Performance of Portland Limestone Cements

Barrett, Sun, Nantung, and Weiss 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43

1

Performance of Portland Limestone Cements

Timothy J. Barrett Graduate Research Assistant Purdue University West Lafayette, IN 47907 USA E-mail: [email protected]

Hongfang Sun Post Doctoral Associate Purdue University West Lafayette, IN 47907 USA E-mail: [email protected]

Tommy Nantung INDOT Office of Research and Development West Lafayette, IN 47906 USA E-mail: [email protected] And

W. Jason Weiss (Corresponding Author) Jack and Kay Hockema Professor, Director of the Pankow Materials Laboratory School of Civil Engineering Purdue University West Lafayette, IN 47907 USA E-mail: [email protected] Submitted on July 31st, 2013, Resubmitted November 13th, 2013 Number of words: Body: 4945; Figures and Tables: 2,000 (8×250) = Total: 6,945 (7,000 +ref)

Barrett, Sun, Nantung, and Weiss 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64

2

ABSTRACT Portland limestone cements (PLCs) have recently been approved as a part of the ASTM C595/AASHTO M240 specifications. These cements are designed to enable more sustainable concrete production by replacing up to 15% of the clinker with interground limestone particles. The PLC’s represent a potential method to reduce the CO2 embodied in built infrastructure and extend the life of limestone quarries. This paper presents a comparison of the performance of three commercially available PLCs meeting the ASTM C595/AASHTO M240 specifications and three ordinary portland cements (OPCs) made from the same cement clinkers. An additional OPC was blended with limestone that had two different mean particle sizes. One OPC and PLC were used with a Class C fly ash. Each of these cementitious systems were used to produce typical concrete paving mixtures with the performance of these materials being quantified through a series of standardized tests. The results of the study show that the mechanical properties of PLCs have negligible changes as they relate to design practices and implementation. The early age volume changes of a PLC ground to levels consistent with achieving similar 28 day strengths were shown to be similar to that of the corresponding OPC, while PLCs that are ground significantly finer may lead to increased early age shrinkage. The transport properties show behavior that is +/- 30% of the conventional OPC, while the results for PLC systems containing fly ash show a synergistic effect with improved performance. The overall results of this study show that PLCs conforming to ASTM C595/AASHTO M240 can achieve similar performance as the OPC they would replace.

Barrett, Sun, Nantung, and Weiss 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110

3

INTRODUCTION In 2009, ASTM and AASHTO allowed the use of up to 5% interground limestone in ordinary portland cement (OPC) as a part of a change to ASTM C150/AASHTO M85. When the research described in this paper was initiated in 2010, a proposal was being discussed that would enable up to 15% of interground limestone in cement to be considered in ASTM C595/AASHTO M240. In 2012, this initiative was passed and portland limestone cements (PLCs) are included in ASTM C595/AASHTO M240 [1], however they are still not widely used. The proposal for increasing the volume of limestone that would be permitted to be interground in cement is designed to enable more sustainable construction, which may significantly reduce the CO2 that is embodied in the built infrastructure [2] while also extending the life of cement quarries [3]. Research regarding the performance of cements with interground limestone has been conducted, as these cements became widely used in Europe over three decades ago; much of this work has been chronicled in recent state-of-the art reports [3-5]. These reports identified additional research necessary for application of PLCs in North America. Specific areas of ongoing research include low temperature sulfate attack, potential interaction with deicers, and early age shrinkage and cracking potential. At present, studies are ongoing at the University of Toronto and the University of New Brunswick to better understand whether PLC has a different sulfate susceptibility than OPC. Additional studies are ongoing at Purdue University to assess the potential for issues due to chloride attack and early age volume change [6][7] of interground and blended limestone. It should also be noted that field trials [8, 9] have been performed on PLCs produced under the CSA Standard A3000-08 in Canada. In addition to these studies, recent work has focused on the potential application of using more finely ground limestone (e.g., nano-limestone) particles to accelerate hydration reaction of OPCs containing high volumes of supplementary cementitious materials (SCMs) [10, 11] as well as regulating the setting time of high volume fly ash systems [12]. Further, recent work has examined the use of jet-mill grinding as an alternative to conventional grinding when adding limestone to concrete [13]. The present work focuses on North American PLCs, which are specifically designed to have similar performance to the OPCs they would replace. This paper presents a study in which the potential application of commercially available cements containing up to 15% limestone was systematically assessed. Three commercially produced PLCs were obtained and compared to three commercially produced OPCs made from the same clinker. An additional cement was tested where the limestone was blended (i.e., not interground), enabling variation of the mean size of the limestone particles. In addition, one of each of the commercially produced OPCs and PLCs was used with fly ash. RESEARCH OBJECTIVES This project examines the potential impact of using cement with up to 15% of the cement replaced by limestone on the potential performance of concrete pavements and concrete in transportation structures. In order to quantify this potential impact, a series of standardized tests were performed to address the following objectives: 1. Examining the influence of limestone size and method of cement replacement (interground or additive) on composition and early age reactivity. 2. Examining the influence of limestone on the mechanical properties of concrete. 3. Examining the influence of limestone on the transport properties of concrete. 4. Examining the influence of limestone on the early age volume changes of concrete.

Barrett, Sun, Nantung, and Weiss 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142

4

The present work provides a condensed overview of these studies (a full report is available at [14]) with a focus on performance and application to practice. Objective 1 presents results on the size, shape and composition of the OPCs and PLCs. Objective 2 includes results on the compressive strength, elastic modulus, and flexural strength of concretes made with these materials, while objective 3 focuses on the porosity and chloride diffusion coefficients of these same concretes. Finally, objective 4 presents restrained shrinkage cracking performance on selected systems as measured by the dual ring test. CHARACTERIZATION OF THE OPC AND PLC Scanning Electron Microscopy and X-Ray Diffraction Back-scattered electron (BSE) imaging and element mapping techniques were used in scanning electron microscopy (SEM) to determine the shapes, sizes, and elements of the cementitious systems. The samples were prepared by blending approximately 10 g of powder with an epoxy resin to form a viscous paste, which was transferred into a 10 mm diameter x 55 mm tall glass container and cured at 23+/-1 °C for one day. The cured specimen was then cut with a lowspeed saw to obtain a plane surface and then subsequently polished using 15, 9, 3, 1, and 0.25 μm diamond paste for 4 min each on top of a polishing cloth. Before loading the samples into the SEM chamber, the samples were gold coated to form a conductive surface. The SEM was performed on a FEI Quanta 3D with field emission gun working at 12 kV. The collection time for element mapping was 2 h. The BSE images of one of the limestone systems and its corresponding OPC can be seen in Figure 1. The limestone was identified by using elemental mapping in conjunction with gray levels, which was done to avoid confusion between the similar grey levels of gypsum and limestone during the analysis. Using the sulfate to distinguish the gypsum from the limestone, the calcite present in each cementitious system was identified and is indicated as yellow on the images. In general, a smaller particle size can be observed in the interground PLCs in comparison to corresponding OPCs. Furthermore, it can be noticed that the limestone particles are generally smaller than the calcium silicate particles and the majority of the limestone occupies the smallest particle size regions. This is due to the softer nature of the limestone and is consistent with previous findings [15].

Barrett, Sun, Nantung, and Weiss

5

(a)

(b) FIGURE 1 BSE image of (a) OPC 1 and (b) PLC 1. (The limestone is indicated as yellow, all other cement phases appear as grey, and the black indicates epoxy) 143 144 145 146

X-ray diffraction (XRD) combined with Rietveld fitting was used to determine the phases present in each cementitious system and the content of each phase in the unhydrated state. The testing was conducted with a Bruker D8 instrument with a CuKα source (λ = 1.54 Å) at 40 kV

Barrett, Sun, Nantung, and Weiss 147 148 149

6

and 40 mA, with the Rietveld fitting being performed using Maud software [16]. composition of each of the OPCs can be seen in Table 1.

The

TABLE 1 Compositions of the OPC materials. Cement OPC 1 OPC 2 OPC 3 OPC 4

150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174

C3 S (%) 54.08 42.42 59.21 65.95

C2 S (%) 27.14 34.08 15.07 9.37

C3 A (%) 2.13 3.28 2.66 6.07

C4AF (%) 11.6 12.38 16.43 14.71

MgO (%) 0.52 2.28 1.79 2.18

Gypsum CaCO3 (%) (%) 2.81 0.00 2.48 1.73 2.63 0.00 1.25 0.00

Particle Size Distribution, Blaine Fineness, and Limestone Content The mean particle size, density, Blaine fineness, and CaCO3 content of each material is listed in Table 2. For each material designation, the number represents the clinker used to produce the cement, numbered in no particular order (e.g., OPC 1 and PLC 1 were each made from the same clinker, etc.). The CL and FL designations denote coarse limestone (mean size of 10.8 μm) and fine limestone (mean size of 1.3 μm) respectively, each of which was blended with OPC 4 at the time of mixing, while FA designates a Class C fly ash. The particle size distribution (PSD) of each material was measured by a Coulter LS32 laser sizer with high reproducibility (< 1%), with ethanol being used as a dispersant. The density required for PSD calculations was determined using a MicroMeritics AccuPyc 1330 Pycnometer. The Blaine fineness was also determined in order to compare the associated increase in surface area of interground PLCs. A Blaine permeability apparatus was used to perform this test following ASTM C204-11 [17]. It should be noted that it is generally necessary to grind the PLCs finer by approximately 8 to 10 (m2/kg) per 1% additional percentage of limestone in order to achieve similar 28 day strengths as the OPCs they are intended to replace [18]. For this study, PLC 1 and PLC 2 had an increase in surface area of approximately 15% (relative to each control, i.e. PLC/OPC Control) while PLC 3 had an increase in surface area of more than 30%. The content of CaCO3 was determined from the mass loss between 600 to 780 °C using thermogravimetric analysis (TGA) and differential thermal analysis (DTA) [19]. The interground PLCs were determined to have CaCO3 levels ranging from 7.5 to 10.5 %, however it should be noted that the limestone used in the productions of these PLCs is permitted to have CaCO3 contents as low as 75 %. These impurities in the limestone suggest that the actual limestone filler content in the PLCs ranged between 11 to 13%.


7

TABLE 2 Mean particle size, density, Blaine fineness, and calcite content for each material. Material OPC 1 PLC 1 OPC 2 PLC 2 OPC 3 PLC 3 OPC 4 CL FL FA

175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193

Mean size

Density

Blaine fineness

CaCO3

μm 9.4 7.9 7.9 7.7 7.1 5.7 9.9 10.8 1.3 9.8

g/cm3 3.12 3.12 3.16 3.11 3.17 3.13 3.17 2.70 2.70 2.74

m2 /kg 384 452 376 430 392 518 377 321 1069 331

wt% 2.9 7.5 2.0 8.4 1.9 10.6 0.5 97.2 99.5 -

MIXTURE PROPORTIONS, SPECIMEN PREPARATION, AND TESTING METHODS The concrete mixtures used in the study are shown in Table 3. This study assessed eleven different cementitious systems at three water-to-powder ratios (w/p) for a total of thirty-three different concrete mixtures. As previously mentioned, a total of four ordinary portland cements (OPCs), three complimentary portland limestone cements (PLCs), and two blended limestone cements were used to assess the effects of ground limestone additions in cementitious systems. Both of the blended limestone cements (OPC 4-FL and OPC 4-CL) had 15% of the cement replaced (by volume) with ground limestone. In order to assess the effects of fly ash on PLC’s, two additional cement systems were created by replacing 20% of the cement (by volume) in OPC 2 or PLC 2 with a class C fly ash (FA). All eleven of these cement systems were evaluated at water-to-powder ratios (w/p) of 0.38, 0.42, and 0.46 (the approximate range of permissible INDOT mixtures), corresponding to water-to-cement ratios (w/c) of approximately 0.41, 0.45, and 0.49 for PLC’s and blended limestone cements. The concrete mixtures had an aggregate volume fraction of 75%, with a coarse-to-fine ratio of 55:45. The naming convention for each mixture follows the format of [w/p – (OPC/PLC) # – (FA/CL/FL)], where FA signifies a fly ash replacement, CL denotes a blended limestone-cement system with coarsely ground limestone, and FL denotes a blended limestone-cement system with finely ground limestone.


8

TABLE 3 Mixture proportions and naming conventions. (1lb/yd3 = 0.593 kg/m3) (1 fl oz/100 lb powder = 65 mL/100 kg of powder) Mixture 0.38-OPC 1 0.42-OPC 1 0.46-OPC 1 0.38-PLC 1 0.42-PLC 1 0.46-PLC 1 0.38-OPC 2 0.42-OPC 2 0.46-OPC 2 0.38-OPC 2-FA 0.42-OPC 2-FA 0.46-OPC 2-FA 0.38-PLC 2 0.42-PLC 2 0.46-PLC 2 0.38-PLC 2-FA 0.42-PLC 2-FA 0.46-PLC 2-FA 0.38-OPC 3 0.42-OPC 3 0.46-OPC 3 0.38-PLC 3 0.42-PLC 3 0.46-PLC 3 0.38-OPC 4 0.42-OPC 4 0.46-OPC 4 0.38-OPC 4-CL 0.42-OPC 4-CL 0.46-OPC 4-CL 0.38-OPC 4-FL 0.42-OPC 4-FL 0.46-OPC 3-FL

194 195 196 197 198 199 200 201 202

Water/ Water/ Powder Cement (w/p) 0.38 0.42 0.46 0.38 0.42 0.46 0.38 0.42 0.46 0.38 0.42 0.46 0.38 0.42 0.46 0.38 0.42 0.46 0.38 0.42 0.46 0.38 0.42 0.46 0.38 0.42 0.46 0.38 0.42 0.46 0.38 0.42 0.46

(w/c) 0.38 0.42 0.46 0.41 0.45 0.49 0.39 0.43 0.47 0.47 0.52 0.57 0.41 0.46 0.50 0.50 0.56 0.61 0.38 0.42 0.46 0.41 0.45 0.49 0.38 0.42 0.46 0.44 0.48 0.53 0.44 0.48 0.53

Ordinary Portland Cement (lbs/yd3 ) 604 571 542 604 571 542 494 467 443 604 571 542 604 571 542 521 493 468 521 493 468

Portland Limestone Cement (lbs/yd3 ) 598 566 537 598 566 537 486 460 437 598 566 537 -

Ground Fly Ash HRWRA Limestone (By Volume) (By Volume) (%) (%) (fl oz/100 lb powder) 11.4 5.7 1.4 11.4 5.7 2.8 11.4 5.7 1.4 20 7.1 20 2.8 20 0.7 11.4 5.7 1.4 20 8.5 20 5.7 20 2.8 11.4 5.7 1.4 11.4 5.7 1.4 11.4 5.7 1.4 15 11.4 15 5.7 15 2.8 15 11.4 15 5.7 15 2.8

Mixing Procedure The mixing procedure was conducted in accordance with ASTM C192 [20]. The concrete was made in 0.056 m3 batches using a dual action, 0.085 m3 capacity pan mixer. The materials were batched at a temperature of 23±2°C, with the aggregates being prepared in the oven dry state. The fine and coarse aggregate were combined in a “buttered” mixer first, adding a portion of the batch water to control dust and ensure proper water absorption for the aggregate. Next, the cement and any additional supplementary materials (ground limestone and/or fly ash) were added to the mixer and combined with the aggregates until a uniform distribution was achieved.

Barrett, Sun, Nantung, and Weiss 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248

9

It should be noted that the coarse aggregate was sieved then recombined at the time of mixing to ensure consistent gradations between batches. The remaining batch water was then added and the time of water to cement contact was noted. Immediately following the addition of water, the HRWRA was slowly added directly to the concrete mixture. The concrete was mixed for three minutes, rested for three minutes, and then mixed for an additional two minutes. Compressive Strength and Young’s Modulus of Elasticity The compressive strength was determined in accordance with ASTM C39 [21]. The static modulus of elasticity was determined using the procedure in ASTM C469 [22]. A set of 100 mm diameter x 200 mm tall cylinders were cast to study the compressive strength and modulus of elasticity up to one year, with testing ages of 1, 3, 7, 14, 28, 90, 180, and 365 days. The cylinders were cast in two lifts, being vibrated and rodded 25 times after each lift. After one day of curing, the cylinders were demolded, sealed, and stored in a 100% relative humidity (RH) chamber at a temperature of 23±1 °C until tested. For each day of testing, three cylinders were tested to determine the compressive strength. Upon testing for the modulus of elasticity, the cylinders were fitted with a compressometer equipped with a linear variable differential transformer (LVDT) displacement transducer. The cylinders were then loaded to 40% of their ultimate strength two separate times. The resulting slope of the stress-strain curve from the second loading was taken as the static modulus of elasticity. For each day of testing, two cylinders were tested for every mixture with no cylinder being tested at more than one age. The cylinders for both tests were loaded at a rate of 241±34 kPa/s in a 3115 kN capacity hydraulic compression machine, utilizing neoprene end caps. Flexural Strength The flexural strength of the mixtures was determined in accordance with ASTM C78 [23]. For each mixture, two 150 mm tall by 150 mm deep by 530 mm long beams were cast. The beams were cast in two lifts, being vibrated and rodded after each lift. After one day of curing in the molds, the beams were demolded, sealed in plastic, and stored in a 100% RH chamber at a temperature of 23±1 °C. At six days of age the beams were unsealed and placed in a saturated lime water curing tank for 24 hours, at which point they were removed and wrapped in wet burlap until tested. The flexural strength of each mixture was determined under third-point loading, being loaded at a rate of 155±20 N/s. Restrained Shrinkage The dual ring test was used to quantify the autogenous deformations of two equivalent mortar mixtures. The dual ring testing device consists of two instrumented concentric invar restraining rings that operate in an insulated chamber [24, 25]. In this test, a mortar specimen was cast between the inner and outer rings in two lifts, being vibrated with a handheld vibrator after each lift then trowel finished upon completion. After casting, a copper tube and plate that is connected to a circulating water bath was loosely placed on top of the rings to maintain a constant temperature. Strain measurements were automatically recorded at five minute intervals and used to determine the stress that accumulates in the sample. Migration Cell and Stadium Lab The diffusion coefficients for ionic species were measured using Stadium Lab and a migration cell. In this test, the intensity of electrical current passed through a 100 mm diameter by 50 mm

Barrett, Sun, Nantung, and Weiss 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293

10

thick cylindrical specimen is monitored over a 14 day period. The samples used for this test were cut from a set of 100 mm diameter by 200 mm long concrete cylinders that were sealed and placed in a chamber at 100% RH and 23 ± 2 °C for 90 days. After the samples were cut, the sides of the samples were sealed with an epoxy after which they were vacuum saturated with 0.3 M NaOH for approximately 18 hours. Once saturated, the samples were mounted between a cell filled with 0.3 M NaOH solution (downstream) and a cell filled with 0.5 M NaCL + 0.3 M NaOH solution (upstream). A constant DC potential of 20V was maintained across the specimen for 14 days while the voltage, current, and temperature were measured and recorded at 15 minute intervals. In conjunction with the migration cell testing, the volume of permeable voids of the samples was determined in accordance with ASTM C642 [26] (with the exception that boiling was replaced with vacuum saturation). For this test, additional 100 mm diameter by 50 mm thick samples were cut from 100 mm diameter by 200 mm tall cylinders at an age of 90 days, sealed on the lateral sides, and placed in an environmental chamber at 50±1 %RH and 23±1 °C. The mass change of the samples was monitored until a mass equilibrium of ±0.5% was reached, at which point the samples were oven dried then vacuum saturated. Using the oven dry mass, saturated mass, buoyant mass, and conditioned mass, the volume of permeable voids was able to be determined. The results from the migration cell and the volume of permeable voids were entered into STADIUM Lab software to evaluate the ion diffusion coefficients and the tortuosity of the samples [27].

RESULTS AND DISCUSSION Mechanical Properties The intent of introducing portland limestone cements (PLCs) under ASTM C595 is to address the growing concerns of sustainability in cement and concrete production by providing an alternative for ASTM C150 Type I/II ordinary portland cements (OPCs) that achieves similar performance while reducing clinker content. As such, the implementation of these materials is subject to the present engineering design practice, which focuses on the strength of materials and the prediction of these properties using codified equations. To address these concerns, the following series of figures were prepared. FIGURE 2(a) shows the compressive strength of the concretes made with PLCs plotted against each corresponding OPC made using the same clinker. The dashed line on the plot shows a oneto-one relationship, while the solid lines indicate ± 10%. As it can readily be seen, these mixtures containing limestone fall mostly within 10% of a one-to-one strength ratio in comparison to the OPC system they are intended to replace. FIGURE 2(b) shows the modulus of elasticity for each concrete mixture tested, with the dashed line on the plot indicating the result of ACI 318-11, Section 8.5.1 [28]. The results indicate a close relationship in modulus of elasticity between OPCs and PLCs, while all mixtures lie 55% above the ACI 318-11 equation. FIGURE 2(c) shows the flexural strength as a function of the square root of compressive strength for each concrete mixture tested, with the dashed line indicating the result of Equation (9-10) in ACI 318-11. As it can be seen, each of the mixtures tested lie well above the code equation, with an average increase of 55%. These results collectively indicate that the use of PLC instead of

Barrett, Sun, Nantung, and Weiss 294 295 296 297

11

OPC will have minimal effect on the mechanical properties for use in the design of transportation structures.

PLC Compressive Strength (psi)

12000 w/p = 0.38 w/p = 0.42 w/p = 0.46

8000

4000

0 0

298 299 1.2E+7

Elastic Modulus (psi)

1E+7 8E+6

4000 8000 12000 OPC Compressive Strength (psi) (a) 0.38 - OPC 0.38 - PLC 0.42 - OPC 0.42 - PLC 0.46 - OPC 0.46 - PLC ACI 318

6E+6 4E+6 2E+6 0

300 301

0 40 80 120 Sqrt. Compressive Strength (psi) (b)


12

Modulus of Rupture (psi)

1500

1200

OPC PLC ACI 318

900

600

300

0

302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330

0 40 80 120 Sqrt. Compressive Strength (psi) (c)

FIGURE 2 (a) PLC compressive strength versus reference OPC strength (dashed line indicates a 1:1, solid lines indicate ± 10%) (b) Elastic modulus of each concrete mixture (dashed line indicates result from ACI 318, Section 8.5.1 [28]) and (c) Flexural strength of each concrete mixture (dashed line indicates result from ACI 318, Equation (9-10) [28]). (1 psi = 47.9 Pa) It should also be added that, when the data presented in Figure 2(b) is compared in the relative sense, the modulus of elasticity of each PLC shows in general a negligible difference compared to their OPC reference. The maximum reduction in elastic modulus of any PLC at any age or w/p is 7%. This slight reduction is likely attributed to the replacement of hydration products with a ground limestone powder that is slightly less dense than the cement it replaced (specific gravity of 2.70 versus 3.15, approximately). Likewise, a relative comparison of the flexural strength of each mixture (PLC/OPC) shows an average increase of 3% across all w/p for the PLCs relative to their control mixture, with a maximum reduction of 6% for all mixtures. The most significant change observed was with blended limestone cements, where an average reduction of 9% was observed. This reduction in flexural strength increases with decreasing limestone particle size, (from 6% for the CL to 12% for the FL) and is greater with increasing w/p. Compressive Strength A closer comparison of the normalized compressive strength of all limestone systems relative to their OPC counterparts (PLC/OPC) can be seen in Figure 3. The first trend that can be noticed is the relative performance of the PLC systems. In general, an initial increase in relative compressive strength is seen in the PLCs at early ages (as expected from an increase in fineness) but this improvement diminishes with age due to the dilution of cement. At 28 days, the PLC 1, PLC 2, and PLC 3 cements exhibited an average increase in strength of 1.5%, with a maximum

Barrett, Sun, Nantung, and Weiss 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350

13

reduction of 6% at any w/p. In contrast to the interground cements, the blended cements (OPC 4–FL, OPC 4-CL) exhibit an average reduction at any given age or w/p of 8.5% with a maximum strength loss of 20%. In general, the influence of the w/p on the relative compressive strength of the PLCs is negligible. Likewise, the OPC 4-FL systems show a consistent average reduction in strength of 5% across all w/p. The OPC 4-CL systems, however, show an increasing reduction in strength with increasing w/p. When the effects of limestone fineness in the blended cements is analyzed, the general trend seen is that finer limestone additions result in a smaller reduction in strength (from 5% for OPC 4-FL to 12% for OPC 4-CL). This is mostly a consequence of the difference in particle packing of each of these systems, where the presence of the CL causes larger initial particle spacing [29, 30]. This results in a slower rate of hydration which leads to reduced strength development. When the influence of fly ash in these systems is assessed, two observations can be made. The first is that the inclusion of fly ash in either OPCs or PLCs leads to higher absolute strengths after 7 days. The second observation can be made from Figure 3, where the PLC 2-FA system shows a relative increase in strength of more than 10% at early ages for w/p of 0.42 and 0.46. This suggests that the increased rate of hydration of the PLC is able to overcome the slow early age reactions typically seen in mixtures containing FA. The result is a mixture that has the benefit of improved long-term performance due to the presence of FA without the slow onset of strength gain.

Barrett, Sun, Nantung, and Weiss 120

14

w/p = 0.38

Normalized Compressive Strength (PLC/Control)

110 100 90 80 0

Age (d)

(a) 120

1 3 7 14 28 90 180 365

w/p = 0.42

110 100 90 80 0 (b) 120

w/p = 0.46

110 100 90 80 0

PLC 1

PLC 2

PLC 2-FA PLC 3 (c)

OPC 4-CL OPC 4-FL

FIGURE 3 Normalized compressive strength of limestone cement systems with w/p of (a) 0.38, (b) 0.42, and (c) 0.46. 351 352 353 354 355 356 357 358 359 360 361 362 363 364

Restrained Shrinkage As previously discussed, it is frequently necessary to grind PLCs finer than OPCs in order to overcome the reduction in hydration rates due to the dilution of cement. The finer nature of the PLCs increases the early age hydration, but studies have also shown that the cement fineness influences the rate of shrinkage development and cracking potential. Bentz et al. studied several cements of varying finenesses and observed that the finer cements showed higher shrinkage and stress development [29]. Chariton and Weiss examined the restrained shrinkage cracking of a commercial cement at different finenesses and reported that finer cements resulted in higher stress development and increased cracking damage with earlier ages of visible cracking [31]. One of the central arguments for the increased shrinkage in finer cements is related to the suggestion that capillary stress develops in the pore fluid as a pore dries, as described by the Young-Laplace relationship which states that the capillary stress is inversely proportional to the radius of the pore being emptied. Finer cements have been measured to have more smaller

Barrett, Sun, Nantung, and Weiss 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390

15

pores, suggesting higher levels of capillary stress development, and therefore shrinkage [32]. In the case for PLCs, the intergrinding results in a reduction of larger cement particles with the majority of the limestone particles occupying the smallest size ranges [7]. The resulting pore size distribution in PLCs shows an increase in the very small pores (