Chapter 1 Introduction

Motivation for the Project

The deterioration of the bridge infrastructure in the United States is evident. Structures produced during the construction of the U.S. Interstate Highway System just after World War II are reaching the end of their design lives. Approximately 50 percent of the nearly 600,000 highway bridges in the United States were rated as substandard (Ritter 1990).  At the same time, bridge loads from vehicles and trains have increased significantly.

Many short span highway bridges have a deck-only configuration. They may be comprised of either a longitudinal timber deck of dimension lumber only or a solid reinforced concrete slab construction.  Typical construction of a timber bridge deck may consist of two- or four-inch wide dimension lumber connected vertically to the adjacent boards with nails or bolts.  The decks usually range from four to twelve inches deep.  The shallower timber decks could be strengthened with the addition of a retrofit concrete layer attached to the existing timber layer with a series of mechanical connectors.

Compared to steel, wood decks have the advantages of light weight, ease of construction, high tensile strength in bending and use of a low unit cost, renewable material.  They have the disadvantages of a low stiffness material, decay issues, and low compression strength in bending compared to plain concrete.

Reinforced concrete slabs have advantages of a higher stiffness and compression strength than wood, wearability, and ease of construction. There are important disadvantages. Typically, 40-60 percent of the cross-section develops hairline tensile cracks from supporting only dead loads.  This cracked portion is rendered ineffective with the exception of holding the reinforcing steel in place.  Subsequently, the exposed steel is subjected to corrosion and is a potential fire hazard because of its efficient conduction of heat.  Cracking of the concrete also can lead to spalling and eventual deterioration of the concrete.

The research described herein is an investigation of the feasibility of constructing short span deck-only bridges using a composite wood-concrete layered deck.  A composite wood/concrete cross-section would be suitable for the strengthening of existing timber deck bridges and for new construction applications.  This would combine the advantages and remove or diminish the disadvantages of each material, and likely result in lower initial and long-term cost.  The report summarizes the research work.  Extensive details are available in thesis work associated with the study (Brown, 1998, Koike 1998, Etournaud 1998).  Several technical papers also were presented at various conferences (Gutkowski, et al. 1999, Gutkowski, et al. 2000, Gutkowski, et al. 2001).

Objective

The research described herein is part of a comprehensive project to analytically predict and experimentally verify composite action in layered wood-concrete decks.  The objective is to configure a layered deck, which is structurally effective for use in bridges.  The hypothesis is that a high degree of composite action will be achieved by using the anchored notch connection to provide interlayer slip resistance.  Key tasks in the project included conduct of:  a) withdrawal tests of the anchor connecotor, b) interlayer load-slip tests of connection specimens, c) preliminary flexural tests of layered beam specimens, and d) tests of full-scale deck specimens.

Chapter 2 Concept of the Research

Behavior of Composite Construction

The behavior of a composite wood concrete cross-section is bounded by two extreme limits.  The upper stiffness limit (Figure 2.1a) is that of "composite action."  In this case the cross-section has a single neutral axis and the two material strains are identical at the material interface. The method of transformed sections applies for analysis.  When the layers are not rigidly connected, relative motion (termed "slip") occurs at the interface of the materials.  The single neutral axis splits and when the slip between the layers propagates, the now two neutral axes move farther and farther apart. When some interlayer shear resisting force is present the composite action is referred to as "partially composite" (Figure 2.1b).  The lower limit is that of "non-composite action" interaction, the condition of no shear transfer between the two layers (Figure 2.1c).  The material layers have individual neutral axes and discontinuous strains at the material interface. There is neither mechanical bond nor friction between the two layers.  Independent action of the layers results in both layers experiencing tensile and compressive strains and stresses about their individual neutral axes.  The non-composite limit is the least stiff and results in the largest deflection.

Anchored Shear Key Connection

The degree of composite behavior of a layered member is largely dependent on the behavior of the connection functioning to resist the tendency of the layers to slip relative to each other.  For this project it was proposed to connect the wood and concrete layers with an anchored notch. The concept of this anchor connection was developed in Switzerland (Natterer et al. 1996).  A free body diagram identifying load paths of the notch connection and its components is shown in Figure 2.2.  The design of the notch connection is such that the slipping action between the material layers is resisted by a bearing force between the concrete and wood layers rather than by a shear force in a mechanical connector.  The vertical component of the bearing force is by the tensile force in the dowel.

Performance of the anchored notch is dependent on the unique design of the Hilti dowel shown in Figure 2.3.  The Hilti connector is composed of a threaded dowel, a plastic sleeve, a common nut, and a conical-shaped washer.  The dowel is attached to the wood with an adhesive prior to the placing of concrete.  The washer allows for the nut to be recessed below the concrete surface necessary for vehicle travel or an overlay of some type.  A common deficiency in various types of shear layer connectors is the concern of gaps from shrinkage in both materials.  The Hilti connectors design allows for small gaps on all three surfaces of the notch to be reduced by tightening the connector nut after the concrete hardens.  The plastic sleeve prevents permanent attachment of the hardened concrete to the steel dowel so that the tensioning nut can be tightened freely.

Slip-Modulus Concept

The resistance to interlayer slip is commonly experimentally quantified through the use of a material property termed the "slip-modulus." This property is obtained by conduct of an interlayer shear test of a full-scale sample of the layered system.  In wood/wood layered systems connected by mechanical fasteners or glue (no notch), the load vs. interlayer slip behavior is nonlinear (Figure 2.4).  In such applications, various methods have been used to establish a slip modulus value for approximate linear analysis.  A secant modulus is based on a line from the origin to a pre-determined slip value.  An alternate method is to assume the slip modulus to be the tangent to the measured curve at a specified load or slip.

At Colorado State University (CSU), Thompson (Thompson 1974) conducted interlayer shear tests on three different connections, one of which was the notched connection with the Hilti dowel.  For the anchor notch connectors, the load vs. interlayer slip behavior differs as shown in Fig. 2.4. Typically, the response is linear up to initial brittle failure, followed by drop off to extreme low resistance.  The results of Thompson's work (Thompson 1974) verified the increased stiffness provided by the Hilti dowel compared to the other two steel connectors.