Load Testing Of Continuous Span Bridges

Dr. Rajeev Goel

As soon as the bridges are constructed, they start deteriorating due to environmental impacts and various loads. A significant number of bridges are either structurally deficient or functionally obsolete, requiring immediate attention. Load testing is an effective tool to measure the structural response of a bridge under various loading conditions and to determine its structural integrity. Study revealed that vertical deflections under load test are comparable to the theoretical vertical deflections and percentage recovery of deflections on removal of loads fulfils the acceptance criteria. The nature of the strain on the bottom face of the deck slab and inside the box on the soffit slab of the box section is appropriate and exhibited the desired trend. Bridges are a critical component of a road transport system that allows the safe movement of people and goods from one place to another. Bridges are continuously subjected to environmental effects as well as vehicular loading that lead to their deterioration. Therefore, inspection and load testing are needed to obtain good estimate of bridge response and helps in taking decisions for proper maintenance and rehabilitation.

The purpose of conducting load tests on existing bridges is to evaluate their structural response without causing damages. The objective of load testing is to compare field response of the bridge under test loads with analytical response

Nowadays, various types of sensors are also fixed in some of the important bridges for the load testing. The data obtained from these sensors can give meaningful information regarding the present condition of the bridge. For the load testing of one of the four-span continuous unit of North carriageway of Second Thane Creek Bridge in Mumbai, vibrating wire strain gauges were fixed inside the box-girder to continuously monitor its behaviour during the load test.

The actual response of a bridge under load testing may differ from the theoretical response. Factors that contribute to this difference may include distresses in bridges, composite action, load distribution effects, presence of parapets / railings/ kerbs/ utilities, material property differences, unintended continuity, participation of secondary members, effects of skew, portion of load carried by deck, improper functioning of bearings, etc.

The purpose of conducting load testing on existing bridges is to evaluate their structural response without causing damages. The objective of the load testing is to compare field response of the bridge under test loads with analytical response. Guidelines for load testing of bridges are given in IRC: SP-51-1999. Load testing on bridges can be classified in:
i) Behaviour load test
ii) Proof load test
iii) Stress history Load tests
iv) Ultimate load test
v) Diagnostic load tests
The behaviour load test is carried out to verify the results of any method of analysis or design. The test load could be equal to or lower than the design load. The proof load test is adopted for operational rating of bridge and is done on new structures, which had design/ construction problem or for rating of an existing bridge. The stress history load test is carried out to establish the distribution of stress ranges in fatigue prone areas of bridges. The data obtained from the passage of regular traffic is used to assess the fatigue life. The ultimate load test is performed to understand the global behaviour when sufficient theoretical knowledge is not available to predict the response. This test provides valuable information regarding the sequence and mode of failure. The diagnostic load test is meant to monitor the behaviour of a component of a bridge either to establish the cause of damage or its share in transfer of loads when enough theoretical analysis is not developed.

Activities involved in load testing typically include determination of testing objectives and load configuration, selection of span for load testing, the selection and placement of instrumentation, the visual inspection before, and during load test, the adoption of appropriate analysis techniques, and the evaluation and comparison of test results and analytical results.

Second Thane Creek Bridge, Mumbai
The North carriageway of Second Thane Creek Bridge was constructed in the year 1994 on Sion-Panvel Road, near Vashi village in Mumbai. It comprises of total 23 spans as shown in Fig.1. The total length of the bridge is about 1,840 m. It comprises of total 23 spans, which consists of six continuous Units (i.e., one unit of three-spans from Vashi end, one unit of four-spans at Mumbai end and four intermediate units of four-spans each). The single cell prestressed concrete box girder superstructure of the bridge was constructed by the free cast in-situ cantilever construction method. The maximum depth of box girder over the pier support is 7.0 m, reducing to 3.5 m at the mid span. The superstructure is supported over Pot-PTFE bearings. The carriageway width is 11 m (3-lane) with 600 mm wide footpath on North side and 890mm central median on South side. Fig.2 shows the general view of North Carriageway of the bridge. General view of inside the box girder is shown in Fig.3. The piers are of reinforced concrete solid piers supported over open foundation from pier P1 to P21 and well foundation under pier P22. The RCC abutment at Vashi end is supported on open foundation whereas Mumbai end abutment rests on well foundation.

Load Testing
Load testing of one unit of North carriageway of this bridge, which is in between Pier P11 and Pier P15 was carried out in the month of April 2013. The superstructure of this unit is supported over piers through Pot-PTFE bearings. The bearings over piers P11, P12, P14 and P15 allows transverse as well as longitudinal movements while the bearings over pier P13 did not allow movement in any direction. Finger type expansion joints have been provided at the locations of pier P11 and P15. The span P12-P13 (Fig.4) was loaded to produce bending moment equivalent to 3-lane of IRC Class A loading.

Load testing was carried out to assess the flexural capacity of the bridge superstructure at working load in the elastic range, through measuring the vertical deflections of the superstructure. The objective of the load test was to monitor and to quantify the global behaviour of the bridge. The general acceptance criteria of load testing are:
i) Concordance between measured and calculated vertical deflections
ii) Crack width should be within limits, and
iii) Percentage recovery of vertical deflections of superstructures on the removal of the applied loads should be at least 85%

The bridges in India are normally tested for an IRC class of loading which is used for their design. As the hypothetical vehicle corresponding to 3-lane of IRC Class A loading are not available and the bridge lies on a busy National Highway and therefore, it was not feasible to conduct a static load test, which needed a closure of traffic for a minimum period of 4 to 5 days. In view of these constraints, it was decided to conduct a representative load test by using pre-weighed loaded trucks to produce the equivalent bending moment. To conduct a load test, the important parameters to be considered are as follows:
? Instrumentation
? Loading Scheme and Arrangements
? Testing Procedure

This Unit was instrumented for the following observations during the load testing:
? Measurement of vertical deflection at road level and inside the box-girders using Total Station (of least count as 1mm) at the locations is shown in Fig.5.
? Measurement of strain at the mid-span of P12 - P13 and P13 - P14 in deck slab as well as soffit of the box girder using Surface mounted Vibrating wire strain gauges at the locations is shown in Fig.6.
? Monitoring of the longitudinal movement of bearings using dial gauges (of least count as 0.01mm) at four supports i.e. P11, P12, P14 and P15 is shown in Fig.7.
? Monitoring of existing cracks by fixing glass strips across them.
Initial readings of all the installed instruments were taken before the loads were placed, to provide a baseline. Readings of the installed instruments were taken again to determine the change due to specified loads.
Loading Arrangements
As it was not feasible to produce the desired moments in each span of this unit by the placement of vehicular load, therefore, the maximum sagging moment in the span P12-P13 was considered for load testing and theoretical analysis was carried out accordingly.
The load test was planned to be conducted by placing three-axle TATA make trucks with GVW about 16 Tonne each over span P12 - P13, to simulate the bending moment as obtained from analytical study for IRC loads.
Testing Procedure
Step 1: Before carrying out any load testing of a bridge, condition survey by visual inspection is essential. Material deterioration is often indicated by surface cracking and spalling of concrete. The crack pattern indicates the cause of cracking. Hence, the visual inspection is not confined to the surface only; it also includes examination of the condition of the Pot-PTFE bearings and expansion joints, walkway and any other utility services.
Step 2: Fix the instruments for measurements of ambient temperature, vertical deflection of superstructure, movement of bearings, VW strain gauges and glass strips across the cracks at the designed locations.
Step 3: Before conducting a load test on any bridge, it is essential to monitor the vertical deflection of the superstructure as well as movement of bearings of the bridge due to the variations in the ambient temperature, without any load over the bridge.
Step 4: Prior to loading the deck, first, wheel positions of the loaded trucks shall be marked on the road surface for the accurate positioning of the trucks as per the loading scheme.
Step 5: The initial readings of various gauges and levels shall be taken without load on the road level (deck) of the carriageway and inside the box section. The initial data of dial gauges used for monitoring the function of the bearings shall also be recorded simultaneously.
Step-6: Loads then shall be placed in several loading-stages as per the loading scheme and after the completion of loading of each stage, readings of the installed instruments shall be taken.
Step-7: After placement of 100% of planned load, readings of installed gauges shall be taken at regular interval for next 24 hours.
Step-8: Unloading shall then be started in the reverse way loads were placed over the bridge. At the end of each stage of unloading, readings of the installed instruments shall be taken.

Analytical Study
This unit was modelled and analysed theoretically by using software RM 2006. The appropriate boundary conditions as per existing bearings were taken. In modelling, this unit was divided into 292 segments of box each having a length of 1.6719 m. Fig.8 shows the different views of modelled unit in RM - 2006.

This unit was then analysed for IRC Class - AA Tracked, IRC Class - AA Wheeled, IRC Class - 70R Tracked, IRC Class - 70R Wheeled and IRC Class - A vehicle loading as per relevant provisions of IRC -6. From the analysis, it was found that three-lanes of IRC Class - A vehicle with three trains 20 m apart (Fig.9) produced the maximum positive (sagging) bending moment of 23,557 kN - m inclusive of impact factor and reduction of live, at the mid-span of span P12 - P13 and negative moments in spans P13 - P14 and P11 - P12, respectively.

The load test was conducted by using fifteen numbers of three - axle TATA Trucks with GVW about 16 Tonne each, to simulate the bending moment equivalent to 3 - lane of IRC Class A loading. The simulation of these trucks enabled to produce an equivalent moment of about 92% of 3 - lanes of IRC Class - A loading and the same was considered for load testing as the precise loading in trucks would have not been possible for remaining load. The bending moment due to 3-lanes of IRC class - A loading and this equivalent vehicular loading is shown in Fig.10.

Load on the span P12 - P13 was applied in several stages. The theoretical magnitude of bending moment and vertical deflections corresponding to 3 - lane of IRC Class A loading and different stages of equivalent vehicular loading at various locations of span P11 to P15 are shown in Fig.11 and Fig.12 respectively.

Field Investigations
1. Visual inspection (condition survey)
Tell-tales (glass strips) were fixed across the observed cracks in concrete top deck slab of box girder, webs of span P11 to P15. The crack width was measured at different locations on the deck slab and on the webs. The Pot - PTFE bearings (sliding, guide) were found to be lacking in lubrication. Trough along the finger type expansion joints was found badly damaged. The steel walkway which was fixed to the superstructure did not have any provision to accommodate the expansion/contraction of the superstructure.
2. Load testing
Fifteen numbers of pre-weighed Tata trucks LPK - 2516 TC 6x4 BS - II were positioned on span P12 - P13 to produce the bending moment equivalent to that of 3 - lane of IRC Class-A loading during the load test. The total weight of 15 Tata Trucks corresponding to 3 - lane of IRC class loading was 240.92 Tonnes say 241 Tonnes, which was about 8% less than the GVW of 3 - lanes of IRC Class A train (one train) of vehicles loading inclusive of an impact factor of 8.9%.
This Unit was monitored for temperature displacement measurements induced due to variations in the ambient temperature on April 12, 2013. The levels of deck slab and soffit slab were recorded with the help of four Total Stations at all the measurement locations. The longitudinal movement of bearings over the supports was also recorded simultaneously. The individual loaded trucks were gradually brought to the marked positions.
The loaded trucks were placed over the span P12 - P13 in four load stages. In load stage 1, 2, 3 and 4, total numbers of loaded trucks placed over the span P12 - P13 were five, nine, thirteen and fifteen respectively. Placement locations of trucks at fourth stage are shown in Fig.13.

The data collection during each stage of loading took an average time of about 45 minutes. The loading was commenced on April 13, 2013 at about 1030 hrs and completed at 1530 hrs. The applied test load i.e., 15 trucks were kept in position for 24 hours and the hourly data was recorded. Unloading was started on April 14, 2013 at about 1530 hours and the unloading sequence was reverse of the loading stages i.e., unloading of Stage-4 followed by the remaining stages. The parameters that were monitored during various stages of loading were same even during the various stages of unloading. The deflection and recovery on 100% loading and unloading along U/S (South) and D/S (North) of each span is shown in graphical form in Fig.14 and Fig.15.

The strain variation at mid-span of span P12 - P13 during loading and unloading is shown in Table - 1.

The maximum crack width was observed to be about 0.1mm in deck slab and in the range of 0.10 to 0.15 mm in both the webs of box-girder. Neither new cracks were developed, nor the existing cracks propagated/widened during the loading and unloading operations. The glass strips did not exhibit any sign of cracking or debonding from the concrete surface.

The movement of bearings at pier P11, P12, P14 and P15 during loading and unloading on April 13-14, 2013 is shown in Fig.16.

Discussion of Test Results
The maximum instantaneous deflection values at 100% of applied load at the mid-span of the loaded span P12-P13 is 48.50mm. As per clause No. 12.4 of IRC-112-2011, the limiting value of deflection under live load is Span/800, which works out to be 133.75mm for span of 107m. Thus, the observed deflection of the loaded span P12-P13 is within permissible limit. Further, the observed instantaneous deflection values from field test is comparable to the theoretical deflection due to 3-Lane of IRC Class A.

As the maximum deflection occurred at the mid span of span P12-P13, therefore, the percentage recovery of maximum deflection for acceptance pertains to the loaded span. This Unit has exhibited recovery more than 88 per cent.

The bearings on support P12 exhibited some degree of restraint. Due to the restraint of the bearings on pier P12, the bearings on pier P11 moved out ward i.e., towards P10 on removal of applied test load. The bearings on the supports P14 and P15 have exhibited a similar trend. This has affected the recovery of movement during unloading.

The strain variation at different stages of the applied loading and unloading is shown in Fig.17. Maximum transverse strain on the bottom face of the deck slab in span P12-P13 at 100% loading was 164.94 µ¤ (Tension) and on the top of the soffit slab was 22.65 µ¤ (compression) in span P12-P13. The strain in the longitudinal direction recorded from L-B & T-B were124.81 µ¤(Tension) and 60.01 µ¤ (compression), respectively. From above, it was inferred that the recorded strain values were less than the permissible maximum strain values for a concrete grade 40MPa.

Based on the visual inspection and results of load testing of this Unit of North Carriageway of Second Thane Creek Bridge, the following conclusions were drawn:
? The vertical deflection values of bridge superstructure under load test are comparable to the theoretical deflection due to 3 - Lane of IRC Class A loading.
? The percentage recovery on removal of loads full fills the acceptance criteria.
? Neither the existing cracks got activated nor did any higher order deformations (deflection) occurred at any stage of loading / unloading.
? The bearings exhibit restraint in their functioning.
? The recorded strain values are fund to be less than the strain values corresponding to the limiting strain of concrete to initiate cracking.
? The Unit behaved elastically during the load test.

1. D. M. Frangopol, A. Strauss, and S. Kim (2008), “Bridge reliability assessment based on monitoring,” Journal of Bridge Engineering, Vol. 13, No. 3, pp. 258–270.
2. Hubo Cai, Osama Abudayyeh, Ikhlas Abdel-Qader, Upul Attanayake, Joseph Barbera, and Eyad Almaita (2012), “Bridge Deck Load Testing Using Sensors and Optical Survey Equipment”, Advances in Civil Engineering, Vol. 2012.
3. IRC: 6-2010, “Standard specifications and code of practice for Road bridges (Section II) Load and stresses”, Indian Roads Congress, New Delhi
4. IRC: SP 37-2010, “Guidelines for evaluation of load carrying capacity of bridges”, Indian Roads Congress, New Delhi
5. IRC: SP 51-1999, “Guidelines for load testing of bridges”, Indian Roads Congress, New Delhi.
6. Project report, Load testing of Unit-4 (P11 to P15) of North carriageway of second Thane creek bridge, Vashi, Mumbai, CRRI report No. CRRI/QSP/BAS/01/CNP-1892, June 2013.
7. Raina VK (2002), “Bridge inspection (why and what to look for)”, IRC Journal, Vol. 63, No.3, pp. 543-624.