Research began in the laboratory with the screening of various aggregate sources, cementitious combinations, and chemical admixtures. Both Lafarge North America and Grace Construction Products were extremely helpful throughout this process. Different combinations and quantities of supplementary cementitious materials were tested, including fly ash, silica fume, slag, and even metakaolin.
We considered the effects these materials would have on fresh properties, early age, and ultimate strengths, as well as heat generation. The final mix design utilizes a ternary blend of Type I/II cement, class F fly ash, and silica fume. We selected a polycarboxylate-based, high-range water reducer to produce a workable and pumpable mix at a very low water-to-cement ratio.
Although the mix was not initially designed as a self-consolidating concrete mix, it exhibits very similar properties, with a typical spread of 25 inches. We also incorporated a hydration-stabilizing admixture to aid in slump retention and control the set time, especially during the summer.Reaching elasticity
Once we determine the required strength could be met, discussions with the structural engineer focused on the modulus of elasticity (MOE). Initial tests concluded that the desired MOE for the exterior columns could not be achieved with the locally available granite aggregates.
We then focused on using a limestone aggregate to achieve the required MOE. We selected Salem Stone Corp., of Salem, Va., to supply the limestone aggregate for the exterior columns. Martin Marietta Aggregates supplied the coarse aggregate for the interior columns and normal weight structural concrete.
The initial laboratory work also focused on concrete testing methods. We conducted a study to investigate the effect of different cylinder end preparation methods and studied using a high-strength capping compound, grinding the cylinder ends, and pad caps. Using a capping compound created the greatest chance for variability and human error, so we quickly excluded this.
We also compared ground ends to both 70-durometer and 90-durometer pad caps and found that testing variability was reduced when 90-d pad caps were used, while the measured compressive strength was slightly higher than those cylinders whose ends were ground and tested directly on the platens. The final specification allowed cylinders to be ground or 90-d pad caps to be used.
Field-testing, consisting of four-yard test batches, began in fall 2006. Initial field tests were conducted with bagged silica fume while each of our batch plants was being modified to handle bulk silica fume. We dedicated three central mix plants, all within a 5-mile radius of the jobsite, to the project. Trucks were typically batched and held at the plant yard for one to two hours to measure slump loss and determine the mix's response to re-dosing with high-range water reducers.Prep work
Internal training and a number of meetings with the contractor, engineer, and pumping company were held at the beginning of 2007. We worked closely with Chuck Gaston, the project superintendent for Batson-Cook, to make sure everyone understood how the concrete was going to perform. After several successful trial batches, everyone agreed it was time to put the mix to the test in a real-world application.