"Evaluation of a High-Performance Steel Bridge Using"
Evaluation of a High-Performance Steel Bridge Using Field Test Data Kelsey Miner NSF Research Experiences for Undergraduates in Bridge Engineering University of Delaware Intro Location / History of Churchman’s Bridge Previous Research Bridge Model Strain Gage Data Results / Discussion Location - Map University of Delaware Churchman’s Road Bridge Location - Pictures New bridge looking from the east. Aerial view of old bridge (notice skew). 27º History of Churchman’s Bridge Work began in 2004 to replace existing overpass 773’ long steel bridge with five spans Sharply skewed at an angle of 27º Cross frames (to connect the girders) are parallel to the abutments (with the skew and not normal to the girders) Previous Research Eight girders and two cross frames have been instrumented with Geokon vibrating wire strain gauges. Gauges have been continuously recording data since before construction began. Data has been removed from the data loggers approximately every two weeks, and then stored as text files. Previous Research – Strain Gauges Two cross Cross Frame Detail frames have gauges on all members. All eight girders have strain gauges on the bottom flange and two also have them on End View top. Research Objectives and Importance Very few skewed steel bridges have been instrumented before construction. Currently they require significant design experience due to cross-frame and girder rotation issues. This project aims to increase the understanding of dead load stresses in these bridges. Modeling Modeling Program: STAAD 2003 Easy to learn, available and recommended 2 1 n Girder Details - Side View n tio tio c c Se Se Using STAAD, a line model was G2 to G7 PL16'x13 4" PL20'x118" Top Flange G1 & G8 PL16'x15 8" PL20'x118" created. Members are treated 2-d lines, as compared to a 3-d model in Web 7 PL72"x 8" (TYP.) 110' which they are solids. Girder cross-sections produced in Field Splice 1 Section Wizard and imported into G2 to G7 PL16'x13 4" PL20'x118" Bottom Flange STAAD. G1 & G8 PL16'x15 8" PL20'x118" AutoCad used to organize the 26 different sizes of girders on this bridge. Modeling - Simplifications Cross-frames modeled as single members. Models were created in STAAD and their effective areas determined. W-sections with the same area were then chosen to represent these cross-frames. Modeling - Cross-Frames, Utility Supports By modeling the cross-frames as single members, their true weight was not being represented. The utility supports were not modeled, and instead loads were added to represent them. Gir r der Girde Cro ss-f ram e Girder Cros s-fra me Utility Supports Modeling - Supports and Errata Middle pier modeled as fixed support, with all other piers being modeled as rollers. Modeled this way since k-values for springs would be difficult to determine. Also, STAAD has a node limitation of 500 nodes, so only the west half of the bridge was analyzed, although entire model was created. Model of Entire Bridge Model of West Half of Bridge Modeling - Construction Process Each step in the construction process was modeled, with one shown below. Done so that the evolution of stresses in the girders could be analyzed. Strain Gage Data Strain data was plotted versus time for months before construction to weeks after construction Girder placement In High Steel’s In transit / fabrication yard At construction site Strain Gage Data – Three Types Three types of strains were determined. 1. Overall Strain on Bottom Of Girder 1 During Construction 2. Construction 250 150 Maximum Strain 3. Maximum 50 -50 Construction Microstrain Extreme Maximum Strains Construction Strains -150 Strain Overall Strains -250 Overall Strain -350 -450 -550 05 05 05 05 05 05 05 05 05 05 05 - - - - - - - - - - - ar ar ar ar ar pr pr pr pr pr pr -M -M -M -M -M -A -A -A -A -A -A 22 24 26 28 30 01 03 05 07 09 11 Results Stresses from gages compared to the results from the models. West End of Bridge Stresses Agreed on stress development, such as both having largest stresses after girder six was placed. Did not agree on magnitude of stresses with gages reading 12-14 KSI and model predicting 5 KSI Discussion Discrepancy in stresses is most likely due to loads not being taken in account. From plan drawings and bridge visits it is not clear what these loads would be. Possibly welds, safety devices, stiffener plates Results could also be improved by better modeling of supports and of cross-frames. Determine k-values for all the supports Take into account other properties besides effective area for the frames. Further Research 3-D modeling of the bridge. Would allow accurate modeling of the cross- frames, since could attach members to the tops and bottoms of girders Relationship between temperature and strain in the girders, since temperature causes significant strain variation. Acknowledgements Professor Chajes Gary Wenczel Jason Winterling National Science Foundation This material is based on work supported by the National Science Foundation under Grant No. EEC 0139017 “ Research Experiences for Undergraduates in Bridge Engineering” at the University of Delaware.