Asphalt Paving of Treated Timber Bridge Decks
The dimensions of timber fluctuate almost exclusively because of changes in moisture content. Thermal expansion of wood is minimal. Most moisture-induced dimensional change occurs perpendicular to the grain. Dimensional change perpendicular to the grain is about nine times more than the dimensional change parallel to the grain (Forest Products Laboratory).
Glued-laminated timber is fabricated at a maximum moisture content of 16 percent (American Institute of Timber Construction 1994). Studies have shown that moisture contents in glued-laminated timber decks average between 15 and 23 percent (Gutkowski and McCutcheon 1987). In most environments, moisture levels in glued-laminated timber decks treated with oilborne preservatives remain relatively constant. However, if deck panels are improperly stored, the wood's moisture content could increase, resulting in significant shrinkage after installation.
Glued-laminated deck panels for bridges are normally treated with oilborne preservatives that minimize moisture penetration, moisture loss, and the associated volume changes. Waterborne treatments, or in some locations, light oil solvent treatments, do not provide the same level of protection against moisture change. Another potential problem of the waterborne treatment is that the moisture content of the wood increases significantly. If the treated wood is not redried before installation, drying can cause significant deck shrinkage and asphalt pavement cracking at the panel joints. Horizontal movement of an 1/8-inch-per-panel joint causes asphalt pavement to crack. The loss of 1-percent moisture content in a 48-inch-wide glued-laminated deck panel can cause 1/8 inch of shrinkage.
Example 6—The Standish Avenue Bridge in Petoskey, MI, was constructed in the fall of 1999. The 80-foot-long, two-lane bridge was constructed with transversely placed, glued-laminated timber panels treated with chromated copper arsenate, a waterborne treatment. The panels were attached to glued-laminated timber beams spaced every 52 inches with aluminum clips. The deck panels were not mechanically interconnected.
The deck panels fit tightly against each adjoining deck panel when installed. The asphalt pavement was laid down immediately after the deck panels. The pavement showed reflective cracking (figure 11) within days after the bridge was opened to traffic. The cracks continued to grow during the first year of operation. When the bridge was inspected during the summer of 2000, gaps up to 1/4 inch (figure 12) were observed between deck panels. The early cracking of the asphalt paving on this bridge may have been caused by differential deflection of the deck panels. Shrinkage of the deck panels enlarged the cracks and contributed to the failure of the pavement.
Figure 11—The pavement cracked about
1 year after construction of the
Standish Avenue Bridge in Petoskey, MI.
Figure 12—Shrinkage of deck panels
contributed to pavement cracking
on the Standish Avenue Bridge.
Expansion of glued-laminated deck panels because of increased moisture content is unlikely to cause pavement damage, because the deck-to-beam connections restrict expansion. In extreme situations, glued-laminated deck panels have expanded on timber bridges in Alaska, buckling and damaging backwalls.
![]() |