Many of today's bridge construction projects are erected in stages to limit traffic interruptions or to minimize the environmental impacts. Typically, one half of the bridge superstructure is constructed in the first stage and the other half constructed in the second stage. The final stage is to cast a closure pour to join the deck slabs of the two structures together.
Matching the final deck elevations of the first two stages of construction has created numerous problems during construction. Most common are construction delays caused by the need to reanalyze the structure and reset the screed and buildup elevations. Improperly aligned deck elevations between the two stages may require the need for grinding of the deck surface after placing the closure pour. In addition, reduction of the thickness of the deck may result in reduction of the concrete cover leading to possible deterioration of the bridge deck. The misalignment of the two deck slab surfaces may occur due to differential non-composite deflections of the adjacent girders in different stages of construction.
Deflection calculations are normally based on a single girder line with no accommodation for the varying transverse distribution of the loads. The variation in the transverse load distribution occurs because of frame action exhibited by the cross-frames connecting the girder lines, the effects of bridge skew, and variations in non-composite dead load due to different overhang conditions.
The solution to this problem is to create a three-dimensional finite element computer model that is more accurately representative of the three-dimensional behavior of the structure. Creation of the computer models is a very time consuming task that must be performed by an engineer with considerable experience in computer modeling. This represents an extremely laborious task for the NCDOT to perform on all staged-construction projects.
The primary objective of this research is to develop an empirically based method to predict the non-composite deflection of steel plate girders in staged-constructed bridges. The effects of bridge skew, girder length, girder spacing, cross-frame stiffness, in-place deck slab thickness, and composite action on a portion of the girder length will be included in the method. Simple span, two span continuous, and three span continuous bridge configurations will be considered. The empirically based method will utilize a series of simple modifiers to adjust the traditionally predicted single girder line deflection to the expected staged-construction deflection. The formulation of the simplified method will require a combination of field-measured data and extensive three-dimensional analytical simulation.
The research will provide a direct benefit in cost savings to the NCDOT both in the long-term and short-term. These benefits will be realized by decreasing construction delays and design time required to predict the deflection during staged construction of the girders. In addition, the possible deterioration of the bridge deck and the needs for repairs due to insufficient concrete cover of the reinforcement will be eliminated."
