2. Experimental

Three mortar mixtures were prepared: 1) a control low w/b mixture with neither SRA nor fine LWA, 2) a mixture with fine LWA prewetted with a 10 % by mass fraction solution of the SRA (FLAIR delivery), and 3) a mixture with fine LWA prewetted with only water while the same quantity of SRA as in the previous mixture was now added directly to the mixing water (conventional delivery). The LWA used in this study was an expanded shale, passing through a #8 sieve, but retained on a #16. It had a saturated surface dry specific gravity of 1.67 and a water absorption of 20 % per unit dry mass LWA. The prewetted LWA was prepared by first oven drying, cooling to room temperature, and then mixing in a sealed plastic container with the appropriate volume of either water or the SRA solution. The prewetted LWA in the sealed container was then placed in a chamber maintained at 23 oC for a minimum of 24 h. The details for the prepared mortar mixture proportions are provided in Table 1. Cement and Concrete Reference Laboratory (CCRL) proficiency cement sample 152 [8] was used for all three mixtures. The mortars were prepared in an epicyclic mixer, with the water and chemical admixtures being placed in the mixing bowl first. Then, the cement was added and mixing performed on low speed for 30 s. All of the sands were added during 30 s of further mixing on low speed, followed by 30 s of mixing on medium speed. After a rest of 90 s, final mixing on medium speed for 1 min was performed. Mixtures 2 and 3 were prepared with the prewetted LWA replacing an equal volume of the S15 coarse sand.

The air content of the fresh mortar was determined according to ASTM C185 [9], and the following specimens were prepared: 50 mm cubes for compressive strength testing, 25 mm diameter by 400 mm corrugated tubes for evaluation of autogenous deformation [10], and 25 mm by 25 mm by 285 mm prisms for measurement of drying shrinkage. All specimens were placed in an environmental chamber maintained at 23 ºC. The capped corrugated tubes were removed periodically and their length determined using a digital dilatometer [10]. The cube and prism molds were first sealed in two plastic bags and then placed in the environmental chamber. After 1 d of curing, the cube and prism specimens were demolded and placed in a new set of double plastic bags. After 3 d of sealed curing, the prisms to be used in the evaluation of drying shrinkage were removed from their plastic bags and placed in a 23 ºC, 50 % relative humidity environment. Both their mass and length were periodically recorded.

For each of the three mortar mixtures, three cubes were prepared for each compressive strength evaluation (after 7 d, 28 d, and 91 d of sealed curing), four prisms were prepared for the measurement of drying shrinkage, and two tubes were prepared for the measurement of autogenous deformation. After the cubes were broken at the appropriate age, small portions of one per set were retained and ground to a powder with a mortar and pestle, for the subsequent measurement of degree of hydration by loss-on-ignition (LOI) techniques, correcting for the LOIs of the original cement powder and the sands present in each mixture. Based on a propagation of error analysis, the estimated uncertainty in the calculated degree of hydration was 0.01, assuming a coverage factor of 2 [11].