Once the casings were sealed and grouted, RCD was used to excavate each 6.5 ft (2 m) rock socket. The pile top rotary table spun a 67.5 ton (61 tonne) drill string, including a full-faced roller cutting bit attached to the bottom. In the reverse circulation process, compressed air is injected into the bottom of the drill string, creating a pressure differential that airlifts cuttings back to the surface. Flowrates on the order of 2,000 gpm (7,570 Lpm) are typically generated by this type of system. However, strict environmental regulations prohibited any water to be supplied from or discharged into the nearby Newtown Creek without first being treated to bring heavy metals, pH and solids down to acceptable levels. Instead, all flushing water was supplied from nearby hydrants and recirculated in a continuous loop, settling out the rock cuttings in a lined detention pond. Further complicating the installation, this phase of the work had to be performed as a two-shift operation during the dead of winter in order to maintain the project schedule. Temperatures typically rose above freezing for only a few hours each day, requiring a significant effort to keep pumps running and prevent supply and discharge piping from icing up. Due to the depth of the shafts, rein- forcing cages were constructed in three sections, using internal template rings attached to “hourglass” stiffening frames. The fully assembled cage contained 18 No. 18 75 ksi (517 MPa) vertical bars, with bundled bars in the upper 60 ft (18.3 m) and weighed nearly 50 tons (45.5 tonnes). Custom designed and built lifting rings were used to pick and suspend the cage sections during splicing and concrete placement. Although the concrete supplier’s ready- mix plant was located only about 2 mi (3.2 km) from the site, local Brooklyn traffic would frequently push delivery times to 45 minutes or more. A 6,000 psi (41.3 MPa) self-consolidating concrete (SCC) mix, developed for previous Skanska pro- jects, was chosen for its ease of worka- bility and extended placement time. Quality Control Since the shaft design relied on end bearing to provide nearly 80% of its capacity, ensuring cleanliness of the rock socket bottom prior to pouring concrete was paramount. A miniature shaft inspection device (Mini-SID) was employed to determine the depth of sediment present at the tip of each completed shaft. Five individual bottom locations per shaft were probed, with the requirement that 50% of the base of each shaft have less than 0.5 in (13 mm) of sediment and no location exceeding 1.5 in (38 mm) at the time of concrete placement. To meet these criteria, each shaft was airlifted, removing any fines leftover from the RCD and exchanging the entire fluid column with fresh water. Further, during construction of the first shaft, after a successful Mini-SID inspection was achieved, a 72-hour waiting period was imposed and the shaft inspection process repeated. This demonstration alleviated concerns that the shafts could fall out of specification if a weekend or unexpected downtime period occurred between the time of the Mini-SID inspection and concrete placement. A “Chicago Method” Osterberg Cell (O- cell) load test, performed on the first completed production shaft, was then carried out to confirm the design end- bearing assumptions. This variation on the traditional O-cell test places the hydraulic jack at the shaft tip with a circular bottom plate of a reduced diameter compared to the rock socket. Unit end-bearing values are able to be verified by a single 6,000 kip (26,689 kN) capacity O-cell jack, which applies load onto the smaller diameter plate using the available side shear capacity and dead weight of the shaft above as a reaction. Picking the bottom section of the O-cell cage RCD settlement pond DEEP FOUNDATIONS • SEPT/OCT 2015 • 89
