Rating and Posting of Bridges

Article by Sandipan Goswami 1 The procedure for full scale load testing of bridge super-structure is discussed in this article and that includes recommendations for the acceptance criteria. Now-a-days the bridge load testing is considered as a routine requirement across the world, to meet the requirement of the designs, constructional quality of forthcoming bridge structures and as an acceptance test conforming to the provisions of the standard Code. The purpose of the load test is mainly to assess the flexural capacity, and the required parameters can be measured directly and accurately. The bridge deck-girder super-structures are rarely tested for shear strength evaluation due to absence of any reliable method. The load testing envisaged in general is only for assessing the strength and evaluating the load carrying capacity for purposes of rating. Whereas, the load testing, that is discussed here is for assessing the behaviour of a bridge by application of design live loads over a period of 24 hours for confirmation of the elastic performance of the super-structure. This discussion deals with ‘Proof Load Test’ which covers testing of super-structures, excluding arches for evaluation of their flexural capacity and the shear capacity is not considered. This test is not intended to assess ultimate load carrying capacity of bridge deck-girder super-structure. Unless specified otherwise, the details of the method shall generally follow the recommendations given in this article. Bride Load Test is more appropriate to decide on the need for replacement or rehabilitation of existing bridges when construction drawings and specifications originally followed are not available, and the condition survey warrants for load test. For carrying out the detail design for new bridges, when data from design standards and site situations by studying other bridges on the route don’t suffice for design, the bridge load test on existing bridges provide the required information. Following circumstances are to be considered to conduct the Load test for rating purposes and for posting purposes: a) Load Test for Rating i) Load Test for rating may be done in situation when it is not possible to determine the rated capacity of a bridge due to lack of essential details. ii) Load testing for rating is always recommended for masonry arch bridges. Test Vehicles for Rating Rating is essentially done to verify which of the loadings described in the standard in combination with other loads can be assigned to the bridge. Use of mobile test vehicles duplicating the axle loads for various classes of loadings should be preferred as compared to the use of equivalent static load which are difficult, time consuming and needs longer closure of traffic on the bridge. 2 The advantage of using such mobile vehicles is that they can be quickly positioned in the exactly required locations. Also being rolling loads, all cross-sections of superstructure are tested without having to workout special loading patterns to represent envelope diagrams of the span for bending and shear. In exceptional cases if commercial vehicles as specified in standards are used, the number and spacing of such vehicles need to be worked out so as to produce equivalent B.M. and shear at critical sections on those due to the standard loading. b) Load Test for Posting Load Test for posting is conducted if is not possible to verify the strength of all elements of existing structure by analytical methods due to lack of reliable data. Test Vehicles for Posting The test vehicles are selected from the commercially available vehicles as specified in Standards and are chosen as the next heavier vehicle than the predominant heavy vehicles presently plying over the bridge. After the load testing with the first vehicle is complete and found to be satisfactory, the second next heavier vehicle may be considered for testing, if that testing is required, permitted and available for testing. The test vehicle used for purpose of posting shall allow for appropriate overload factor based on actual traffic data in the region. Positioning of Test Vehicles To produce the maximum bending moment effects on girders the test vehicles should be placed at marked locations on the bridge. The vehicles will preferably be moved from both directions leading to their final positioning at the desired location on the deck. The response of the structure to loading may be checked at a few critical selected locations. The following checks are usually considered as adequate for posting purposes: a) For slabs/girders/ box section bridges the mid-span region and 1/4th span for sagging B.M. b) For cantilever bridges, continuous bridges and the bridges with over hangs the support section for hogging B.M. c) For effect of shear forces at support and at points of changes in web thickness. Arch bridges In case of arch bridges, the rear axle of a standard truck shall be placed on the crown. In case of twin tandem rear axle, the rear twin tandem axles shall be placed symmetrically about the transverse centre line of the bridge. 3 The load for rating for girder bridges shall be taken as the least of the followings: (i) For simply supported spans, the load causing a deflection of either 1/1500 of the span in any of the main girders or for cantilever spans the load causing a deflection of 1/800 of the cantilever span in any of the main girders, may be considered for the rated load. For calculating the deflection, the rotation of the piers on either end should be accounted for. (ii) The load that is causing tension cracks of width more than 0.3 mm in any of the girders for normal cases and 0.2 mm for structures exposed to very severe and adverse conditions may be considered for the rated load. (iii) The load causing appearance of visible new diagonal cracks of width more than 0.3 mm for normal cases and 0.2 mm for structures exposed to very severe and adverse conditions or opening/widening of existing cracks close to the supports in concrete girders may be considered for the rated load. (iv) The load at which recovery of deflection on removal of test load is not less than 75 percent for R.C.C. structures, 85 percent for pre-stressed concrete structures may be considered for the rated load. Temperature correction be considered as per provisions contained in the relevant standard. For rating of bridge structures, the following three are the generally followed methods. (i) Analytical Method may be followed when either design or as-built drawings and specifications are available, or when such drawings can be prepared by site measurement to an acceptable level of accuracy (e.g. for steel, masonry or composite bridges). In case the drawings are available, their correctness is to be verified at site, as commonly, the As-built drawings/data are not available. (ii) Load Testing Method may be followed when construction/As-built drawings and specifications are not available, or when data used in design cannot be obtained and by the condition survey, the extent of corrosion and loss of strength cannot be assessed. (iii) Correlation Method may be followed in certain cases, when it is possible to ascertain the safe carrying capacity of the bridge structure by correlating the sectional details of the structure with those bridges having identical sectional details, construction specifications and whose safe load carrying capacities are known. In case (iii), it is however, necessary to know the differences and relative deterioration of the bridge under question vis-a-vis the details of those structures whose safe carrying capacities are known. This may help for proper assessment by correlation and factoring the known differences and relative deterioration. This presumes that the physical condition of the bridge under question is otherwise satisfactory. The method is not further discussed in these article. 4 The methodology for rating of existing bridges will be for standard live loads as specified in the design standard, and/or for the modified loading suitable for local conditions as specified in the standard or as directed by the engineer. Therefore, for new bridges, the load for rating of the bridge would be the highest class of design live load as applied in the original design. Thus the description will be rated for highest loading Class of the design standard. The requirements of the design codes covering the type of bridge under investigation are to be fully satisfied. In case of recently built bridges, some local defects such as honeycombing, spading of concrete, onset of corrosion, local deep scour etc. may observed during inspection. Such defects can be taken care of by localized repair, and it is not necessary to re-evaluate the rating of the bridge due to such repaired defects. It should be realized that the design standards have built-in margins for time dependent loss of strength, and for localized reduction of strength at sections, which are not critical, which are not explicitly stated in the design codes and standards. However, local reduction up to 10 percent estimated on the basis of calculations does not require re-rating of the entire bridge. When the original designs and drawings of old bridges are not available the Rating is required for reevaluation of many complex factors and conditions of previously rated structures in situations where bridge had suffered deterioration of strength of any of the main components from superstructure to foundations. Items that need to be included in this evaluation are briefly discussed n following paragraphs. ! The loads considered in analysis of bridge for rating purposes, are of three types: i) Design live loads at the time of construction ii) Design live loads in force at the time of re-evaluation iii) Changes in loads in loading standard such as wind and seismic loads iv) Changes in loads based on field observations such as design flood level " # The allowable stresses are to be taken as per the relevant design standards, governing the type of structure under review. The allowable stresses normally take into account the long terms effects. However, these may need to be downgraded in case of specifically observed deterioration of materials and structural components. 5 $ i) The details about the original material specifications used such as grades of concrete/steel etc. have to be obtained from the records. The year of construction shall help for reference to be made to codes and specifications which were in force at that time and allows one to make an assessment of materials most likely to have been used and their design strengths considered at that time. If the information is not available, samples of materials taken from the bridge itself can be examined/tested to determine the type and strength and grades of concrete/ steel etc. While making assessment of the strength of concrete based on the core samples, essential factors need to be taken into account. ii) Affected by corrosion etc., the loss of effective section of reinforcement, or pre-stressing steel has to be based on the investigations specifically made to assess the same. Long term pre-stressing losses are known to be higher than those estimated using earlier codes and standards. Similarly higher effects of creep and shrinkage are to be considered. The original warning about likely higher loss of pre-stressing force may be found from actual observations during inspection or unsatisfactory performance such as excessive deflection, vibration etc. Such defects may be observed in only a part of the bridge, which may be one or two spans out of all spans, or in a few piers/foundations. The re-rating in such cases need not be based on the strength of the/ weakest span/pier/foundation, but can be carried out after repairing/strengthening the affected portion so that better overall rating could be maintained. This approach is termed as improving the strength up to that of the next weakest link after repair. % & Without knowing the loading history of road bridges, it is difficult to assess the fatigue history, or expected future fatigue cycles. The only possible assistance for assessing damage by fatigue or assessing the remaining life is laboratory testing of material samples taken from the bridge. Fatigue is generally relevant for steel bridges. The effects of fatigue on joints and connections is not possible to estimate, and one has to depend on visual inspection for observations such as cracking of welds, loosening of bolts, extent of corrosion etc., and thus the assessment for the effect of fatigue can be made. ' !( ) ) Limit State Design is based on defining various "limits" which should not be exceeded by the structure during the expected design life of the structure subjected to loading from usage, and caused by natural environment. The aggressive elements of the environment have deteriorating effects on the structure. The risk of exceeding the limits may be maintained very low, but generally cannot be made zero. Without making the structure uneconomical or unaffordable, the assessed probability of risk is maintained sufficiently small and acceptable. The design codes of various countries have based on this approach. In assessing the strength of existing structures, these methods are very useful in a rational way examining the increased risk that can be taken while continuing the use of the structures instead of replacing it at a high cost. This is more relevant for the assessment of effects of overdimensioned/overweight vehicles. 6 *+ , # * All postings for bridges should be made in terms of equivalent axle loads and/or gross vehicle weights (GVW) of the commercial vehicles plying on the roads and satisfying provision of the Motor Vehicles Act of the country. For over-dimensioned/over-weight commercial vehicles the limits may also be indicated in the ‘Motor Vehicles Act’. - . () Bridge structure is rated for vehicles classes as per the standard and should be posted for the commercial vehicles by comparing the forces caused by GVW of commercial vehicles with the design forces imposed by the standard class of moving load. Bridges with 3-lane and 4-lane superstructure with simply supported spans from 10 to 75 m visa-vis GVW of commercial vehicles may be considered specially. Only the main total span moments and shears need to be checked. The transverse distribution of load and resulting increase of forces and deflections in different girders or portions of slab bridges has to be done for each bridge as per its geometry. The transverse analysis by effective width method is used in software ASTRA Pro (www.techsoftglobal.com) for the analysis of deck slab and used in the limit state method of design () By the assessment of the bridge condition, the rating/posting engineer would decide the necessity of the following traffic restrictions on the bridge from safety considerations until the exercise of posting is completed. i) Speed Restriction – This should to be into effect until the detailed investigations, strengthening or rehabilitation work and load test (if required) on the repaired bridge are completed. The speed limit of vehicles over the structure will be decided by the bridge authority depending upon the physical condition of the structure. ii) Geometrical Restriction - This would involve reduction of the carriageway width on the bridge to ensure lesser extent of live load on the bridge at a particular time by installation of side and height barriers on either end approaches to restrict passage of large number, overloaded, and oversized commercial vehicle on the bridge. iii) Footpath Loading - Observing the structural condition of the footpath slab, restriction on load on footpath may be imposed until the distressed part is rehabilitated, in order to reduce the total load on the bridge superstructure. 7 " # Postings for bridges should be made on both sides of the bridge and the load regulatory and advance warning signs should be installed on both ends of the bridge on the approaches and at all road junctions leading to the posted bridge. " ! # These signs should be placed at a sufficient distance (not less than 100 m) advance from the bridge abutment, at both ends of the bridge so that truck drivers can use detours or limit their loads to the maximum weight allowed on the bridges. " - / 0 # For all bridges to be posted, an advance informative warning sign indicating a "Limited Load Bridge Ahead" should be placed at least 200 m from the abutments on either end of the bridge and at all road junctions leading to the posted bridge starting from the earliest major junction. $ - - 1 ) The rating of the bridge based on the hypothetical trains of live load cannot be directly restricted by putting controls to traffic actually plying on the bridge. For this purpose, the maximum effects of loads by moving vehicle plying on bridges need to be calculated instead of following standard class Live Loads, and should be combined with loads from other sources such as, flood, wind, etc. These effects can then be compared with the 'rating capacities' obtained from load test which is the directly assessed strengths based on actual observations. For this purpose the live loads to be considered are described below: % !/ ! 1) The design standard should mention about design-permits for use of 'any specific variation' to cover special load conditions. Deviation from the standard loading classes may be made under approval of such special clause in the design standard. 2) The data of actually plying vehicles in the country should be available with the motor vehicles department. The 'design train of vehicle' may be selected based on the following considerations: 8 a) Bridges should be posted for one of the Nominal classes of moving live loads, except over dimensioned vehicles. b) Special studies have to be made for over-dimensioned and overweight container carrying multi-axle vehicles on case to case basis. However, on specially selected routes such as port connectivity routes or routes serving heavy industrial complexes maximum permitted load can be estimated based on the maximum number of axles that can occupy the critical length of the bridge, multiplied by the maximum permitted standard axle load. If the axle load and placement of tyres is different from the standard axles individual case based studies have to be carried out. The analysis may be based on influence line diagrams for bending and shear c) After selecting the reference standard class of vehicle, distance between two successive vehicles, the impact factors and the overload factor are to be taken as below: i) Reference class vehicles at 4.0 m spacing between last axle and first axle of next vehicle without impact factor with appropriate overload factor. ii) Reference class vehicles at 20.0 m spacing between last axle and first axle of next vehicle with impact factor with appropriate overload factor. The above is further clarified as below: i) Moving Traffic Condition The road on which the bridge structure is situated may have been designed for vehicle speed of 80 to 100 km/hr. However, at more normal (lower speed) of about 60 km/hr, the reasonable distance between two successive vehicles is considered as two to four times the length of vehicles (approx. 10 m to 25 m). For simplifying the calculations, on an average distance of 15 m is specified between two vehicles for all classes of standard vehicles. At this distance full impact factor should be considered as per relevant standard for moving live load on the bridge. ii) Crowded Loading or Traffic Jam Condition Vehicle standing 'bumber-to-bumber' with spacing of 4.0 m between back axle of vehicle to front axle of vehicle behind it is observed in practice. Number of vehicle or part there of that is causing the load effect on a span of all the girders, should be considered in design of a section, and is decided by the influence line diagrams for bending and shear for that section. It is to be noted that, ‘No impact factor’ should be considered in this crowded loading/traffic jam case. The overload factor that is to be considered should be based on the actual traffic survey data, but should not be less than the average overload factor indicated in the standard. The overload factor is to be considered only for the purpose of checking structural safety of a bridge, and is not for an official approval for overloaded vehicles. The laws of the country will govern in this matter. 9 d) Transverse Spacing of Vehicles The transverse spacing of all vehicles should be taken as mentioned in the standard. The design of deck slab supported between girder webs, or cantilever slab should be verified considering the effect of rear axle loads of the design standard class vehicles with impact factor. The concept of ‘Transverse analysis of bridge deck slab by effective width method’ (Ref. Bridge Deck Slab Design by Software ASTRA Pro, www.techsoftglobal.com) may be used for the same as a simplified method, instead of analysis made of full span using grid analysis with finite element method. For non-standard distribution of wheels on an axle exact analysis taking into account actual positions of wheels is recommended. "2 3 ( ! ) ! The load testing of bridges serves for many purposes, such as a design aid, for verification of satisfactory performance, for new bridges to meet the acceptance criteria, identification of defects and subsequent repair, rating and posting of bridges, retrofitting to meet new requirements of standards etc. The basic features of load testing of bridge deck-girder super-structures are as follows: 1) The test is planned to best achieve the purposes of the testing 2) The deck-girder component of the bridge which represents the super-structure is selected for the test as required. 3) The test member in the super-structure may either be the normal type of element, or the defective which is adversely affected element due to presence of some known or suspected shortcomings. The design detailing errors, construction defects, deterioration of steel connections, corrosion of reinforcement, and inadequate strength of concrete are some of such shortcomings. 4) The design standards and construction specifications are followed to achieve the safety and serviceability in the process of design and construction. For obvious reasons the ultimate strength of the bridge components is not feasible to test. However, by using test loads at a level equal to the design service load, the serviceability performance can be tested duly considering the aims of the test and limitations in the procedure. 5) In load test the load is applied in steps and the response of the structural element is recorded using appropriate instruments. The strains, rotations, deflections and vibrations may be the various types of responses. 6) Each expected response is calculated by using advanced analytical methods, such as stage analysis (by using software such as ASTRA Pro), than the methods used in the normal design activity. Each observed response is to be compared with the expected response. 10 7) In all load-tests the action effects by the environment, such as flood, normal wind and temperature effects are restricted to the unavoidable minimum. In case of temperature effects, temperature corrections are to be invariably made in respect of standard temperature to the observed response to test load, measured in terms of strains and/or deformation. 8) The maximum possible information is to be extracted from the test to decide further course of action. " ) ) / ) For the bridge load tests the type and details of are decided and planned by the purpose of the test, or a series of tests. Following are the purposes and respective types of tests briefly described below: " + ! Diagnostic load tests are carried out to determine certain response characteristics of a bridge. Load tests are categorized as static or dynamic and both are typically used for diagnostic load tests. Diagnostic load tests are typically used to measure load effects on a bridge and compare them with analytical models or theoretical calculations. Calibrated finite element models (FEM) can be developed based on diagnostic tests to predict bridge response to different loading scenarios. Bridge and structure analysis software ASTRA Pro (www.techsoftglobal.com) provides the necessary modelling and analysis by producing the detail analysis report. Load transfer mechanisms in the bridge structure changes over the period of time due to geometrical changes such as, settlement of bearings and supports, deterioration, and hence the requirements for repair/rehabilitation measures. Therefore, it becomes necessary to verify the extent to which the load carrying capacity is affected, and the type of restoration measures required. Since the method and details of testing depend on the nature of change and its effects, the test has to be specifically designed to obtain the required information and data. " ! On the Bridge or its constituent elements the ‘Proof Load Test’ is to be carried out as per relevant standard. This helps ensuring that the design intents are satisfactorily met by measuring the response thereof. Proof tests are used to evaluate a bridge’s safe load carrying capacity and static load tests are primarily used for proof load tests. Proof load tests are more useful when a bridge cannot be analytically load rated (e.g., when material or section properties are unavailable and cannot be accurately determined). The Load Test may be done either for a superstructure span or part thereof, which is subjected to combination of self load, imposed loads and some percentage of live loads can be carried out as specified in the contract for procurement and construction. " - + On the following circumstances the test may be carried out to assist the design of forthcoming bridges: 11 • In case doubt about the adequacy of the analysis models • Where similar components are used are large in number • To check the correctness of the assumptions made in the design The method adopted for interpretation and use of test results should fulfil the requirements of the design. Relevant literatures may be concerned for further information. " " - 3 1 / # ( To verify the results of any method of analysis or design the ‘Behaviour tests’ is carried out as mentioned in the following cases. The test load could be equal to or less than the design load. " • In case the design model is not based on the working loads, in 'Limit State Method', design of some components with detailing of reinforcement is based only on the Ultimate Limit State (ULS). • In some cases the design model may not be sufficiently accurate, e.g. stresses around the openings on the bridge deck, covered with iron grating, to send the storm water to drainage pipes. • The load test may be required to determine the load path and load sharing between alternative load paths, which is too complex or uncertain. $ . ( ) These tests are to measure the improvement after repair and the nature and purpose of these tests are similar to the Proof Load Test, Behaviour Test, or repetition of diagnostic test. " % , # & The member stresses are closely predictable in steel structures. Based on ensuring adequacy on plastic redistribution, redundancy, over design, testing or combination of these strategies, the design and detailing of joints are done. In this approach, the stresses are non-linear even for load levels for which the main structural members are in the elastic range and exact level of stresses remains unknown. In welded connections the effect of locked-in stresses is significant, which have not been stress-relieved as a part of fabrication and after erection. For conducting fatigue load tests on such joints, strain gauges may be provided at typical joints in the structure and the strain and stress range can be established. " " ! # #) For existing bridges the critical span/spans for load test shall be decided, after studies of structural details (either design or as-built or measured), repairing records and detail inspection. 12 " - / A detailed activity schedule including method statement is to be prepared, programmed for the timely completion of load testing. The following details with the method statement are essential for conducting the load test, i) ii) iii) iv) v) vi) vii) viii) ix) x) xi) xii) xiii) The Bridge must be having expansion joints in place, The wearing course must be laid on the bridge deck, On critical locations of bridge whitewash should be applied, At site mobilization of testing personnel, With notes and photographs visual inspection of bridge is to be recorded, With notes and photographs existing status is to be recorded, Proper fixing of instrumentation, Thermal response of the structure is to be recorded, System for temperature correction measurements, Position of the load are to be marked by painting on the bridge deck, Measurement recording, Note on visual inspection during and after load test, Preparation and submission of load test report with all recorded notes. For satisfactory completion of the load test, a ‘flow chart’ may be prepared with the above activities and any other additionally required activities. " # + ( ! Static load tests are carried out to compare, verify and ascertain the theoretical analysis results with actual structural behaviour of the bridge. Dynamic load tests are carried out for special structures if required. The primary objective of the load tests is to understand bridge's response to static and dynamic loading and to verify the dynamic behaviour of bridges. Specialist literature may be concerned for details of carrying out such tests. " # ! It is performed to measure vertical deformation of bridge at mid span or at any predetermined locations. Measurements of recovery of deflection and crack width are the part of static load testing. In special long span bridges such as the suspension and cable stayed, the measured and theoretical deflected shapes of the superstructure during and after the static load tests are also recorded. The deck-girder model of the bridge superstructure may be tested for static, stage and dynamic analyses using bridge and structure analysis software ASTRA Pro (www.techsoftglobal.com) with AASHTO-LRFD or BS Eurocode or IRC standard class load to know the translation and forces developed in the various components. For composite bridges with steel plate girders and reinforced concrete deck slab the orthotropic analysis may also be carried out, which takes into account of the steel and concrete material properties for relevant members of the model. The individual analysis shall give the results separately and to be noted. 13 As the software uses SAP processor, the results should be close to actual tested results, however depends on the correctness of the model developed with appropriate boundary conditions defining nodal degrees of freedom and support conditions. " + ( ! The measurements the dynamic response of the structure as a whole or of some of its members is a new category of tests and is being increasingly used. The bridge and structure engineering software ASTRA Pro (www.techsoftglobal.com) has the facilities to obtain the dynamic analysis results for Eigen values, Response Spectrum and Time History of the bridge deck girder superstructure or the substructure models. In field the structural elements which are susceptible to damage due to dynamic loading are subjected to appropriate dynamic excitation and their response in terms of the frequency, amplitude, accelerations, etc. are measured. By comparing the test results with the model analysis results the strength of the structure may be understood. By using suitable recording instruments, when the response is plotted as a function of time during the test, produces dynamic behaviour of the member. These tests can be used for wind sensitive, long span bridge structures, such as suspension bridges, cable stayed bridges, Arch Suspension Bridges etc. For railway bridges the dynamic loading is often used, and may be replicated in similar manner for steel road bridges in suitable cases. "" ! The Test Load is mainly either of the following two forms: (i) (ii) "" Static Loads Mobile Test Vehicles # ! The load effect on a span can be created by building up pre-weighed units on loading blocks/slabs spaced as per code provisions. The blocks/slabs are built either with steel plates or brick masonry or concrete. Rolled steel sections are placed across a pairs of blocks/slabs, so that platforms could be built on a group of four blocks/slabs for placing another set of pre-weighed units. The area of each platform depends on the magnitude of the load and weight of individual unit. A pre-weighed unit normally made by sand or soil filled gunny bags, concrete cubes, bricks etc, which can be carried manually. Otherwise, large concrete blocks, containers of water or stone-ballast or steel ingots could be used if mechanical handling facilities are possible to load and unload them. The loads are placed eccentrically on the carriageway of a bridge in such a way that maximum bending moment is produced in any longitudinal member. 14 & ( "" # 1 ! & # ! ! By using mobile test vehicles the creating of axle loads for various standard classes of vehicles is convenient and should be preferred as compared to the use of equivalent static load which is difficult, time consuming and requires longer duration for closure of traffic on the bridge. The main advantage of using such mobile vehicles is that, they can be quickly positioned in the exactly required locations. By rolling the wheels and thus the loads, all cross sections of superstructure can be tested without working out special loading patterns to represent envelope diagrams of the span for bending and shear. In cases of using commercial vehicles, the number and spacing of axles and loads of such vehicles need to be worked out, so as to produce equivalent B.M. and shear at critical sections in the structural members as caused by the Standard class live loads. & 4 ! 1 ! & "4 1 5 15 "$ ( 4* ( For achieving a satisfactory test and desired results the correct type, number and location of instrumentation used on a structure are critical during a load test. Deflections, Strains and Inclinations may be measured with following devices. a) b) c) d) e) f) g) Linear Variable Displacement Transducer (LVDT) system with least count of 0.01 mm. Dial gauges, with least count of 0.01 mm Strain gauge and measuring system, with load-count of 1 micro strain Inclinometer, with least count 0.1° Precision digital levelling instrument with bar coded staff and with least count 0.1 mm Total station, with least count of 0.1 mm. Thermometers, Digital or Analogue with least count of 0.5°C $2 ! Theoretical deflection at critical locations of the span to be tested shall be worked out, with design load. This may be done by defining additional nodes in the bridge deck-girder model, defining boundary conditions with fixed and standard class of moving live loads applied on the model. The model may be analysed by software ASTRA Pro (www.techsoftglobal.com). This is to be done before start of loading. The test load shall be applied in stages, so that in case of any untoward distress is observed at any stage a timely action, such as stopping the test, can be taken. The dead load is already acting and the test load is usually a specified factored design live load, with the factor of value more than one. The load placement stages are commonly 50 percent, 75 percent, 90 percent and 100 percent of the full value of the test live load. If required these stages can be altered judiciously when loading is by vehicles. While the structure is loaded in stages, the next incremental loading should be added once the deflections under the previous load have stabilized. The time is also called cooling time and is normally about one hour. Unloading should also be in the same stages as that of loading. At every stage of loading, the test span shall be constantly monitored for appearance and widening of cracks. During nights flood lit must be applied on the entire span under test, i.e. underneath the bridge deck for easy visibility of cracks or distress, if any, at all critical locations. At every increment of loading, the load-deflection characteristics should normally be linear, in case of any abnormal behaviour that will be reflected in the load versus deflection data. In case, the observed deflection exceeds the theoretical deflections (obtained from model analysis by software ASTRA Pro www.techsoftglobal.com) of the structural component (girder) at any stage, further loading, shall be stopped and span is to be allowed to stabilize for longer durations. Sometimes after placing 100 percent load, immediate deflections at a section in the structural component, may exceed theoretical deflections marginally and deflections may get retrieved to the extent, less than the theoretical deflections in less than 2 to 3 hours. If the deflection is not retrieved then the load shall be removed and is to be placed again after 2 to 3 hours. Subsequent action should be taken after consulting all concerned Engineers. 16 Sometimes, cracking sounds at the location of expansion joints are heard, when the rotation capacity is exceeded, commonly, in balanced cantilever bridges. The spalling of delaminated concrete also occurs during the load test, which needs to be investigated before continuing the load test further. $ 3 i) All visual defects shall be measured, mapped and plotted. ii) The functional condition of the Bearings shall be ensured. iii) The functional condition of the Expansion joints and gaps shall be ensured. $ ! Following precautions are to be taken at the time of Load Test: i) All staging shall be stable and safe. ii) Staging for instruments and that for observers shall be independent. iii) Staging for the instruments shall be rigid. iv) Depending on site condition, wherever required a safety staging is to be erected in order to prevent any eventuality to traffic and/or pedestrian below the span under test. v) The superstructure may tend to hog or sag due to temperature variation therefore it is to be ensured that, contact with the spindle of the LVDT/dial gauges is not lost. Spindle extension is to be fixed to take care of this aspect. $ ! 6 Loading operation in stages from 0 percent, 50 percent, 75 percent, 90 percent to 100 percent of test load is to be completed within 24 hours, similarly unloading operations from 100 percent, 90 percent, 75 percent, 50 percent, 0 percent is to be completed within 24 hours. $ After complete loading the structure for 100 percent of its test load, it is to be retained for 24 hours on the structure. $ .2 The structure is to be unloaded immediately after 24 hours in the decremented stages of loading. $ .3 After 24 hours of unloading of 100 percent load from the structure, all required measurements are be recorded. $" ! In general the test procedure shall be as per relevant standard for Load Test of Bridges. Any specific requirements arising out of earlier observations about defects/cracking etc., during inspection shall be taken into account. The following points are important to consider: (i) A whitewash should be applied at the critical sections before the load test for facilitating the observation of behaviour of existing cracks and new formations of cracks during the test. 17 (ii) Observations for any crack in concrete girders in the structure are to be made prior to load test. Any cracks, if observed, are to be measured for their width and marked by painting. The section in terms of external dimensions of the concrete sections and properties of concrete may be used for computation of theoretical deflection by software analysis. (iii) The preferred time to start the load test is generally in early hours of morning or late evening, during such period of the day, the variation in temperature is low. (iv) The test load should be applied in stages at 1 hour interval, with values of 0.5W, 0.75W, 0.90W, and 1 .OW, where "W" is the gross laden weight of the test vehicle. (v) For each load stage, the correspondingly loaded test vehicle is to be brought to the intended marked position. The observed deflections are to be recorded immediately on loading and after five minutes. As the test vehicle is removed from the bridge, instantaneous deflection recovery and deflection recovery after 5 minutes are to be recorded. (vi) $$ After placement of the load, observation is also to be made for development of any new crack and widening of the existing cracks. - Measured deflections and strains at critical locations is to be compared with theoretical deflections by software analysis for the purpose of acceptance/rejection. $$ / + The percentage recovery is to be calculated for values of deflection and is calculated at 24 hours after removal of the load, the analysis is carried out as follows, after applying temperature correction, bearing displacement correction and or rotation corrections to deflection data: a) Initial value - deflections before commencement of first stage loading = R1 b) Deflections after one hour, since placement of 100 percent test load = R2 (If the loading starts at 6:00 am in the morning, then first stage load at 6:00am 0.5W (50% of full load), second stage load at 7:00am 0.75W (75% of full load), third stage load at 8:00am 0.9W (90% of full load) and fourth/final stage load at 9:00am 1.0W (100% of full load), where "W" is the gross laden weight of the test vehicle) c) Deflections at 24 hours after placement of 100% test load = R3 (At 9:00am on the next day) d) Deflection measurements immediately after removal of test load= R4 (removal of first stage load at 9:00am 0.1W (10% of full load), removal of second stage load at 10:00am 0.15W (15% of full load), removal of third stage load at 11:00am 0.25W (25% of full load) and removal of fourth/final stage load at 12:00pm 0.50W (50% of full load) e) Deflection measurements at 24 hours after removal of test load = R5 (removal of fourth/final stage load at 12:00pm) 18 f) Total deflection = R3 - R 1 g) Total recovery of deflection after 24 hours after removal of test load = R3-R5 h) Percentage recovery of deflection 24 hours after removal of test load = (R3-R5) x 100 / (R3-Rl) (If this value exceeds 100 percent, it shall be restricted to 100 percent) The following points are to be noted: $$ • The recovery of deflections shall satisfy the requirements as per relevant clause of the specifications for various types of bridges. • Measured deflections at critical locations shall be compared with theoretical deflections obtained from software analysis, for the purpose of acceptance/rejection. • Measured strains at critical locations shall be compared with theoretical strains obtained from software analysis, for the purpose of acceptance/rejection. - ) . Prior to starting of testing, the theoretical deflections for various stages of loading are to be computed and plotted at points of interest. On this graph, the actually observed deflections shall be plotted during the progress of testing. The plot obtained by the load test results, generally shows linearity of load deflection. If two successive readings show excessive deflection of more than 10 percent from the extended linear behaviour, it can be due to onset of non-linear (plastic) behaviour. The test is to be temporarily discontinued, and the design office is to be concerned for review of the entire procedure. However, load test is to be continued for next 24 hours and deflections are to be recorded. 4 + ! / 1 ) )) "7 ) ( ( + 1. Reinforced Concrete Pre-stressed Concrete # " . () / ( / '$ 8$ 8$ '$ ! 19 %2 ! ( &70- Load ratings are determined in the United States as per the guidelines in the ‘Manual for Bridge Evaluation (MBE)’ by American Association of State Highway and Transportation Officials (AASHTO) [1]. The theoretical approach mentioned in the MBE, has been used to load rate the bridges. As per the MBE, the objective of a load rating is to evaluate the safe live load carrying capacity of a bridge based on as-built construction plans and material properties upon observing any structural damage. Historically bridges have been designed using several different design truck and lane loads. Load ratings are expressed as a Rating Factor (RF) or in terms of a particular truck weight in tons. Two different ratings are determined: (1) Inventory Rating - The Inventory Rating is the live load that can travel over a bridge an indefinite number of times; and (2) Operating Rating - The Operating Rating is the maximum permissible live load that can move across a bridge safely. When a bridge of insufficient capacity is identified, truck loads are restricted by load posting. Due to many of the assumptions made in the calculation process, the theoretical load rating calculated based on the MBE approach tends to be conservative. The live load distribution in longitudinal beams is one assumption evaluated based on recommendations in ‘Specifications for Highway Bridges’ by AASHTO [2]. The MBE mentions for establishing load ratings by using non-destructive load testing, due to its inherent advantages. The MBE identifies two types of load tests: (1) diagnostic tests and (2) proof tests, which are discussed earlier in this article. The MBE [1] lists several factors that increase the live load capacity of a bridge and which can be evaluated through the load test, including: 1. Unintended composite action 2. Unintended continuity/fixity 3. Participation of secondary members 4. Portion of load carried by deck All or most of these factors have influence on live load capacity of non-composite steel girder bridges and if such bridges are analytically load rated these factors are neglected. Field load tests provide accurate load distribution factors for the main beams, which are otherwise based on conservative design load distribution factors. Field test also identifies the possible structural deficiencies and damages which are not observed during routine inspections. 20 The bridge load testing can take several days, for setting up equipment, conducting load testing, and data reduction and analysis. Rapid and efficient load testing of bridges is growing with advances in strain gauge and instrumentation technology. New commercially available gauges greatly reduce the time required to fix sensors across a bridge, which is one of the most time-consuming exercises, and thus making field load test a much faster and more feasible option for load rating of bridges. 6.1 Bridge Load Rating by Analytical Method The MBE provides the details for three rating methods: • • • Allowable Stress Rating (ASR), Load Factor Rating (LFR) and Load and Resistance Factor Rating (LRFR) The LRFR is a most recent development and provides a more uniform safety margin in terms of reliability. For the bridges built before the introduction of LRFR, the researchers carried out the load rating by using LFR. An analytical load rating of a bridge should be carried out before doing any load testing. The MBE specifies the following equation to calculate the RF based on LFR [1]: RF = (C – A1 x D) / (A2 x L) / (I + 1) (1) Where: C = Capacity of the member D = Dead load effect on the member L = Live load effect on the member I = Impact factor A1 = Factor for dead loads A2 = Factor for live loads The factors A1, A2 vary depending on the level of rating performed — Inventory or Operating. For analytical load ratings the dead and live load effects used are typically bending, shear, or axial stresses. The capacity of the rated member is based on the material and sectional properties as well as the rating level. Based on the deck type and beam orientation, the design load distribution factor for analytical rating is calculated based on guidelines provided in ‘Specifications for Highway Bridges’ by AASHTO’ [2]. Bridge and structure analysis software ASTRA Pro (www.techsoftglobal.com) provides the facilities for stage analysis. If the design life is considered for the bridge is 100 years or whatsoever, then it is divided into five stages, considering each stage is of 20 years. The analysis for each stage is performed by adding the deflections at nodes/joints by load of every previous stage to the node/joint coordinates and the analysis is done once again for the current stage and producing the detail analysis reports for each stage. This gives more information on deflection of nodes/joints for a bridge of age 20 or 40 or 60 or 80 or 100 year old. 21 6.2 Bridge Load Rating by Load Testing To produce maximum forces and effects on a bridge, the initial calculations determine the required load of the test truck and the positions of loads. For diagnostic load testing, the test truck should be heavy enough to represent anticipated service loads. No portion of the bridge is considered under nonlinear behavior. Load rating used to be evaluated using Eq. 1 based on field load test results. Load test data for Dead and live load effects are utilized typically as, strain or corresponding stress values. The load rating by using field load test results can be determined based on guidelines in Manual for Bridge Rating through Load Testing by NCHRP [8], the summary of which is given in Chapter 8 of the MBE [1]. The manual provides an adjustment factor (K), based on field test results and other criteria, to modify the RF calculated by Eq. 1 previously. The Eq. 2 is the modification to the rating equation. RFT = RF x K (2) Where: RFT = Load rating factor based on field test RF = Rating factor from Eq. 1 K = Adjustment factor The adjustment factor, K, is calculated using Eq. 3: K = 1 + (Ka x Kb) (3) Ka accounts for both (i) the benefit derived from the load test, if any, and (ii) the consideration of the section factor resisting the applied test load. The general expression for Ka is given below: Ka = (εC / εT) - 1 (4) Where: εT = Maximum member strain measured during load test εC = Corresponding theoretical strain due to the test vehicle and its position on the bridge Kb takes into account for the understanding of the load test results when compared to those predicted by theory, the type and frequency of follow-up inspections, and the presence or absence of special features such as non-redundant framing and fatigue-prone details. The Kb factor is as follows: = 1× 2× 3 (5) By using the tables provided in Manual for Bridge Rating through Load Testing by NCHRP the three compounded factors are obtained [8]. This is to be noted that the MBE [1] does not include Kb1 and Kb2 for calculating Kb. The factor Kb1 considers for the behavior of the rated member, it is linear elastic at 1.33 times the rating vehicle and could be extrapolated. This also takes into account for the ratio between the test vehicle effect and the rating vehicle effect, providing a range of values for Kb1 between 0 and 1.0. Kb2 considers 22 for the type and interval of bridge inspection; it varies from 0.8 to 1.0. Kb3 considers for the presence of critical features that may lead to fatigue or fracture-induced failure; it varies from 0.7 to 1.0. The live load distribution factor has significant influence on the live load effect ‘L’. Provided that all the girders have the same section properties and sufficient strain gauges are fixed at the same location on these girders, then for a particular girder the actual distribution factor can be calculated with the following equation: (6) Where: DFTk εk εi n = Test load distribution factor for girder k = Strain measured in girder k for a given load case = Strain measured in girder i for a given load case = Total number of girders Based on the actual live load distribution factor, the initial analytical load RF from Eq. 1 can be revised, replacing the live load distribution factor taken from Specifications for Highway Bridges by AASHTO [2]. By modifying Eq. 2 with the adjustment factor K, the revised RF should be used to evaluate the rating factor based on the field test. In case the distribution factor calculated with Eq. 6 using the strain values observed during the field test is less than that obtained from the AASHTO Specifications [2], an immediate increase in load rating will be considered. ! ! 9. : (Source Reference: Research Report KTC-19-16/SPR06-423-1F, Bridge Load Testing Versus Bridge Load Rating Abheetha Peiris, Ph.D., Research Engineer and Issam E. Harik, Ph.D., Professor of Civil Engineering, Kentucky Transportation Center, College of Engineering, University of Kentucky, Lexington, Kentucky, In Cooperation With Kentucky Transportation Cabinet, Commonwealth of Kentucky.) AASHTO standard trucks, including the HS20-44, are all hypothetical trucks developed for standardizing the design and load rating of bridges. Pursuant to local state regulations, bridges are typically load rated for several different truck types so trucks with different axle configurations are included. During the project, four additional hypothetical trucks were conceptualized by the Kentucky Transportation Cabinet (KYTC) and were included in the rating analysis. If required, the load posting would be based on the lowest of these ratings. The details of truck axle positions and loading are described in Table 1. Multiplying the RF by the weight of the posting truck, the posting load is calculated. -)) # ; --#7 ) ! 90 < & ! . 2: 2 Area As : 38.3 in Depth d : 33.1 in Web thickness tw : 0.58 in Flange width bf : 11.5 in 23 Flange thickness tf : 0.855 in Nominal Weight ws : 130 lb/ft 4 Moment of Inertia Ixx : 6710 in Elastic section modulus Sxx : 406 in 3 Plastic section modulus Zxx : 467 in 4 3 4+ 5 ! ) Modulus of Steel Es : 29000 ksi Yield strength Fy : 33 ksi (assumed) Concrete density wc : 150 pcf Concrete strength f’c : 3.5 ksi Span Length L : 61 ft Deck height hd : 6.5 in Effective deck width be : 6 ft 24 . 5 () (AASHTO Standard Specifications 10.48.1) (a) Compression Flange OK, (b) Web thickness OK, D= 31.39 in. is the clear distance between the flanges Nominal Flexural Strength Mn = Fy Zxx = 33 × 467 = 1284 kip-ft ! Deck weight = 787.5 lb/ft (includes curb and wearing surface) Steel Beam Weight = 130 lb/ft Total weight wT = 787.5 + 130= 917.5 lb/ft Dead load moment = = 426.75 kip-ft Impact factor = = 0.27 Distribution factor = DF = Ss/5.5 = 6/5.5 = 1.09 Live and Impact load moments • Truck Type HS20 Load per beam P1 = 16 kips (½ the center axle weight at mid span) Live load ML1 = 244 kip-ft 25 Load per beam P2 = 16 kips (½ the rear axle weight 14’ from mid span) Live load ML2 = 132 kip-ft Load per beam P3 = 4 kips (½ the front axle weight 14’ from mid span) Live load ML3 = 33 kip-ft ML = (244+132+33) × 1.09 = 446 kip-ft (AASHTO Manual for condition evaluation of bridges 6.5.1) where; RF = Rating Factor for the live-load carrying capacity. The rating factor multiplied by the rating vehicle in tons gives the rating of the structure. C = Capacity of member D = Dead load effect on member L = Live load effect on member I = Impact factor to be used with the live load effect A1 = Factor for dead loads A2 = Factor for live loads A1 = 1.3 for Inventory and Operating levels A2 = 2.17 for Inventory and 1.3 for Operating levels (AASHTO Manual 6.5.3) Truck Type HS20 Inventory level RF = [1284 – (1.3×426.75)]/(2.17×1.27×446) = 0.594 Operating level RF = [1284 – (1.3×426.75)]/(1.3×1.27×446) = 0.991 -)) ;+ ! ! . ! Field load test Distribution Factor: DF = Σ (εinterior beam/ εall beams) = 114 / (114+111.6+71+45) = 0.334 Live load Moment of Test Truck: Load per beam P1 = 43.44 kips (rear axle weight at mid span ) ML1 = 662.46 kip-ft Load per beam P2 = 18.26 kips (front axle weight 14’ from mid span ) ML2 = 132.38 kip-ft ML = (662.46+132.38) × 0.334 = 264.68 kip-ft The maximum strain due to ML: εc = ML/(E×Sx) = 264.68×12 /(29,000 × 406) 26 = 270 microstrains εT = 159 microstrains (from field test readings) Therefore from NCHRP Manual for Bridge Rating through Load Testing Eq. (6-4) Ka = (εc/εT) – 1 = (270/159) – 1 = 0.7 From NCHRP Manual for Bridge Rating through Load Testing Eq. (6-6) Kb = Kb1 × Kb2 × Kb3 = 1 × 0.8 × 1 = 0.8 From NCHRP Manual for Bridge Rating through Load Testing Eq. (6-3) K = 1 + Ka × Kb = 1 + (0.7 × 0.8) = 1.56 Truck Type HS20 (from calculations in Appendix A) Inventory level RF = 0.594 Operating level RF = 0.991 From NCHRP Manual for Bridge Rating through Load Testing Eq. (6-1) RFT = RF × K Inventory level RFT = 0.594 × 1.56 = 0.92 Operating level RFT = 0.991 × 1.56 = 1.54 Field load test Distribution Factor: DF = Σ (εinterior beam/ εall beams) = 94.6 / (69.1+94.6+79.5+32.7+18.3+6.5) = 0.315 Live load Moment of Test Truck: Load per beam P1 = 27.92 kips (rear axle weight at mid span) ML1 = 172.2 kip-ft ML = (172.2) × 0.315 = 54.2 kip-ft The maximum strain due to ML: εc = ML/(E×Sx) = 54.2×12 /(29,000 × 88.9) = 252 microstrains εT = 98.95 microstrains (from field test readings) Therefore from NCHRP Manual for Bridge Rating through Load Testing Eq. (6-4) Ka = (εc/εT) – 1 = (252/98.95) – 1 = 1.55 27 From NCHRP Manual for Bridge Rating through Load Testing Eq. (6-6) Kb = Kb1 × Kb2 × Kb3 = 0.8 × 0.8 × 1 = 0.64 (HS20, KY1, KY2, KY3) = 1.0 × 0.8 × 1 = 0.8 (KY4) From NCHRP Manual for Bridge Rating through Load Testing Eq. (6-3) K = 1 + Ka × Kb = 1 + (1.55 × 0.64) = 1.99 (HS20, KY1, KY2, KY3) = 1 + (1.55 × 0.8) = 2.24 (KY4) Truck Type HS20 (from calculations in Appendix B) Inventory level RF = 0.916 Operating level RF = 1.529 From NCHRP Manual for Bridge Rating through Load Testing Eq. (6-1) RFT = RF × K Inventory level RFT = 0.916 × 1.99 = 1.83 Operating level RFT = 1.529 × 1.99 = 3.05 Summary and Conclusions The current method for load rating bridges — based on AASHTO specifications — can underestimate the capacity and behavior of bridges. Analytical equations do not account for the degree of rigidity in supports, unintended composite action due to friction between girders and the slab, and other factors. Load testing of individual bridges can produce a load rating that much more accurately reflects the capacity of a non-composite bridge. However, current methods of load testing require significant time commitments to instrument a bridge profile to record data, rendering it less feasible. New types of commercially available strain gauges, however, greatly reduce the time required to instrument a location. This report discussed the load rating of two bridges using field load test data. Researchers evaluated two types of strain gauge in field load tests to determine how effectively they minimize deployment time while maintaining accuracy. Magnetic Sensormate QE-1010 strain gauges and reusable BDI ST350 strain gauges were outfitted with wireless data transmission capabilities for rapid field deployment to determine if using their use would significantly reduce the amount of time required to load test a bridge. For bridges with characteristics such as unintended composite action or end fixity, this would increase the feasibility of load testing bridges, leading to a more favorable load rating. Compared to the theoretical load ratings, ratings based on load tests are expected to be more accurate. Both gauges were tested in the laboratory under flexural loads. Their readings were compared to those obtained from traditional foil-type strain gauges prior to their deployment on two bridges in Kentucky. Laboratory tests demonstrated the magnetic strain gauges and BDI reusable strain gauges are very accurate at low strains. At higher strains (i.e., more than 400 microstrain) the magnetic strain gauges 28 slipped. The wireless data transmission capability of both systems made it possible to carry out data acquisition without being close to the gauges. This significantly reduced the amount of wiring typically associated with strain gauge data acquisition. While magnetic strain gauges performed well in the field and gauge installation time was reduced, due to the rugged requirements of field testing, they will not be considered for future deployments given the current status of the technology. Reusable BDI strain gauges coupled with wireless transmitters balance rugged performance with short installation times. Each bridge was load posted because the load rating factor for several truck types was less than one. Table 4 lists the AASHTO load rating and field load testing results for the KY 1068 and KY 220 bridges. Field load tests revealed the load rating factor for strength was adequate for the KY 220 Bridge in Hardin County, while the load rating for the KY 1068 Bridge in Lewis County could be increased by 68%. Table 4 Load Rating Results References [1] American Association of State Highway and Transportation Officials. The Manual for Bridge Evaluation, Second Edition, AASHTO, Washington, D.C., 2011. [2] American Association of State Highway and Transportation Officials. AASHTO Standard Specifications for Highway Bridges, 17th Edition, AASHTO, Washington, D.C., 2002. [3] Nowak, A.S. and Saraf, V.K., “Load Testing of Bridges”, Research Report UMCEE 96-10, University of Michigan, Ann Arbor, MI,1996. [4] Schiff, S.D., Piccirilli, J.J., Iser, C.M., and Anderson K.J., “Load Testing for Assessment and Rating of Highway Bridges”, Research Project No. 655, Clemson University, Clemson, SC, 2006. [5] Jeffrey, A., Breña, S.F., and Civjay, S. “Evaluation of Bridge Performance and Rating through Nondestructive Load Testing”, University of Massachusetts Amherst, Amherst, MA, 2009. [6] Hosteng, T. and Phares, B. “Demonstration of Load Rating Capabilities through Physical Load Testing: Ida County Bridge Case Study”, Part of InTrans Project 12-444, Bridge Engineering Center, Iowa State University, Ames, IA, 2013. [7] Hag-Elsafi, O. and Kunin, J. “Load Testing For Bridge Rating: Dean’s Mill Over Hannacrois Creek”, Report FHWA/NY/SR-06/147, Transportation Research and Development Bureau, Albany, NY, 2006. [8] National Cooperative Highway Research Program. Manual for Bridge Rating Through Load Testing, NCHRP Project l2-28 (13) A, Transportation Research Board, Washington D.C., 1998. [9] AISC Steel Construction Manual (14th Edition). 29