Like in life, there are construction jobs where everything must go according to plan, when second chances are limited or non-existent. The recent uprighting of a seriously tilted 150 year-old masonry bridge in Puerto Rico was that sort of occasion.
Completed in 1853, the Rio Piedras bridge is the oldest surviving part of the Central Highway from San Juan to Caguas. The brick arch bridge is listed in the National Register of Historic Places and had been in continuous use until one of its three spans collapsed. Now, the U.S. Department of Transportation is funding a $4 million, two-year restoration.
On the Rio Piedras Bridge, one of three spans is collapsed, and the column at right is badly canted.
The badly canted east column adjacent to the collapsed span presented a major challenge to crews. Some of the timber piles supporting the column sunk from their original level. As a result the column, roughly 7 by 25 ft. (2 by 7.6 m) at its base, had settled 14 in. (35.6 cm) at the upstream end and 1 in. (2.5 cm) along one side.
Suitable hydraulic jacks could easily supply the force necessary to lift the column out of the river-bottom mud and reposition it. The real challenge was how to apply the lifting forces to the column without fracturing it.
The four lifting beams project through the canted column, and a pedestal for jacks sits beneath them. A similar pedestal is situated on the other side of the column. (Diagonal braces had been installed to prevent further tilting.)
As might be expected, core samples from the column confirmed a low tensile strength for the 150 year-old mortar. (That posed no threat to the bridge itself, since all elements of the bridge are in compression.) It was clear that the lift points would have to be located as low as practicable on the column.
Just above the column’s pier, crews cut four square holes for lifting beams through from side-to-side. Next, a hollow-core mold was fabricated within and around each hole. Then a low-viscosity, slow-curing grout was injected at 500-600 psi and maintained under pressure for several days to permeate the column material surrounding each hole.
Meanwhile, crews poured large concrete footings topped by pedestals for hydraulic jacks parallel to each side of the column. After curing was complete, 16 by 16 in. (40.6 by 40.6 cm) lifting beams were inserted through the four holes.
As an additional measure, an I-beam was run across the four lifting beams on each side of the column and secured to the original column pier at each end The extra cross-beams would serve to distribute the lifting forces and also provide some lifting force at each end of the pier. At this point, the column was ready to be lifted.
An Enerpac synchronized lifting system provided and orchestrated the lifting forces at the ends of the lift beams. This particular system uses eight, 150-ton (136 t), 6-in. (15.2 cm) stroke jacks. Each jack is equipped with a locknut and accompanied by a pressure gauge, a flow control valve, a solenoid valve and a position sensor.
The position sensors provide input signals to a PLC-based control system, which operates the solenoid valves feeding the cylinders.
The sensors were simply C-clamped to each lift beam. They operate by means of a flexible extension line that unwinds from the sensor housing in a manner similar to a tape measure. The end of the sensor extension line was held in place on the concrete pedestal by a weight.
The hydraulic system was powered by a 10,000 psi, 1.5 hp, radial piston, 115 V electric pump with a 10 gallon reservoir.
With force and displacement monitored at all eight lift points, the lift began.
The control system gradually applied pressure to the cylinders in proportion to the distance each lift point had to be raised. Great care was exercised to maintain a preload at all lift points, including those that required no movement.
As the cylinders began to extend, it became clear that the west side of the column was not moving. Instead, the jacks on that side were pushing their footing into the riverbed. Attention now focused on applying jacking forces along the west side until that footing either stopped moving or sank too far.
Fortunately, the west footing sank quite uniformly and stopped moving after it had dropped several inches. Lifting of the column resumed.
The lift had to be done in stages in order to achieve the 14 in. displacement required at the south end of the column. Each time the cylinders were extended, jack stands with shim plates supported the load while the cylinders were retracted and repositioned for the next lift stage.
Instead of using double-acting or spring-return cylinders, engineers equipped the hydraulic pump system with a venturi-and-valve arrangement to provide a quasi power-return capability for the single-acting jack cylinders. This economical arrangement served very well to retract the cylinders between lift stages.
To the crew’s relief, the column did not fracture. With a little extra lift applied to the west side to compensate for the “soft” footing, the jacking was completed in about three hours. The lift exposed the original timber pile butts, and they were found to be in excellent condition, having been fully submerged. The presence of anaerobic bacteria in the long-undisturbed muck beneath the column was undisputable — quite a stink emerged.
Finally, the newly positioned original pier was stabilized by pouring concrete under and around it. Whereas brute-force jacking might well have reduced the 19th century bridge column to a pile of fractured ruble, synchronized lifting successfully restored it to a true upright position.