Improving Accuracy of the Camber Predictions for Precast Prestressed Concrete Bridge Girders

Sponsor: Iowa Department of Transportation
01 August, 2011 to 01 August, 2013

Objective:

The primary objective of this research is to provide accurate methods for predicting short-term and time dependent camber during design and, if desired, means of increasing camber for prestressed beams fabricated for Iowa bridges. The approach is to evaluate existing data and models as well as to systematically understand instantaneous and time dependent components of camber from casting of the PPCBs to construction of the actual bridge and beyond by quantifying the most significant parameters affecting camber of beams used in Iowa. A set of prestressed beams will be identified and monitored from fabrication through final erection. Samples of the concrete used in these beams from different precast plants will be tested for strength gain, progressive change of modulus of elasticity, creep and shrinkage behavior in the laboratory. Through this systematic evaluation and measurement of camber immediately after fabrication and continuing through construction, recommendations is made in a final report for improving predictions of camber in design such that any unnecessary construction delays and costs can be avoided.

Methodology:

This research report focusses on improving both short-term and long-term camber predictions. Also, to minimize the error between the expected and actual camber of PPCBs, especially at the time of erection. To achieve the project goal and minimize the influence of uncertainties, the research involved three precast plants who supply PPCBs to bridge projects in Iowa. An evaluation of current camber measurement methods adopted at the precast plants was found to contribute to discrepancies between measured and predicted camber. Thus, modifications to camber measurement techniques were explored to obtain consistent and accurate measurements of the instantaneous camber. The behavior of three normal concrete and four high performance concrete mix designs were first evaluated and their engineering properties and long-term behaviour can be accurately characterized. Then, using the material properties including creep and shrinkage, long-term camber of PPCBs were analytically estimated with the aid of Finite Element Analysis.

Material Characterization:

A total of seven different concrete mix designs, representative of three precast plants were investigated for concrete material properties such as, modulus of elasticity, and creep and shrinkage. Four of the seven mixes were high performance concretes (HPC) and are currently used for casting prestressed bridge girders, and the rest of them were normal concretes (NC) used in girders in the recent past. Based on the test results, average creep and shrinkage curves were proposed to be used for long-term camber prediction of PPCBs.

Instantaneous Camber:

Problems with predicting camber are typically evident at the bridge site after girders have been set on the piers. Although camber prediction problems are present at the bridge site, they have originated from inaccurate predictions with the design of the girder, in the precast plant during fabrication, or with the transfer of prestress. Investigating inaccurate instantaneous camber measurements was accomplished by examining past camber measurements as well as measuring camber at three precast plants and five bridge sites. A combination of the measurement techniques used by precasters and researchers along with new methods were explored to determine a consistent, accurate way to measure the instantaneous camber. While some previous measuring methods neglect measurement errors caused by bed deflections, inconsistent beam depths, and friction between the beam and bed, the measurement method used to gather data on over 100 PPCB, accounted for these factors.

Long-term Camber:

To further investigate time dependent camber, the 88 beams fabricated for five different bridges in Iowa were monitored for camber measurements from precasting yard to the bridge site. Different types of IDOT girders with various lengths were selected for long-term camber measurement. Camber was measured from top flange when the girders were erected on the piers before and after deck cast. The rotary laser lever was used for all long-term camber measurement. Also, the effects of support conditions, and thermal effects on camber variability were inspected. The measured data was subsequently used to validate simplified model, and Finite Element model developed using MIDAS Civil software.

Major Findings:

  • Sealed specimens of concrete represented the behaviour of creep and shrinkage of the full-scale prestressed bridge girders much better than the unsealed specimens.
  • The AASHTO LRFD 2010 creep and shrinkage models were found to give the best estimates when compared to the measurements taken from four HPC and three NC mixes over one year.