Вы находитесь на странице: 1из 20

BRIDGE RETROFITTING TO CORRECT UPLIFT OF A CURVED GIRDER BRIDGE THROUGH FIELD TESTING AND ANALYSIS

Sreenivas Alampalli New York State Department of Transportation State Campus, Albany, NY 12232-0869 Tel/Fax: (518) 457-5826/7535 E-mail: salampalli@gw.dot.state.ny.us and Thomas A. Morreale HDR Engineering Inc. 5700 Lake Wright Drive, Suite 300 Norfolk, VA 23502 Tel/Fax: (757) 222-1516/1515 E-mail: tmorreal@hdrinc.com

BRIDGE RETROFITTING TO CORRECT UPLIFT OF A CURVED GIRDER BRIDGE THROUGH FIELD TESTING AND ANALYSIS Sreenivas Alampalli and Thomas Morreale

ABSTRACT The Ramp C Bridge, at the New York State Thruway interchange with I-287 and New York State Route 17 in Rockland County near the New Jersey state line, is a continuous four-span bridge with a reverse horizontal curve alignment. It consists of five steel girders with radial abutments and skewed pier supports (61). During construction immediately after the girders were erected and the counter weights were poured, at low temperatures, it experienced unanticipated girder uplift at one corner of an end span. Physical testing and analysis were undertaken to investigate this behavior and to estimate the uplift forces. The data were used to design a tie-down system to correct the uplift problem, thus avoiding delays in opening the bridge to traffic.

INTRODUCTION The Ramp C Bridge connects the New York State Thruway (I-87) to southbound I-287 and NY State Route 17, in Suffern, Rockland County, NY (Fig. 1). This ramp crosses over Route 59 and 9 to 10 railroad tracks, and then rises to cross over the Thruway mainline. The continuous, four-span, horizontally curved bridge has five plate girders supporting a 216 mm reinforced-concrete slab with a 63 mm asphalt wearing-surface. The design was constrained due to the complex geometry of the interchange, site restrictions, railroad tracks, and physical clearances. As a result, the bridge has several undesirable details/features, which bridge engineers normally avoid during design. Some unavoidable structural details include: 1) a severe sag curve (3.1% grade at east and sag camber), 2) reverse compound horizontal curve (radii of 915 m and 457 m, reversing to 213 m at west end), 3) highly skewed (61-deg) piers with abutments at zero skew, 4) poor span ratios (42 to 91.4 m), and 5) unequal girder lengths in Span 1 (due to the zero-skew abutment) varying from 30.5 m at south fascia to 54.3 m at the north fascia (Fig. 2). The girders are supported on steel/elastomeric, multi-rotational (POT) bearings at all supports. For proper performance, these bearings require a minimum compressive load of 10 % of bearing capacity 133 kN in this case -- to be present at all times ( 1).

The original design analysis indicated that uplift forces would occur at the east abutment under live loading. The upward 133 kN live-load reaction would not be fully counteracted by the bridges dead load. A concrete counterweight, to be placed between girders at the southeast corner of the steel superstructure, was provided for in the original contract plans. An unanticipated uplift problem developed during construction, under dead load only, requiring immediate attention. Field tests were conducted and reactions measured for 48 hrs to identify the cause, and estimate loads for the design of a retrofit to correct the uplift condition. Following testing, a detailed three-dimensional, finite-

element analysis was conducted to determine the effects of variable thermal loadings on the superstructure, and to investigate the effect of a tie-down restraint on the force distribution to other structural components .

PROBLEM During construction, immediately after the girders were erected and the counterweight was poured, field personnel observed that several girders at the east abutment had lifted off the supports, with a visible gap occurring between the bottom girder flanges and top bearing plates (which were not yet welded to the flanges) as seen in Fig. 3. The gaps varied from a maximum of 35 mm at the south fascia girder (G1) to 0 mm over the next two or three girders. Although the remaining girders were in full contact with the bearings, the vertical reaction they were providing was questionable. Contractor personnel also observed that the gaps varied with changing temperature. In mornings when it was cold, the largest gap was 11 mm and then grew to 35 mm in the sunny afternoon. It was suspected that these changes in gap and uplift were due to temperature difference between the steel girders and concrete deck.

To keep the girders securely seated on the bearings, the contractor installed temporary uplift restraints composed of threaded rods and channels anchored to the footing and clamped to the girder flanges. The girders were drawn down to minimize the gaps, but they could not be completely eliminated, because the force required to achieve this couldnt be obtained with the contractors equipment and this simple system. The concrete deck was poured, anticipating that the additional dead load would close the gaps and ensure sufficient compressive forces on the bearings. After the concrete deck was poured, the south fascia girder still exhibited a 3.2 mm gap between the bottom flange and bearing top

plate. Although no visible gap was observed at the other adjacent girders, they were not in full contact with the bearing plates, as evidenced by sliding the plate around under the flange, by hand. Contractor personnel also observed that the visible gap still seemed to vary as the temperature changed throughout the day.

With the temporary tie-downs still in place, the bridge was opened to traffic on the scheduled date. However, the uplift condition caused vertical misalignment at the roadway joint between the bridge and approach slab. As vehicles passed over that joint, drivers felt a mild impact and a noticeable thump was heard at the bearings. This indicated not only separation between the girders and supports, but also the bearings were being subjected to continuous impact loads, literally banging the bearing components together with each live load cycle. This was an unacceptable condition that would surely lead to degradation and ultimate failure of the bearings and joint. It also presented a safety hazard to motorists using the bridge. A solution was to provide permanent tie-downs at the east abutment bearings, that would eliminate the upward deflection, provide a downward force sufficient for proper operation of the bearings, and that would counteract the upward forces due to live loads.

INVESTIGATION The two alternatives considered were additional counterweights, or permanent tie-downs. For the design of either alternative, the downward compressive force required to counteract the uplift had to be determined. At this point, it was known that it would take at least 133 kN of downward force on the multi-rotational (pot) bearings for their proper operation, plus 133 kN additional load to counteract live load effects. Therefore, a concrete counterweight placed between each girder would have to weigh at least 266 kN, and would be 11 cubic meters in volume, or 3.8 cubic meters of solid

steel. Most of the weight would have to be concentrated at the ends of each girder, and this would also impose additional bending stresses in the flanges. The sheer physical requirements and added stress effects of the counterweight alternate made it a poor, if not impossible, choice. The permanent prestressed tie-down system was the best choice, because it was lighter, easier to construct under traffic, transferred the loads through bearing and shear on the girder ends rather than incurring additional bending stresses. Also, the actual system load could be tailored to counteract the expected forces resulting from temperature and the elastic forces developed internally in the girders by forcing the superstructure down to close the gaps and contact the bearing surface.

Although the reason for the uplift was initially attributed to live load effects in combination with possible construction errors (e.g., improper camber and/or misaligned girder splices), there was also a serious concern about what role temperature played in causing the variable gaps that were observed during construction. This was as yet an unquantified effect, that was unacceptable for the development of a safe, effective retrofit design. Also, there were concerns that too much force in the tie-down system could damage the structure, by overstressing other structural components, and/or cracking the bridge deck.

Field-testing was conducted to obtain quantitative measurement of all uplift forces occurring at the supports, and to determine the correlation with temperature. Weather conditions in New York State vary greatly with the season, and the bridge will experience extreme cold and hot temperatures throughout the year. Because of time constraints (i.e., schedule for opening the bridge to traffic), testing could not be conducted throughout the full temperature range that the structure would experience during the year. Also, for any given day, substantial differences in temperature can occur

between the bridge deck above and the supporting steel below. This is due to the difference in thermal conductivity of concrete and steel, and due to the uneven exposure of the two structural components to wind and sun (i.e., bridge deck surface open to these environmental effects, while the steel girders underneath are somewhat insulated by the deck). The field tests collected the data necessary to determine whether ambient temperature or differential temperature (between the deck top surface and the top steel flange) was the controlling factor. A detailed finite-element analysis was also performed. Using the field data for model calibration, the analysis investigated the observed uplift mechanisms, determined design loads for the tie-down system, and evaluated the effects of the tie-down system on the structure. The contractor installed a temporary tie-down system and kept the bridge operational during load testing, design of permanent retrofit, and subsequent traffic use.

FIELD TESTING Instrumentation was installed to measure the net uplift and/or compression exerted by the fascia girder at the bearing location. It also measured temperatures of the ambient air, steel-girder bottom and top flanges, and top surface of the concrete bridge deck. Type T thermocouples (270-4000C range) measured the temperatures of the top and bottom girder flanges and of the top deck surface directly above these flanges (see Fig. 4). A load cell with 110 kN capacity, capable of measuring both compressive and tensile forces, was used to measure the force exerted by the south fascia girder (G1). The load cell was attached on one end by anchors to the abutment, and the other end was rigidly attached to the girder, as shown in Fig. 5. The north fascia was also instrumented to obtain additional temperature data for future use, if necessary. A computer-based data-acquisition system sampled and recorded the load and temperature data at 60-second intervals, continuously for 48 hrs.

These tests confirmed that in addition to live and dead loads, thermal forces were also causing uplift. They also indicated that the thermal uplift forces were independent of ambient temperature, and correlated more or less linearly with the temperature difference between the deck and the top flange of the girder (Figs. 6 and 7). Temperature ranges of the deck surface, girder top flange, ambient air, and the difference of flange and deck temperatures were, respectively, 12 to 400C, 18 to 240C, 10 to 290C, and -25 to -0.440C. The temperature differences between the top and bottom flanges of the girders were relatively negligible. It was also observed that uplift occurred when the concrete deck temperature was higher than the top girder temperature. During this 48-hr test period, the maximum uplift force measured was about 7 kN and the maximum compression downward force was about 22 kN.

ANALYSIS Two independent finite-element models were developed (2) to determine design loads for the tiedown system, and the effect of the restraint system on other components of the bridge structure. In each model, dead loads of structural components were included in stages, according to construction sequence. Effects of dead load, live load, thermal load, and initial uplift deflection were investigated. Appropriate combinations of the load effects were studied. A temperature difference of 28.50 C between the top and bottom portions of the deck was assumed in the analysis based on the field test data. Based on the worst combined load case, a 445 kN design load was recommended for the tiedowns. The dominant load case determined by analysis included forces from thermal effects and the elastic forces required to counteract the initial uplift (which had occurred after completion of steel erection).

SOLUTION Several tie-down systems active and passive, custom fabricated and proprietary systems -- were investigated. Based on the force and structural requirements, and the ease of specifying and construction, a DWYIDAG post-tensioned thread-bar system, with shop-fabricated anchorages was designed (Figs. 8 and 9). Fatigue, secondary stresses due to longitudinal expansion and contraction, long-term durability, and clearances between components all were also considered in the design.

The post-tensioned rod system consists of: (2) - 35 mm diameter, A-722 post-tensioning rods (one on each side of the web for symmetrical loading), 44 mm thick load plates at the top and bottom anchorages, and special hardened, spherical load bearing nuts at either end. The load plates have milled; hemi-spherical shaped countersinks to accept the spherical load nuts. Loads are transferred at the superstructure through a shop-fabricated stiffener box-assembly attached to each side of the girder web. It is composed of a 12.7 mm web plate and (2) - 229 mm x 12.5 mm stiffeners welded to a 25.4 mm thick tapered base plate. The assembly is bolted to the girder web with (10)- 19 mm diameter, high strength field-drilled bolts. The stiffener spacing is wide enough to allow the post-tensioning jack to be mounted above the lower girder flange, in between the stiffeners, so that the tensioning was performed from above. The anchorage to the abutment is shop-fabricated using a 38 mm horizontal bearing plate, (3)-19 mm triangular stiffener plates, and a 19 mm vertical mounting plate, which is placed flush against the abutment face. It is anchored with (8)- 25.4 mm diameter, A449 anchor rods that are drilled and grouted, horizontally, into the abutment.

Long-term durability is provided by means of a sealed grease-filled plastic sheathing system that encases the rods and load nuts. Slotted holes were field-drilled in the girder flanges and base plates

to provide clearance for the rods to flex without binding or damaging the rod or girder flange, when the beam undergoes longitudinal expansion (thermal) or end rotation (live load).

The system was constructed under the traffic, however traffic was diverted during the actual tensioning operations. The post-tensioning was performed in stages, until each girder was snug against the bearing top flange. Then another 133 kN (66.5 kN per rod) was applied in stages, equally to both rods of each girder. The final cost was about $55,000 for the five restraints . The system was inspected 11/2 years after installation and was found to be functioning well.

CONCLUSIONS Field-testing was conducted to estimate the forces required for a permanent tie-down system. Uplift forces such as those reported here may be caused by faulty erection of girders, poor distribution of dead and live loads, and from thermal forces developed when temperature differentials occur between structural components. These thermal forces may be estimated by finite-element analysis, using thermal parameters measured by direct field-testing. The results may then be used to design a permanent tie-down system to counteract the uplift forces from the combined effects of temperature, live and gravity loads.

ACKNOWLEDGMENT The authors gratefully acknowledge assistance many New York State Department of Transportation personnel during this study. Everett Dillon assisted in the field-testing. Gongkang Fu, Jyotirmay Lall, and Ruijia Mu conducted the finite element analysis required for the study. The views expressed here are those of the authors and not necessarily of the Department. This work was partially funded by the

Federal Highway Administration, U.S. Department of Transportation.

REFERENCES 1. Standard Specifications for Highway Bridges. American Association of State Highway and Transportation Officials, 15th ed., Washington, D.C., 1992.

2. Mu, R., Lall, J., and Fu, G. Finite-Element Analysis to Counteract Uplift of A Doubly Curved Steel Bridge. Client Report 79, Transportation Research and Development Bureau, New York State Department of Transportation, Albany, NY, 1997.

Figure 1. South fascia (note horizontal curvature)

25 Uplift (kN) : - tension, + compression 20 15 10 5 0 -5 -10 -25 -20 -15 -10 Temperature Difference (oC) -5 0

Linear fit

Note: Data were collected at 1-min intervals; each point represents an average of 1 hr of data (60 records)

Figure 6. Uplift force with temperature difference between deck and Girder 1 flange top.

25

0
Temp Dif Bet Deck and G1 Fl. Top (oC)

Load (kN) : - tension, + compression

20

-5
15
Temp Diff.

10 5 0

-10

-15

-20
-5
Uplift Force

-10 0 10 20 30 40 50

-25 Time (hours)

Note: Data was collected at one minute intervals. Each point in this graph represents an average of one hour data (60 records).

Figure 7. Time history of uplift force and temperature difference.

Вам также может понравиться