U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590

Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

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

Federal Highway Administration > Publications > Research > Structures > Covered Bridge Manual

Publication Number: FHWA-HRT-04-098
Date: APRIL 2005

Previous | Table of Contents | Next

Chapter 4. Types of Longitudinal Trusses

This picture shows a side view of a covered bridge with a protruding portal and approach rails over a white water river running high, and the autumn colors still visible.

Figure 23. Salisbury Center Bridge-Herkimer County, NY.

This picture shows the entry to a covered bridge with the approach road and side rails. The portal has a diagonal cut opening and the Town lattice casts shadows on the inside of the bridge where the running planks can be seen clearly.

Figure 24. Brown Bridge—Shrewsbury, VT.

This chapter describes the key engineering features of the timber truss types introduced in chapter 3. Terminology and illustrations are included to facilitate comparisons and contrasts among the truss types. Special details are highlighted.

The truss types described here are presented in the order of their span length, starting with the shortest. The first three truss types (kingpost, queenpost, and multiple kingpost) are ones used in the earliest North American covered bridges. No patents were ever taken on their configurations, and no individual is specifically credited with their development. The other truss types that follow were developed and ultimately named after enterprising early builders/engineers (usually in recognition of a patent obtained for the details of the truss).


The most elementary heavy timber truss configuration is the kingpost (see figure 25). The inclined members of a kingpost truss serve both as the top chord and as the main diagonals, and resist compression forces. The horizontal member, along the bottom of the truss, is the bottom chord and acts in tension. A central vertical member (the kingpost), also acts in tension to support the floor loads and serves as the connecting element between the opposing main diagonals. The kingpost truss configuration has two panels. A panel is that portion of the truss that lies between any two vertical components.

This diagram shows the most elementary timber truss configuration. The kingpost with the tail that goes through the bottom chord is a hanger that takes the loads in the center of the bottom chord up to the top of the kingpost brace, which is the diagonal support that takes the load down to the abutment and also ties into the bottom chord at the heel connection. The bottom chord that acts in tension to support the floor loads connects the opposing main diagonals and sits on bearing blocks or bed timbers that tie into the abutments.

Figure 25. Diagram of kingpost truss.

In addition to resisting the tensile forces generated by the opposing diagonals, the bottom chord almost always supports the floor beams. In most kingpost truss bridges, the floor beams are located only at the ends of the bridge and next to the center kingpost. The floor beam point loading does not coincide with the intersections of the theoretical centerlines of the truss members. This connection eccentricity induces bending stresses in the bottom chord that may be large or negligible, depending on the distance of the floor beams from the joints and the depth of the bottom chord.

The dead and live loads are applied differently to kingpost trusses. Live traffic loads are carried to the truss through the central floor beam, while much of the bridge dead load is carried in the rafter plate, along the eaves of the roof. As a result, almost half of the bridge weight is carried to the end posts of the bridge, which transfer their loads directly to the foundation. The kingpost truss carries the centerline floor beam(s) and the inner ends of the four eave plates. Technically, the end posts and the eave struts are not structural members of the kingpost trusses, and their connections are not intended to transfer axial loads within the truss; they are simply members of the associated framework.

The inclination angle for the kingpost diagonals is restricted. Generally, steeper diagonals are more efficient at resisting shear forces in a truss. There are, however, compromises to consider when laying out the members in any truss. For instance, given a set span for a two-panel kingpost truss, steeper diagonals make taller trusses. Beyond the aesthetic issues of building unusually tall, but short-span structures, there are practical limits to the height of the bridge involving bracing and its connections. Hence, the span limit for this simplest truss is quite short, typically only about 7.6 to 9.1 m (25 to 30 ft).

Longer kingpost trusses have been built by including subdiagonals. These members act as braces, from the bottom of the kingpost up to the midpoint of the main diagonals, thereby producing a minitruss within the larger kingpost truss. Short struts often extend above this junction to support the load from the roof eave plate. Vertical metal rod hangers may also be used from the intersection of these subdiagonals downward to the bottom chord, allowing installation of floor beams at this quarter point of the bridge. These modifications allowed builders to increase kingpost spans out to about 10.7 to 12.2 m (35 to 40 ft).

This diagram shows a longer kingpost truss configuration. The kingpost with the tail that goes through the bottom chord is the vertical member that supports the center of the deck (and the center floor-beam). It connects to the kingpost brace (or end diagonal), which ties into the bottom chord at the heel connection. In addition, subdiagonals act as braces for the midpoint of the main diagonals. Vertical rods run from the midpoint of the main diagonal down to the bottom chord, where they pick up force from the floor beams. The connection near the ridge is called either a seat, step, or corbel. The bottom chord that acts in tension to support the floor loads connects the opposing main diagonals and sits on bearing blocks or bed timbers that rest on the abutments.

Figure 26. Diagram of kingpost truss with subdiagonals.

Most kingpost trusses were built with single member components, usually large sawn or hand-hewn timbers. The most critical connection in kingpost trusses is the heel connection of the main diagonals to the bottom chord. These connections are prone to several weaknesses discussed in more detail later.

The kingpost truss is not very common in the extant United States covered bridge population. There are only about 30 kingpost covered bridges1 remaining in the United States, with spans ranging from 6.7 to 21.3 m (22 to 70 ft).[1] It is very unusual for a kingpost bridge to span 6.7 m (70 ft)-approximately 15.2 m (50 ft) would be the more common upper limit. The extant kingpost bridges were built between 1870 and 1976. [1]


The next range in span lengths commonly includes trusses developed from a simple modification of the kingpost. The queenpost truss is, conceptually, simply a stretched-out version of the kingpost truss, accomplished by adding a central panel with extra horizontal top and bottom chords (see figure 27). Classic examples of queenpost trusses do not have any diagonal web members in the central rectangular panel. Therefore, the most simple queenpost trusses are not true trusses at all,; but rather frames (although this distinction is not relevant to this discussion). The vertical members are termed queenposts. These trusses are considered to have three panels.

This diagram shows the next range in span lengths, a central panel with longer horizontal top and bottom chords. Instead of the middle kingpost, this configuration has two vertical queenposts that go to the bottom chords with main diagonal or queenpost braces that tie into the bottom chord at the heel connection. In addition, counter braces and struts stiffen the centers of the top chord and the main diagonals. Rods support beams in the center of the middle and end panels.

Figure 27. Diagram of queenpost truss.

The member forces and behavior in queenpost trusses are very similar to those found in kingpost trusses: therefore, the design considerations for these two basic truss styles are equally similar. A number of similarities exist between kingpost and queenpost trusses:

The span lengths of queenpost truss bridges range from about 12.2 to 18.3 m (40 to 60 ft), although there are a few examples that are longer. The longer span requires that many of their bottom chords be spliced longitudinally from separate timbers. This tensile connection is another area of weakness in the truss and is discussed in more depth later.

There are approximately 101 bridges supported by queenpost trusses, or slightly more than 10 percent of all the surviving covered bridges in the United States. Their spans range from 7.6 to 39.6 m (25 to 130 ft), and they were built between 1845 and 1985.[1]

Multiple Kingpost

A straightforward way to stretch the span capability of the queenpost truss is to add panels to the kingpost truss to create what is known as multiple kingpost trusses (see figure 28). Accordingly, the basic kingpost truss is sometimes referred to as a simple kingpost truss. (The image depicted in figure 28 demonstrates verticals that have been cut down to accept the diagonal; —some refer to these as gunstock verticals. The verticals depicted in figure 29 are, perhaps, more commonly notched to accept the diagonal.) Most of these trusses were built with an even number of panels so that all the diagonals are in compression and all the verticals are in tension under normal loading. Very few multiple kingpost trusses have an odd number of panels, with opposing (or crossing) diagonals in the center panel.

This diagram shows a way to stretch the span capability of the queenpost truss by adding panels to the kingpost trust. The configuration is basically rectangular with a main kingpost and additional vertical posts, diagonals and end posts. The drawing shows the face of the tie beams connecting the roof and on the bottom surface the deck planking set on the floor beams that rest on the bottom chord.

Figure 28. Diagram of multiple kingpost trusses.

There is a lack of tensile capacity of the connection of the diagonals to posts. In this instance, the compressive force in the diagonals under the influence of the dead load of the bridge is usually much larger than the tensile force resulting from the passage of vehicles. Hence, under normal circumstances, the diagonals remain in compression under all combinations of loading, and the tensile connection is unnecessary.

The longer spans of the multiple kingpost truss, without increasing truss depth significantly, generate higher member forces, which require more capacity. Multiple kingpost truss chords are often comprised of twin members that sandwich a central plane of single web (vertical and diagonal) members. The longer chord members also usually require splices that typically are staggered along the truss length. This critical detail is meant to ensure that, at any particular cross section along the bridge, there is at least one unspliced bottom chord (tension) member in each longitudinal truss; —more specifically, there should be 1 m (3.28 ft) separation between splices of adjacent members of the bottom chord.

The panels in multiple kingpost trusses are often quite short, which means that the transverse floor beams could be located abutting each vertical member. This minimal eccentricity between load application and truss joint location greatly reduces bending stresses in the bottom chord. In addition, these more closely spaced web members tend to have smaller member forces in the diagonals due to their geometry, so that the connection forces are somewhat smaller than those associated with kingpost or queenpost trusses.

The truss diagonals bear on shoulders cut into the sides of the vertical tension members. This means that the verticals must be made from substantially wide timbers. Unfortunately, this joint eccentricity means that the shoulders of the verticals are significantly overstressed in shear along the grain. Many truss verticals have failed in shear; it is common to find evidence of separation and slippage of the shoulder relative to the main portion of the vertical. This can happen at either the top or bottom of the post. Figure 29 provides an example of a shear failure at the notch for the diagonal. Note the vertical shift of the right half of the post above the notch, most noticeable at the top of the post. This is from the Mill Bridge in Tunbridge, VT, before it collapsed due to flooding-borne ice impact in 1999.

The truss diagonals bear on shoulders cut into the sides of the vertical tension members that are subject to overstressed shear forces along the grain. The picture shows two white arrows pointing to the vertical shift of the right half of the post where separation and slippage shows at the notch for the diagonal.

Figure 29. Shear failure of a vertical at a notch—Mill Bridge, Tunbridge, VT.

Another truss component that suffers a common weakness is the bottom tail of the vertical member. It is subject to the same high shear stresses as discussed and illustrated above. The tails are also subject to impact by floodwaters, debris, and/or ice floes. In many instances, the tails have been broken off, as illustrated below. Unfortunately, the tails hold the chords in place vertically, and collapse of the floor is probable when the tails are broken. Figure 30 presents an example of a complete fracture of a tail from ice impact, which was subsequently repaired. Figure 31 provides a view along the same bottom chord showing the bowing due to the impact to the inside of the bottom chord from ice floes, from right to left. The broken tail is just outside of the photo. Looking closely, one can see that the chord has been pushed out from under the floor beams at midspan. Only the longitudinal timber decking, spiked to the floor beams, kept the floor from falling into the river.

Another truss component that suffers damage is the bottom tail of the vertical members from shear stress and floodwaters, debris or ice floes. This view of the outside underside of the bridge shows a white arrow pointing to the tail broken off from ice impact. Because the tails hold the chords in place vertically, this condition can lead to floor collapse.

Figure 30. Example of a broken tail from ice impact—South Randolph Bridge, VT.

The underside longitudinal view of this bridge shows an arrow pointing to the bowing of the inside of the same chord after the tail has broken off. The chord has been pushed out from under the floor beams at midspan. Only the longitudinal decking spiked to the floor beams holds the floor in place.

Figure 31. Bowing of bottom chord due to impact from ice floes —South Randolph Bridge, VT.

About 95 bridges using multiple kingpost trusses remain, or a little more than 10 percent of all covered bridges in the United States. Multiple kingpost trusses have spans that range from 11.0 to 41.1 m (36 to 124 ft), and they all seem to have been built between 1849 and 1983.[1] Interestingly, comparing the span ranges and the construction dates between queenpost and multiple kingpost trusses, one may observe the similarity of these two features.

Burr Arch

As noted in chapter 3, Theodore Burr obtained the first U.S. patent issued for a specific timber truss configuration in 1806. The Burr arch is, basically, a combination of a typical multiple kingpost truss with a superimposed arch (see figure 32). The arch was added to allow heavier loads on the bridges and to stretch their span capabilities to greater lengths. Surviving examples of Burr arch bridges have spans of up to 67.7 m (222 ft).[1]

Burr's development was immediately popular with bridge builders and has proven durable. More existing North American covered bridges use the Burr arch than any other type. The classic, or conventional, Burr arch supports the ends of the arch components at the abutment, with no connection between the bottom chord and arch as they pass each other (the chord is supported by the abutment directly separated from the arch end). A variation of the Burr arch (sometimes referred to as a modified Burr arch) terminates (and ties) the arch with a connection directly to the bottom chord, which is supported on the abutments.

The conventional and modified Burr arch has the same components as the multiple kingpost truss: a rectangular configuration with a main kingpost and additional vertical posts, diagonals and end posts. The drawing shows the face of the tie beams connecting the roof and on the bottom surface the deck planking set on the floor beams that rest on the bottom chord. In addition, the drawing also shows a superimpose ed arch. The classic or conventional arch support on the right side of the drawing supports the ends of the arch at the abutment with no connection between the bottom chord and arch while the modified Burr arch terminates and ties the arch with a connection directly to the bottom chord, which is then supported on the abutments.

Figure 32. Diagram of conventional and modified Burr arch.

The actual arches of most Burr arches are in pairs; these sandwich a single multiple kingpost truss between them. The most common connection uses a single bolt to join the arches through each of the vertical members of the truss. This means that the load sharing between the truss and the arch components is largely dependent on the relative stiffnesses of those bolts. The floor beams carry the live loads to the truss bottom chords, and the roof loads bear on their top chords. For these vertical loads to be distributed into the arch, the bolts must resist significant vertical shear forces. The initial, traditional Burr arches used arch components sawn from large, single timbers that were lap-spliced to each other at the verticals. Later, use of continuous but laminated (multiple-layer) timber arches became popular with some builders.

In addition to the critical areas of interest cited above for the multiple kingpost truss that comprises the central portion of the Burr arch structure, special attention should be paid to the ends of the arches and the interconnections of the arch to the truss. Figure 33 shows the connection of timber arch with post using only a single bolt. This Burr arch happens to have a dual timber arch, —one above the other.

The picture shows the diagonal supports in the background with the arch superimposed and connected to the vertical post.

Figure 33. Connection of arch to post—Wehr Bridge, Lehigh County, PA

There are about 224 remaining bridges supported by the Burr arches and its multiple variations (about 25 percent of all covered bridges).[1] The Burr arch has individual spans that range from 10.0 to 67.7 m (33 to 222 ft); this longest span is 10 percent longer than the next rival configuration of truss (the Howe). The extant Burr arches were built between the early 1800s and 1988.[1]

Town Lattice

Ithiel Town, an architect by education, obtained his first patent for a unique type of timber truss in 1820 (see figure 34). All the other trusses mentioned above, and those that follow this subsection, principally rely on large and heavy timbers that require skilled artisans to properly craft the rather elaborate joinery between the various components. Town sought a means of constructing bridges that would rely on an easily adapted design and would require less skilled labor. His patented truss developed a configuration that could be extended to a wide range of span lengths with relatively little modification of the configuration. In the opinion of many informed bridge aficionados, his patented truss represents arguably the most important development in the history of covered bridges, and one that remains a popular and enduring style. Later portions of this manual will examine the merits of this truss configuration.

Town's lattice configuration relies on assembling relatively short and light planks that were available and easy to handle. He connected the overlapping intersection of members with round timber dowels or pegs, termed treenails—pronounced trunnels (and so spelled hereafter in this manual). The plank intersections in the web may have from one to three trunnels. Where chord members intersect with web or lattice members, the overlapping zone may contain as many as four trunnels. The dowels are often 38 to 51 millimeters (mm) (1.5 to 2 inches) in diameter. The parallel and closely spaced web members are joined to chords along both the top and bottom of the trusses. Two levels of chords commonly are used as the bottom chords. The top chords may have one or two levels of members. The lowest bottom chord provides the seat for the transverse floor beams.

This drawing looks like a lattice fence. In contrast to the other bridge configurations, the Town lattice has an upper or upper top chord and an upper secondary or lower top chord. The lattice is short, light planks, connected to the overlapping intersection of members with round timber dowels called tree nails. The end post and tie beam are vertical members. On the bottom the lower secondary chord or upper bottom chord and the lower chord or lower bottom chord are attached to the deck planks and floor beams.

Figure 34. Diagram of Town lattice truss.

Town, or lattice, trusses are most commonly comprised of thin members with pairs of chords on each side of the lattice webs. In this case, the truss is sometimes termed a plank lattice. The chord members generally are not spliced to abutting pieces at their ends, but the terminations are staggered so that any panel of chord has at least one unspliced member. A few Town lattice trusses were fabricated of heavier components using single chord members on each side of the lattice. In this case, the truss is termed a timber lattice. The chord members require splices at their ends.

There remain about 135 bridges supported by Town lattice trusses.[1] Town lattice trusses support varying span lengths, from relatively short (only 7.6 m (25 ft)), up to some of the longest covered bridge spans in the world. Individual Town lattice trusses span up to 49.4 m (162 ft).[1] The oldest surviving Town lattice bridge (the Halpin Bridge in Middlebury, VT) was purportedly built about 1824.[1] New examples of Town lattice covered bridges are still being built.

Long Truss

Colonel Stephen H. Long first patented a truss configuration in 1830. His focus was on a parallel chord truss made with heavy timbers and with crossed diagonals in each panel (see figure 35). A special feature of his bridge included the use of timber wedges at the intersections of the chords, posts, and diagonals. The wedges allowed builders and maintainers to adjust the shape of the panels, and provided the opportunity to adjust the initial camber.

The Long truss features top and bottom parallel chords with vertical posts, crossed with diagonals at each panel. Timber wedges are used at the intersections of the chords, posts and diagonals to increase the strength of the connection between the horizontal component of the load in the diagonal and the chord.

Figure 35. Diagram of Long truss.

In today's jargon, the wedges allowed builders to induce forced loads in the diagonals in a way that is described as pretensioning. It is extremely difficult to predict the amount of the induced prestressing force. Long's patent applications included images of wedges between the vertical and the chord (as shown in figure 36) and between the counter and the chords (as indicated in figure 35).

However, the wedges do increase the strength of the connection between the horizontal component of the load in the diagonal and the chord. The transfer of load without wedges flows from the end bearing on the diagonal to the cross grain bearing in the post, then from the cross grain bearing at the shoulder of the post back to the end grain bearing at the shoulder of the chord. Introducing the wedge distributes the bearing load from the chord over a much larger area of the post through the wedge in direct cross grain bearing.

Figures 36 and 37 clarify how Long wedges work. The image in figure 36 is from the outside of the bridge (siding and outside chord stick removed) looking back toward the inside of the bridge. The wedge on the right side normally is hidden from view by the floor beam. As the wedge is driven downward, the post is moved with respect to the chord along the shoulders cut in the chord stick. An important engineering aspect of the wedge is to distribute large edge stresses along the vertical face of the shoulder across a wider face of the post at the interface with the wedge.

This picture shows a white arrow pointing to a wedge located on the bottom chord of a Long truss. The wedge takes some of the bearing load and has large metal bolts in it.

Figure 36. Long truss bottom chord wedge—Downsville Bridge, Delaware County, NY.

This picture shows a white arrow pointing to a block of wood between a thick floor beam and a diagonal counter member.

Figure 37. Wedges between counter and floor beam in a Long truss—Hamden Bridge, Delaware County, NY.

The Long truss was adopted by many builders for use in highway and railway bridges, but the timing of its introduction meant that it was destined to be overtaken quickly in popularity by the Howe truss, as discussed in the following section.

There are about 40 surviving bridges supported by the Long truss, with individual spans that range from 15.5 to 51.8 m (51 to 170 ft).[1] The oldest extant Long truss was built in 1840, and the newest was built in 1987.[1]

Howe Truss

William Howe (1803--52) of Massachusetts was granted his first truss patent in 1840 and a second one later in the same year. His second patent used metal rods as the vertical members of what was otherwise a simple timber parallel-chord, cross-braced truss. This was the first truss patent granted with some major structural components made with metal. The configuration used easy-to-erect and readily prefabricated components that could be assembled on site and adjusted via threaded connections at the rod ends. Little skilled labor was involved in assembling and erecting this truss type, and it became an immediate success (see figure 38).

The picture shows a Howe truss that is similar to a parallel chord, cross-braced truss except the vertical structural members are metal to allow longer spans. In addition to the top and bottom chords, end posts, diagonal pairs and counters, floor beams and deck planking, the drawing shows tension rod pairs with metal angle blocks at the top and bottom chords.

Figure 38. Diagram of Howe truss.

Another factor in the success of Howe's truss type was his inclusion of a detailed structural analysis with the patent application. Up to this time, the selection of member sizes, materials, and overall geometry, was generally left to the judgment of the individual bridge builder. The fledgling structural engineering profession was developing rules and relationships to govern such matters, but no consensus had been attained at the time of Howe's patent.

The initial Howe truss bridges had wooden blocks cut to fit at the connections at the ends of the diagonal members against the chords. Later versions converted to the use of cast iron angle blocks. These blocks were simple to construct and install, and they were a major factor in the popularity of this configuration.

The Howe truss is second only to the Burr arch in popularity of extant covered bridges in the United States. There are about 143 bridges supported by the Howe truss, or about 15 percent of all covered bridges.[1] The Howe truss has individual spans that range from an unusually short 6.1 m (20 ft) up to an impressive 61.0 m (200 ft), the longest being only 10 percent shorter than the longest Burr arch.[1] The oldest extant Howe truss was built in 1854, and the configuration remains popular with new authentic examples built today.[1]


The preceding seven truss configurations support the vast majority of covered bridges. There are many other truss configurations, however, that were patented with a few representative examples still standing, including those identified as Smith, Paddleford, Pratt, Childs, and Partridge trusses. Each of these trusses contains some technical nuance to differentiate it from others, but the basics of their behavior follows those described above.

The Pratt truss deserves special note because it was the precursor of the very popular metal truss of this configuration. In the initial form, Pratt used metal rods for the diagonal tensile elements and timber in the compression posts, taking advantage of the respective strengths of those materials. Very few Pratt timber truss bridges remain, in large part due to the difficult connection of the diagonals to posts, but a very large number of Pratt metal trusses survive, in which the connections with metal were simplified.

While very few exist, the Paddleford trusses (see figure 39) are remarkable in that the assembly of interconnected timbers requires exceptional skill for a proper fit. These structures behave more like frames than trusses, involving shoulder bearing at the frame connections with much of the resistance due to shear and bending stresses in the elements, in addition to the axial forces. The analysis of these structures is especially complex and challenging.

The Paddleford truss looks similar to a multiple kingpost truss with upper and lower chords. In addition to the vertical kingposts and end posts, the drawing shows diagonal braces and counter braces running in the opposite direction. Note that the counter braces are at a flatter slope than the diagonals and cross into the adjacent panels.

Figure 39. Diagram of Paddleford truss.

There are also a number of covered bridges supported by tied arches (technically not trusses at all). The tied arches are labeled as such due to a horizontal tension element that connects the ends of the arches. The roof and siding are supported by rafter plates and columns above the arches. Rods suspend the floor from the arches.

Previous | Table of Contents | Next

Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101