Work Description

First a survey of the bridge was conducted to ascertain which existing or non-historic materials would not be saved. These included extensively deteriorated historic materials as well as recent accretions.

Dismantling of the bridge was conducted by experienced private contractors overseen by RIDOT. Materials that could not be saved were removed and discarded, including weight-limit signs, chain link fencing, asphalt paving, and the timber deck. Once this material was removed, the timber deck stringers and iron web braces could be dismantled.

At this point a temporary bracing system was built to stabilize the trusses before construction work proceeded (see figure 5). This wooden system was extremely important--without it, the trusses might have twisted or overturned during the remaining disassembly work.

Figure 5. A temporary bracing system assured the protection of the bridge during disassembly.

With the trusses stabilized, the deteriorated floorbeam hangers were removed. The floorbeams themselves were lifted, one at a time, by crane onto a flatbed truck, and transported to the shop. With the temporary bracing removed, the trusses were then attached to the crane and lifted in one piece, then placed on their sides on the truck (see figure 6). The plans had specified that the trusses be placed on the truck in an upright position, with protective bracing. But when the first truss was mistakenly laid on its side, the deteriorated lattice girder stiffener was severely bent. Fortunately, this lattice girder was slated for disposal before the damage occurred. The other truss was removed without damage.

Figure 6. Although plans specified that the trusses be placed in an uprightposition, they were placed on the truck on their sides. The lower truss was damaged during the loading operation. Photo: Beta Engineering.

Fabrication/Rehabilitation

Once at the shop, the bridge was completely taken apart. All pins, beams, bars, nuts, and bolts were labeled and inspected to determine their structural condition (see figure 7). The parts were dismantled using a variety of methods. Threaded parts were unscrewed; some rivets had to be drilled out. Other parts were heated with a torch until they expanded, then cooled; the resulting contraction allowed them to be removed. The bridge's pinned construction made disassembly easier, as fewer pieces were actually riveted together.

Some parts were so deteriorated that they could not be preserved and re-used as planned, including the end posts, lattice girder stiffeners, and all four vertical (lattice girder) stiffeners. This was somewhat disappointing, as the original specifications called only for the removal of the floorbeam hangers, and the intent was to save as many of the original parts and fasteners as possible. The new, matching members would be more cost-effective and actually stronger than repairing the original pieces. The other drawback was that the new elements would need to be fabricated of steel, rather than higher-priced wrought iron. Once completed, however, the steel members would be visually indistinguishable.

Figure 7. The pieces were disassembled in the shop then cleaned of rust and paint. The worn condition of the eye bars can be seen in the center of the photograph. Photo: Beta Engineering.

The salvageable parts were placed in a Wheelbrator, a machine that removes rust through repeated vibration. The iron members were then sandblasted to remove paint and any remaining rust, then subjected to magnetic particle testing to determine which pieces had unseen damage or decay. The test was useful in determining wear in some of the parts, particularly the eyebars of the lower chord and the middle panel tension rods. In addition, some of the pins had worn by more than 1/4" over time.

In order to reuse as much of the historic material as possible while ensuring sufficient structural integrity, it was decided to "build up" through welding those pieces that were worn down. In order to match the metallurgical composition of each part, a chemical test was undertaken. This testing allowed for an exact metallurgical match between the new built-up welds and the original metal of each part. Such a match reduced the possibility that a harmful physical or chemical reaction would take place between the weld and the wrought iron.

Meanwhile, the endposts, lattice girder stiffeners, and vertical posts were being fabricated from steel (see figure 8). The original end posts had been constructed as a box section; one plate and two angle iron pieces were riveted together to form a column with one lattice side.

While some old rivets were eventually located, they were hardly ready for use. Only the rivet heads could be found, and there seemed little use for them. But after some thought, a creative solution was formulated. The shanks of the rivet heads were threaded, and the plates screwed together. For structural strength, the pieces were also welded, but this weld is invisible from the outside. The result is a welded box girder that looks, from the outside, like an authentic riveted member. Had no rivets been found, a similar procedure could have been followed using hack bolts. (These are bolts with heads that look like rivets but have nut fasteners.)

Figure 8. New end posts (foreground) and lattice vertical posts (background, center) were fabricated from steel. The rivets are actually threaded; hidden bolts help keep the pieces together. Welding provides additional strength. Photo: Beta Engineering.

The final challenge lay in the floor beams. The new site was nine to ten feet narrower than the original site. To fit the new foundation, the bridge had to be reduced in width. This was done by cutting the ends of the floorbeams by four to five feet on each side. Fortunately, the cut was made near a vertical brace in the floorbeam, matching the original ends, which also terminated near a vertical brace. The lateral bracing attachment points were removed from the excess lengths and re-installed on the shorter floorbeams. New U-bolts of round stock (as opposed to the original square stock) were also fashioned.

Once the built -up welds and fabrications were completed, the entire bridge was sprayed with an epoxy primer, followed by two coats of red iron oxide paint. Because none of the early paint remained, a compatible color was chosen in keeping with the bridge's historic character.

The final problem was that the old bridge foundation at the new site was intended to support a shorter span. Since reducing the length of the span was not possible, a new foundation would need to be built for the relocated bridge. The old foundation of the bridge that once occupied the site was to be retained and preserved. First, cut granite blocks from the Providence River Relocation Project were obtained and transported to the new site. The new foundation was laid so that the bridge, as installed, would clear the old foundation by approximately one foot.

When the new foundation was completed, the trusses were lifted by crane to the site, installed in place, and secured with temporary wooden bracing (see figure 9). The rest of the bridge (floor beams, hangers, stringers, etc.) was then
assembled, in the reverse of the dismantling process (see figure 10), and a new hardwood deck was installed.

Figure 9. The trusses were lifted by crane and set into positionon in the new site. Photo: Beta Engineering.

For safety, a contemporary steel pedestrian railing was installed on each side. Whether the historic bridge had a railing or not is unknown, but any such early railing would probably have consisted of only one or two horizontal members. The pipe railing on the Stillwater Bridge was chosen because it matched a railing design used throughout the park and met applicable code requirements. To lessen the visual impact and distinguish it as a new feature, the railing was painted black. As noted, the historic bridge members were painted red.