U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
|This report is an archived publication and may contain dated technical, contact, and link information|
Publication Number: FHWA-HRT-04-098
The Fitch's Covered Bridge was originally built in 1870 on Kingston Street over the West Branch Delaware River in the Village of Delhi to replace a previous covered bridge destroyed by floods. The bridge was built by James Frazier and James or Jasper (conflicting information) Warren. The bridge proper was built for about $1,900 and the stone abutments were erected by M. Hathaway and W. A. Cummings at a cost of $725. When the Town decided to replace the covered bridge with a modern iron bridge in 1885 by the popular Groton Bridge Company, the decision was made to relocate the 15-year old covered bridge to a site upstream in East Delhi, named for a previous crossing of Fitch's Bridge. The relocation of the span is credited to David Wright and a town crew. Figure 188 depicts the bridge early in this project. The hole in the roof and siding removal allowed the steel piles to be driven through the roof for support of a steel falsework system.
The bridge is supported by Town lattice trusses, so named for Ithiel Town who received his first patent for this truss configuration in 1820. Unlike the majority of Town lattice trusses, the Fitch's Bridge contained splayed lattice members at the end of the trusses (at the time of our new rehabilitation). The splayed lattice members are the result of an alteration of the bridge that was originally built without the splay, typical of the traditional Town lattice. The splay arrangement may have resulted from trying to fit the bridge from Delhi into existing abutments at Fitch's Crossing that supported a shorter span (the existing lattice members, prior to our work, still contained the original hole pattern compatible with traditional parallel lattice. Comparison of the open original holes demonstrates how the members were rotated about one of the holes in the top chord). A recently obtained photograph depicts the bridge prior to its rehabilitation in 1976 and the splayed lattice members are evident then. Further, their condition had deteriorated significantly and they were supplemented with large steel plates, apparently installed in 1976. In any event, the splayed lattice members are not unique to the Fitch's Bridge, and are found on other extent covered bridges, especially in this geographic area. Also note that the bridge contains only a single level of top chords, which is also somewhat unusual, but found in many geographic areas. Figure 189 depicts a view of splayed lattice members prior to this project. Note the empty holes of the original construction, prior to the splay of the members. And note the extreme deterioration of the ends of the lattice, a condition that was hidden behind steel plates.
The original interior knee braces were too weak to maintain the dimensional stability of the bridge. At some time external bracing was added; some identify these features as buttresses or "elephant ears". These forms of external bracing exist on other bridges; they were not original on typical historic covered bridges, but were found on some covered bridges. It is not known if they were original to this bridge.
The original floor was hung beneath the lowermost chord via iron rods. This is supported by a recently obtained photograph of unknown date (figure 190). The photo also shows a ramp floor at the north end of the span but inside of the bridge. The ramp transitions to a floor that would appear to be at the correct elevation to have been supported beneath the lower bottom chords. In 1976, the floor was modified to install it onto the lower bottom chord in accordance with the more conventional approach for this type of bridge.
The 1976 replacement floor utilized floor beams comprised of three 3 by 12 (75 by 300 mm) members spiked together that supported longitudinal stringers on top of the floor beams and transverse timber nail-laminated decking. The majority of the floor beams were placed on an 8 foot- (2.4 m) spacing, although there was a panel near mid-span with a 12 foot (3.6 m) spacing. The stringers were positioned at about 2 foot (610 mm) spacing. The floor system was deteriorated and too weak for retention. Figure 191 depicts the partially removed 1976 floor system. The near edge of the floor is the transverse nail-laminated decking. The longitudinal stringers are obviously not uniformly spaced. Don't be deceived by the transverse diaphragms between stringers; the floor beams are beneath as visible on out in the span.
The bridge was covered with metal roofing, last replaced in 1976, and was in bad condition.
The dry-laid stone abutment is still exposed on the South Abutment, although a concrete cap exists atop the stone, while concrete has been installed in front of the stone of the main portion of the North Abutment (refer to figure 190). Steel sheet piling has been installed in front of the downstream wing wall. This work is believed to have been performed in 1976. The North Abutment with a concrete facade in front of the dry-laid stone and steel sheet piling along the wings is depicted in figure 191. The rusty steel frame at the corner of the backwall and transverse across the top of the stem was left over from the 1976 work; it was used for erection and not for support of the bridge and was removed in this project.
The lattice splay feature was not retained, since it was not original to the bridge and it leads to stress concentrations that are nearly impossible to accurately analyze. Further, the members were so badly deteriorated as to be unusable. Since the abutments were judged to be sound in their current condition, it was decided to remove a portion of each corner to allow installation of a longer truss (see figure 194). The longer structure resembles the original structure when built for the Delhi site (to the extent we can deduce that appearance.)
The chord elements of the bridge were deteriorated at many locations, especially along the bottom chord at the intersection of the lattice elements, and along the top chord. The entire upper and lower bottom chord elements were replaced due to deterioration. A number of the existing top chord elements were replaced due to deterioration, along with all new elements at the ends to make a longer truss. Similarly, several of the lattice elements were replaced due to deterioration. In all cases, the chord and lattice elements were replaced with similar-sized sawn Southern Pine or Douglas Fir members. Traditional wooden pegs (trunnels) were used to connect the truss elements and those existing pegs in good condition were reused.
The Elephant Ears were removed and a stronger internal bracing system was installed.Most of the original upper lateral members were retained. The tie beams and knee braces were replaced with stronger components. Traditional wooden peg connectors and matching timber wedges similar to those in the original construction were used.
Another feature related to the internal bracing system was the use of larger rafters (3 inches by 12 inches) (75 by 300 mm) adjacent to the tie beams and knee braces; these members are identified as "Principal Rafters." The common rafters were replaced using 2 inches by 8 inches (50 by 200 mm) sawn members. Figure 195 depicts the reworked internal bracing system. Note the retained darker upper lateral members.
The floor was completely replaced, using the common configuration of floor beams and longitudinal decking, typical of most Town lattice Bridges. Floor beams are often spaced at 2 feet (600 mm) on centers in many Town lattice bridges, yet analysis work determined that the floor beams in this bridge could be spaced at 4 feet (1.2 m.) The floor beams are comprised of glue-laminated components, to gain sufficient reserve capacity to be able to handle overweight vehicles. The longitudinal decking is conventional 4 inches by12 inches (100 by 300 mm) Douglas Fir timber plank. We also installed sacrificial 2 inches- (50 mm-) thick oak running planks on top of the primary deck planks. An image of the installation of the longitudinal planks is depicted in figure 194.
New horizontal metal rods were installed to hold the floor system and bottom of the trusses together. Such rod systems have become quite common installations over the past many years for those locations where adequate underclearance is not possible and flood waters and debris can pull the downstream truss out from under the floor beams. Similarly, metal rod hold-down anchors were installed at each corner to prevent the truss from floating upward during extreme events.
Based on community preference, the metal roofing material was replaced with traditional Western Cedar wood shakes (see figure 197) supported by rough cut 2 inches by 4 inches (50 by 100 mm) nailers.
The bridge was previously covered with vertical rough-cut siding without battens. The siding was replaced with rough cut 1 inch by 12 inches (25 by 300 mm) vertical Hemlock, but battens were used to lessen weather damage to the truss components (see figure 198).
Modifications of the abutments was limited to removal of a portion of each corner to accommodate the new longer trusses, (about 8 feet (2.4 m) longer to eliminate the splayed lattice arrangement.) The floor system maintains the same distance between abutment backwalls.
The entrance to the bridge on the North approach involves non-standard geometry, yet it has been acceptable to the local users of this bridge for a very long time. Accordingly, the non-standard geometry has been maintained. As a consequence, and due to the lack of accident history at the entrances to the bridge, an approach railing system comprised of heavy timbers was installed. Its appearance meshes well with the timber covered bridge. The poor alignment at the North approach tends to make vehicle speeds extremely low. For these several reasons, these vehicle guidance systems are judged to be prudent and acceptable (refer to figure 199.)
To the extent possible and practical, the new bridge maintains traditional construction materials and techniques, including:
The superstructure of the bridge was restored to its original full-length and parallel lattice element configuration. The external bracing, added during the life of the bridge, was removed, requiring more reliance on the internal system, similar to the original construction.
A feature of the bridge that was altered relates to the use of glue-laminated floor beams. Today's design specifications and minimum loading requirements routinely demand heavier and/or stronger floor beams than were typical at the time of original construction. Few preservationists object to such modifications due to the fact that floor systems have routinely been replaced at least once during the life of the bridge, sometimes several times.
A number of issues were encountered which demanded unanticipated solutions, including:
Many of the elements of Town lattice trusses are positioned beside mating elements (e.g., chord elements along the top chord or chord/lattice elements at their intersections). Over the life of the structure (in this case more than 130 years), deterioration from roof leaks led to significant section loss of elements. There is no currently available practical means of identifying all such deterioration in advance of disassembly of the truss during its reconstruction. Accordingly, almost all Town lattice trusses are found to have more deterioration during reconstruction than anticipated during the engineering phase of the project.
Similarly, powder post beetles had infested many of the hardwood pegs and surrounding primary element material. The extent of the damage by insects was more extreme in some instances than anticipated. In one unusual case, the initial damage by insects led to an entrance into the chord area by a rodent that hollowed out the pair of mating chord members, leading to only a shell of the members remaining. No outward appearance of distress was evident and no inspections had ever identified same.
Isolated top chord elements were sufficiently deteriorated to warrant replacement, primarily from rot due to long-term leaks in the roof. However, the trusses had attained a permanent distortion such that in many cases the transverse wooden peg connectors were no longer straight or horizontal. Accordingly, if one wished to replace an inside element, but not the corresponding exterior element, then the procedure needed to use existing holes as a template for drilling the new holes in the new material. In some cases, the process led to misplaced holes in the new material such that the new material would not be properly positioned vertically. In this case, reuse of the existing holes would not have satisfied the requirements and some existing elements had to be replaced also, in order to have acceptable peg locations in the new material.
While it is desirable to have a smooth curvature of the trusses when finished, in this case, we were locked into the position of the existing trusses. Otherwise, we would not have been able to reconnect intersections of lattice and chords using existing holes. Accordingly, the camber cannot be adjusted to any significant degree during the rehabilitation of a Town lattice truss without replacement of more material, and/or replacement of existing trunnels with oversized pegs in reamed holes.
In our case, the downstream truss had more camber, and more uniform curvature, than the upstream truss. Fortunately, both trusses retained positive curvature of sufficient amount as to be acceptable without having to do more extensive replacement. However, the roof lines were able to be made smooth by adjusting the birds mouth of the rafters.
Two types of physical research were conducted as a part of this project.
Field instrumentation of portions of the existing Town lattice trusses was performed to assess the distribution of forces around chord interruptions and trunnel (wooden peg) connections of chords and lattice elements. To our knowledge, this work was the first of its type for this kind of application. The work was performed by a contractor.The testing included installation of 46 strain transducers at various locations on the bridge and recording information from the passage of a 10-ton (9 MT) vehicle. An evaluation of the information indicated that actual strains were less than predicted, and generally supported implications from finite element modeling of similar lattice trusses.
A spare transducer was applied at the bottom of a transverse floor beam during the field instrumentation. The results of that particular measurement were especially interesting and substantially less than expected. As a follow-up, that floor beam (scheduled for replacement due to inadequate strength) was tested to failure. The beam failed at a load much higher than expected. While this particular test could not be used in any conclusive way, due to the lack of multiple samples or repetitions, the information gleaned from the work was nonetheless interesting.
The rehabilitation of the Fitch's Covered Bridge was undertaken without comparison to the costs of replacement, due to its status as an historic bridge and eligible for inclusion on the National Register of Historic Places (a status it has since received). The project included an extensive rehabilitation of the timber bridge and modification of the abutment caps. All design and construction work was performed by County forces. The cost of the work was about $425,000.