FIELD TESTING AND LOAD RATING REPORT: BRIDGE OVER RUSH RIVER CASS COUNTY, ND

Transcription

1 FIELD TESTING AND LOAD RATING REPORT: BRIDGE OVER RUSH RIVER CASS COUNTY, ND PREPARED BY: REVIEWED BY: BRIDGE DIAGNOSTICS, INC th Court North, Suite 100 Boulder, CO HOUSTON ENGINEERING, INC ST Ave North Fargo, ND July 2012

2 BRIDGE LOAD RATING PREPARED FOR CASS COUNTY HIGHWAY DEPARTMENT & NORTH DAKOTA DEPARTMENT OF TRANSPORTATION COUNTY ROAD 32 OVER RUSH RIVER BRIDGE NO FARGO HIGHWAY DISTRICT CASS COUNTY, ND DATE OF LAST INSPECTION: 5/2010 DATE OF RATING: 6/2011

3 EXECUTIVE SUMMARY In March of 2012, Bridge Diagnostics, Inc. (BDI) was contracted by Houston Engineering Inc. to perform diagnostic load testing and Non-Destructive Evaluation on two bridges in Cass County, North Dakota in order to develop current load ratings for the North Dakota Department of Transportation (NDDOT). Bridge was found to be a single-span post-tensioned concrete girder bridge and was the second of the two bridges tested. During the field testing phase, the superstructure was instrumented with a combination of strain transducers, deflection sensors, and tiltmeter rotation sensors. Once the structure was instrumented, controlled load tests were performed with a 3-axle dump truck along three lateral positions. Data obtained from the load tests was evaluated for quality and subsequently used to verify and calibrate a finite-element model of the structure. During the structural investigation, all available geometric data was recorded and compared with the previously collected data. Additionally, beam details including the location and sizes of both the stirrups, Post-Tension (PT) ducts, and the deck reinforcement were determined using GPR techniques. All of the information obtained from this investigation was compiled into the As-Inspected drawings that have been provided with this report. Although much of the crucial information was determined from BDI s structural investigation, it was still necessary to make some educated estimates on certain parameters that were critical in the calculations of the structure s capacity; namely the design material properties and the number of PT wires in each duct. Using design and fabrication information obtained from structural plans from a similar structure (e.g., same bridge type built in the time frame in the same geographic area), it was assumed that this bridge was designed for H-15 loading and that the design was based on allowable stress. A required area of PT steel was then backcalculated based on the required midspan flexural stress. The result was a total of ksi 0.25 diameter stress-relieved PT wires (22 wires in each of the three post-tensioning ducts). It was verified that 22 wires could fit in the ducts, which were measured on-site. It should also be noted that this area of steel is near the practical limit that can fit within the ducts. This steel configuration was then used for all subsequent load rating calculations. Load ratings were performed according to the AASHTO LFR method for the standard rating vehicles. The strength based load rating results were controlled by the ultimate flexural capacity of the girders at midspan; while the serviceability based ratings were controlled by allowable concrete tension stress at midspan of the girders. The results indicated that the bridge essentially met the HS-20 Inventory criteria for strength but failed to meet serviceability limits. Note that the only slightly unsatisfactory Inventory level flexural rating of 0.97 was for the HS20 loading under the two lanes loaded condition. The following tables provide a summary of load rating values for both the strength and serviceability limit states. The strength based load ratings were significantly greater than the original design load ratings due to improved load distribution and the use of a different limit state and applied load factors. Serviceability load limits were similar to the assumed design load (H-15) because while the more accurate analysis reduced the beams live-load effects, the assumed pre-stress losses prescribed in AASHTO Standard Specification were greater than those used in the original design procedures. Therefore, from a serviceability perspective the gains from the load test were cancelled by the change in design/rating practices. BDI considers the strength based load ratings to be the more realistic bridge assessment since they are based on fewer assumptions. Serviceability ratings are heavily influenced by assumed FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND III

4 pre-stress losses which are known to be highly subjective. It may be more appropriate to base load postings on strength limits and adjust serviceability limits with information from current inspection records. If the bridge owner decides that the strength based ratings are used to determine the posting requirements, the current load posting on this structure could be removed. Additionally, since the concrete allowable stress may be an issue for the structure s extended life, future inspections should focus on the detection and growth of flexural cracks in the concrete beams. Critical load rating factors & weights for standard rating vehicles Strength. RATING VEHICLE HS-20 Type 3 Type 3-3 Type 3S2 LOCATION/LIMITING CAPACITY Midspan of Interior Beam / Flexure Midspan of Interior Beam / Flexure Midspan of Interior Beam / Flexure Midspan of Interior Beam / Flexure INVENTORY RATING FACTOR INVENTORY RATING WEIGHT, TONS OPERATING RATING FACTOR OPERATING RATING WEIGHT, TONS Critical load rating factors & weights for standard rating vehicles Serviceability. RATING VEHICLE LOCATION/LIMITING CAPACITY RATING FACTOR - SINGLE LANE RATING WEIGHT - SINGLE LANE, TONS RATING FACTOR - MULTI-LANE RATING WEIGHT - MULTI-LANE, TONS HS-20 Type 3 Type 3-3 Type 3S2 Midspan of Interior Beam / Concrete Tension Midspan of Interior Beam / Concrete Tension Midspan of Interior Beam / Concrete Tension Midspan of Interior Beam / Concrete Tension This report contains details regarding the instrumentation and load testing procedures, a qualitative review of the load test data, a brief explanation of the modeling steps, and a summary of the load rating methods and results. The load test, structural investigation, and load rating results presented in this report correspond to the structure at the time of testing. Any structural degradation, damage, and/or retrofits must be taken into account in future ratings. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND IV

5 Submittal Notes: This submittal includes the following files on CD: 1. Bridge _Testing_Documents.pdf This file provides pertinent details about the instrumentation plan and testing scenarios/procedures. 2. Bridge _As-Inspected_Plans.pdf This file provides pertinent details about the instrumentation plan and testing scenarios/procedures. 3. BDI_ Bridge _Submittal_V1.pdf This is the BDI report in pdf format. It contains details regarding the testing procedures, provides a qualitative data evaluation, displays response histories for each sensor, and discusses any notable observations and/or conclusions arising from the testing process. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND V

6 TABLE OF CONTENTS EXECUTIVE SUMMARY... III TABLE OF CONTENTS... VI 1. STRUCTURAL TESTING PROCEDURES PRELIMINARY INVESTIGATION OF TEST RESULTS STRUCTURAL INVESTIGATION RESULTS BACKGROUND OF INVESTIGATION METHODS STRUCTURAL INVESTIGATION RESULTS MODELING, ANALYSIS, AND DATA CORRELATION MODELING PROCEDURES MODEL CALIBRATION RESULTS LOAD RATING PROCEDURES AND RESULTS APPROXIMATION OF STRUCTURAL CAPACITIES - PROCEDURES/ASSUMPTIONS LOAD RATING PROCEDURES/ASSUMPTIONS LOAD RATING RESULTS CONCLUSIONS APPENDIX A BRIDGE INVENTORY (05/2010) APPENDIX B CAPACITY APPROXIMATIONS APPENDIX C- REFERENCES FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND VI

7 1. STRUCTURAL TESTING PROCEDURES Bridge is a post-tensioned concrete beam structure that carries two lanes of County Road 4 over Rush River in Cass County, North Dakota. The overall width was approximately 27-4 (26-0 roadway width) and the overall length was approximately 60-0 (57-0 clear span). All of the important readily-available geometric details were recorded during the field visit. Additionally, a through non-destructive evaluation was completed by BDI in order to determine an approximate capacity of the structure. See Section 3 Non-Destructive Evaluation for a detailed description of the information gained through this investigation. The structure was instrumented with 28 reusable, surface-mount strain transducers (Figure 1.1 through Figure 1.3, 5 cantilevered displacement sensors (Figure 1.3), and 6 tiltmeter rotation sensors (Figure 1.1). The final instrumentation plans, including sensor locations and IDs, have been provided in Figure 1.5 through Figure 1.8 and are also provided in the drawing file labeled CR4_Testing_Documents.pdf. Once the instrumentation was installed, a series of diagnostic load tests were completed with the truck traveling at crawl speed (3 to 5 mph). During testing, data was recorded on all channels at sample rate of 40 Hz as the test vehicle (3-axle dump truck) crossed the structure in the eastbound direction along three different lateral positions, referred to as Paths Y1, Y2, and Y3 (Figure 1.9). The truck s longitudinal position was wirelessly tracked so that the response data could later be viewed as both a function of time and vehicle position. During the actual live-load test procedures no other vehicles were allowed on the bridge. Information specific to the load tests can be found in Table 1.1. The test vehicle s gross weight, axle weights, and wheel rollout distance (required for tracking its position along the structure) are provided in Table 1.2. A vehicle footprint is also shown in Figure The vehicle weights were obtained from certified scales at a local gravel pit, and all vehicle dimensions were measured in the field at the time of testing. BDI would like to thank Houston Engineering for their help in scheduling, planning, and organizing the testing project. BDI would also like to thank the Cass County Public Works field team for their excellent field support and assistance in bridge access and traffic control at each bridge site. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 1

8 Table 1.1 Structure description & testing info. ITEM DESCRIPTION STRUCTURE NAME Bridge BDI PROJECT NUMBER ND TESTING DATE May 22, 2012 CLIENT S STRUCTURE ID # LOCATION/ROUTE County Road 32 over Rush River, Cass County, ND STRUCTURE TYPE Post-Tensioned Concrete Beam Bridge TOTAL NUMBER OF SPANS 1 SPAN LENGTHS 60-2 SKEW N/A STRUCTURE/ROADWAY WIDTHS Structure: 27-4 / Roadway: 26-0 WEARING SURFACE Concrete Deck SPANS TESTED 1 TEST REFERENCE LOCATION (BOW) (X=0,Y=0) TEST VEHICLE DIRECTION TEST BEGINNING POINT LOAD POSITIONS NUMBER/TYPE OF SENSORS SAMPLE RATE NUMBER OF TEST VEHICLES 1 STRUCTURE ACCESS TYPE STRUCTURE ACCESS PROVIDED BY TRAFFIC CONTROL PROVIDED BY TOTAL FIELD TESTING TIME South-west corner of the structure along the inside edge of the curb Eastbound Front axle 15.3 ft west of test reference location (BOW) See attached testing documents - 28 Strain Transducers - 5 Deflection Sensors - 6 Rotation Sensors 40 Hz Snooper Cass County Cass County 1 day TEST FILE INFORMATION: FILE NAME LATERAL POSITION FIELD COMMENTS OTHER TEST COMMENTS: CR4_1 Y1 Good test. Slight Node issue but resolved. CR4_2 Y1 Bad test. Nodes dropped. CR4_3 Y1 Good test. CR4_4 Y2 Good test. CR4_5 Y2 Good test. CR4_6 Y3 Good test. Y3 CR4_HS Y2 Good test. 35 mph. Weather Sunny, ~80 F, Very Windy FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 2

9 Figure 1.1 Surface mounted strain transducer and rotation sensor near girder end (typical). Figure 1.2 Strain transducer on top of curb at midspan (typical). FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 3

10 Figure 1.3 Displacement and strain sensor installed at midspan (typical). Figure 1.4 Test Reference Location Beginning of World (BOW). FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 4

11 Figure 1.5 Instrumentation Plan Plan view with sensor locations, sensor IDs, and corresponding channel IDs. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 5

12 Figure 1.6 Cross-sectional View of Section A-A Including sensor and channel IDs. Figure 1.7 Cross-sectional View of Section B-B Including sensor and channel IDs. Figure 1.8 Cross-sectional View of Section C-C Including sensor and channel IDs. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 6

13 Figure 1.9 Breakdown of lateral truck paths Y1, Y2, and Y3 relative to the BOW. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 7

14 Table 1.2 Test vehicle information. VEHICLE TYPE TANDEM REAR AXLE DUMP TRUCK GROSS VEHICLE WEIGHT (GVW) 48,440 lbs WEIGHT/WIDTH - AXLE 1: FRONT 15,980 lbs 7-0 WEIGHT/WIDTH - AXLE 3: REAR TANDEM PAIR 32,460 lbs (~15,830 lbs each) 7-2 SPACING: AXLE 1 - AXLE SPACING: AXLE 2 AXLE WEIGHTS PROVIDED BY WHEEL ROLLOUT DISTANCE Cass County 10.6 per wheel revolution Figure 1.10 Test truck footprint. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 8

15 2. PRELIMINARY INVESTIGATION OF TEST RESULTS All of the field data was examined graphically to provide a qualitative assessment of the structure's live-load response. Some indicators of data quality include reproducibility between tests along identical truck paths, elastic behavior (strains returning to zero after truck crossing), and any unusual-shaped responses that might indicate nonlinear behavior or possible gage malfunctions. This process can provide a significant amount of insight into how a structure responds to live-load, and is often extremely helpful in performing an efficient and accurate structural analysis. RESPONSES AS A FUNCTION OF LOAD POSITION: Data recorded from the wireless truck position indicator (BDI AutoClicker) was processed so that the corresponding strain and displacement data could be presented as a function of vehicle position. This step was critical in comparing the measured and computed responses. REPRODUCIBILITY AND LINEARITY OF RESPONSES: The structural responses from identical tests were very reproducible as shown in Figure 2.1, Figure 2.2, and Figure 2.3. In addition, all strains appeared to be linear with respect to magnitude and truck position, and all strains returned to nearly zero, indicating that the structure was acting in a linear-elastic manner. The majority of the response histories had a similar degree of reproducibility and linearity, indicating that the data was of good quality. CONSISTENT COMPOSITE BEHAVIOR: Measurements taken along the girders verified the presence of composite behavior between the deck and the girders. The field measured neutral axis locations at midspan were found to be consistently near the top portion of the AASHTO girders. Figure 2.4 shows the top and bottom strains for Girder 4 at midspan under a truck path Y2 loading. These responses are consistent with composite behavior and are a good indication that the structure was designed to be composite for live-load, likely through the use of shear studs and/or shear keys. Based on this finding, all subsequent load rating calculations were based on composite behavior for positive moment. OBSERVED END-RESTRAINT NEAR SUPPORT LOCATIONS: A fair level of end-restraint was observed in the sensors placed near the supports, as shown in Figure 2.5. End-restraint was detected as compression strain or negative flexure near the supports while the test truck was still on the structure. A slight variation was detected between the degree of end restraint generated by the flat and rounded bearing plates. This observed behavior was carefully considered during the modeling calibration process. STRAINS MEASURED ON STRUCTURE S CURBS: Strains were taken on the curbs of the structure near midspan. These responses were recorded to evaluate the curb s participation in resisting the truck load. Figure 2.6 shows strain histories from truck path Y1 along the south curb (top and bottom gages). The significant magnitudes of these responses indicate that the curbs were providing a fair amount of edge stiffness and contributing to the stiffness of the exterior beams. This observed structural behavior was considered during the modeling. LATERAL LOAD DISTRIBUTION: When evaluating a bridge for the purpose of developing a load rating, the bridge s ability to distribute load to other beam-lines is an essential characteristic to quantify. Lateral distribution can easily be observed by plotting the responses from an entire lateral cross-section, as done in Figure 2.7 and Figure 2.8. Figure 2.7 displays the midspan displacements from all three truck paths, while Figure 2.8 displays the corresponding midspan strains. The response values shown in these figures correspond to the FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 9

16 longitudinal load positions producing the maximum midspan responses for each truck path. From these figures, it was observed that the structure exhibited a significant level of lateral load distribution across its cross-section and that overall the responses were very symmetrical, which is a good indication that no signs of distress were present during testing. As previously stated, all test data was initially processed and assessed for quality. Then, one set of test data for each truck path was selected for having the best apparent quality. This selected data was then used to calibrate the finite-element (FE) model of the structure, which was in turn used to produce the load ratings. Table 2.1 provides a list of the data files that were used in the FE analysis. Table 2.1 Selected truck path file information. TRUCK PATH Y1 Y2 Y3 SELECTED DATA FILE CR32_3.dat CR32_5.dat CR32_7.dat FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 10

17 Truck Path Y3 tests Truck Path Y2 tests Truck Path Y1 tests Figure 2.1 Example of strain response reproducibility. Truck Path Y3 tests Truck Path Y2 tests Truck Path Y1 tests Figure 2.2 Example of displacement response reproducibility. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 11

18 Truck Path Y1 tests Truck Path Y2 tests Truck Path Y3 tests Figure 2.3 Example of rotation response reproducibility. Bottom Gage Response Top Gage Response: Indicating that the N.A. was near the top of the girder Figure 2.4 Observed Composite Behavior Girder 3 at midspan. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 12

19 Negative response denoted compression along bottom of girder near abutments during loading, which is a direct indication of end-restraint. Figure 2.5 Observed End-Restraint Behavior- Girder 3 near west abutment. Gage along bottom of curb/deck Gage along top of curb Figure 2.6 Strain Response History from South Curb at midspan indicating structural participation. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 13

20 Figure 2.7 Lateral load distribution observed in maximum displacement responses. Figure 2.8 Lateral load distribution observed in maximum strain responses. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 14

21 3. STRUCTURAL INVESTIGATION RESULTS Due to the lack of As-Built plans for Bridge , BDI performed an investigation that was used to create As-Inspected plans. These As-Inspected plans were utilized to approximate capacities, which in turn were used calculate ratings for this structure. During the investigation, all available geometric data was first recorded, followed by an examination of the AASHTO girder reinforcement details. This reinforcement examination consisted of ground penetrating radar (GPR) techniques coupled with physical verification at remote locations. Below are brief descriptions of the investigation methods used and the subsequent results. BACKGROUND OF INVESTIGATION METHODS Ground Penetrating Radar The Ground Penetrating Radar (GPR) method involves the transmission of high frequency electromagnetic radio (radar) pulses into the object of interest and measuring the time elapsed between transmission, reflection off a buried discontinuity, and reception back at a surface radar antenna. The discontinuities where reflections occur are typically produced by metallic reinforcing, voids or by changes in electrical properties of the underlying materials. GPR was used to locate the position and spacing or reinforcing steel, including the posttensioning ducts and shear stirrups. GPR was also used to investigate the thickness of the concrete deck. The equipment used was a Geophysical Survey Systems Incorporated (GSSI) SIR-3000 unit with both a 2.6 MHz antennae and Palm 2 GHz antennae units. Both antennae units provided high resolution scans that enabled the testing crew to locate the shear reinforcing and post-tensioning ducts in the concrete beams. Figure 3.1 shows a BDI engineer performing a scan along the bottom of one of the girders near the end of the structure. Verification and Calibration Verification of reinforcing diameter was done by drilling small holes to intercept the reinforcing steel and then physically measuring the diameter of the stirrups/ducts and cover of the exposed steel as shown in Figure 3.2 and Figure 3.3. Holes were later filled with a nonshrink cementitious patching material. Calibration of the GPR system was typically done at the location of a known thickness of concrete (i.e., web of AASHTO Type II girder). Here, the time of reflection of a known discontinuity was compared to the measured thickness of the concrete at this location to establish a velocity of the radar energy in the concrete. Once this velocity had been determined, the depth to subsequent reflections was estimated provided that the radar energy as traveling through a similar material. This calibration was completed before the thickness of the deck was estimated. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 15

22 STRUCTURAL INVESTIGATION RESULTS The structure was found to have an overall length of 60-0 (57-0 clear span) and an overall width of 27-4 (26-0 roadway). The superstructure consists of 5 post-tensioned (PT) AASHTO Type II girders spaced at approximately 6-0 on center that were found to act compositely with the 6 thick reinforced concrete (RC) deck. The structure had PT diaphragms at third points that had approximate dimension of 31 x9. The general structural details have been provided in Figure 3.4 thru Figure 3.6. Note that no haunch was found on this structure, which differed from the 1 haunch found on Bridge The girders were found to have three PT ducts that were approximately 1.75 in outer diameter, and the center of these ducts was found to be an average of 4 from the bottom surface of the girders. An exception to this duct location was found along the center PT duct. This duct was draped starting approximately 8-6 from the girder s midspan to the center of bearing. At the end of the drape it was found that the distance from the bottom of the girder to the center of the draped duct was approximately 2-0. These girder PT duct details have been provided in Figure 3.7 and Figure 3.8. During the investigation of the shear reinforcing, it was found that the shear stirrups were ½ diameter #4 rebar, as shown in Figure 3.2. The approximate shear stirrup locations found during the investigation have been provided in Figure 3.9. The stirrup locations shown in this figure were found to be symmetrical about the girder s midspan in multiple beams. Three dimensional GPR scans were performed on the bridge deck in order to verify the reinforcement patterns present at the top and the bottom of the deck. These 3D scans involved first marking out a 4 x4 deck section with a line increment of 4 both longitudinally and transversely. Then a series of scans along these 4 spacings was performed so that a 3D profile of the deck could be rendered. From these scans, it was determined that transverse reinforcement was spaced at 6 along both the top and bottom of the deck, as shown in Figure The longitudinal steel was found to have a 12 spacing along the bottom mat while having a 16 spacing along the top reinforcing mat. It is important to note that this reinforcing pattern appears to match what can still be gleaned from the As-Built details found of the deck. Since the deck reinforcing was not considered critical for the rating of this structure, no verification of the size of this reinforcing was performed. However, it appears that from provided deck plans that this reinforcing is most likely #4 rebar. As additional reference a few example GPR scan images have been provided. First, Figure 3.11 shows transverse scans along three different girders bottom surface that highlight the variable depth of the three PT ducts. As shown in this figure, the depth of the PT ducts were found to vary along the girder s length by approximately ±2. Next, Figure 3.12 displays a GPR scan taken longitudinally along a girder s web and highlights the depth, location, and spacing of the shear stirrups found in the girders. Figure 3.13 shows a longitudinal scan centered along the bottom of a girder that illustrates the observed depth to the draped PT duct. Lastly, Figure 3.14 shows the top reinforcement layout in the deck from the rendered 3D scan of a 4 x4 deck section. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 16

23 Important Investigation Notes: A variation of up to 2 in stirrup spacing and duct locations were observed during the investigation. This variation was expected given the construction practices used in the time period this structure was built. The stirrup and duct locations that have been provided are averaged values based on collected data and engineering judgment. Due to limitations caused by the antennae unit size and overall access limitations, some assumptions and/or educated guesses were required to complete the As- Inspected plans. For example, due to lack of access near the beam ends it was required to assume that the shear stirrup spacing found closest to the end diaphragm continued to the girders center of bearing. Since it appeared that both the top and bottom transverse deck reinforcement was placed in-line with each other, it was assumed that the both the top and bottom reinforcing mat was spaced at 6 in the transverse direction. The reason that this assumption was necessary was because the reflection of the top mat made it impossible to see the bottom mat. However, the provided deck plans offered verification that this assumption (both the top and bottom transverse reinforcing had a 6 spacing) was reasonable. Figure 3.1 BDI engineer performing a GPR scan along the bottom of a girder. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 17

24 Figure 3.2 Example Location of Verification of Shear Stirrup Diameter. Figure 3.3 Example Location of Verification of Post-Tension Duct Diameter. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 18

25 Figure 3.4 General Structural Details Plan View. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 19

26 Figure 3.5 General Structural Details Cross-Sectional View. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 20

27 Figure 3.6 General Structural Details Elevation View highlighting support details. Figure 3.7 Girder Details Cross-sectional view highlighting post-tensioned duct locations. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 21

28 Figure 3.8 Girder Details Elevation View highlighting approximate post-tensioning duct locations. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 22

29 Figure 3.9 Girder Details Elevation view highlighting shear stirrup layout. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 23

30 Figure 3.10 Deck Reinforcement Details View of reinforcement in scanned 4 x4 area. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 24

31 1 st Girder Scan 2 nd Girder Scan 3 rd Girder Scan Reflection of a Single PT Duct Vertical Scale designates the calibrated depth (in inches) from the bottom face of the girders to the PT ducts Figure 3.11 Example GPR Scan Transverse scans along multiple girder bottoms highlighting variability of PT duct depth. Approximate spacing between stirrups (24 away from beam ends) Reflection of shear stirrup Reflection of opposite side of web Figure 3.12 Example GPR Scan Longitudinal scan along single girder web highlighting location, depth, and spacing of shear stirrups. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 25

32 Reflection of shear stirrup Reflection of center PT duct Figure 3.13 Example GPR Scan Longitudinal scan along center of girder bottom highlighting draped PT duct location. Figure 3.14 Example 3D GPR Scan Plan view (Left) and elevation view (Right) of 4 x4 deck section, highlighting the top reinforcement. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 26

33 4. MODELING, ANALYSIS, AND DATA CORRELATION This section briefly describes the methods and findings of the Bridge modeling procedures, and a list of modeling and analysis parameters specific to this bridge is provided in Table 4.1. MODELING PROCEDURES First, geometric data collected in the field and insight gained from the qualitative data investigation were used to create an initial, two-dimensional finite-element model using BDI s WinGEN modeling software which is illustrated in Figure 4.1. Once the initial model was created, the load test procedures were reproduced using BDI s WinSAC structural analysis and data correlation software. This was done by moving a two-dimensional footprint of the test truck across the model in consecutive load cases that simulated the designated truck path used in the field. The analytical responses of this simulation were then compared to the field responses to validate the model s basic structure and to identify any gross modeling deficiencies. Figure 4.1 FE model of superstructure with modeled test truck load starting to cross. The model was then calibrated until an acceptable match between the measured and analytical responses was achieved. This calibration involved an iterative process of optimizing material properties and boundary conditions until they were realistically quantified. During the optimization process, various error values are computed by the analysis program that provides a quantitative measure of the model accuracy and improvement. The error is quantified in four different ways, each providing a different perspective of the model's ability to represent the actual structure; an absolute error, a percent error, a scale error and a correlation coefficient. In the case of this structure, the majority of the calibration effort was spent modeling the observed load distribution behavior, both longitudinally and transversely. Accurate modeling of the girder s boundary conditions at the abutment locations was also critical for reproducing the longitudinal load distribution. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 27

34 Table 4.1 Analysis and model details. - Linear-elastic finite element - stiffness method. ANALYSIS TYPE MODEL GEOMETRY NODAL LOCATIONS MODEL COMPONENTS LIVE-LOAD DEAD-LOAD TOTAL NUMBER OF RESPONSE COMPARISONS MODEL STATISTICS ADJUSTABLE PARAMETERS FOR MODEL CALIBRATION - 2D composed of shell elements, frame elements, and springs. - Nodes placed at the ends of all frame elements. - Nodes placed at all four corners of each shell element. - Shell elements representing the slab elements. - Frame elements representing the curb. - Springs representing the boundary conditions at the abutment walls. - 2-D footprint of test truck consisting of 10 vertical point loads. Truck paths simulated by series of load cases with truck footprint moving at 1.5 ft increments along a straight path. - Self-weight of structure strain gage locations x 165 load positions = 4,290 strain comparisons - 5 displacement gage locations x 165 load positions = 825 displacement comparisons - 6 rotation gage locations x 165 load positions = 990 displacement comparisons Nodes Elements - 11 Cross-section/Material types Load Cases - 37 Gage locations Member Stiffness, E [ksi] 1. Deck Slab/Diaphragms 2. AASHTO Girders 3. Curb Member Eccentricity, Ecc [in] 4. Curb 5. AASHTO Girders Eccentric Axial Spring Resistance, Fx [kips/in] 6. Fixed Bearings 7. Rocker Bearings Member Torsional Resistance, J [in4] 8. Abutments/End Diaphragms FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 28

35 MODEL CALIBRATION RESULTS Following the optimization procedures, the final model produced a 0.98 correlation with the measured responses, which can be considered an excellent match for any structure. The parameters and model accuracy values used in the initial and final bridge models are provided in Table 4.2. The final model was found to closely match the member displacements, strains, and rotations; as shown in the comparison plots provided in Figure 3.2 through Figure 3.6. Additionally, the model s midspan lateral distribution of strain and displacement closely matched that of the actual structure as shown in Figure 3.7 and Figure 3.8. The following is a summary of the conclusions from the calibration results: The girders were found to act completely composite with the deck. An eccentricity term for the girders that helped model these members compositely with the deck was found to greatly increase the correlation between the measured and computed responses. This composite action was considered for rating. The structure was found to be very stiff in both the longitudinal and transverse direction, as indicated by the large stiffness values calibrated for the girders, diaphragms, curbs and deck (Young s Modulus of ksi). It should be noted that these stiffness values are effective values and are not necessarily to be considered actual material stiffness values. These effective stiffnesses account for variations between the structure s actual complex behavior and necessary model assumptions. For example, possible variations could have come from parameters such as the effective span length, assumption of uniform deck thickness, lack of a modeled haunch, member interaction, etc. In general, it was found that this structure was acting slightly stiffer than the other tested structure, Bridge along CR4. This was not expected since this structure did not have as much observed end-restraint and did not have a 1 haunch between the deck and the beams as Bridge did. A fair amount of end-restraint was verified during the calibration process. This endrestraint was found to reduce the midspan responses and was primarily caused by horizontal friction between the embedded beam bearing plate and the steel plate on the abutment wall. Additional end-restraint was likely induced by soil pressure applied along the end diaphragms over the abutment walls. While this type of resistance is typically not considered in design, a portion of the end-restraint was utilized for load rating since this structure has retained this behavior over the 47 years of its service. To maintain a conservative assumption, 50% of the observed end-restraint was allowed for the live-load analysis. No end-restraint was applied during the dead-load analysis. The curbs were found to cause a substantial amount of edge-stiffening. These elements were allowed to increase in stiffness to account for the interaction with the guard rails. The additional stiffness beyond the deck modulus was eliminated for the load rating model since the influence of the guard rails would not be reliable with heavier loads. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 29

36 Table 4.2 Model accuracy & parameter values. MODELING PARAMETER Member Stiffness - Deck Slab/Diaphragms E [ksi] INITIAL MODEL VALUE (ALL GAGES) 3,200 FINAL MODEL VALUE 4,500 - AASHTO Girders E [ksi] 3,200 5,500 - Curb E [ksi] 3,200 5,500 Member Eccentricity - Curb E cc [in] AASHTO Girders E cc [in] Eccentric Axial Spring Resistance - Bearings along west support F x [kips/in] Bearings along east support F x [kips/in] Member Torsional Resistance - Abutments/End Diaphragms J [in 4 ] ,000 INITIAL MODEL FINAL MODEL FINAL MODEL ERROR PARAMETERS VALUE VALUE VALUE (ALL GAGES) (ALL GAGES) (NO TOP GAGES) Absolute Error 544, , ,566.9 Percent Error % 7.0% 5.9% Scale Error 244.3% 12.0% 6.7% Correlation Coefficient FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 30

37 Figure 4.2 Final model example strain comparison plot at gage point on A-A. Figure 4.3 Final model example rotation comparison plot at gage point on A-A. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 31

38 Figure 4.4 Final model - example strain comparison plot at gage point on B-B. Figure 4.5 Final model - example displacement comparison plot at gage point on B-B. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 32

39 Figure 4.6 Final model example strain comparison plot at gage point on C-C. Figure 4.6 Final model example rotation comparison plot at gage point on C-C. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 33

40 Figure 4.7 Final model Lateral load distribution displacement comparison along B-B. Figure 4.8 Final model Lateral load distribution strain comparison along B-B. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 34

41 5. LOAD RATING PROCEDURES AND RESULTS APPROXIMATION OF STRUCTURAL CAPACITIES - PROCEDURES/ASSUMPTIONS Although the results of the structural investigation provided much needed information about the structure s capacity, some of the most important details, like number of PT wires present in each duct and design material properties, were still unknown. Therefore these important unknown parameters had to be approximated. The following is a brief description of the approximation methods used to come up with the missing information, and in turn calculate the approximate moment and shear capacities for this structure. Based on the observed PT duct size, provided plans of a similar bridge built around the same time, and sound engineering judgment; the following assumptions were made in order to approximate the beam capacities: Plans for a similar PT girder bridge located along I-94 in Cass County were provided as reference to possible design procedures and subsequent design material properties. From analytical investigation of these plans it was determined that this highway bridge was likely designed for the AASHTO H20 loading. Since this I-94 bridge likely had a higher importance than Bridge (along County Road 4) and since the I-94 bridge had larger AASHTO type beams, it was considered necessary to assume that Bridge was designed for the AASHTO H15 loading, two lanes loaded. From these plans and general information on materials used around 1965, it was assumed that the design material properties for this structure were as follows: o Post-Tension Wires: 250 ksi 0.25 diameter stress-relieved wires o Beam Concrete: An initial concrete strength of 4 ksi and a final design compressive strength of 4.5 ksi. These strengths were primarily based on the core test values provided on the I-94 bridge plans. It was assumed, based on the very consistent composite behavior, that the structure was designed to be composite for live-load; which resulted in non-composite and composite section moduli of S nc =3,220 in 3 and S c =5,618 in 3, respectively. A simple beam analysis with the AASHTO live-load wheel distribution factor equation S/5.5 was utilized in determining the required design live-load capacity. It was assumed that the centers of the PT ducts were likely designed to be closer to 3.5 from the bottom of the beams. This is a noteworthy assumption since a significant variation in duct depth was found during the structural investigation and the average depth ranged from 3 to 4 inches. Note that for the capacities utilized during BDI s rating, the ducts were conservatively assumed to be between from the bottom of the beams. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 35

42 Based on the 1.75 outer diameter of the PT duct measured in the field, it was assumed that the inside diameter was likely closer to 1.5. Using a drafting program, it was determined that up to approximately wires could fit in this assumed inside diameter. However due to realistic limitations of pulling these wires through the ducts, it was decided that the practical limit was likely 22 wires per duct, shown below. Due to the age of the structure and details provided in the I-94 bridge plans, it was comfortably assumed that the structure was designed according to ACI or AASHTO ASD standards with load applications similar to AASHTO specifications. For the beams with un-bonded reinforcement, a concrete tension stress limit of 3 f`c was used to approximate the needed number of PT wires, according to AASHTO Based on the information gathered and the assumptions listed above it was determined that a total of diameter PT wires were needed to meet the stress limit under an AASHTO H15 loading. This number of wires fits well with the approximate maximum number of wires per duct using the assumed 1.5 PT duct inner diameter. This calculation has been provided in Appendix B. Therefore it was assumed that each PT duct contained ksi 0.25 diameter stressrelieved wires. From this approximation, moment and shear capacities were determined along the beams length according to AASHTO Standard Specifications for Highway Bridges, 17 th Edition Note that the moment capacities were determined along the length of the beams using a Microsoft Excel spreadsheet while the shear capacities were taken from a model created in the AASHTO Virtis Rating Program. Virtis was used to calculate the shear capacities for ease of determining the effects of the prestressing and applied load effects on these capacities. Summaries of the critical moment and shear capacities have been provided in Table 5.1 and Table 5.2, respectively, while example calculations for these capacities have been provided in Appendix B. Table 5.1 Critical PT girder moment capacity (ΦMn, k-in). LOCATION AREA OF PT WIRES, IN 2 AVERAGE PT STRESS AT ULTIMATE, KSI DISTANCE FROM STEEL CENTROID TO COMPRESSION FACE, INCHES DESIGN MOMENT CAPACITY, KIP- INCH Midspan ,575 FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 36

43 Table 5.2 Critical PT girder shear capacities (ΦVn, k-in). LOCATION At Beam Taper ~0.16 Span (24 spacing) AREA OF SHEAR STIRRUPS, IN 2 SHEAR STIRRUP SPACING, IN SHEAR STRENGTH OF CONCRETE, KIPS SHEAR STRENGTH OF STEEL, KIPS DESIGN SHEAR CAPACITY, KIPS LOAD RATING PROCEDURES/ASSUMPTIONS Load rating was performed on all appropriate bridge elements in accordance with the AASHTO LFR guidelines. Structural responses were obtained from a slightly modified version of the final calibrated model, and member capacities were determined as discussed in the previous sub-section. The rating methods used in BDI s approach closely match typical rating procedures, with the exception that a field-verified finite-element model analysis was used rather than a typical AASHTO slab-strip analysis. This section briefly discusses the methods and findings of the load rating procedures. Once the analytical model was calibrated to produce an acceptable match to the measured responses, the model was adjusted to ensure the reliability of all optimized model parameters. This adjustment involved the identification of any calibrated parameters that could change over time or could become unreliable under heavy loads. Due to different conditions present at the application of dead and live load, different adjustments to the model were made when determining these two types of loads. In the analysis of the Bridge , the following adjustments were made before using the model for dead load and live load calculations: Dead Load Model: o The end-restraint was removed by reducing the eccentric axial spring values to zero. o The deck, curb, and diaphragm stiffnesses were reduced to a very low value to simulate the pouring of these elements. Live Load Model: o The end-restraint was reduced by 50% to insure a conservative load response results. o The curb stiffness was reduced to that of the deck since these elements were poured monolithically with the deck and it was assumed that the guardrail influence would be unreliable. Load ratings were performed on the final calibrated model according to the AASHTO Manual for Bridge Evaluation, Second Edition 2011 (see Table 5.3 for applied rating factors). Given the 26 wide roadway, both one lane and two lane loaded conditions were considered for the rating. Figure 5.1 shows the load configurations for the standard load rating vehicles. All structural dead loads were automatically applied by the modeling program s self-weight function. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 37

44 Table 5.3 Applied AASHTO LFR rating factors and limits. FACTOR TYPE DESCRIPTION FACTOR VALUE AASHTO Load Factors AASHTO Strength Reduction Factors Dead Load (Strength) 1.30 Live Load Inventory (Strength) 2.17 Live Load Operating (Strength) 1.30 Serviceability Load Factors (Dead and Live Load) 1.00 Impact Factor 27% Flexure (Moment) in PT Sections 1.00 Shear in PT Sections 0.90 AASHTO Stress Limits Unbonded PT girder Concrete Tension 3 f`c = 201 psi Figure 5.1 AASHTO standard load rating vehicle configurations. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 38

45 LOAD RATING RESULTS The following is a summary of the load rating factors for the four AASHTO rating vehicles. The critical flexure and serviceability ratings were controlled at midspan while the shear ratings were critical at the point where the stirrup spacings increase to 24 (~9.4 from center of bearing). Summary tables for the critical moment and shear ratings have been provided in Table 5.4 and Table 5.5; while critical serviceability rating factors (for concrete in tension) for both the single and multiple lanes loaded conditions have been provided in Table 5.6 and Table 5.7. As shown, the bridge essentially met LFR rating criteria (RF>1.0) for all standard design and posting loads under the strength limit states considering the two lanes loaded condition; however failed to meet the rating criteria for the serviceability limit state considering both the one and two lanes load conditions. The strength based HS-20 Inventory level rating (RF=0.97, under the two lanes loaded condition, was slightly below 1.0 but all legal load configurations were well above 1.0. Note that if only the single lane loading is considered that only the HS20 serviceability rating is unsatisfactory at It is very important to note that the serviceability ratings were heavily influenced by the assumptions made about the structure s original design and allowable stress design procedures in the AASHTO Standard Specifications. The greatest variables and largest effects on the serviceability ratings are the assumed prestress losses and the assumed allowable tension stress in the concrete. The discrepancy between serviceability and strength ratings is primarily a function of the AASHTO specified prestress losses being 25% greater than those indicated in the design procedure obtained from the I-94 design plans. Therefore the serviceability ratings are considered to be more subjective and adjustable based on current and future inspection results. Since the structure did not exhibit any cracks during the field visit, the serviceability ratings might serve better as an indication that the beams need checked thoroughly for tension cracks near midspan during their routine inspection. Table 5.4 Critical rating results - AASHTO rating vehicles LFR Strength - Flexure. RATING TRUCK HS-20 CONTROLLING LOCATION Interior Beam - Midspan DEAD LOAD MOMENT, KIP INCH LIVE LOAD MOMENT, KIP INCH CRITICAL INVENTORY LEVEL RF CRITICAL OPERATING LEVEL RF Type 3 Type 3-3 Type 3S2 Interior Beam - Midspan Interior Beam - Midspan Interior Beam - Midspan * Both the Live Load and Dead Load Responses are unfactored responses. FIELD TEST AND LOAD RATING REPORT - BRIDGE OVER RUSH RIVER: CASS COUNTY, ND 39