Evaluation of In-Place Strength of Concrete By The Break-Off Method. Tarun Naik Ziad Salameh Amr Hassaballah

Evaluation of In-Place Strength of Concrete By The Break-Off Method By Tarun Naik Ziad Salameh Amr Hassaballah

Evaluation of In-Place Strength of Concrete By The Break-Off Method By Tarun R. Naik Associate Professor of Civil Engineering P.O. Box 784 University of Wisconsin - Milwaukee Milwaukee, WI 53201 Phone: (414) 229-6696 Ziad Salameh Graduate Student University of Wisconsin - Milwaukee Amr A. Hassaballah Formerly Graduate Student University of Wisconsin - Milwaukee March 1988

2 TABLE OF CONTENTS LIST OF FIGURES...5 LIST OF TABLES...7 ABSTRACT...9 CHAPTER l INTRODUCTION...11 l.l General...11 l.2 Objective...13 l.3 Scope and Research Significance...16 l.3.l l.3.2 l.3.3 l.3.4 l.3.5 Aggregate Shape...16 Concrete Strength...17 Age...17 Slab Thickness...18 Methods of obtaining Cylindrical B.O. Test Specimens...18 l.4 Review of Published Studies...20 CHAPTER 2 THE BREAK OFF METHOD...24 2.l The Principle of the Method...24 2.2 The B.O. Test Apparatus...27 2.3 The B.O. Tester Calibration Procedure...30 2.4 The B.O. Test Procedure...36 2.5 The B.O. Specimen Testing...37

3 CHAPTER 3 CONCRETE TESTING AND PREPARATION OF SLABS AND TEST SPECIMENS...39 3.l General...39 3.2 Concrete...41 3.3 Fresh Concrete Testing...46 3.4 Casting of Slabs...50 3.5 Inserting Sleeves...50 3.6 Curing of Slabs...54 3.7 Cylinders Making, Curing, and Testing...54 CHAPTER 4 RESULTS...56 CHAPTER 5 DISCUSSION OF RESULTS FOR THE CRUSHED AGGREGATES CONCRETE...61 5.l General...61 5.2 Strength Development vs. Age...62 5.3 Within Test Coefficient of Variation...62 5.4 Profile Analysis...71 5.5 Regression Analysis...74 CHAPTER 6 DISCUSSION OF RESULTS FOR THE ROUNDED AGGREGATES CONCRETE...78 6.1 General...78 6.2 Strength Development vs. Age...79 6.3 Within Test Coefficient of Variation...79 6.4 Profile Analysis...86 6.5 Regression Analysis...88 CHAPTER 7 CONCLUSIONS, LIMITATIONS AND RECOMMENDATIONS AND POSSIBLE FURTHER INVESTIGATION...92 7.l Conclusions...92 7.2 Limitations and Recommendations...95 7.3 Possible Further Investigation...97 LIST OF REFERENCES...99 APPENDIX: SAMPLE DATA SHEETS...102

4 LIST OF FIGURES l.2.l Schematic of a Concrete Cylindrical Specimen Obtained by Inserting a Sleeve or Drilling a Core, and Location of Applied Load...14 l.2.2 l.3.5.l 2.l.l 2.l.2 2.2.l B.O.Manometer Reading vs. Concrete Compressive Strength as Provided by the Manufacturer...15 Inserted Sleeve and Drilled Core Specimen...19 Tubular Plastic Sleeves for Inserting in Fresh Concrete...25 Core Drill Pit for Drilling a Core for B.O. Testing...26 The B.O. Test Equipment (l) Load Cell, (2) Manometer, and (3) Hydraulic Hand Pump...28 2.2.2 Low and High Range Settings of the B.O. Tester Load Cell...29 2.2.3 Sleeve Remover...31 2.2.4 The B.O. Tester Calibrator...32 2.2.5 Two Speed Electrically Driven Core Drilling Machine...33 2.3.l Calibrator Placed in the Load Cell...34 2.3.2 Calibrator Chart as Provided by The Manufacturer...35 3.l.l 3.3.l Schematic of Location of Inserted Sleeves and Drilled Cores B.O. Specimens in the Slab...40 The Density Bucket...48 3.3.2 The Air Meter...49 3.5.l 6 in. String Grid System...51 3.5.2 Inserting Sleeves by Rocking Action...52 5.2.l Cylinder Compressive Strength and B.O. Reading vs. Age for Mix l...63 5.2.2 Cylinder Compressive Strength and B.O. Reading vs. Age for Mix 2...64 5.2.3 Cylinder Compressive Strength and B.O. Reading vs. Age for Mix 3...65 5.3.l Coefficient of Variation of Concrete vs. Percentage of

"True" Strength...70 5 5.4.l Profile of Slab Thickness vs. B.O. Readings...72 5.4.2 Profile of Method of Obtaining B.O. Specimens vs. B.O. Readings...72 5.4.3 Profile of Shape of Aggregate vs. B.O. Readings...72 5.5.l Plots of Regression Equations for Inserted Sleeve Specimens for Mix l, 2, and 3...75 5.5.2 Plots of Regression Equations for Drilled Core Specimens for Mix l, 2, and 3...76 6.2.1 Cylinder Compressive Strength and B.O. Reading vs. Age for Mix A...80 6.2.2 Cylinder Compressive Strength and B.O. Reading vs. Age for Mix B...81 6.2.3 Cylinder Compressive Strength and B.O. Reading vs. Age for Mix C...82 6.4.1 Slab Thickness Profile vs. B.O. Reading...87 6.4.2 Method of Obtaining Cylindrical Specimen Profile vs. B.O. Reading...87 6.4.3 Aggregate Shape Profile vs. B.O. Reading...87 6.5.1 Sleeve Specimen B.O. vs. Cylinder Compressive Strength...89 6.5.2 Drilled Specimen B.O. vs. Cylinder Compressive Strength...90

6 LIST OF TABLES 3.2.l Concrete Mix Data...42 3.2.2 Concrete Mix Proportions...43 3.2.3 Chemical Analysis and Chemical Composition for Type I Dundee Cement...44 3.2.4 Physical Tests for Type I Dundee Cement...44 3.2.5 Chemical Composition of Class C Fly Ash...44 3.3.l Fresh Concrete Test Results, Crushed Aggregates...47 3.3.2 Fresh Concrete Test Results, Rounded Aggregates...47 4.l Results for Cylinder Compressive Strength and B.O. Tests for Mix l...57 4.2 Results for Cylinder Compressive Strength and B.O. Tests for Mix 2...57 4.3 Results for Cylinder Compressive Strength and B.O. Tests for Mix 3...58 4.4 Results for Standard Cylinder Compressive Strength and B.O. Tests for Mix A...59 4.5 Results for Standard Cylinder Compressive Strength and B.O. Tests for Mix B...59 4.6 Results for Standard Cylinder Compressive Strength and B.O. Tests for Mix C...60 5.3.l Standard Deviation ( ) and Coefficient of Variation (v) for Standard Cylinder Tests and B.O. Tests for Mixes 1, 2, and 3...66 5.3.2 Average Within Test Coefficient of Variation, v, percent, for Mix No. 1, 2, and 3...68 6.3.1 Standard Deviation ( ), and Coefficient of Variation (v) for Cylinder Tests and Break-Off Tests for Mixes A, B, and C...83 6.3.2 Average Within Test Coefficient of Variation, v, percent...84

Evaluation Of In-Place Strength Of Concrete By The Break-Off Method BY Tarun R. Naik Ziad Salameh Amr A. Hassaballah ABSTRACT The overall objective of this research was to investigate the reliability of the Break Off (B.O.) method as a tool in determining the in-place strength of concrete, and to investigate the sensitivity of this method to different types of concrete. Effects of several significant parameters were investigated to accomplish this objective. The parameters considered were: (l) concrete strength; (2) aggregate shape; (3) age of concrete; (4) slab thickness; and, (5) method of obtaining cylindrical B.O. test specimens (either by inserting plastic sleeves or by drill cores). A total of 524 B.O. tests and 90 standard cylinder tests were carried out. Evaluation of the results has indicated that the B.O. test readings show a similar trend of strength development versus age as that for the standard cured compressive strength of concrete. The average coefficient of variation for the B.O. test was eight percent. The B.O. test results for crushed aggregates concrete were on the average l0% higher than that for rounded aggregates concrete. Slab thicknesses of 5 and 7 in. did not have any significant effect on the variability or the average value of the B.O. reading. The drilled cores B.O. test results were on the average nine percent higher than the inserted sleeves B.O. test results. In order to develop empirical relationships correlating the cylinder compressive strength to the in-place concrete strength, as a function of the B.O. reading, regression analysis was performed. The regression equations obtained showed a high degree of correlation between the B.O. readings and the compressive strength of concrete. It was concluded from the results of this research project that the B.O. test is an accurate, fast, and easy way of determining the in-place compressive strength of concrete.

CHAPTER l INTRODUCTION l.l General In North America, the standard 6" x l2" cylinder is generally used to determine compressive strength of concrete. Such test cylinders only measure the strength of the sampled concrete as a representative of a particular batch of concrete. The test cylinder would indicate if the potential strength of the concrete batch is satisfactory or not. The question regarding the actual compressive strength of the concrete in the structure (in-place strength) is not answered by the test cylinder. Destructive and/or nondestructive tests are conducted for determining the strength of the in-place concrete. For many years questions have been raised regarding reliability of concrete quality assurance test methods based upon standard cylinder tests in order to determine the in-place strength of concrete. In-place concrete strength is not the same as the cylinder concrete strength because the in-place concrete is placed, compacted, and cured in a different manner than the cylinder specimen concrete. It is increasingly being recognized [1] that strength of concrete in structures should be measured by in-place testing. Nondestructive test methods available for determination of in-place concrete quality include the Hardness Test [19,20], the Penetration Test [19,20], the Pulse Velocity Method [19,20,23], the Maturity Method [22,26], the Pull-out Test [12,20,25], the Pulse-Echo Method [19,20], and the Break-Off Test [10,11,18]. Clearly strength tests based upon placing and curing conditions other than those existing in the structure may be misleading. Determination of accurate in-place strength is most critical in form removal and prestress or post-tension force release operations, because the structural element should not be stressed before a certain level of in-place strength is achieved. Fast construction schedules are emphasizing the need for in-place concrete strength techniques also. Currently about two dozen different methods exist for determining the in-place quality of the concrete [8]. Out of all these methods only the Break Off and the Pull Out tests measure a direct strength parameter. The Break Off test was developed in Norway. It consists of breaking off an in-place cylindrical concrete specimen at a failure plane parallel to the finished surface of the concrete. The break-off stress at failure can then be related to the compressive strength of the concrete using a predetermined relationship which relates the compressive strength of concrete measured by conventional test specimens, cylinders or cores, to the break off strength for that particular concrete. Traumatic construction failures, such as the cooling tower failure in West Virginia in l978, have affected the current quality assurance practice in North America. Recently more and more attention is being devoted to developing and adopting methods for determining in-place concrete strength. A former president of the American Concrete Institute (ACI) R. E. Phillieo, states that: "I'm not aware of an example where collapse followed the verification of concrete quality by in-situ testing" [27].

It is the opinion of this writer that cylinders will still be in use for decades as a recognized method because reinforced and prestressed concrete design methods are based upon measuring cylinder strength as the design strength for the concrete element. Therefore, calibration curves for in-place strength methods are generally related to the cylinder strength. l.2 Objective The Break Off (B.O.) method is a new procedure for determining the strength of the in-place concrete. A force required to break off an in-place concrete cylinder, 2.l7 in. (55 mm) diameter and 2.76 in. (70 mm) long, by cantilever action, is measured, Fig. l.2.l. This force can be related to the compressive and the flexural strength of the concrete, for example see Fig. l.2.2. The primary objective of this research project is to evaluate the reliability of the B.O. method as a tool in determining the in-place strength of concrete, and to investigate the sensitivity of this method to different types of concrete. The secondary objective is to find out advantages, disadvantages, applications, and limitations of this method. l.3 Scope and Research Significance In order to achieve the objectives of this project, comprehensive laboratory work was planned. The following variables were chosen as most significant: (l) aggregate shape; (2) concrete strength; (3) age; (4) slab thickness; and, (5) methods of obtaining the cylindrical B.O. test specimen, either by inserting a plastic sleeve in the fresh concrete or by drilling a core after the concrete has hardened. Each of these test variables should give a good indication of this test method's sensitivity as explained in the following sections. l.3.l Aggregate Shape Two types of aggregates, rounded and crushed, were used in the concrete mixes. Crushed aggregates concrete generally gives a higher flexural strength for a given value of the compressive strength [21]. This is due to adhesion and mechanical bond in the transition zone between the crushed aggregate and the mortar matrix. On the other hand mostly the adhesion bond exist in the case of the rounded aggregates. The better interlocking action of crushed aggregates provide higher flexural strength. Since the B.O. tester is mainly a flexural test, this variable is essential in distinguishing concrete of potentially different flexural strength but same compressive strength.

l.3.2 Concrete Strength Compressive strengths of 4, 6 and 8 ksi at the 28-day age were chosen as the more prevalent concrete strengths currently used in North America for high rise concrete construction. This covers a wide range of structural elements, from 4 ksi for concrete floors and beams to 8 ksi for highly stressed precast or post-tensioned members. l.3.3 Age This variable is important in order to determine the variation in the B.O. strength with the concrete age. The strengths at the age of l, 3, 5, 7 and 28 days were investigated. This provided a good picture of the strength variation with time from the very early age to the design strength age. In the case of drilled core B.O. tests only 5, 7, and 28-day ages were investigated for the 4 ksi and the 6 ksi concrete to avoid the problem of drilling cores from young, low strength, concrete. For the 8 ksi concrete, 3-day age cores were also drilled for B.O. tests since the compressive strength of this concrete at the 3-day age was sufficiently high enough for drilling. l.3.4 Slab Thickness Based upon data obtained in Norway [l4], the equipment manufacturer recommended 4 in. as the minimum depth of the test member. Five in. and 7 in. slab thicknesses were selected for this research project to determine if there are any statistical differences in the B.O. test readings between these two depths. This variable would also show if there should be a minimum depth requirement for inserted sleeve as well as for drilled core test specimens. Furthermore, 5 in. and 7 in. slab thicknesses represent typical slab thicknesses used in high rise construction in North America. l.3.5 Methods of Obtaining Cylindrical B.O. Test Specimens This variable was chosen to investigate the potential difference in the quality of concrete between an inserted-sleeve test and a drilled-core test, Fig. l.3.5.l. Different types of test specimens should also show different within test variation for each type of test specimens. From preliminary trial experiments it was noticed that drilled cores often gave slightly higher test readings because the concrete had more uniform quality. On the other hand, the inserted-sleeve test specimens concrete is more disturbed because of the sleeve insertion process in the fresh concrete. The drilled core B.O. test sample is relatively undisturbed because the test specimen is obtained from hardened concrete of sufficient in-place strength. For the drilled-core B.O. test specimens, however, some of the aggregates on the core surface are only partially embedded in the surrounding mortar matrix.

l.4 Review of Published Studies In l977, researchers at the Norwegian Technical University (NTH), and the Research Institute for Cement and Concrete in Norway, developed and patented the Break-Off tester as a method for determination of the compressive strength of the in-place concrete [l3]. The instrument was further developed in a cooperation between NTH and A/S Scancem Company in l98l-82. A/S Scancem is a private company in Norway which is now providing technical support for the tester. Several papers have been published in Europe about this method but few investigated the reliability and applicability of this method to North American concrete industry, particularly for high strength concrete. Johansen published the first paper on the Break Off tester in l976 [l7]. He indicated in his paper the main use of this test as a very efficient way of determining the in-place concrete strength for form removal. In l979, Johansen and Dahl-Jorgensen published a paper on the use of the B.O. method to detect variation in the concrete strength and curing conditions [ll]. In their research a comparison was made between the B.O. method and the Pull-Out test method. The compressive strength of cores obtained from the B.O. tests and the standard cube compressive strength was also compared. They concluded that the Pull-Out test method and the cores compressive strength values obtained from the B.O. tests have a better ability to differentiate between concrete qualities than the standard cube test. On the other hand, the B.O. test results and the cores compressive strength results demonstrated their ability in detecting variation in curing conditions, while the Pull-Out test method did not register some of the curing differences demonstrated by the B.O. and the core results. Also, they show that both the B.O. and the Pull-Out test methods are very suitable for testing young concrete. Also, in l979, Johansen published another paper [l8] on the use of the B.O. method, with a particular reference to its application to airport pavements made of vacuum concrete. He concluded that the variation of the concrete strength detected by the B.O. method is of the same order of magnitude as the variation detected by conventional flexural beam test. Furthermore, the B.O. strength was about 30% higher than the conventional modulus of rupture because of deviations in the load configurations and geometric parameters between the two testing methods. He also detected a high sensitivity of the B.O. method to sense the influence of the ambient air temperature on early strength. He obtained a good relationship between the B.O. test readings and the compressive strength of the concrete obtained by standard cube testing. In l980, Byfors tested concrete at early ages, using the Break Off method [9]. In his research he tested the concrete with different water to cement ratios, and different aggregate sizes (5/l6", 5/8", l-l/4"). The conclusion was that the B.O. method is well suited to detecting low strength concrete. After developing the second generation of the B.O. tester in l982, Dahl-Jorgensen published two reports on the use of the B.O. method [ll,l2]. In his study, he investigated the use of the new equipment in testing epoxy to concrete bond strength, and compared the results of the B.O. and the Pull-Off methods. He concluded that the B.O. test provided results with

smaller variations between individual tests than the Pull-Off method. Also, fewer tests were rejected for the B.O. method as compared to the Pull-Off method. Nishikawa published his work in l983 [24], after conducting laboratory research on the use of the B.O. method for determining flexural strength of concrete. He concluded that the relationship between the B.O. test results and cylinder compressive strength tests is "complex and practically useless". Therefore, no attempt was made to correlate these two test results. He further concluded that the change in the shape of the aggregates was not sensed by the B.O. method. These two conclusions are further discussed in chapter 5 since the data in this research, as well as most other research, do not support these conclusions. Nishikawa indicated a relatively high within test variation for the inserted sleeve B.O. tests as compared to cylinder and beam tests. With respect to other variables, he found that B.O. test results were affected by water-cement ratio, age, curing conditions, and cement type. In l984 Carlsson, Eeg, and Jahren published a paper on field experiences with the use of the B.O. tester [l0]. Six case histories were discussed. The authors concluded a trend towards greater acceptance of the B.O. test method in the field as the need for in-place testing increases in the future. The Break-Off method has been standardized recently in Norway [l3], New Zealand [28], and England [8]. Two other publications appeared in l987 in Japan, but these were not pursued because of difficulty in obtaining them.

2.l The Principle of the Method CHAPTER 2 THE BREAK OFF METHOD The B.O. method is presently the only available test method for directly determining flexural strength of in-place concrete. This method is based upon breaking off a cylindrical test specimen of in-place concrete, Fig. l.2.l. The test specimen has a diameter of 2.l7 in. (55 mm) and height of 2.76 in. (70 mm). The test specimen is effectively formed in the concrete either by means of a disposable tubular plastic sleeve, which is cast into the fresh concrete and then removed at the desired time of testing, or by drilling the hardened concrete at the time of the B.O. test. Fig. 2.l.l and 2.l.2 show the tubular plastic sleeve and a drill bit, respectively. Both the sleeve and the drill bit are capable of producing a 3/8 in. wide groove at the top of the test specimen, Fig. l.2.l, for seating a load cell (Section 2.2). A force is introduced at the top of the test specimen with the load cell until rupture occurs. Fig. l.2.l illustrates the loading configuration. The force required to break off a test specimen is measured by a mechanically operated manometer. The break off stress can be calculated as: B.O.= M/s where M = P x h h = 2.57 in. P = break off force s = 3.1416 * (d**3)/32 d = 2.l7 in. The Break Off method assumes that the ultimate flexural stress is reached at the extreme outside fiber. In this case, the circular area would restrict the ultimate stress fibers theoretically to a point, and a crack is initiated at this point. The exact location of the rupture plane is determined by the loading arrangement, Fig. l.2.l, at a distance of 2.76 in. (70 mm) from the concrete surface. Away from the extreme outside fiber the stresses are successively changing in the direction of the neutral axis from tension to compression. In essence, the procedure is the same as a cantilever beam with circular cross section, subjected to a concentrated load at its free end. Time studies from this project have shown that a set of five sleeves takes 20 minutes to insert, 45 minutes to drill a set of five core test specimens, and l0 minutes to test a set of 5 inserted sleeve or drilled core specimens. The test is easy and fast to perform. 2.2 The B.O. Test Apparatus A Break Off tester, Fig. 2.2.l, consists of a load cell manometer and a manual hydraulic pump capable of breaking a cylindrical concrete specimen having the specified dimensions in Section 2.l. The load cell has two measuring ranges: low range setting for low strength concrete up to approximately 3000 psi, and high range setting for higher strength concrete up to about l0,000 psi, Fig. 2.2.2. A tubular plastic sleeve, with internal diameter of 2.l7 in. (55 mm) and geometry shown in Fig. 2.l.l, is used for

forming cylindrical specimens in fresh concrete. A sleeve remover, Fig. 2.2.3, is used for removing the plastic sleeve from the hardened concrete. A calibrator is provided with the test equipment for calibration and adjustment of the Break-Off tester, Fig. 2.2.4. The B.O. tester was calibrated each time before using, Section 2.3. A diamond tipped drilling bit was used for drilling hardened concrete cores, Fig. 2.l.2. The bit is capable of producing a cylindrical core as well as a reamed ring in the hardened concrete at the top with dimensions similar to that produced by using a sleeve. A two speed electrically driven machine with water swivel for core drilling, suitable for coring horizontal and vertical surfaces, Fig. 2.2.5, was used for drilling core test specimens. 2.3 The B.O. Tester Calibration Procedure The B.O. tester should be calibrated periodically. To calibrate the tester, the calibrator gauge is set to zero and the calibrator is placed in the load cell, Fig. 2.3.l. Using the high range setting on the load cell, force is applied to the calibrator via the load cell until the load cell manometer reading of l00 is obtained. At this point, the dial gauge reading for the calibrator should be within 4% of the expected value which is obtained from the manufacturer's calibration chart, Fig. 2.3.2. Adjustment of the test system is necessary if error in the reading obtained is greater than plus or minus 4 percent of the expected reading from the chart. Through out this project the B.O. tester was checked each time before use, and, no adjustment was needed. A similar procedure can be followed with the low range setting on the load cell. 2.4 The B.O. Test Procedure At the time of test the load cell is placed on the groove on the top of the concrete surface so that the load is applied according to Fig. l.2.l. The data recorded for inserted sleeve and drilled-core test specimens included the B.O. reading, maximum and minimum length of the broken cylindrical test specimen, time of the test, and remarks about the nature of the failure plane, etc. In case of drilled-cores, the diameter was also recorded to the nearest l/8 in. See Appendix for sample of actual data sheets for concrete inspection, standard cylinder test, inserted sleeve B.O. test, and drill core B.O. test. For each slab, at each test age, all test locations were chosen randomly. Six sleeve specimens were tested at the 1 and 3 day ages, while five sleeve specimens were tested at the 5, 7, and 28 day ages. For the sleeve specimens, the plastic tubular sleeve was removed by applying a coupling force by hands using the sleeve remover, Fig. 2.2.3. Five drill specimens were tested at 5, 7, and 28 day test ages for Mixes 1, 2, 3, A, B, and C. For Mix 3 and C, five drill specimens were also tested at the 3-day test age. The preparation for drill specimens involved drilling cylindrical core specimens using the special drill bit, Fig. 2.1.2. At each desired test age, the following procedure was followed for drilling drill specimens:

(1) Location of five cores for drilling was chosen randomly in each slab, based upon the originally established 6 in. grid pattern, Section 3.5. (2) The drilling machine and the drill bit were aligned perpendicular to the slab surface using a spirit level. (3) Water was used to cool the drill bit. The drill bit was driven down to exactly 2-3/4 in. in the slab. The actual drilling time for each core was about 5 minutes on average after setting the drilling machine. 2.5 The B.O. Specimen Testing For testing the Break-Off specimens, the following procedure was followed after calibrating the load cell, Section 2.3, at each test age: (1) To ensure that no loose concrete has entered the slit, Fig. 1.2.1, compressed air was used to blow and clean the slit. (2) The oil pressure in the load cell, if any, was relieved by turning the valve shown in Fig 2.2.1, approximately half a turn counter clockwise. (3) Testing level was set at the high range on the load cell throughout this project, since all concrete mixes were expected to give more than 3000 psi compressive strength at the 28-day test age. (4) The valve was turned clockwise and closed finger tight. (5) The manometer red pointer for the load cell reading was set to zero. (6) The load cell was held in its proper position after aligning its axis with that of the Break-Off specimen. (7) Approximately one stroke per second was applied on the hydraulic pump, about 70 psi per second. The pumping was continuous until the Break-Off specimen snapped off. (8) The load cell pressure reading on the manometer, the test time, and dimensions of the test specimen were recorded. (9) Each specimen was numbered. They have been kept in the concrete laboratory environment for future investigation.

CHAPTER 3 CONCRETE TESTING AND PREPARATION OF SLABS AND TEST SPECIMENS 3.l General In order to determine the effect of each variable on the B.O. test results, a total of six different types of concrete were obtained from a local ready mix supplier. Each mix contained crushed limestone or rounded aggregates, and it represented one strength level. Mix l: 4000 psi, Mix 2: 6000 psi, and Mix 3: 8000 psi for crushed stone at the 28-day age. Each of these mixes were non-air entrained. For the rounded aggregates Mix A, 4000 psi, air entrained; Mix B, 6000 psi, and Mix C, 8000 psi, non-air entrained. From each mix one 4' x 4' x 5" and one 4' x 4' x 7" slab was made. Each slab provided a maximum of 48 test locations (at a nominal 6 in. center to center spacing), Fig. 3.l.l. A nominal center to center and edge distance of 6 in. was maintained in the process of locating sleeves or drilling cores. At each test age 5 or 6 B.O. specimens were planned for testing. Forms were made from l/2 in. thick plywood sheets. They were sprinkled with water and then form oil was applied prior to casting of slabs. From each mix a total of l7 cylinders were prepared in a standard manner as specified in the ASTM. Three cylinders were planned to be tested at each test age. 3.2 Concrete Table 3.2.1 gives the crushed aggregates concrete mix data per cu. yd. of concrete. Table 3.2.2 shows the chemical analysis of cement and fly ash used for Mix 1, 2, and 3. Maximum size of coarse aggregates used for all six mixes was 3/4 in. Each ready mixed truck load consisted of 2 cubic yards of concrete. In the case of Mix l, the concrete had no slump when it arrived. Ten gallons of water were added to the concrete in the truck and it was allowed to mix for 7 minutes. Slump was then 4-l/2 in. and the concrete was accepted. The concrete was workable and bleeding was normal. The amount of water recorded for each mix in Table 3.2.1 includes the amount of water added at the laboratory test site. For Mix 2, 7 gallons of water were added at the laboratory site and the truck drum was also allowed to rotate at the transit speed for 7 minutes. This brought the slump from l in. to 4-3/4 in. which was about the slump required for the desired workability. This concrete was cohesive and little bleeding was noticed. In the case of Mix 3, superplasticizer was added at the laboratory test site. The first slump value in Table 3.3.l was the slump after adding the superplasticizer. Before the superplasticizer was added the slump was 4-l/2 in. This concrete was wet and very workable. Normal bleeding was noticed. Mix proportions per cu. yd. of concrete for the three rounded aggregate concrete mixes, Mixes A, B, and C, are given in Table 3.2.3. Chemical and physical analysis for Type I cement used, for these mixes, are given in Tables 3.2.4 and 3.2.5, respectively. Chemical analysis for Type C Fly Ash used, for these mixes, is given in Table 3.2.5. All the information given in this section was obtained from the ready mix company that supplied the concrete.

In the cases of Mixes A and B, the first measured slump at arrival was 6-1/2 in. and 6-3/4 in., respectively. This was considered acceptable since the minimum slump requirement for this project was decided to be 4 inches. Therefore, no additional water was added. For Mix C the first measured slump at arrival was 2-3/4 inches. Therefore, three gallons of water were added. This increased the slump to 3-3/4 in., then two gallons of water were added and that increased the slump of Mix C to the required 4 inches. The amount of water recorded in the Table 3.2.3 includes the amount of water added at the laboratory test site. It should be noted that the workability of Mix C was very poor due to low water to cementatious ratio used in this mix, while Mixes A and B were very workable. It should also be noted that excessive bleeding was noticed in the case of Mix B, while normal bleeding was noticed in the case of Mixes A and C. 3.3 Fresh Concrete Testing To insure that a good representative sample of the concrete was obtained for testing from each truck load, at least one or two wheelbarrows full of concrete (approximately 5 to 10 cu. ft.) were usually discarded before taking concrete for testing, making cylinders, and casting slabs. The following tests were performed on each concrete mix: (l) concrete temperature: a metal thermometer was inserted in the fresh concrete for at least 5 minutes. The concrete temperature was then recorded, Tables 3.3.l and 3.3.2. (2) slump test: slump was measured in accordance with the ASTM Test Designation C-l43 [4]. Slump was measured twice; first l0 to l5 minutes, after the arrival of the ready mixed truck, and second 30-40 minutes after the first test. Tables 3.3.l and 3.3.2 show slump measurements for each mix. (3) density test: a half a cu. ft. volume bucket that weighed, when empty, 20.5 lb was used, Fig. 3.3.l. ASTM Test Designation C-l38 [3] was followed for the density test. (4) air content; the air content of the fresh concrete was measured in accordance with the ASTM Test Designation C-23l, [6], Tables 3.3.l and 3.3.2. The air meter used for this test was calibrated each time before use, Fig. 3.3.2. 3.4 Casting of Slabs Concrete was placed from the truck into the wheelbarrow. The concrete was then placed in the plywood forms for the slabs. Shovels were used to move the concrete in the forms. Concrete was placed until about half the thickness of the slab, then it was vibrated by an internal vibrator for 2 minutes. After the full thickness of the slab was achieved, concrete was vibrated again for 2 minutes. Concrete was allowed to bleed, then the surface was finished using a steel trowel.

3.5 Inserting Sleeves The exact position of each sleeve center was determined by a grid system, (strings 6" center to center and edge distances), Fig. 3.5.l. Sleeves were carefully inserted at each grid point. They were pushed in place by a rocking and twisting action, Fig. 3.5.2. Concrete inside the sleeve was then tapped by fingers to insure good compaction for the B.O. specimen. The sleeves were then moved up and down in place and brought up to the same level as the concrete surface as it's final position. This procedure was chosen after some trial and error tests were performed for best results, (prior to starting this project). For stiff mixes, i.e; low water to cementatious ratio mixes, a depression occurred within the confines of the sleeve during the insertion process. In such cases the sleeve was filled with concrete, tapped with fingers, and slightly jiggled from side to side. On the other hand, for wet concrete mixes, sleeves tended to move upward due to bleeding. In this case sleeves were pushed back in place, as necessary, to the level of the finished concrete surface. Some times this process had to be repeated until the uplift movement stopped after the initial setting had occurred. The slab surface was smoothed again after this. The concrete adjacent to the sleeve was tapped by fingers to insure good contact between the concrete and the sleeve. It was noticed at the time of testing that inserted sleeve specimens had some honey-combing on their surface. This honey-combing did not cause the failure plan to shift from its intended place at the 2.76 in. (70 mm) from the top surface of the slab to the place of the honey-comb. Thus, the honey-combing on the sides of the inserted sleeve cylindrical B.O. test specimens did not affect the B.O. readings. For all inserted sleeve specimens, the top diameter was about 0.l8 in. less than the bottom diameter. This is due to the design of the plastic sleeve insert, which made it easier to remove the sleeve insert at the time of the B.O. test. Even though the inserted sleeve specimens had more of a trapezoidal and not a cylindrical shape, top diameter l.99 in. (50 mm) and bottom diameter 2.l7 in. (55 mm) the B.O. reading was not affected because the bottom diameter at the failure plane was always maintained at 2.l7 in. (55 mm). Heavy grease, similar to that used for vehicle wheel bearings, was used for easier removal of the plastic sleeve after the concrete had hardened. 3.6 Curing of Slabs Slabs were covered with plastic sheets, immediately after the final finishing operations. The plastic sheets were kept in place for 7 days. Plastic sheets were removed from the slabs at the time of testing which was approximately 2 hours at each test age. Throughout the 28-day period, slabs were soaked with water once or twice a day. All holes in each slab resulting from the B.O. tests were also kept filled with water. After the seventh day testing was over, each slab had at least a total of 3l holes distributed over the entire area of the slab.

3.7 Cylinders Making, Curing, and Testing The ASTM Test Designation C-l92 [5] was followed in making and curing of concrete cylinder test specimens. The specimens were compacted in three layers by rodding each layer 25 times. The tops of the cylinders were smoothed with a steel trowel and then covered with plastic bags and stored for about 24 hours in the concrete laboratory environment, approximately 73oF + 3oF and 50% plus or minus l0% Relative Humidity. The next day after casting, the cylinders were demolded, marked, and placed in a lime saturated water tank for curing until the time of test. All specimens were capped with a hot sulfur compound in accordance with ASTM Test Designation C-6l7 [7]. Then they were tested in surface dry condition in accordance with the ASTM Test Designation C-39 [2]. The data recorded for the cylinder tests included cylinder diameter, failure load, time of test, type of failure, and remarks about the nature of the failure. The diameter for the cylinders were measured before testing them.