth stop was used to ensure that the holes were drilled perpendicular to the surface of the concrete and to the correct depth. Figure 21: Drilling rig and hammer drill. The spoil at the concrete surface was removed with a vacuum prior to cleaning the holes. The holes were cleaned according to the MPII which generally included blowing with oil-free compressed air, brushing with a steel brush provided by the manufacturer, and then blowing again with compressed air until no dust was discharged. Durations and numbers of brushing/blowing cycles varied by manufacturer, but for each case the holes were cleaned according to the MPII. Details of the full cleaning procedure are listed in Table 3.

C-22 Table 3: Full hole cleaning procedures per MPII. Adhesive A Adhesive B Adhesive C Blow with compressed air (4x) Brush with drill (4x) Blow with compressed air (4x) Blow with compressed air (2x) Brush by hand (2x) Blow with compressed air (2x) Blow with compressed air (4x) Brush with drill (1x) Blow with compressed air (4x) Brush with drill (1x) Blow with compressed air (4x) Brush with drill (1x) Blow with compressed air (4x) Brush with drill (1x) Blow with compressed air (4x) To prevent dust from blowing into the operator’s mouth and eyes, an adaptor for the vacuum (Figure 22) was used to capture the dust ejected from the hole when blowing with compressed air. This adaptor attached to the vacuum hose and allowed the compressed air nozzle to be easily inserted and removed. Once clean, masking tape was placed over the hole to ensure that dust and humidity did not enter the hole prior to installation of the adhesive anchor. In all cases the time between cleaning and installation was not more than a few minutes. A hole was gently cut in the masking tape prior to installation. The adhesive products were dispensed with a manufacturer supplied cartridge gun. According to the manufacturer’s specifications, several squeezes of adhesive were discharged and disposed of before dispensing into the holes to ensure that the adhesive was of uniform color and consistency indicating that it was properly/thoroughly mixed. The anchors were wiped clean with acetone and allowed to air dry. The anchors were then attached to an “embedment depth chair” (Figure 23) set for the appropriate embedment depth of 3-1/8”. The chair rested on the face of the concrete test specimen ensuring the proper embedment depth and did not interfere with the adhesive squeeze out. The anchor rod was rotated counterclockwise and jiggled while it was installed in the hole until the legs of the “chair” came to bear on the concrete. The anchors were left undisturbed during the specified gel/working time and the adhesive was allowed to cure for seven days prior to conditioning. Excess adhesive was carefully chipped away from around the anchor prior to conditioning. The masking tape left around the hole prevented the concrete from being removed during chipping.

C-23 Figure 22: Vacuum adaptor. Figure 23: Embedment depth chair. Test Series 7 (Moisture during Installation) Installation Procedure This installation procedure was adapted from ACI 355.4 section 7.10 and 7.6. The holes were initially drilled to roughly half the final diameter using a 3/8” diameter carbide tipped concrete bit. A water dam (Figure 24) constructed out of 2x4 dimensional lumber was secured to the top of the concrete test specimens with silicon. The holes were filled with water and the test specimens were covered with 3” of water for a minimum of eight days (192 hours). Following eight days of saturation, the water was drained and the standing water was vacuumed out of the holes and the installation procedure then followed the above described standard baseline installation procedure. Figure 24: Water dam for test series 7 installation.

C-24 Test Series 9 (Reduced Hole Cleaning) Installation Procedure ACI 355.4 section 7.5 defines a reduced hole cleaning effort as 50% of the full hole cleaning procedure. The standard installation procedure was followed except that the cleaning effort was modified to the procedures described in Table 4. Table 4: Reduced hole cleaning procedures. Adhesive A Adhesive B Adhesive C Blow with hand pump (2x) Brush with drill (2x) Blow with hand pump (2x) Blow with hand pump (1x) Brush by hand (1x) Blow with hand pump (1x) Blow with compressed air (2x) Brush by hand (1x) Blow with compressed air (4x) Brush by hand (1x) Blow with compressed air (4x) Test Series 13 (Type of Drilling) Installation Procedure The holes were created with a 3/4” (11/16” for adhesive A) diamond core bit using a Hilti model DD130 core drill. A drilling rig (Figure 25) was used to ensure that the holes were drilled perpendicular to the surface of the concrete. The drilling rig was secured to the concrete specimen with ratchet tie-down straps. The cores were wet-drilled by use of a water jacket attached to the chuck of the drill. Efforts were made to reduce the amount of excess water on the concrete specimens. A water collector (Figure 26) connected to a wet vacuum surrounded the bit to collect water during drilling.

C-25 Figure 25: Core drill for test series 13. Tape was placed on the core drill to indicate the proper depth during drilling. The holes were drilled to a depth of 4-½” to ensure that the core cylinders would break below the required embedment depth. The cylinders were broken off by inserting a small screwdriver in the hole and gently prying the cylinders loose. An extraction tool (Figure 27) was used to remove the cylinder pieces from the hole. Standing water was removed using a wet vacuum with a narrow hose attachment. The holes were then flushed twice using a ½” diameter rubber hose at normal street water pressure and the excess water was captured with the water collector, then brushed twice, and then flushed twice again until the water ran clear. Finally, the standing water was removed with a wet vacuum. The holes were cleaned according to the MPII as presented in Table 3. The holes were then dried with compressed air by inserting and removing a special air nozzle tip (Figure 28) two times.

C-26 Figure 26: Water collector. Figure 27: Extraction tool. Figure 28: Air nozzle. The holes were covered with masking tape to ensure that dust and humidity did not enter the hole prior to installation of the adhesive anchors. The anchor installation proceeded as described earlier in the standard installation procedure. Specimen conditioning Upon completion of the 7 day adhesive curing period, the test specimens were wheeled into the 110ºF (43ºC) 35% humidity environmental test chamber on dollies for conditioning. The temperature of the concrete test specimen and the environmental chamber as well as the humidity in the environmental chamber were monitored and recorded. Testing began upon completion of the 24 hour conditioning period in the environmental test chamber. Testing procedure The standard testing procedures for the short-term and sustained load (creep) tests are described below which was followed for test series 1, 7, 9, 12, 13, 14, and 15. Exceptions to these standard testing procedures as used in test series 16 follow. Short-Term Test Procedure A 0.03” thick PTFE confining sheet and steel 5/8” thick confining plate were placed over the anchor and the non-rigid coupler was attached to the anchor. A 1/16” to 1/8” gap was left between the confining plate and the coupler to allow for rotation of the coupler in order to prevent bending forces from being transferred between the anchor and the loading rod. The short-term test apparatus was placed over the anchor as discussed earlier. Steel spacers were placed under the linear potentiometers so that the initial position reading was in the 0.300”–

C-27 0.500” range (this was done because the position readings at the far extremes of the instrument are less accurate). The Enerpac model RCH-603 Holl-O-Cylinder (60 ton) hydraulic ram was placed on the frame and connected to the Enerpac model P802 (10,000 psi) hydraulic hand pump. The loading rod was then connected to the coupler. The Houston Scientific Model 3500 100-kip load cell was placed on top of the ram sandwiched between four 1/4” plates (two above and two below). The loading rod nut was hand tightened to remove slack in the system. The LabVIEW 8.6 program was started to confirm that the program was functioning correctly and that the linear-pot values were within acceptable ranges. The program was then reset and the test was started. Pumping did not start until after a few seconds in order to read the initial load reading and to allow the program to zero out the initial load cell and linear-pot readings in order to calculate load and displacement. The anchors were loaded at a constant load rate. The operator adjusted the pump rate to conform to an ideal pump rate that would cause failure at the expected load within 120 seconds by following the ideal load rate curve on the load versus time plot on the screen. The operator was only in the environmental chamber to disconnect and connect the testing apparatus to the anchors. The pumping and test observation was conducted outside the chamber. The LabVIEW 8.6 program automatically recorded the test data in a MS Excel spreadsheet. Test Series 16 (Unconfined) Test Procedure The above procedure was followed with the following exceptions: The PTFE confining sheet and 5/8” thick steel confining plate were eliminated. The test frame supports were placed no closer than two times the embedment depth from the anchor. Standard Sustained Load (Creep) Test Procedure The tests began by placing the 0.03” thick PTFE confining sheet, 5/8” thick steel confining plate, coupler, and linear potentiometers as described in the short-term test procedure. The compression springs were compressed in an INSTRON System 3384 150KN universal loading machine and the load was monitored with the on-screen display from the

C-28 universal testing machine. Once the desired load was obtained, the four corner bolts on the test frame were hand tightened to maintain the load. The compression spring frame was placed over the anchor and the loading rod was connected to the coupler. Two ¼” steel plates and a washer were placed on top of the test frame and a nut was loosely placed on top. The entire assembly was rolled into the testing chamber on a dolly for conditioning. Once the 24 hour condition period elapsed, the hydraulic jack chair (Figure 4) was placed over the loading rod and a ½” steel loading plate was placed on top. The Houston Scientific Model 3500 100-kip load cell was placed on top of the loading plate sandwiched between four ¼” plates (two above and two below). Another nut was placed on top and hand tightened. The LabVIEW 8.6 program was started to confirm that the program was functioning correctly and that the linear-pot values were within acceptable ranges. The program was then reset and the test was started. The data acquisition system initially entered a loading cycle in which the load was monitored by the load cell. The load was applied by pumping the ENERPAC P-14 hand pump, which displaced the top plate of the test frame, causing the load to be transferred from the corner bolts to the loading rod. After the desired load was reached, the nut at the top plate of the test frame was hand tightened and the program exited the loading cycle. The pressure was released from the hand pump and the hydraulic jack chair and load cell were removed from the test frame. The load was thereafter calculated from the spring stiffness and anchor displacement. If it became necessary to add load during the duration of the test, the hydraulic jack chair and load cell were placed on top of the test frame as described above. The program entered a tightening phase in which the load was monitored once again by the load cell and the spring stiffness and displacement. The greater of the two values was used as the load on the anchor. Once the desired load was achieved, the nut on the top plate of the test frame was hand tightened, the pressure was released from the pump, and the test continued as described above. The LabVIEW 8.6 program automatically recorded the test data in an MS Excel spreadsheet.

C-29 Test Series 16 (Unconfined) Test Procedure The above procedure was followed with the following exceptions: The PTFE confining sheet and 5/8” thick steel confining plate were eliminated. The test frame supports were placed no closer than two times the embedment depth from the anchor. The compression spring frame was placed on top of a steel plate that rested on the test frame supports. Post-Test Procedure A few of the anchors were cored with a 2-½” diameter concrete cylinder core bit using a Cincinnati Bickford coring machine. The resulting cores were saw cut on each side to the depth of the anchor and then split open. The resulting concrete core provided a more detailed investigation of the failure mode and is discussed in Chapter 3. Photos were taken of the cores and are presented in Appendix H.