IMPACT OF INCREASE IN AXLE LOADS ON TRACK & BRIDGES ON INDIAN RAILWAYS SHIV KUMAR* 1.1 BACKGROUND The Indian economy entered the tenth plan with an expectation of 6% to 7% annual growth in the GDP and consequently 7.2% to 8.0% growth in the transport sector. These expectations placed heavy demands on the already saturated road and rail transport system which coupled with the inadequacies in the power sector posed a major constraint in the realization of the projected economic growth. With airways, coastal shipping and inland waterways being in the fringes, freight transport in India is basically shared between road and the rail sectors. The road network in India has grown from 4 lakh km in 1951 to over 30 lakh km now, second largest in the world. Post independence, Indian Railways (IR) made a flying start almost doubling the transport output in the first 5-Year Plan. There was, however, a perceptible slowing down from 1968 to 1980 followed by a revival in the last two decades climaxing to introduction of heavier axle load trains (22.82 tonnes) on certain routes on IR with effect from May, 05. Volume - I 1.2 WAKE UP CALL • Freight traffic on IR has grown 90 times from 5.5 to 500 billion tonne km from 1951 till now. • Passenger traffic on IR has grown 80 times from 23 to 1800 billion passenger km in the same period. • National and state highways comprising only 8% of the road network carry 80% of the traffic • IR’s share of freight traffic has declined from 89% in 1951 to 38% now. • Golden Quadrilateral road network and induction of multi axle road vehicles will further make a serious dent on the share of freight traffic carried by railways. • Even heavy bulk freight may not remain the exclusive preserve of the IR. Volume - I * Director /IRICEN/Pune 1.3 COMPARATIVE EVALUATION OF INDIAN RAILWAYS WITH OTHER RAILWAYS basically aimed at network expansion, doubling, creation of dedicated passenger and freight corridors. The world’s heaviest and longest freight trains run in Australia. With a payload of 82,000 tonnes and gross load of 99,734 tonnes, the train is formed of 682 wagons, hauled by eight 6000 HP diesel locomotives. 1.4 The 7.2 km long train transfers minerals in bulk from one part of Australia to other crossing thousands of miles of largely uninhabited and desert areas. Theoretically, just 20 such trains are enough to carry the entire volume of about 1.8 million tonnes of freight moved every day by IR, which deploy 5,000 trains of varying capacities to do the job. Accordingly, Integrated Railway Modernisation Plan (2005-10) has been formulated on IR with an objective to enhance capacity, improve railport connectivity, introduce higher axle load wagons to carry bulk material and development of Dedicated Freight Corridors (DFC) and running of freight train @ 100 kmph on the high density Golden Quadrilateral and its diagonals connecting the four metropolitan cities. A comparison of the railway systems in China and India makes interesting study. In the decade 1992 to 2002 the route Km on the Chinese Railways (CR) has grown from minus 6% to plus 14% in comparison to that of IR. The two railways carried almost the same volume of passenger traffic both in 1992 as well as 2002. However, in respect of freight traffic, the volume carried by CR was four and a half times that of IR. They have achieved these results through more efficient exploitation of track, locomotives and wagons, and by assigning lower priority to passenger services. CR has a larger proportion of double line and has adopted automatic signalling more aggressively than India. As a result, CR operates roughly twice the number of trains on electrified double line tracks than the Indian Railways. In 1993 the Ministry of Railways, China decided to increase maximum freight axle load from 20.5 to 25 tonnes over the next 10 to 15 years at the same time when mechanized track maintenance was being introduced. Theirs is a large railway system with high capacity utilization, which has track materials and components that are not as strong and durable as they should be. The key to successful mechanization on lines which are being used to full capacity, where maintenance today is performed manually between trains, is the ability of mechanized gangs to obtain a sufficiently high quality of work so that less time is required for track occupancy and maintenance over a period of years. Stronger, more durable materials and track components would be essential. Thus, CR planned to lay all major heavy haul trunk lines with 60 kg or heavier rails. They would lay 29% of main tracks with continuous welded rail. Higher strength concrete sleepers and matching higher-capacity elastic fasteners would be introduced. New stone quarries have been opened, and existing quarries either have been reconstructed to provide quality ballast or closed. A track maintenance planning system with new technical standards for track condition management is being implemented. CR are planning an investment of US$ 200 million in the mega plan period from 2004 to 2020 Volume - I 2 MODERNISATION AND EXPANSION OF INDIAN RAILWAYS IR have embarked upon a path of modernisation and expansion in a big way as per vision of the Honourable Prime Minister of India. At present, predominantly running axle load on IR is 20.32 tonnes. Heavier axle loads will enable carrying more payloads in one train, which in turn will improve throughput substantially. However, before a heavier axle load is permitted to run, the safety of infrastructure has to be ensured as it carries passenger traffic also. Running of such heavy axle load trains on the existing routes would cause very high stresses on the track and bridge structure, on the requirements of its components, their maintenance and service life. As a precursor to the introduction of heavier axle loads, IR took a bold decision in May,05 to introduce a pilot project of running freight trains with BOXN wagons loaded up to CC+8+2 tonnes on iron ore routes ( where CC is designed carrying capacity of wagon). This resulted in introduction of heavier axle load of 22.82 tonnes and Track Loading Density (TLD) of 8.51 tonnes/m. Introduction of freight trains with 25 tonnes axle loads has also been planned on two routes on IR by end of the year 2006-07. 1.4.1 ENHANCING TRANSPORT CAPACITY ON INDIAN RAIWAYS We are in transport business. Trailing loads and operating speeds are our principal efficiency indicators. Within the limitations of a loop length of 686 meters and the existing track loading density of 8.25 tonnes per meter, the options for enhancing transport capacity of IR are as under:• Introduction of higher axle loads • Increasing number of axles per wagon Even though about 12% higher throughput with 22.82 tonnes axle load and about 23% with 25 tonnes axle load could be achieved, yet Volume - I 3 track loading density increases to 8.51 tonnes/m with axle loads of 22.82 tonnes and 9.33 tonnes/m with 25 tonnes axle load. Since rail wear is already a matter of concern, it may be aggravated by higher axle loads. Higher operating speed may not be possible with higher axle loads. Increase in number of axles in a wagon means redesigning and revamping the fleet of wagons which is cost intensive and immediate solution is not possible. Therefore, to carry more traffic cost effectively, IR went for the option to introduce higher axle load trains being run with certain type of wagons (BOXN) and carrying bulk consumables like iron ore and coal. Current loading standards for bridges (MBG 1987) permit axle loads up to 25 tonnes in case of locos and track loading density of 8.25 tonnes/m. The trailing loads of 8.25 tonnes/m translate into trailing axle loads of 20.32 tonnes at the wheel spacing as in BOX-N wagons. With introduction of higher axle loads of 22.82 tonnes, IR has been poised into a select group of Heavy Haul Railways. This would be the stepping stone for moving ahead towards the regime involving 25 tonnes axle loads and track load density of 9.33 tonnes/m on the existing network without any hiccups. The axle loads running on the Heavy Haul routes of American, Australian, China and other advanced Railways are ranging from 30 to 40 tonnes. However there is major difference in scenario prevailing on Indian Railway as unlike the other Railways where heavy haul freight trains run on a dedicated heavy haul lines, in Indian Railways same infrastructure has to carry both goods and passengers traffic. Golden Quadrilateral and its two diagonals constituting 16% of route km (25% of running track km) carry 55% of passenger and 65% of freight traffic of the IR and are saturated on most lengths. 1.4.2 DEDICATED FREIGHT CORRIDORS ON INDIAN RAILWAYS Dedicated Freight Corridors (DFCs) have been planned on IR with the capability of carrying 32.5 tonnes axle load wagons in train formations of over 15,000 tonnes hauled by multiple units of modern freight locomotives at speeds of 100 kmph. The loop lengths would be 1500 meters or longer to permit accommodation and crossing of train lengths of 120 wagons. RITES has done a feasibility study for providing dedicated freight corridor Volume - I 4 on two routes. The 819 km double track Eastern Corridor will run from Khurja to Sonnagar via Mughalsarai, Fatehpur, Etawah. The 1493 km double track Western corridor will run from Jawahar Lal Nehru Port in Mumbai to Dadri in U.P. via Vadodara, Ahmedabad, Palanpur, Phulera, Rewari & Tughlakabad. Eastern corridor will have an extension which will run from Khurja to Ludhiana via Meerut, Saharanpur & Ambala. Western corridor will have an extension to Dandarikalan Container Depot near Ludhiana via. Rewari, Hissar & Jakhal. Detailed planning is already underway on the above mentioned two routes, which were approved in principle by the government in March, 06. They are intended to form the first phase of a network totalling around 10000 km, which is to be developed over the next decade at a cost of more than Rs.700 billion. Built to a larger loading gauge and capable of accepting wagons with axle loads of 32.5 tonnes (Table I), the corridors will permit the gradual segregation of freight and passenger services, allowing existing mixed-traffic routes to be optimised as highquality passenger corridors. Table 1. Proposed operating parameters on DFC Train length 120 to 130 wagons Trailing load 14000 to 16000 tonnes Maximum permissible speed 100km/h Length of loops 1500 m Average distance between stations 45 km 1.4.3 THE BENEFITS OF HIGHER AXLE LOADS ON INDIAN RAILWAYS Fewer wagons will be needed to haul the same load, leading to lower capital cost and possible reduction in wagon maintenance cost, fewer locomotives, lower fuel consumption per net tonne, reduction in train wagon kilometre operated, and fewer crew deployment entailing savings in wages. The railways in North America, Australia, South America, South Africa and Sweden have all increased axle loads to obtain significant savings in operating cost. These savings have been achieved despite increased cost of maintaining track and bridges, greater component damages and shorter component lives. The increase of the axle load from 22.5 tonnes to 30 tonnes yielded 40 per cent savings in transportation cost in the US. This in turn helped the railways in that country achieve significant reduction in Volume - I 5 operational cost of transporting containers and introducing customised wagons to win back traffic from the roadways. Boosting wagon productivity that is, how to carry more load per wagon, or how to achieve higher payload per wagon, has become important for the IR in view of the increasing threat from the various other modes of transport, particularly roadways. Overall, IR now accounts for 38 per cent of the country’s total freight movement compared to more than 89 per cent half a century ago. An analysis of the commoditywise market share shows that between 1991-92 and 2000-01, there was a sharp drop in the rail coefficient for cement, POL, food grains and iron steel while it improved for coal, iron ore and fertilisers. Also, while the drop has been significant, not so the extent of improvement. But, then, the lower axle load presents only one issue. There are several other issues that need to be tackled along with the raising of the axle load. Thus, along with higher axle load, the track load density, that is, the maximum load permissible per metre length of track (TLD) too needs to be increased. Any increase in axle load without corresponding increase in TLD will have a marginal effect on the throughput. Maximum Moving Dimension having a width of 3.66 m and height of 6.81 m has been proposed for DFC.This would permit double-stacking of 9'-6" high containers. However, it will be important to ensure a reasonable degree of interoperability between the DFCs and the existing network, in particular on feeder routes. Accordingly,MMD having a width of 3.50m and height of 6.81m on routes where double stack containers will be running and 4.385m on other routes has been proposed to ensure interoperability of wagons on feeder routes . In the long term, IR would be expected to upgrade as much as possible of its existing network to the new standards. During the initial stage, however, the priority will be to enhance the carrying capacity through marginal relaxation of the loading gauge coupled with an increase of axle loads to 25 tonnes on existing network. Proposed wagon dimensions are as under:(i) Width of wagon Height above rail level (in mm) Width(in mm) 1000 3500 945 3250 <945 3135 The other issues that deserve careful consideration in this connection are track friendly bogies, smaller wheel size, and enhancement of Maximum Moving Dimensions (MMD). One of the biggest constraints on rail productivity is the restrictive Maximum Moving Dimensions in IR. This has essentially remained unchanged since 1913, and today railway systems operating on both 1000 mm and 1435 mm gauges offer higher cubic capacity and payloads than IR’s 1676 mm gauge vehicles. As a result, according to a study conducted by consultant David Burns, the unit cost of bulk freight movement on the IR is many times that of the most efficient US railroads when considered on the basis of purchasing power parity. The payload to tare ratio of IR wagon fleet varies from 2 to 2.6, with only the BOXNLW(2.96) and BOY (3.42) designs approaching the range of 3.5 to 5 achieved on other railways. Apart from the restrictive vehicle profile, a large wheel diameter of 1000 mm, a coupling height of 1105 mm and a relatively low axle load of 20.32 tonnes have not helped. The construction of DFCs presents an opportunity to effect a quantum leap in productivity by modifying the design parameters of freight stock and liberalising the loading gauge. Volume - I 6 Note: (a) It has been decided to reduce the height of goods platform to 840 mm for a width of 1850 mm from track centre on routes where these wagons will be running. (b) The width 3500 mm as proposed above is for door less wagons.The width for wagons with doors will be 3250 mm. (ii) Height of wagon For open wagons 4025mm For covered wagons 4025 mm at side 4385mm at centre (iii) Wheel Diameter Maximum 1050 mm and minimum 840 mm when new. Table II shows the proposed design parameters for wagons able to run on both new and existing lines. When handling bulk commodities such as coal, iron ore or cement, it may not be possible to make full use of the 32.5 tonne axle load within the existing height limits of Volume - I 7 4265 mm in the centre and 3735 mm on the sides. Therefore, proposed height for open wagons would be 4025 mm and for covered wagons 4025mm at side and 4385mm at centre. This will call for selectively cutting back the platform canopies on the existing network which would be a major task but offers immense benefits in terms of enhanced productivity. In conjunction with a wagon width of 3500 mm and smaller wheels this would conveniently permit the operation of freight wagons with gross weight in the range of 110 to 130 tonnes. Table II. Proposed design parameters of wagons Proposed Wheel diameter* mm Maximum 1050 Minimum 840 (Authority: Rly Bd’s Track Policy Circular No. 2 of 2006 dt 19th July 2006) DFC • 68 kg (90 UTS or greater) Rail • New design PSC sleepers, 1660 per km • 350 mm cushion of hard stone ballast • New PSC sleeper for 30 tonnes axle load to suit 68 kg rail are being developed by RDSO • Flash butt welding joints to be used • E-clip type elastic fastening with improved rubber pad, being developed by RDSO shall be used 1000 Bogie wheel base** mm 1830 2000 Floor height **mm 1091 1260 936 1105 • Thick web switches and swing nose crossings in turnouts Axle load tonnes 32.5 20.32 • Track loading density tonnes/m 12.0 7.67 PSC sleepers have also to be designed for points & crossings, level crossings, SEJs, bridge approaches 4.2 2.5 • Formation width in embankment/cutting and centre to centre spacing will be as follows: Coupler height** mm Payload to tare ratio 2.1 Existing PSC sleepers with 1540/km, 25 tonne axle load my be permitted at a restricted speed of 60 kmh. * When new ** A final decision is yet to be taken. Single line 6.85 m (as existing) Double line 12.35 m TRACK STRUCTURE ON HIGHER AXLE LOAD ROUTES Centre to centre spacing of DFC tracks 5.50 m Proposed track structure is as under:- Centre to centre spacing of DFC track from existing track shall be 6.00 m. Feeder routes • Suitable thickness of sub-ballast/blanket will have to be provided. Judicious use of reinforced earth construction will have to be made at those locations where land width is not adequate to accommodate slope in formation. Flash butt welding joints to be used • Curvature will be limited to 700 m (2.50 degree) • E-clip type elastic fastening with improved rubber pad, being developed by RDSO shall be used after rail seat design of PSC sleepers is modified (in future relayings) • Ruling gradient will be 1 in 200 compensated. • Thick web switches with weldable CMS crossings in turnouts. To start with, curved switches on PSC layouts can be permitted. 2.2 TRACK MAINTENANCE ON HIGHER AXLE LOAD ROUTES • • For sections where 52 kg, 90 UTS rails have been recently laid on A new approach based on mechanised track maintenance will have to be formulated and implemented consisting of • 60 kg (90 UTS rails) • Existing design PSC sleepers, 1660 per km • 300 mm ballast cushion • Volume - I 8 Volume - I 9 TRC based track geometry measurement to decide need for tamping Need based mechanised track maintenance Spurt car based USFD testing Rail grinding On foot inspection by PWM built as per various loading standards are given in Table III. Table III Results of analysis of standard span bridges STANDARD SPAN(m) • All maintenance will need a daily track possession of 4 hours. • A service road will be required, all along for facilitating maintenance. 3.1 BRIDGES ON HIGHER AXLE LOAD ROUTES BGML LOADING Axle load 22.9 tonnes TLD 7.67 tonnes/m T.Effort 47.6 tonnes Proposed loading standards for bridges on DFC is as under: • MBG LOADING Axle load 25.0 tonnes TLD 8.25 tonnes/m T.Effort 100.0 tonnes 1-20 60 kmph 60 kmph 60 kmph 20-63 60 kmph 55 kmph 60 kmph 78.7 50 kmph 55 kmph 55 kmph 3.2.2 SUBSTRUCTURE Axle load 30.0 tonnes Tractive effort 60.0 tonnes per loco for double headed locomotives 3.2 RBG LOADING Axle load22.5 tonnes TLD 7.67 tonnes/m T.Effort 75.0 tonnes Locomotives: 45.0 tonnes per loco for triple headed locomotives • SPEED POTENTIAL OF BRIDGES FOR DIFFERENT LOADING STANDARDS OF EXISTING BRIDGES Braking force 25.0 tonnes Train load: Axle load 32.5 tonnes Track loading density(TLD) 12.0 tonnes/m EXAMINATION OF EXISTING BRIDGES ON FEEDER ROUTES 3.2.1 SUPERSTRUCTURE Existing structures have been examined by RDSO on the following premises: Substructure designs are based on local geographical features. Therefore, substructure analysis has to be done of all bridges by the zonal railways. 3.2.3 BEARINGS • Bearings of standard spans spans designed for MBG loading have been found fit for 25 tonnes axle load. • Bearings of standard spans designed for BGML/RBG loading having spans 31.9 m, 47.25 m, 63.0 m and 78.8 m (all effective spans) needs strengthening. • One time inspection of bearings apart from schedule inspection may have to be done by zonal railways before permitting 25 tonnes axle load. 3.2.4 BRIDGE INSPECTION & MONITORING Following locations/members will have to be monitored as being critical from fatigue consideration: Connection of cross girder with stringer. • Actual axle load and axle spacing for moving loads of WAG9 locomotive and 25 tonne BOXN wagons. Outstanding leg of the top compression flange of the stringer at the junction with web rivet. • Coefficient of Dynamic Augment (CFA) for limited speed of 60 kmh. Rivets connecting bottom flange angle of the stringer with web at mid span. The results of analysis for superstructures of standard spans of bridges Vertical members at connection with top chords. Volume - I 10 Volume - I 11 Rivers connecting bottom flange of cross girders with web at mid span. Hot Wheel Detector (HWD) Truck Hunting Detector (THD) Rivets of splice joints of bottom flange in plate girders. Wheel Profile Monitoring System etc. • Physical condition of bridges needs to be certified by zonal railways. • More intensive bearing & substructures inspection will have to be done by zonal railways. In this connection, directives for checking superstructure of non standard spans including arch bridges, bearing of non standard spans and substructure of all bridges have been issued to zonal railways by RDSO. 4. CONCLUSIONS • Experience of increased axle load on other major railway systems in world has shown that fatigue life of rail comes down drastically with increase in axle loads and it may take 5 to 10 years for heavy axle loads to tell its effect on rails. • Rail grinding is required to be undertaken to prevent Rolling Contact Fatigue (RCF) and increase rail life. • Till rail grinding is implemented, there is a need to safeguard against rail fractures due to RCF by extensive and effective USFD. • The rail life obtainable under present axle load itself is in range of 350400 GMT which is way below the prescribed life of 800 GMT for 90 UTS rails. • Fatigue life of rails also needs a review. Life of rails subjected to heavier axle load needs to be realistically brought down till grinding is introduced. • Installation of way-side vehicle condition monitoring systems for identifying poorly performing wagons will go a long way in reducing stress state of track by identifying defective rolling stock and removing them from service. Some of these are: Wheel Impact Load Detector 9WILD) Truck Performance Detector (TPD) Hot Bearing Detector (HBD) Dragging Equipment Detector (DED) Volume - I 12 • Development of track friendly, self steering bogie will be required in the long run. • A review on the conclusion of the pilot project to run 22.82 tonnes axle load trains needs to be done in June 07 to decide its universalisation all over IR. Volume - I 13 RATIONALE BEHIND INCREASE IN AXLE LOAD ON INDIAN RAILWAYS & ROAD AHEAD Sweden is 30%, which is double when compared with 15% of Europe. After the deregulation of road haulage, net permissible pay load of Swedish trucks has increased from 31 to 40 tonnes, which has cut the average trucking cost per tonne by approximately 22%. V.K. JAIN* This has posed further challenge to Railways. To accept this challenge from road sector, Swedish Railway has planned:- 1.0 INTRODUCTION Capturing this extra traffic will not be an easy task for IR. There are two basic problems. First, there is considerable improvement and augmentation in road sector. NHAI is adding state of the art road network at a fast pace. High capacity trucks are available, which provide door to door services. In the present scenario, even bulk movements are opting for road. Railway’s market share have fallen during last 10 years. IR is to gear up to find out technical solution of these constraints and to accept the challenge from other mode of transportation, especially road. 2.0 WORLD SCENARIO 2.1 SWEDISH RAILWAYS Sweden is located between latitudes 550 to 690 and faces very harsh weather conditions. Labour cost is very high. Main freight traffic is iron ore, steel, timber and paper. Most of the companies are global and will prefer to shift the business in case cost is high. Therefore, to keep the transportation cost low, Swedish Railways introduced 25 T axle load on ore lines about 40 years back. Railway market share in Volume - I * Executive Director Civil Engg. (P), Railway Board Introduction of 30 T on ore lines. (ii) Universalisation of 25 T. Two studies have been conducted to work out the economics of running 30T axle load; one by ZETA-TECH Associate, Inc. of USA and second by internal team of Banverket/Jerbaneverket. These studies have shown:- After economic reforms and liberalization, Indian economy has shown a robust growth; the rate of growth 7.5% during 2004-05, 8% during 2005-06 and projected growth for 2006-07 is 7.5% – 8%. This rate of growth will result in about 10 – 12% growth in transport sector. IR must gear up to carry this extra traffic. Secondly, golden quadrilateral along with its two diagonals and grand chord, which is only 16% of the network, carries 65% of the freight traffic. This network is saturated in many sections, part of the sections is still single line or non-electrified. Other important constraints are inadequate development of terminals and warehousing capacity, less axle load, poor tare to payload ratio and priority to passenger traffic. (i) (i) Increase in track maintenance cost (including early renewal cost) will be 13 – 15%. (ii) Saving in overall operation cost will be 27-30%. These studies have recommended:(i) Rail grinding and lubrication. (ii) Use of movable crossing. (iii) Use of head hardened rails. (iv) Strengthening of bridges. Railway is adopting line by line approach to upgrade the network. 2.2 US RAIL ROADS Total track kms is about 2 lac 75 thousand kms. Track gauge is mainly 1435 mm. Most US Railways are private owned and run only freight trains. Operation of long distance passenger trains is monopoly of AMTRAK. Rail road employment has fallen drastically from 4 lac 50 thousand employees in 1982 to one lac 55 thousand in 2003. Safety regulation is the responsibility of the Federal Rail Road Administration (FRA). Traffic on US Rail Road can be divided into four general categories. (i) Unit train traffic (coal, ore, grain) (ii) Inter-modal traffic (ocean containers and truck trailers) (iii) Mixed freight traffic (car load, plastic, automobile). (iv) Passenger. Volume - I 15 Coal, most of which move in unit trains, account for 42.9% of US Rail Road traffic originates and about 20% of revenues. Inter-modal traffic have gone up rapidly in recent years. Operation of double stack container was introduced in 1977 and most container trains are double stack. Main axle load is 30 T, although a number of rail roads have begun purchasing stock with 32.5 T and 35.5 T axle load. Hunter Valley is a dedicated freight line. Entire track is track circuited and with bi-directional signaling. Total length of the track in Hunter Valley is 600 kms. Rails are 60 kg, out of which, 395 kms are head hardened, 90% sleepers are concrete sleepers. Track is inspected by Keyman twice a week and by TRC once in 3 months. Average length of train is 7000 feet which is going to be extended to 8000 feet. Trains load is 10,000 to 11,000 tonne and locomotives are typically of 3000 to 5000 horse power. Now 6000 horse power locomotives are being offered by both General Motors and General Electric. Rail grinding is a normal feature and done through Service Contracts. For sharp curves of radius less than 400 m, grinding is done after passage of 10 GMT and for straight track, grinding is done after passage of 20 GMT. Average life of 60 kg rails is 1200 MT. To enhance safety, ARTC has provided – Structurally 60 kg rails are sufficient to support 30 T axle load, but there is considerable maintenance problem, USA is mainly using 67 to 71 kg rails on timber and concrete sleepers. Ballast cushion is 300 mm with hard ballast. Rail hardness is 380 BHN and more. Rail grinding is must for higher axle load operation. One machine with 120 stone is able to grind about 10,000 km in a year. Operational speed is 20 kmph. Grinding work has been mainly outsourced. On-board rail lubrication is used on curves sharper than 40. New turn outs are mainly Thick Web Switches with swing nose crossing. On freight routes, track is inspected by a staff twice a week and by TRC @ 3 to 6 months depending on traffic density. No regular inspection by Supervisors or Officers. 2.3 AUSTRALIAN RAILWAYS Australian Railways has been reorganised into number of small units, three important Railways are Australian Rail Track Corporation (ARTC), BHP Billiton and Queensland Railways. ARTC is operating 30 T axle load in Hunter Valley, BHP Billiton is operating 36 T to 40 T axle load on ore lines and Queensland Railway is operating 26 T axle load on coal lines. Salient features about ARTC are – Volume - I ARTC was formed in 1997. It is a Government owned company. Total length of track under ARTC is about 10,000 kms and total number of employee is about 25,000. It operates 30 T axle load in Hunter Valley, which carries mainly coal traffic. 30 T axle load was introduced in 1994 -1996. Maximum train length is 1550 m and maximum train load is 7,500 T. 16 (i) In-motion weigh bridges at loading points, (ii) Hot box detector. (iii) Wheel USFD tester. (iv) Wheel Impact Load Detector (WILD). 2.4 Only one derailment in last 18 months. SOUTH AFRICAN RAILWAYS South Africa has a well developed transportation system. The South African Railways and Harbour Administration, established in 1910, managed the operations of most of the nation’s transport network; in 1985, it became the South African Transport Services (SATS). In 1990 SATS reorganized as the Public Commercial Company, Transnet. Transnet has six business divisions – Spoornet to operate the rail roads; Portnet to manage the country’s extensive port system; Autonet, a comprehensive road transport service; South African Airways (SAA); Petronet to manage petroleum pipelines; and a parcel delivery service, known as “PX”. The Government is the sole share holder in Transnet. Spoornet, the National Rail Authority, manages a network of 21,303 km of 1065 mm gauge track and 314 km of 610 mm gauge track. Spoornet has restructured its operation to compete with road sector. COAL link and OREX, manage Spoornet’s coal and iron ore traffic over the Richards Bay and Sishen – Saldanha lines respectively. The general freight services divisions are grouped into 15 industry based segments. Since 1991, the Spoornet work force has dropped from nearly 1,20,000 to 43,736 in March 1999. Volume - I 17 Improvement of other infrastructures such as rolling stock is also a continuous process. Based on day to day experience, the rolling stock is also improved continuously to ensure better availability, reliability and safety. BOXN wagons have been provided with elastomeric pads and composition brake blocks which has improved its suspension system and reliability. Other salient features : Rails 60kg (1950 km), 57kg (4,370 km) and rest 48kg, 40 kg & 30 kg. Sleepers Concrete, Steel & wood. Sleeper density 1538/km & 1440/km. Fastenings (Pandrol) Fist, E 3131 chairs. Maximum axle load Locomotive - 29 tonnes. Wagons 26 tonnes. Traffic:- - Considering these improvements, Railway has been increasing the loading capacity of wagons from time to time. In past the 2 T loading tolerance was provided for BOXN for loose commodities. The Permissible Carrying Capacity (PCC) or Chargeable Carrying Capacity of BOXN wagon was enhanced by 2T in July, 1997. About 210 million tonnes. BOY, BOBS and BOBSN wagons have been introduced on Indian Railways with an axle load of about 22.9 T. These wagons are running for a considerable time and no adverse effect has been noticed on the track and bridges. Fast freight trains operate on eighteen routes nation wide at a maximum speed of 120 kmph. However, common speed is 60 kmph. In 1989, Spoornet created a World record by running a 71,600 – ton train at a speed upto 80 kmph on 861 – kilometer Sishen – Saldanha ore lines. 3.0 3.1 HIGHER AXLE LOAD ON INDIAN RAILWAYS BOXN was introduced in 1982 with an axle load of 20.32 T and with a speed potential of 75 kmph in loaded condition and 80 kmph in empty condition. Minimum track structure prescribed was 90 R rails with M+4 sleeper density. Rails were mainly 3-rail panels with large number of bolted joints. Sleepers were mainly metal sleepers, which used to be maintained manually. Over the years, not only the track structure has improved but maintenance standard has also improved considerably. 90 R rails have been replaced with 52 kg and 60 kg rails. Section modulus, an indicative parameter of structural strength, of 90 R rails is 235.65 cu cm while that of 52 kg rail is 285.50 cu cm and that of 60 kg rails is 377.4 cu cm, i.e. section - modulus of 60 kg and 52 kg rails are higher than 90 R by 60.15% and 21.15% respectively. In addition to improvement in rail structure, sleeper technology has also changed completely. Metal sleepers have been replaced with PSC (Pre-stressed concrete) sleepers. These sleepers are provided 25 cm - 30 cm ballast cushion and are maintained with the state-of-art track machines. With the introduction of 52 kg/60 kg rails on PSC sleepers, track modulus, a parameter, which indicates ability of track as a whole to carry the axle load, has improved considerably. Volume - I 18 Taking the advantage of improved track structure, rolling stock (BOXN) and experience of running of BOY, BOBS & BOBSN, Indian Railways took a historic decision in May 2005 to increase the axle load of BOXN to 22.82 T (CC+8+2 T) on selected iron ore routes as a Pilot Project. In November 2005, Board has permitted CC+6+2 T for ‘E’, ‘F’ & inferior grade of coal in BOXN on identified routes as Pilot Projects. Further, CC+6+2 T has been permitted on identified routes for other type of wagons i.e. BOXNHS, BOBR, BOBRN, BCN, BCNA, BCNAHS & BOST for all type of commodities. Presently, CC+8+2 T is running on 42 routes covering eight Railways (ER, ECOR, SR, SER, SCR, SECR, SWR, WCR). CC+6+2 T is running on about 50 routes covering 14 Railways (except SWR & SR). 3.2 PRECAUTIONS While introducing this Pilot project, following additional precautions have been taken: (i) 90 R rails are to be replaced on priority. Till renewal, maximum permitted speed of higher axle load trains on 90R track has been restricted to 30 kmph. (ii) Maximum permitted speed of higher axle load freight trains has been restricted to 60 kmph in loaded direction on other than 90 R track. (iii) Rails and welds are being tested by Ultrasonic Flaw Detecting Equipment (USFD) for gauge corner fatigue. Volume - I 19 3.3 3.4 (iv) Thorough physical inspection has been done for bridges. (v) Selected bridges are also being monitored by instrumentation to assess the actual axle load and their effects in coordination with the Institute of Learning like SERC, Chennai, and CRRI, Delhi etc. 5.0 ROAD AHEAD 5.1 Detailed commodity-wise analysis of freight traffic carried in 2004-05 is as follows: Commodity % of NTKM tonnage INSTALLATION OF WEIGH BRIDGES To ensure control on adherence of Permissible Carrying capacity, an action plan has been prepared by Railways to install 101 additional electronic in-motion weigh bridges, out of which 58 nos have already been commissioned and rest are likely to be commissioned by December 2006. This will ensure that no overloading beyond what is permitted actually takes place. Coal 45.07 39.74 Foodgrains 7.73 15.36 Iron & Steel 3.05 3.85 Iron ore & other ores 15.99 10.18 Cement 8.93 7.09 MONITORING MECHANISM POL 5.31 5.16 Fertilisers 4.78 5.33 Others 9.14 13.39 The project is being monitored by a Multi-Disciplinary Core group comprising of Heads of Department from Civil, Mechanical & Operating Department and progress is being reviewed at the level of General Managers of Zonal Railways. Quarterly Review Reports are being submitted to Railway Board. 4.0 % of originating 90% of freight traffic is from 8 major commodities i.e. coal, foodgrain, iron and steel, cement, POL, fertilizers, iron ores and other ores. Important type of wagons are BOXN, BRN, BOBR, BOBS, BCN and Tank. 60% of traffic is carried in BOXN wagons. Its carrying capacity is 58-60 tonnes and its pay load to tare ratio is only 2.35, which is very less when compared to international standard. ECONOMIC TURN AROUND OF INDIAN RAILWAYS Indian Railway is passing through best phase in her history. Only in 2001, an expert Committee had put Railway on the verge of financial collapse. It had forecasted an additional financial burden of over Rs. 610 billion to Government of India. This has been proved wrong and Railway has generated enough resources internally and has fund balance of more than Rs.136 billion. The turn around of Indian Railways is real, touchable and backed by record breaking figures. Indian Institute of Management, Ahmedabad & Harword University are so impressed with this turn around that they want to take it as case study for their students. Today Railway has got sufficient fund to finance safety & throughput enhancement works, staff welfare activities and projects of national importance. All this turn around has been possible due to bold decision of Engineering Department under the leadership and guidance of Shri R.R. Jaruhar, Member Engineering to permit CC+8+2 T and CC+6+2 T in BOXN and other wagons. This has resulted in extra freight loading without corresponding increase in operational cost resulting in huge surplus for Railways. This has been well appreciated by one and all, including media. Volume - I 20 5.2 Indian Railways is planning to construct dedicated freight corridors. Phase – I will consist of two corridors i.e. Western and Eastern corridors i.e. Jawaharlal Nehru Port (JNPT) to Tughlakabad/Dadri near Delhi and Ludhiana-Sonepur-Kolkata. Formation and bridges on DFC will be constructed for an axle load of 30 T and initially track will be fit for 25 T axle load which will be upgrade to 30 T as and when wagons and feeder routes are available for 30 T axle load operation. About 24 routes covering 4200 route km have been identified as feeder routes to these DFCs which will also be upgraded for running of 25 T axle load. Subsequently, four more dedicated freight corridors will be planned as follows: North – South : (1800 km) East – West : (2000 km) East Coast : (1500 km) Southern : ( 900 km) Volume - I 21 5.3 In addition, about 25 routes covering about 7000 kms have been identified on iron ore circuits which will be upgraded for 25 T axle load routes. Bhilai – Dallirajhara in SECR and Gua-Barazamda-PadapaharBanspani-Daitari-Cuttack-Paradeep section in SER and ECOR are being upgraded to introduce 25 T axle load within this financial year. Efforts are being also made to universalize CC+6+2 T. 5.4 Field staff working on these enhanced axle load routes, especially on proposed 25 T axle routes have to be careful and alert. Both track and bridges should be kept under close monitoring. To assess the effect of higher axle load on track and bridges, one of the important requirements for 25 T axle load is grinding of rails. Indian Railways is already planning to procure two rail grinding machines. Till these machines are procured and commissioned, track should be monitored intensively with the help of USFD machines for any gauge corner fatigue crack. 6.0 CONCLUSION 6.1 With the economic reforms and liberalization, there is surge in transport sector. Railways has to compete with other modes of transportation to carry this extra traffic. 6.2 Swedish Railways has introduced 25 T axle load on ores line about 40 years back and is now planning to introduce 30 T axle load. 6.3 On US Rail Roads, operation of double stack container was introduced in 1977 and axle load is predominantly 30 T. 6.4 On Australian Railways, ARTC is operating 30 T axle load in Hunter Valley, BHP Billiton is operating 36 T to 40 T axle load on ore lines and Qeensland Railway is operating 26 T axle load on coal lines. 6.5 Spoornet is operating 26 T axle load on coal and iron ore routes. 6.6 Indian Railways has to adopt higher axle load on economic and technical considerations. This will improve wagon productivity and will reduce operational cost. 6.7 Pilot project of running of CC+8+2 T on identified iron ore routes and CC+6+2 T for coal and other commodities has resulted in complete financial turn around of Indian Railways. It has generated a surplus of more than Rs. 136 billion. Now sufficient fund is available with Indian Volume - I 22 Railways to finance safety & throughput enhancement work, staff welfare activities and projects of national importance. 6.8 Since India Railways has got mixed traffic lines, precautions specified for running of enhanced axle load must be adhered to. 6.9 Dedicated freight corridors will completely change the scenario of freight operation in coming years. Volume - I 23 • Locomotive costs (if train net load can be increased within the same gross train weight, there is more revenue for the same locomotive mileage). ENGINEERING JUDGMENT AND RISK MANAGEMENT BASED APPROACH FOR INTRODUCTION OF HEAVIER AXLE LOADS 1.3. In addition, longer track possessions will be possible with reduction in number of trains. ANIRUDH JAIN* 1.4. Heavier wagons impose heavier axle loads on the track, which means more frequent maintenance and shorter track component life. The same is true for bridges. As a result, costs of maintenance and rehabilitation of track and bridges increase. Yet these increases in axle loads are attractive as the savings in operating and ownership costs are significantly higher than the increases in track costs. This is because heavier wagons move long distances on high-quality main line track, and each km traveled means additional savings. 1.5. Many Railways in the world began to increase axle loads to provide more efficient and lower cost transportation of bulk commodities, about 50 years ago. Many systems, particularly North American Railroads, have introduced axle loads of 30 ton or heavier on their network, extensively. In some countries like South Africa, Australia and Brazil, such heavier loads have been implemented on certain identified routes either by upgrading an existing line or by constructing a new dedicated line. 1.6. Indian railways, in order to meet the growing demand of traffic as also to reap the benefits of lower transport cost with heavier axle loads, have decided to construct two Dedicated Freight Corridors, in Mumbai – Delhi and Kolkata – Delhi Sectors. Standards for construction and maintenance for these are to be decided. Also, with the construction of these corridors, many sections of the existing network will be required to feed the corridor. These sections will, thus, be required to carry equally heavy axle loads and up-gradation strategies for them are to be chalked out. 2. THE DEMAND 2.1. Armed with the information that in Australia, USA and to some extent in South Africa, Heavier axle loads have been introduced without much inputs, Indian railway’s policy makers are questioning the engineers “When USA, Australia and SA can introduce 30 ton axle loads on rails lighter than 90 Lbs, why not we?“ SYNOPSIS About 50 years ago, many railways in the world began to increase axle loads to provide more efficient and lower cost transportation of bulk commodities. Serious problems with rail, track, wheels, and rolling stock emerged. Numerous companies and administrations undertook lot of research initiatives to overcome these serious problems. Indian Railways have now decided to increase the axle loads. Similar problems are expected. It is, therefore, imperative that a solution to the anticipated problems is sought by looking into the experience of others. Nevertheless, it is a fact that problems as well as solutions will differ from system to system and lot of engineering ingenuity and Judgment is required to sail through. An attempt has been made in this paper to cover various aspects of Engineering Judgment and Risk Management to analyse suitability of a railway track for higher axle loads. A similar approach can be developed for bridges and rolling stock. 1. INTRODUCTION 1.1. A 30 ton axle load wagon carries almost 60% more commodity than does the conventional 20.32 ton wagon with only a marginal increase in the empty weight of the wagon. This results in an increase in the efficiency of rail freight movement due to an improvement in net/tare ratio. 1.2. If wagons can be loaded more heavily without significantly increasing their tare weight, railroads stand to realize savings in: • Capital costs (fewer wagons needed to move a fixed volume of traffic). • Fuel costs (reduced tare weight means an improved ratio of net load to gross weight). • Crew costs (increased carrying capacity may permit a reduction in the number of trains operated). Volume - I *Executive Director/ Track, RDSO, Lucknow Volume - I 25 2.2. 2.3. Though lot of experience is available on heavy haul with the railway systems which have implemented it, the same can not be straight away copied on to the Indian system. On some heavy haul railroads, heavy haul operations constitute only a small fraction of total traffic on the line. Some other systems have switched over from mixed traffic to a dedicated heavy haul operation. Solutions applicable to the case of a dedicated mine-to-port line with dedicated locomotives and rolling stock are different from those to a railroad with mixed traffic. There is no perfect solution that applies in all the circumstances. 2.4. There are a variety of approaches that should be examined before arriving at an optimum solution for a particular problem. As the approach changes, maintenance practices will have to be up dated. Risk factors will have to be identified and guarded against. 3. THE PRESENT INDIAN APPROACH ON RAIL MANAGEMENT 3.1. Presently, the approach is based on the logic that the section of rail is so chosen that the stresses in rails remain within allowable limits at all times and under all conditions, for the operating axle loads. Rail wears out under repeated action of wheel, looses its section and is required to be replaced due to loss of section and hence its ability to carry the loads. 3.2. As the wheel moves on rail, it applies not only the vertical load but also lateral forces due to various reasons. Due to dynamic effect of moving wheel the vertical wheel load is to be enhanced by a certain factor called dynamic augment. Value of the dynamic augment varies with the type of vehicle, particularly its suspension. Maximum lateral force, to be resisted by rail alone is limited at a certain value because it is presumed that at forces higher than this, entire track frame will move resulting into loss of alignment. 3.3. In addition, the rail is also subjected to stresses due to effects of temperature, residual stresses, flexing of rails at curves etc. 3.4. To calculate the stresses in rail under wheel load the theory in use assumes the track to be a continuously supported beam on elastic foundation. The basic theory has remained the same though better tools are now being employed to solve the equations with the use of Numerical methods. Computers have also enabled a better modeling of the track with more realistic values for stiffness and dampening Volume - I 26 factors being used. 3.5. Thermal stresses and to some extent the residual stresses are added to the Bending stresses calculated due to wheel load. A rail section is presumed to be safe if the stresses, so arrived, are with in the yield stress of the rail. 3.6. Stresses at the point of contact between rail and wheel had not been questioned, till recent past. 4. ACTUAL EXPERIENCE WITH TRACK IN FIELD 4.1. If the above approach is entirely correct, no rail shall fracture except in very adverse circumstances wherein the rail may have some inherent defect. Alternatively, the fracture may be caused if a defective wheel imposes an impact load which is beyond all predictable limits. In practice, it is observed that rails are developing flaws in head and in extreme cases rail failures are taking place. 4.2. Studies have indicated that the contact stresses at the point of contact of rail and wheel are high. As the load is transferred from wheel to rail in a very small contact patch, very high contact stresses are developed, exceeding yield point. Normally observed levels of stress at various levels of track structure are as follows: • Contact stress > 700 MPa > yield stress • Rail bending stress < 180 MPa • Sleeper Bearing stress < 1.4 to 3.0 MPa • Ballast bearing stress < 0.4 to 0.6 MPa • Sub grade bearing < 0.1 to 0.15 MPa 4.3. In extreme cases, contact stresses as high as 1.5 GPa are caused leading to hydrostatic compression resulting in to surface and sub surface cracks due to shear failures. Accumulation of such cracks ultimately leads to rail failure. This phenomenon has not been taken care of in the approach discussed in Para-3 above. 4.4. Similarly, Concrete sleepers are designed against bending failure but heavier axles are causing problems due to rail seat abrasion. 4.5. Thus there is a need to have a new look at the entire process, identify the problem areas, effects, remedies and to set standards for construction maintenance of infrastructure for heavier axle loads. Volume - I 27 Engineers are required to use best of their Judgment to face the challenge. 5. ENGINEERING JUDGMENT 5.1. Looking at the anomalies brought out in Para 4 & 5 above, it comes to mind that there is something more than just the theory. There is a marked difference between theory and practice which calls for a fair judgment on part of the engineer to decide the course of action. Father of Soil Mechanics, Karl Terzaghi had once said, “I produced my theories and made my experiments for the purpose of establishing an aid in forming a correct opinion and I realized with dismay that they are still considered by the majority as a substitute for common sense and experience.” In complex engineering problems, theory is not an end it is just an aid to help one in arriving at a decision. 5.2. 5.3. 5.4. Civil engineers, often face great uncertainties in their work. Uncertainties about the loads that structures will have to withstand, about the properties of the construction materials used, about what kind of computer modeling to do and how faithful the model is to the physical system. This problem is compounded by the fact that the engineer of even a simple structure is faced with a huge number of parameters that can be combined in many permutations, so choices have to be made intuitively. An experienced engineer applies his or her judgment to try and ensure that solutions will really work in practice, allowing for the assumptions, variations and uncertainties of practical systems. The term engineering judgment means how an “Engineering Issue” is analysed, inferred and concluded by an engineer or a committee of engineers. The engineering judgment of different engineers could be different, what really matters is to be able to defend ones judgments on any engineering issue, based on solid engineering reasoning. Engineering Judgment becomes more important in engineering applications where one or more phenomenon is not known with a fair degree of certainty. Problems in the fields of geotechnical engineering and engineering geology require us to work with very limited data about a complex environment where conditions can change radically over a short distance and an engineer has to use lot of judgment even with an element of calculated risk. Volume - I 28 5.5. Railway track engineering is an equally complex subject. Firstly, the forces applied by a moving wheel to the rail are not constant and depend on a number of factors. Then the assumption of track being a continuously supported beam on elastic foundation is not entirely correct. If we consider the contact stresses, contact patch is always moving and its area is varying. Any decision about suitability of a track structure for a given axle load, thus, calls for a lot of engineering judgment. 5.6. Engineering Judgment requires that we use our skills and professionalism to bring about good solutions that recognise the realities of their business context. There is an old adage that ‘an engineer is someone that can do for half a crown what any fool can do for a pound’. It is this statement which is important in the context of increased axle loads on railway track. It is easy to decide that the track is to be replaced but the effort shall be to manage without replacing and ensuring safety at the same time. 5.7. To understand the complexity, I will like to tell a small story: “A Mathematics teacher took his 8th class students out for picnic. A small river was to be crossed. The teacher took a rod, surveyed the river and calculated its average depth. He then calculated the average height of his class, including himself to find that the average height was at least 1½ feet more than the average depth. It was now easy for him to decide that the river can be safely walked through by the bunch of students. Result! Only 20% of the students could cross by walking rest sank. A family of Father, Mother and two kids was waiting by the side. Observing the fate of the school teacher and his class, the parents decided the course of action. Both the adults carried one child each on their shoulders and all 4 crossed the river safely. There was yet another couple waiting to cross the river. Encouraged by the success of the family, they decided to walk through. Result! Husband crossed safely but the lady went afloat and had to be rescued by the man. Remember Archimedes!” An analysis of the above story tells: i. Give reasonable thought to all the aspects, don’t jump to a conclusion based on what comes to mind in the first instance. ii. History and experience of others shall be given due weightage while making important decisions. Volume - I 29 iii. 5.8. 5.9. What happened once may not be true always as the circumstances may be different. Engineering judgment is a complex whole of theoretical knowledge, collection of data, past experience, prediction of behavior. Some people feel that they have lot of theoretical background in a field, some feel they have more experience than any body else in the field hence they are right. Bur the need is, they should learn to respect each other and their technical contributions. Engineering has traditionally used imagination, judgment, reasoning and experience to apply science, technology, mathematics and practical experience to convert concepts into reality. Engineers must be able to work in teams, with colleagues from a variety of disciplines, and must learn how to think across these disciplines in order to solve problems that are affecting them, in our case it is rail wheel interaction. 6. THE IMPLEMENTATION STRATEGY In our scenario the implementation strategy shall be as follows: 6.1. INVESTIGATE PRESENT PROBLEMS Indian Railways are already faced with many problems. Prominent amongst these are Rail Fractures, Weld Failures and in-adequate Toe load. With increase in axle loads, the problems are going to increase. We have to safeguard against the risks due to these problems while simultaneously working for their redressal. 6.2. OBSERVE WHAT OTHERS ARE DOING Railways around the world have increased the axle loads. They have implemented some very good practices about prediction and maintenance. IR must study all these and consider their implementation. Some of the noteworthy practices are: 5.10. The process can be better explained through a flow chart, shown below. This may some times be an iterative process particularly for projects like ours’ where in it is possible to take a corrective course in case it is observed that some decisions have not gone well. Investigate • Guaranteed, daily track possession for fixed hours for reliable inspection and maintenance activities. • Rail Grinding. • Detection and Removal from Service of Defective Rolling Stock. • Track Friendly Bogies. • Rail Lubrication. • Effective Ultrasonic Testing with Data Management, enabling Rail Fracture Prediction as well as Detection. • Track Inspection Aided with High Speed Video Cameras. Analyse 6.3. Predict 6.3.1. When the aim is to increase the through put, first question that comes to mind is why not the train length be increased? We have our own problems of loop lengths, coupler forces, synchronization etc. 6.3.2. Loop length is perceived to be that big a problem that even with increased axle loads of 25 & 30 tons, pressure is on to accommodate the volume in a wagon only as long as the existing one with 20.32 ton axle load. 6.3.3. It is now certain that axle loads are to be increased may be to 25 Ton in first step and 30 tons in the second. We have three options: Try Observe Evaluate Implement Volume - I 30 IDENTIFY ALTERNATIVES I. Relay the track first then allow heavier axle loads. II. Increase the axle loads first then think of any relaying. Volume - I 31 Pooling of their experience and knowledge will help. Possible problems shall be discussed and solutions suggested. A benefit of discussing in a group is that many solutions may be discussed best of them may get selected. III. Yet another option is to increase the axle load without any inputs but then, to ensure safety of passenger trains, slow them down to 50 kmph. 6.3.4. 6.3.5. 6.3.6. 6.4. 6.4.1. A mixed approach is better, lay a criterion for relaying, say all rails lighter than 52 kg 90 UTS to be renewed before any increase in axle loads, all 52 kg 90 UTS rails to be renewed with 60 kg 90 UTS rails in due course and 60 kg rails can be trusted to carry axle loads up to 25 tons. Sleepers and fastenings also have a role to play particularly when heavier axle loads are to be carried. Preferred sleeper type is PSC. Fortunately, most of our heavy density routes are having PSC sleepers. Other sleeper types are also acceptable but then their densities will have to be increased for various reasons. Also consider if the special track works like Points and Crossings, Switch Expansion Joints etc can be trusted to carry the increased axle loads. A sound judgment shows that PSC sleepers underneath these track works have given us a greater confidence and we can begin with the existing layouts. 6.6. The most suitable course shall be decided after completing the above analysis. The requirements, the resources and the speed at which the proposed new systems can be inducted must be given due weightage in deciding the course. 7. 7.1. 7.1.1. We have the benefit of others' experience due to our late start. Some of the problems faced by others are: Increase in rail fractures. II. Increase in weld failures. RISK OF DAMAGES DUE TO OVER LOADING AND DEFECTIVE ROLLING STOCK With increased axle loads prevention of over loading becomes an essential requirement. In addition, wheel and track imperfections cause additional dynamic forces on rail. Other vehicle defects like hot axle, hot bearing, poorly performing bogie (Truck), hunting bogies etc also cause damage to track and may sometimes result into derailments. In order to detect the defects in rolling stock, timely, various way side condition monitoring systems have been developed globally such as: III. Excessive creep. • Wheel Impact Load Detector (WILD), IV. Sleepers going out of square, even PSC. • Truck Performance Detector (TDP), V. Rail seat abrasion on PSC sleepers. • Hot Bearing Detector (HBD), VI. Rounding of ballast and rapid deterioration of track geometry. • Dragging Equipment Detector (DED), • Hot Wheel Detector (HWD), • Truck Hunting Detector (THD), • Wheel Profile Monitoring System etc. VII. Increased incidence of Track buckling. 6.5. SAFEGUARD AGAINST RISKS Having identified the risks, suitable systems shall be implemented to guard against them. Let us take the perceived risks, one by one: PREDICT RISKS I. DECIDE THE MOST SUITABLE COURSE USE COLLECTIVE WISDOM Railway Engineering is a specialized job. A Railway Engineer accumulates lot of knowledge with his observation and experience. Two railway engineers think in a similar way, give them a problem and they will come out with same solution though they might not have consulted each other. Put these engineers together and they will come out with a better solution. An individual may not be prepared to take the risk but collectively, a group of individuals may decide on a course. Volume - I 32 7.1.2. To integrate damage control strategies from different disciplines in railway engineering a vehicle track interaction measuring system to evaluate the performance of the vehicles moving over the track is essentially required for the success of a heavy haul operation. All these systems shall be installed and integrated into a centralized data base so that poorly performing wagons may be identified and taken out of service. Volume - I 33 7.1.3. 7.2. 7.3. To begin with, In-motion weigh bridges and Wheel Impact Load detectors shall be provided at critical locations so that every wagon comes across a weigh bridge almost at the beginning of its journey and across a WILD at least once during a single run. 7.3.2. 7.3.3. 7.3.6. ROLLING CONTACT FATIGUE In controlling risk, the most basic control variable is the test interval. In North American heavy haul railway practice, risk is typically judged to be sufficiently high to merit tightening test intervals when: Rolling contact fatigue due to high contact stresses caused by nonconformal rail and wheel profiles results into rail fractures. Rail grinding has been successfully used by railways to control rolling contact fatigue. We should also implement effective Rail Grinding Regimes. • Service defect rates exceed 0.06 service failures/km/yr (0.1 service failures/mi/yr). • Service plus detected rail defects exceed 0.04 failures/km/million gross ton (0.06 failures/mi./mgt). INCREASE IN RAIL FRACTURES • The ratio of service to detected defects exceeds 0.20. To safeguard against Rail fractures, the strategy shall be to catch the flaws when they are still small and observe its growth so as to predict when a rail is going to fracture. 7.3.1. leads to a broken rail derailment. Rails are typically replaced when total defects are occurring at a sustained rate of 1-2 per rail km. 7.3.7. As a predictive measure, an effective Ultrasonic testing system shall be put in place. The number of in service failures of rail is closely related to the effectiveness of Ultrasonic inspection. In a risk management approach, high inspection reliability is required particularly in lines with mixed traffic. To ensure an effective rail testing program, the test equipment must be properly designed and calibrated to reliably indicate defects. The equipment logic shall be so built that only those locations are indicated to the operator that could be a rail defect. At the same time the operator must be experienced and diligent. Traffic Density (mgt/yr.) With the increased incidence of gauge corner cracks, the Ultrasonic equipment shall be capable of covering almost entire rail head, entire web and the area immediately under the head. In addition to test capabilities, the frequencies must be matched to the growth rate of critical defects so that at least one test, is done in the interval between the development of a rail defect from a minimum detectable size to a size that represents a significant chance of rapid fracture. 7.3.4. Reliability of the Ultrasonic testing regime, effectiveness of the machine and the test interval, can be evaluated by counting the rail failures occurring at locations where no defect was detected and/ or by comparing the ratio of rail fractures to detected rail defects. 7.3.5. North American heavy haul railways detect an average of 0.4 defects per track km, each year while inspecting at intervals of 18 GMT and experience 0.06 rail failures per km. One service defect in two hundred Volume - I 34 Canadian Pacific Rail System uses a risk management approach whereby rail-testing intervals are adjusted according to different categories of risk. The basic testing interval is selected based on tonnage. As the following Table shows, there are six basic testing intervals based upon the level of tonnage. The testing interval for each track segment may then be upgraded to the test frequency corresponding to the next higher risk class if there is an additional element of risk associated with the track segment. Test Class (Time between successive tests) < 0.5 5 yrs. 0.5 – 2.7 3 yrs. 2.8 – 7.2 2 yrs. 7.3 – 13 Annual 14 – 27 > 27 Twice a year. Thrice a year. 4 times a year. 5 times a year. 7.3.8. Traffic Type Rail Type Defects Transportation of Hazardous Materials Non – Cooled (Hydrogen > 2.5ppm) Detected defects > 0.7/km/yr. OR Passenger Traffic at speeds > 70 km/h OR Rail Section < 50 kg/m OR Service failure/ detected ratio > 0.2 For example, a line carrying 10 million gross ton per year would be tested once per year. But if it also carried passenger traffic at speeds >70 kmph, it would be tested twice per year. If in addition to this the line was laid with rails lighter than 50 kg/m (100 lb/yd), it would be tested three times per year, or after every 3 million gross ton of traffic. If it also had a defect generation rate of >0.70 per km per year, the test interval will come down to four times a year. Volume - I 35 7.3.9. On Indian Railways, following test frequencies have been specified: Traffic Density Annual GMT Test Class (Time Testing Frequency, Once in < or = 5 2 yrs. > 5 < or = 8 12 months > 8 < or = 12 9 months > 12 < or = 16 6 months > 16 < or = 24 4 months > 24 < or = 40 3 months > 40 2 months 7.7. 7.3.10. The test intervals are smaller in case of IR mainly due to the fact that it is a mixed traffic scenario and passenger trains are running at speeds over 100 kmph. Based on the rail type and the rates of flaw generation the testing intervals may be tightened to mitigate the risk. 7.4. EXCESSIVE CREEP Creep may occur due to increased tractive and braking forces coming from heavier loads. In case of PSC sleeper track chances of creep shall be nil. Another related problem may be sleepers going out of square. On South African railways, even PSC sleepers have been reported to have gone out of square. Analysis of the problem has revealed that the longitudinal resistance available between rails and sleepers was higher than the ballast resistance. In Broad gauge, with high ballast resistance, this problem shall not occur, still, creep may be measured periodically as also the toe load of the fastening system to safeguard against these aspects. 7.6. RAIL SEAT ABRASION ON PSC SLEEPERS Heavier axle loads are known to have caused problems of rail seat abrasion on PSC sleepers on US and South African Railways. Rail Volume - I 36 DETERIORATION OF TRACK GEOMETRY AND ROUNDING OF BALLAST ETC More frequent track recording may be adopted in sections where disturbance of track geometry is noticed on account of weak formation or any other reason. Based on the TRC results, need based tamping may be carried out. Heavier loads are likely to cause faster pulverization of ballast resulting into rounding of stones. Rounded ballast may not be able to retain packing. Good quality hard stone ballast will have to be provided. 7.8. INCREASED INCIDENCE OF TRACK BUCKLING With introduction of heavier trains, increased tractive and braking forces will have to be encountered by rails. This, along with higher thermal forces on account of heavier rail section will compound the risk of track buckling ahead of a train at the time of braking. Track will have to be maintained properly to guard against this. INCREASE IN WELD FAILURES It has been an accepted fact for years that rail joint is the weakest link. Over the years, such joints have been replaced by welds. Today the position is that thermit weld is the weakest link. We feel comfortable after encasing a thermit weld between joggled fish plates. Joggling of all thermit welds in a higher axle load section wherever weld failure rate exceeds 1 per 100 is a good safeguard against risk of weld failures. 7.5. pad will have to be carefully chosen to overcome this problem. Composite rail pad with softer layer in contact with concrete may provide the answer. Indian Railways have made significant progress in this direction. This, however, highlights the need to increase the rail seat area in the PSC sleeper. A new sleeper, under development for 30 ton Axle load shall provide an increased rail seat area. 8. IMPLEMENT IN STAGES IR has already started its march towards heavier axle loads in stages. Stages are both about increased axle load as well as geographical reach. The axle load is being increased in stages of 22.82 tons, 25 tons and 30 tons. Geographically also it is being implemented in certain identified sections. This is a very good approach as we can gather our own experience while minimizing the risk of irreparable damage. 9. CONCLUSIONS 9.1. To conclude, it can be said that there is a need not only to have a relook at the theory behind track strength assessment but to use a more pragmatic and judgmental approach. Enough experience is available, globally, on heavier axle loads which can be utilized in chalking out the course for Indian Railways. 9.2. In IR's case the most suitable course appears to be: Volume - I 37 9.2.1. All rails lighter than 52 kg 90 UTS to be renewed before any increase in axle loads, all 52 kg 90 UTS rails to be renewed with 60 kg 90 UTS rails in due course and 60 kg rails can be trusted to carry axle loads up to 25 tons. PREPARATION FOR HEAVIER AXLE LOAD 9.2.2. Existing PSC Sleeper can be considered suitable for axle loads up to 25 tons, keep a design ready for a heavier sleeper which may be required for axle loads of more than 25 ton. ATUL KUMAR KANKANE* 9.2.3. Fastenings with increased toe load shall be required in areas having problem of creep. 9.2.4. Rail Grinding is a must even in present scenario in some of the sections. For 25 ton axle loads, it shall be essential to implement it. 9.2.5. More effective methods of rail flaw detection are required to be deployed. Equipment shall be capable of covering almost entire rail head, entire web and the area immediately under the head. 9.2.6. Effective steps shall be taken to control over loading like installation of In-motion Weigh Bridges. 9.2.7. Steps shall also be taken to identify and take defective wagons out of service by installing equipments like Wheel impact load detector etc. 9.2.8. Frequency of USFD shall be related to the weight of rail, GMT and flaw generation rates of the section. 9.2.9. Thermit welds shall be eliminated to the extent possible. SYNOPSIS Indian Railway system is carrying mixed traffic both passenger and freight traffic. With the increase in population and growth potential of country, demand of transporting more traffic by railways have increased. The most economic and feasible proposition to carry more load is to increase the carrying capacity of our goods train which requires upgradation of existing infrastructure. In this paper, an attempt has been made to list out various items for up grading our track and bridges. Precautions for heavy axle load operation have also been identified. INTRODUCTION Benefits of running of heavy axle load specially for transport of high density commodities have been well experienced in North America and Australia where 30T axle load is a common feature. To contain the operational cost and to get more profit, increase in axle load is the only option available. With the more input, better North-South and East-West connectivity and better quality of roads, increasing competition is being faced by Railways. Better trucks with more carrying capacity have further worsened the situation. To maintain our market share of freight transport, increase in the carrying capacity of wagons is the only option available with us to maintain viability. For past 50 years, the growth in Railways could not match with the growth of economy. Our market share in freight segment has reduced to approx. 40% compared to 80% about 50 years ago. Due to constant modernization and upgradation in technology, other modes of transports have threatened us with such a heavy blow that our survival is at stake. If we do not wake up right now and prepare ourselves for heavy haul, we may not survive in this competitive atmosphere. NEED 30 T and more axle load is very common in many of the world railways. Volume - I 38 Volume - I * Dy.G.M. & Secretary to GM, West Central Railway World’s heaviest and longest freight train runs in Australia with a pay load of 82000 tonnes and gross load of 99734 tonnes. The train is formed of 682 wagons, hauled by eight 6000 HP diesel locomotives. The length of train is 7.2 Km used for transport of mineral. At present gross load of a freight train in India is about 4800 tonne with pay load of about 3600 tonne. During 2006-2007, Indian Railways have envisaged Mission 800 MT for freight loading. By 2012, Indian Railway plans to carry 1200 MT of freight. With existing loading capacity and axle loads, it is impossible to achieve the target. Construction of Dedicated Freight corridors (DFC) is one of the right decisions taken in this direction. We have been discussing about higher axle loads for the past several years but nothing perceptible has been done at field level. Because of our slackness and lack of vision in foreseeing the future requirement, we, as Civil Engineers, failed to give desired input to our Permanent Way infrastructure at right time. Nation can not wait for our idleness, therefore decision has already been taken to enhance the carrying capacity of the wagons on existing track as a pilot project on certain identified routes. BENEFITS DISADVANTAGES There are certain inherent disadvantages of increase in allowable axle loads– 1. Increased cost of maintenance of fixed assets like track and bridges because of increased wear and tear of track components such as rails, sleepers, ballast and sub grade. 2. Increased rate of deterioration of geometry of track. 3. Increased wear and tear of turnouts and crossing bodies. 4. Increased maintenance cost of moving assets like locomotives and wagons. 5. Increase in level of noise and ground vibrations. 6. Due to high momentum, longer braking distance will be needed requiring the change in signal overlaps. 7. Reduced life of infrastructure components (Fixed assets as well as moving assets) due to more fatigue damage. 8. Requirement of more powerful engines for hauling extra load. 9. Increase in axle load will increase the risk of derailment due to Rail fracture and increased track and wagon irregularities, which may affect the safety of passenger train also. Running of heavier axle load has many advantages – 1. Better net weight/gross weight ratio leading to better use of hauling capacity of engine. 2 Increased line capacity, requiring less number of trains for hauling the same tonnage. This will also reduce the operational cost per unit tonne of haulage. 3. Fewer wagons will be needed to haul same load. This will increase the wagon productivity. This will also reduce the capital cost and wagon maintenance cost. PRECAUTIONS FOR RUNNING OF HEAVY AXLE LOADS For economic viability of Railway system, heavy axle loads are to be introduced on existing bridges with adequate strengthening, replacement or rehabilitation to increase the structural strength as well as fatigue life of bridge. To safeguard against passage of overloaded, unevenly loaded wagons and to prevent excessive dynamic load due to flat tyre or due to defects in track geometry, certain mandatory safeguards are to be ensured. 4. Less number of trains will increase the terminal efficiency. 5. Less requirement of locomotives for transporting the same amount of tonnage, which will reduce the loco maintenance cost and capital wit. This will also result in lower fuel consumption and less requirement of crews, thus saving highly precious fuel and manpower. 1. Use of in-motion weigh bridges to detect overloading, preparation of RR on the basis of weighment by weighbridge. 2. Use of Wheel Impact Load Detector (WILD) to prevent passage of flat tyre and unevenly loaded wagons on the bridges, Transportation by rail is considered as the most environment friendly means of transport. If we are able to increase our market share, and able to carry more freight by rail, we will be greatly contributing to better environment. 3. Maintain the bridge approaches with better track geometry. 4. Elimination / reduction of rail joints on bridges to reduce dynamic impact on bridges. 5. Preventing over speeding of goods train. 6. Volume - I 40 Volume - I 41 In addition to above, Hot Bearing Detector (HBD) and overload and Imbalance Load Detector (OILD) will prove to be beneficial in preventing the derailment of heavily loaded train and subsequent extensive damage to track. All the above issues can be tackled if we are able to improve our track performance. For better track performance, we have to improve in following areas (i) IMPROVEMENT TO FIELD WELDING PROCESS EFFECT OF INCREASE IN AXLE LOAD • For reducing the failure rate. Running of heavy axle load trains on existing track affects both track and bridges. All the track components like rail, sleeper, fastening, ballast and sub grade are affected due to higher dynamic load. High axle load causes increased wheel wear and higher bending stresses in rail and sleepers, higher bearing stresses in ballast and subgrade. Fatigue life of track component is reduced considerably. In bridges substructure and superstructure both are affected whenever heavy axle load is introduced. Reiff, 1990 suggested from AAR (Association of American Railroads) studies that routine maintenance demands increase by 60% for 20 percent increase in axle loads. Ebersohn et al, 1993 reported that track maintenance increased by 183 percent for a reduction in subgrade stiffness by half. Therefore, maintenance input for running of heavy axle load should be planned meticulously. • For obtaining equivalent performance as of parent rail. • For higher fatigue resistance under heavy axle loads. If we are able to improve the quality of our field welding, introduction of more welds during repair of failed weld can be avoided, Higher wear resistance will reduce battering of heat affected zone of weld. Battering of welds causes sudden impact and more dynamic loading which is very detrimental to rails in the vicinity. (ii) (iii) INFRASTRUCTURE IMPROVEMENTS Strengthening and up-gradation of existing infrastructure (track and bridges) is the only option for increasing the axle loads. The strength of assets and fatigue life both are to be assessed with the help of sophisticated instrumentation and need based attention is to be given at specific location. (a) ATTENTION TO TRACK To bear higher stresses, following issues are to be taken care of – • More resilient design to reduce the sudden impact to reduce the permanent damage. • Use of materials that is more resistant to wear and fatigue e.g. use of better steel for rail to reduce rolling contact failures or fatigue related fractures. (iv) (v) INTRODUCTION OF SUITABLE WIDE GAP WELD TECHNOLOGY • To fill a gap of 75 mm instead of 25 mm. • To prevent introduction of one additional weld while repairing failed weld. IMPROVEMENT IN POINT AND CROSSING AREA • Higher quality of rail steel for manufacture of points and crossings to prevent early failure. • Thick web switches and moveable nose crossings to reduce dynamic load • Better bolt tightening practices. IMPROVEMENT TO SWITCH EXPANSION JOINTS • To reduce discontinuity of rail head at expansion area. • Manufacturing special thick web section for fabrication of new lay outs where expansion is distributed at two places instead of at one place provided in conventional layout. IMPROVEMENT TO SUBGRADE • Better sharing of loads/stresses between track components. • Drainage improvement of embankment. • Improved design of track components for increased elasticity. • • Strengthening or removal of weak components. Increasing the clean cushion of ballast and better drainage through ballast. • Better resistance to gauge widening. Concrete sleepers are better equipped for this. • More ballast cushion which may necessitate widening of embankments. • Reducing the requirement of surfacing and alignment of rails. Volume - I 42 Volume - I 43 • (vi) (ii) Better track maintenance to reduce abrupt discontinuity in movement of wheels on rail ( rail wheel interaction) because of sudden change of stiffness of track at bridge approaches, SEJs and Point and Crossings, Level Crossings etc. ATTENTION TO BRIDGES The methodology for attention to bridge should be able to extend the life of bridge and carrying out the replacement and maintenance works while safely carrying the traffic. An objective decision has to be taken whether it is more economical to replace the bridge or to increase the live load capacity through different means. The decision should be taken on the basis of condition of each bridge separately. (i) Volume - I ATTENTION TO BRIDGE SUB-STRUCTURE • Older bridges may not be able to withstand increase in axle load, therefore more thorough inspection with better technique is required. • Under water inspection of important bridges foundations with most modern techniques where lower portion of pier remains under water throughout the year. • Bridge specific unique repair methodology with use of sophisticated under water maintenance technique. • Strengthening of substructure by grouting or jacketing as per requirement at site. IMPROVEMENT TO TRACK GEOMETRY • (b) Improving the quality of ballast with good interlocking properties. (iii) INCREASING THE LIVE LOAD CAPACITY OF BRIDGES BETTER INSPECTION AND MAINTENANCE PRACTICES • More frequent inspection of critical bridges specially during initial period of introduction of heavy axle load. • Use of better materials, new equipments and gadgets for more safe and comfortable inspection and effective repair. • Better repair techniques for quick and good quality repair/ rehabilitation. • Checking the strength of guard rail for proper functioning during derailment of heavier trains. • Replacing weak fatigue damaged components of bridge. • Strengthening fatigue damaged components. • Reinforcing weak components. • Enlarge bearing area. • Reducing the dead load of bridge by using stronger material in critical components. • • Instrumentation to measure the strength and fatigue life of bridges. Post tensioning of bridge. • Planning adequate time allowance for repair of bridge. • Adding new member to existing structure. • • Converting pin joints to rigid joints. Special inspection strategy of steel bridge to detect fatigue crack initiation in - • Converting beam action to truss or arch action for better load distribution. Eye bars and pin plates. Crack in rivet holes. • Strengthening of riveted connections by replacement with high strength bolts. Welds at end of cover plate. • Welded attachments. Removal of short and discontinuous welds, removal of intersecting and overlapping welds. Corroded locations. • Replacement of welded cover plate with bolted cover plate. Weld defects. • Removal of notches and weld defects. • Reduction of residual stress in the welds. • Removal of corroded members. 44 • Volume - I Measurement of acoustic emissions is one of the such technique which can tell about potential crack formation. Incidence of more acoustic emission near some member is indicative of crack initiation. These techniques should be extensively used in steel structures. 45 CONCLUSION To keep ourselves economically viable, reduction in operating cost is the only alternative in present scenario. Cost and benefit analysis of running of heavier axle loads is to be done. Assessment of cost of rebuilding / rehabilitation and increased maintenance cost due to increased track and structure damage is to be done objectively. This should be compared with the reduction in operating cost. New line construction should be able to withstand even 40 T axle load to cater for future requirement on certain critical routes as cost implication during new line construction may be less than the future rehabilitations. On Indian Railway increase in life of existing bridges is more critical. Replacement and maintenance work has to be done without affecting the safety of traffic. Closer co-ordination is required with Mechanical Directorate for better design of suspension system of rolling stock to reduce the dynamic impact of heavily loaded wagons on track and bridges. If we are able to maintain rail and wheel profile, dynamic impact due to heavier axle load is likely to be reduced considerably. Extensive installation of various rolling stock performance detectors like Hunting Detector (HD), Hot Wheel Temperature Detector (HWTD), Acoustic Bearing Detector (ABD) and finding and fixing vehicle defects by remote vehicle performance monitoring at later stage will go a long way in improving the safety performance of heavy axle load operation. Extensive use of new sophisticated track machines will improve the life of track component as well as better and safe running. Consideration should also be given for running of heavy axle load at higher speeds may be 100 Kmph. This will require more extensive study for rehabilitation planning of existing track and bridges. IMPLICATIONS AND SOLUTIONS FOR RUNNING OF HIGHER AXLE LOAD ON SPECIFIED ROUTES OF INDIAN RAILWAYS K.K.MIGLANI *, JAGTAR SINGH** SYNOPSIS With the increase in freight traffic, it has become necessary for the Railways to increase loading per wagon to achieve fullest capacity of wagons. Slowly, Railways have introduced CC+4+2t, CC+6+2t & CC+8+2t loading. These loadings have enhanced the through put. This has benefited the Railways in a big way. Increase in loading per axle has forced Engineers to think on design of rail section, sleepers, ballast bed, formation and bridges. Because of increased loading, the stresses in rails, sleepers, ballast bed, formation and bridges have definitely gone up, resulting in early fatigue failures. On track front, there will be requirement for increased maintenance inputs, especially rail grinding, ultrasonic testing of rails, and handling rail weld failures. Special attention would be required on fish plated joints, Points & Crossings area & availability of clean ballast cushion. Paper also gives suggestions for various aspects for maintenance of track on higher axle load routes. REFERENCES 1.0 1. Ing Rainer Wenty : The Asian Journal Vol.13 Number 1 July 2006. 2. Duane Otter, Shakoor Uppal ; Railway Age Nov.2003 3. Kalay, Semih : Railway Track and structures, Jan 2002. 4. Reiff S.F. 1990 : Proceedings of Workshop on heavy axle load Pueblow, U.S.A, Paper-21, PP.21.0-21.5. 5. J.F. Unsworth TRB TRB 2003 Annual Meeting CD-ROM. 6. Ebersohn W., Vizo M.C., Sclig E.T., 1993 5th International Heavy Haul conference, Beijing. The Indian economy enters the tenth plan with an expectation of 6% to 7% annual growth in the GDP and consequently 7.2% to 8.0% growth in the transport sector. These expectations place heavy demands on the already saturated road and rail transport system which coupled with the inadequacies in the power sector could be a major constraint in the realization of the projected economic growth. With Airways, Coastal Shipping and Inland Waterways being in the fringes, freight transport in India is basically shared between Road and the Rail sectors. The road network in India has grown from 4-lakh kilometres in 1951 to over 30-lakh kilometres now – second largest road network of the world. OVER VIEW OF TRANSPORT SECTOR Railways made a flying start almost doubling the transport output in Volume - I 46 Volume - I *Dy.Chief Engineer/TO, Northern Railway **Dy.Chief Engineer/Land, Northern Railway the first 5 year Plan. There was however a perceptible slowing down from 1968 to 1980 followed by a revival in the last two decades. 1.1 1.2 1.3 1.4 At present predominantly running axle load on Indian railway system is upto 22.82 tonnes are operating. Heavier axle Loads will enable carrying more payload in one train, which in turn improve throughput substantially. TRENDS IN FREIGHT & PASSENGER TRAFFIC • Freight Traffic has grown 90 times from 5.5 BTKM in 1951 to over 500 BTKM now. • Passenger Traffic has grown 80 times from 23 to 1800 BPKM in the same period. Railway Board has already taken decision to run BOXN and other wagons with CC+8+2 loading on specified routes. This would result in axle loads of the order 22.82t and TLD of 8.51t/m. SHARE OF TRAFFIC WITH HIGHWAYS AND RAILWAYS • National and State Highways comprising only 8% of the network carries 80% of the traffic In this scenario, introduction of trains with 30t axle loads probably is not quite far away. • Railway share of Freight Traffic has declined from 89% in 1951 to 38% now. Running of such heavy axle load trains on the existing track would cause very high stresses on the track structure which would have far reaching implications on the requirements of track components and their maintenance and life. CHALLENGES FOR RAILWAYS • Golden Quadrilateral Road Network and Induction of Multi Axle Road Vehicles will make a serious dent • Even heavy duty Bulk Transport may not remain the exclusive preserve of the Railways. COMPARISON OF FREIGHT TRAFFIC WITH CHINESE RAILWAY A comparison of the Railway Systems in China and India makes interesting study. In the decade 1992 to 2002 the route Km on the Chinese Railways (CR) has grown from minus 6% to plus 14% in comparison to that of the Indian Railways (IR). The two Railways carried almost the same volume of Passenger Traffic both in 1992 as well as 2002. However, in respect of Freight Traffic, the volume carried by Chinese Railway is four and a half times that of India. Chinese Railway operates roughly twice the number of trains on electrified double line tracks than the Indian Railways 1.5 and its diagonals connecting the four metropolitan cities. MODERNIZATION AND EXPANSION ON INDIAN RAILWAY IN A BIG WAY Integrated Railway Modernisation Plan (2005-10) has been made which has objective to enhance capacity, improve rail-port connectivity, higher axle load wagons to carry bulk material and development of dedicated freight corridors, two intercity, corridors Delhi-Patna-Howarh and Delhi – Channai to be developed to run 150 Kmph trains using latest technology high speed coaches.And running of freight train @100Kmph on the high density Golden Quadrilateral Volume - I 48 1.6 TRANSPORT CAPACITY ENHANCEMENT Indian Railway is in Transport business. Trailing loads and operating speeds are the principal efficiency indicators. Within the limitations of a loop length of 686 meters and the existing and proposed Track Loading Density of 7.67 and 8.25 tonnes per meter the options are • Higher Axle Loads Axle Load in tonnes 20.32 22.9 25.0 30.0 No. Of Vehicles 58 58 58 58 Trailing Load in tonnes 4831* 5410 5800 6960 Track Loading Density t/m 7.67 8.25 8.82 10.58 *With CC+2t • Increasing number of Axles per Wagon • 11 % higher throughput with 23t Axle Load and 20 % with 25t Axle Load can be achieved. With introduction of 30t axle load throughput will increase by 30% approx. • With 25 t Axle Load Track Loading Density would need to be relaxed to 8.8 t/m and 10.58 t/m for 30t axle load. • To carry more, cost effectively Indian Railways has gone in for increase in loading per axle and infrastructure development of dedicated freight corridors. Volume - I 49 1.7 continuous beam on closely spaced elastic supports and the bending stresses are determined from the theory of Beam on Elastic foundation (BOEF). The fact that the rail is actually supported on sleepers at some distance apart introduces a very little error and is neglected. The maximum rail stresses calculated for 60 Kg rails with 100 Kmph speed work out to be are as under: BENEFITS OF HIGHER AXLE LOAD Fewer wagons will be needed to haul the same load, leading to lower capital cost and possible reduction in wagon maintenance cost, fewer locomotives, lower fuel consumption per net tonne, reduction in train wagon kilometre operated, and fewer crew deployments entailing savings in wages. The railways in North America, Australia, South America, South Africa and Sweden have all increased axle loads to obtain significant savings in operating cost. These savings have been achieved despite increased cost of maintaining tracks, greater track component damages and shorter component lives. The raising of the axle load from 22.5 tonnes to 30 tonnes yielded 40 per cent savings in transportation cost in the US. This in turn helped the railways in that country achieve significant reduction in operational cost of transporting containers and introducing customised wagons to win back traffic from the roadways. Boosting wagon productivity that is, how to carry more per wagon, or how to achieve higher payload per wagon, has become important for the Indian Railways in view of the increasing threat from the various other modes of transport, particularly roadways. 2.0 DESIGN ASPECT FOR TRACK Computation of stresses has been done, that would be induced in Rail, Sleeper, Ballast and Formation due to introduction of 30 t axle loads and the suitability or other wise of these components. Minimum track structure for running of these heavier loads and its impact on maintenance of P.Way has also been dealt with. On Indian Railway the strength of the for running various locomotives and rolling stocks at different speeds is assessed by calculating rail stresses induced locomotives/rolling stocks running at contemplated speed, using Civil Engg. Report No.C-100 rail wheel contact stresses on straight and curved tracks due to axle load combined stresses in rail head, foot, assuming rail wear of 5% are calculated on 52Kg rail and 60 Kg rail. 2.1. RAIL DESIGN 2.1.1. STRESSES IN RAILS Rail is a very important and costly component of the permanent way. Its failure will affect safety. Therefore, the rail is treated as a Volume - I 50 Permissible Stresses in 72 UTS LWR Track for 19.25 Kg/mm 90UTS 2 25.25 Kg/mm2 It can be seen from the above that even in 60 kg Rail stresses are higher than the permissible limit. In other words, for 30t axle loads, rails of higher pondage will be needed. 2.1.2. RAIL WHEEL CONTACT STRESSES The contact between rail and wheel flange should be theoretically a point. Hertz theory explains that in practice the elastic deformation under higher axle load results in deformation of steel of wheel and the rail creating an elliptical contact area. The dimensions of contact ellipse are determined by the normal force on contact area, while the ratio of ellipse axes a & b depends on the main curvature of the wheel and rail profile. Inside the contact area a pressure distribution develops which in a cross section, is shaped in the form of a semi-ellipse with highest contact pressure occurring at centre. The concentrated load between wheel and rail causes a shear stress distribution in railhead as shown in fig. (5.18 & 5.19) The contact problem is most serious in case of high wheel Volume - I 51 For 72 UTS rail the maximum allowable shear stress will work out to 21.60 Kg/mm2 and for 90 UTS rail, it will be 27 Kg/mm2. It there implies that 90 UTS rail will be required for running 30 tonne axle load. 2.1.3 SLEEPER DESIGN CONTACT STRESSES BETWEEN RAIL & SLEEPER Q = wheel load +load due on loading due to curves. As an extension to the Beam on Elastic on Elastic Foundation model proposed by Zimmerman for arriving at the stresses in the Rail, the contact pressure between rail and sleeper is computed based on based on the principle of discretely supported rail on springs at specified intervals. Using this principle and Dynamic Amplification Factor because of the dynamic interaction between rail and wheel due to the speed of train, the maximum bearing force on a single discrete rail support due to the wheel load is obtained from the formula R = Wheel radius (mm) • loads or relatively small diameters. Eisenmann has devised a simplified formula to calculate the maximum shear stress in rail head, which is as follow Tmax = 4.13(Q/ R)1/2 Where T max = maximum shear stress in rail head Since problem is one of the fatigue strength, the permissible shear stress is restricted to 30% of UTS, which works out to be 21.60Kg/ mm2 for 72UTS rail and 27.00 Kg/mm2 for 90 UTS rails. It is seen that maximum shear stress increases with increase in axle load. It also increases with increase in curvature of track as increase super elevation results in increase on loading of inner rail when goods train ply on mixed traffic routes. The shear stress also increases with wearing of wheels as the wheel radius decreases with the wear of wheel. Thus it may appear that the problem of increase axle load can be solved with increase in wheel diameter but this is not possible as increase in wheel diameter means less carrying capacity because of restricted overhead clearances. Therefore only way to keep the maximum shear stresses within permissible limits is to use the rail with higher UTS. Bearing Force on Sleeper – Fmax = DAF * Pa/2 * (U/4EI)1/4 • Where DAF = Dynamic Amplification Factor, which depends on the track quality, the train speed and a multiplication factor of slandered deviation, depending on confidence interval. • P =Effective Wheel Load (T) • a =Sleeper Spacing (Cm) • U =Modulus of Elasticity of Rail Support or Track Modulus (Kg/ Cm/Cm) • E =Modulus of Elasticity of Rail Steel (Kg/Cm2) • I =Moment of Inertia of Rail Section (Cm4) Based on the charts developed by RDSO in their report no C100 for different Rolling stock based on experimentation, the speed factor for BOX wagons for a speed of 100 Kmph comes to 1.68, However , since the wagon of 30 t Axle load would be The contact stresses for BOY, BOB and BOXNHA wagons would be as under. The diameter of wheel of Casnub bogie is taken as average of new wheel and worn out i.e. (1000+925)/2=962.5 mm. Volume - I 52 Based on Beam on Elastic Foundation Model Volume - I 53 different with different dynamic characteristics , the dynamic effect due to speed is also checked based on the formula proposed by Elisenmann for Dynamic Amplification factor • As per Eisenmann’s Formula, • DAF = 1 + t ø {1+(V- 60)/140} sleeper spacing of 60 cm and 65 cm computed basis on the above formulae are: t = Multiplication Factor Depending on Confidence Interval and Ø = Factor Depending on Track Quality. • For confidence t = 3, Ø =0.2 for average track quality and 100 kmph speed, Permissible contact pressure between the rail and sleeper for concrete sleepers is 4N/mm2. DAF works out = 1 + 3 x 0.2 {1 + (100-60)/140} = 1.78 • For 75 KMPH, DAF works out to 1.66 • Speed factor of 1.68 is adopted for computation of stresses • In the absence of relevant data regarding the type of rolling stock and the speed that would be permitted for the purpose of computation of stresses on sleepers ballast and formation a speed factor of 1.68 is adopted based on the above computed values and RDSO. • For 30 t axle loads, the effective wheel load ‘P’ would be 15t. But as seen from computations in the above table, contact pressure value at the rail seat for a track with 52 kg rail either on 52 kg or 60 kg sleepers would be far excess of permissible value when 30 t axle load rolling stock is introduced. Contact pressures are higher than the permissible value even for a track with 60 kg rail on 60 kg sleepers at 60 cm spacing. As computed above a track with 60 kg rail on 60 kg sleepers at 43 cm spacing only fit for running of 30t axle load rolling stock from contact pressure criterion on PSC sleeper. This poses a very severe restriction, which would have far reaching implications and hence needs to be examined thoroughly before introduction of 30 t axle loads. The Mean Contact Pressure Between Rail and Sleeper on the most heavily loaded sleeper would than be computed from the formula: - Presently, Indian Railways is looking for only allowing upto 25t axle load on the present track. But, if the future is to introduce 30t axle load, then it is necessary to redesign the sleepers. óm = (Fo + Fmax)ó/A Where Fo = A = The total pre-tensioning force of fastenings on rail support (T) 2 Effective rail support area (mm ) For concrete sleepers with elastic rail clips, Fo works out to = 2x1, 000 =2000Kg = 2T for PSC Sleeper “A“ would be 0.125 X the width of the rail foot, since the width of the grooved rubber pad used below the rail is 125mm. For calculation purpose, it is normal to presuppose that the contact force is distributed evenly over the contact surface area. The Contact pressure between rail and sleeper for 52 Kg and 60 Kg rail sections or 52 kg or 60 kg PSC sleeper and also for Volume - I 54 2.1.4 BALLAST BED DESIGN STRESSES IN BALLAST BED Ballast bed and formation are conceived as a two-layer system for the purpose of computation of stresses. Vertical forces on the ballast bed due to wheel loads will be considered as the determining stresses for the load bearing capacity of the layer system. Over loading of ballast bed due to increased axle loads causes rapid deterioration of the quality of the track when heavy axle load trains are introduced. The compresses stresses that the sleepers exert on the ballast bed are considered evenly distributed for the purpose of calculation. It means that the material from which the sleeper is made plays no role. The maximum stress between the sleeper and the ballast bed Volume - I 55 under the wheel load ‘P’ is expressed based on Zimmermann’s theory and by applying a Dynamic Amplification Factor due the speed of the Rolling stock as per Eisenmann’s model. ó sb = { DAF* Pa/2(U/4EI)1/4}/Asb = Fmax/Asb Where Asb = Contact area between sleeper and ballast bed for half sleeper (mm2) 52 Kg and 60 kg sleepers differ only in respect of the distances between inserts so as to accommodate higher rail section and in all other respects, they are identical. Hence, there would be no difference in ballast stresses due to the use of either sleeper and the half sleeper contact area works out to 336,875mm2. Stresses on the ballast bed due to the force on sleeper, competed for 52Kg and 60 Kg rail sections and different sleeper spacing are tabulated as under: 2.1.5 FORMATION DESIGN STRESSES ON FORMATION The loads from rolling stock are finally transferred to the formation through the ballast cushion, where the ballast bed and the formation are conceived as a two-layer system. Introduction of higher axle loads results in imposition of increased compressive stresses on the formation. This would lead to faster deterioration of track and call for more frequent maintenance schedules. The compressive strength on formation should be always kept within the bearing capacity of the formation, which depend on the modulus of elasticity of the formation apart from other geotechnical characteristics of the soil. The stresses transmitted to formation primarily depend upon the depth of ballast cushion and the effective bearing length of the sleeper. From the criteria of the force on individual sleeper due to axle load, compressive stresses on the formation are calculated from the following formula: Pmax = DAF. Pa/ ð DL. (U/4EI)1/4 = Fmax. (2/ðDL) Where, D= Depth of ballast cushion (mm) And L= Effective bearing length of sleeper at rail seat (mm). For PSC sleeper, ‘L’ is taken as 1040mm. Formation stresses for 52 Kg & 60 Kg rail sections and ballast cushion of 250mm & 300mm, on account of introduction of 30t axle Load, computed based on the above formula are tabulated below: The permissible contact pressure on the ballast bed is taken as 0.50 N/mm2. As seen from the values in the above table, for the present track structure, stresses on the ballast bed would be whining the permissible value when 30 tonne axle load rolling stock is introduced. It can be gathered from the above equation that sleeper spacing and the extent of ballast support area have an important influence on the mean stress on ballast bed. A high value for foundation leads to high value of stress on ballast bed, whereas heavier rail profile has a positive effect in this respect. A heavier rail profile has a greater influence on rail stress reduction. The effect of ballast stress however, is approximately half of the effect on the rail stress. Volume - I 56 The modulus of elasticity and permissible stresses on the formation for 2 million cycles of loading as indicated by Coenraad Esveld are reproduced on next page: Volume - I 57 • Therefore, it is paradoxical to say whether the use of head hardened rails/lubrication of rails will actually enhance or reduce the life of rails with heavy haul. Tests at Facility for Accelerated Service Testing (FAST) have also shown that higher wear rates of rail not only reduce surface defects but also suppress the internal defects i.e. detail fractures and shelling. Obviously, with introduction of 30 tonne axle load rolling stock, in most cases, formation stresses would exceed the bearing capacity of the formation. 3.0 IMPACT OF INCREASED STRESSES DUE TO HIGHER AXLE LOADS 3.1 IMPACT ON RAILS If the permissible stresses are exceeded, then : • There will be plastic flow of metal at contact and development of cracks in railhead will take place. These cracks grow gradually due to combined effect of contact stresses resulting is surface breaking. If allowed to grow, they have potential go subsurface and cause failure by combining with already present defect. Another implication is that if the surface cracking is severe, the substantial amount of ultrasonic waves transmitted will be reflected from these surface defects making it impossible for the rail section to be reliably inspected for full depth. • The most prominent defects in rails on the heavy haul routes have been observed as Rolling Contact Fatigue defects predominantly the Gauge Corner Fatigue. • The maximum shear stress is developed not on the contact surface but at a depth of 5-7mm below the railhead. It therefore implies that use of head hardened rails will be effective only if such hardening increases the UTS up to the depth of 6mm or more from the railhead. • It is interesting to know the effect of surface hardening and lubrication in context of maximum shear stresses. If wear is not dictating the life of rail, as on head hardened rail/ lubricated rails, the maximum repetitive shear stress will always occur at same point, thereby increasing propensity of fatigue failure and shelling. Volume - I 58 On the other hand if the rail is allowed to wear, the point of occurrence of maximum shear stresses will gradually shift downwards making it less prone to shear fatigue failures or shelling. RAIL FATIGUE LIFE From the analysis of bending stresses and contact stresses it may though appear that 52 Kg 90 UTS rail may suffice the requirements of increased axle load, but in practice, the above stresses coupled with thermal stresses and residual stresses set up cyclic stresses. From the theory of fatigue, it is evident that such cyclic stresses may result in failure of material at a stress level lower than what would normally require for failure. Allan M Zarembski compared the rail life based on wear limits to rail fatigue life for different axle loading environment and found that in lighter axle Rail Fatigue Life vs Wear 136 RE Rail Volume - I 59 loading environments, rail wear is dominant mode of failure while in heavier axle load environments, the fatigue emerges as dominant replacement criterion. AREA Bulletin No. 685, Vol 83 reports of study made by Dr. Allan.M.Zarembski on the effects of increasing axle loads on tangent track on a continuous welded track. Two independent studies were conducted to determine the fatigue life of rails with different axle loads. The results have shown that heavier axle loads have resulted in a more severe occurrence of rail defects. The study had two conclusions: (i) An increase in axle loading will result in decrease in the fatigue life of rail, measured in terms of cumulative MGT and reduction occurs for both heavier as well as lighter sections. (ii) When the axle loads are increased from 27.5 tonnes to 33 tonnes (corresponding to 70 tonnes and 100 tonnes freight cars), the resulting decrease in life of rail was found to be 40 %. Thus under Indian Railways context, it can be said that with increase in axle load upto 25 tonnes, 52 Kg/m, 90 UTS rail may even though be permitted from the considerations of bending stresses and contact stresses, however, in the interest of long term economy and from fatigue considerations, it will be more appropriate to use a heavier section of 60 Kg/m if further increases in Axle loads are imminent. 3.1.1 IMPACT OF FORCES ON CURVATURE AND TRACK GEOMETRY • ORE 161 studies reports that Dynamic effects of 22.5 tonnes axle loads for different speeds, track quality and radius of curvatures • Dynamic wheel force (DSQ) increases with increase in speed • The lateral rail- wheel force increases with increase in curvature and deterioration of quality of maintenance. • Curves with radius sharper than 400m require a greater care of track geometry with increase in axle loads. • Poorly maintained track will have most pronounced effect, where increase in the wheel force can be up to 22% of axle loads for speed ranging bet. 60 to 100 kmph observed. • Volume - I Track quality was expressed in terms of standard deviation of vertical profile and alignment 60 • Standard Deviation?<1mm very good, 1-2 mm good > 2mm moderate 3.1.2 IMPACT OF WHEEL FLAT • Relationship between the flat size and force is almost linear • On Indian Railways, the permitted sizes of wheel flats are 50mm for locomotives and coaching stock and 60mm for goods stock. • Size of flat will depend on diameter of wheel (C2/8R). • No consideration for size of flat in specifications of wheel flat. The largest loads applied to the track from vehicles are those, which Volume - I 61 rail) in LWR territories, during winter season, when the full tensile stresses are present in rail section. arise from irregularities on wheel such as wheel flat. ORE 161.1/RP 3 reports of the tests carried out on flat tyres measuring the effects of speed, size, sleeper type and axle loads. The results reveal: (i) The forces at frequencies above 500 Hz referred to as P1 forces increases continuously with speed, while the forces at frequencies below 100 Hz, referred to P2 forces are more of less independent of speeds. The P1 forces have bearing on wheel rail contact stresses. This force, which causes most of damage to rails and concrete ties, increases with increase in speeds. (ii) Increase of axle load from 20 t to 22.5 t (12.5%) caused the increased wheel flat force of the order of 0 to 6%. Hence if go from 22.5t to 30 t the increase in wheel flat force will be of the order of 24%. • Studies have also revealed that movement of wheels with flats can generate dynamic forces, as high as six times the normal static load, in extreme situations. • On Indian Railways, the effect of rail/ wheel defects and vehicle suspension, on static wheel load, is represented by a speed factor (Rail stress calculations), which can assume a maximum value of 1.75 for locomotives and 1.65 for wagons. • Volume - I The problem assumes alarming proportions incase of thermit welds (which have the impact strength of 7-10% of parent 62 • Spate of weld failures due to running of flat tyres under these conditions, is not uncommon. 3.1.3 IMPACT ON RAIL/WELD FAILURES With increase in axle load, there will be increase in rail/weld failures. All the AT welds would be required to be supported on wooden blocks and joggled fish plated. The patrolling in rail/weld failure area should be effected so ensure safety. Volume - I 63 The frequency of USFD testing should be increased with latest technology and should be done in the periodicity of 8GMT. still ambiguous, on the basis of AASHO Road Test for Road structures, it is assumed that; • With the increased loading, gauge face corner cracking will increase and will have grater impact on curves. Therefore, there will be need for carrying out ultrasonic testing of gauge face side at shorter intervals. 3.2 IMPACT ON BALLAST AND FORMATION The permissible contact pressure on the ballast bed is taken as 0.50 N/mm2. As seen from the values in the above table, for the present track structure, stresses on the ballast bed would be whining the permissible value when 30 tonne axle load rolling stock is introduced. Obviously, with introduction of 30 tonne axle load rolling stock, in most cases, formation stresses would exceed the bearing capacity of the formation. Blanketing is to be done for whish extra cost of Rs.10 lac/km will be involved. 3.3 While running 30 t axle loads, impact stresses on the ballast will increase and this further lead to crushing of ballast. Therefore, deep screening of ballast is required at close interval of time. 3.5 As per technical instructions no 4 (issued by Member Engineering Sh RR Jaruhar,) on Load carrying capacity of masonry Arch bridges, Test load application with observation or deflection, spread and residual deflection is more appropriate with limiting value deflection as 1.25 mm and spread as 0.38 mm as criterion. It is considered safe for all practical purpose to allow axle loads up to 30t with 60 Kmph speed with proper physical condition of arch being assured. Utmost, an additional ring can help fixed proper skewback on strengthened abutment / pier. Such bridges must be thoroughly examined. There will be higher wear and tear of switches and crossings of turnouts. Loosening of fittings, crushing of rubber pads would be there. IMPACT ON MAINTENANCE OF TRACK • • • Volume - I S.Hammarlund, B. Paulsson, conducted Swedish model for prediction of maintenance costs when increasing axle loads from 25 t to 30 t. The paper shows that by increasing the axle load from 25t to 30t, the deterioration mechanisms on the rails was surface fatigue (60%), engine burns etc (15%), internal defects found by USFD (10%), actual rail and weld fractures and isolating joints (10%). Only 5% of the rail maintenance was found to be due to wear on this part of line. The expected increase of track maintenance cost was 3%, which is substantially lower than the increase in axle load (20%). The possibility to reduce the maintenance cost was suggested by doing rail grinding and better system of lubrication Aasho test showed that, increase in the stress on the ballast bed due to increased axle loads would definitely result in faster deterioration of track geometry quality. Though the relation is 64 IMPACT ON BRIDGES It is required to be thoroughly checked the design of all the bridges in dedicated routes where 30 tonne axle load is to be introduced and make necessary design changes as per HMLS loading. Necessary speed restrictions should be imposed on safety considerations. IMPACT ON SEJ’S AND TURNOUTS SEJ’S will have higher impact loads resulting in pre mature failures. Design developed by Rahi Industries should be used for higher axle loads. 3.4 Decrease in track quality= (Increase in stresses on ballast bed) m Where the value of ‘m’ is generally taken between 3 & 4. Thus, a 10% higher stress on ballast bed leads to 1.2 to 1.5 times faster reduction in track geometry quality and consequent proportionate increase in maintenance. Introduction of 30 tonne axle load rolling stock against the present loading of 20.32t would therefore result in the increase of track maintenance effort 3 times. 3.6 EFFECT ON SCHEDULE OF INSPECTION With the increase in axle load to 30t the schedule of inspections of officials and supervisors needs to have re look. Definitely frequency inspections will be increased. 4.0 WHAT WILL BE REQUIRED - UPGRADATION OF THE TRACK, BRIDGES AND FORMATION, TO WITHSTAND THE INDUCED HIGHER STRESSES ON ACCOUNT OF HEAVY AXLE LOADS 4.1 RAIL SECTION The maximum bending stress in 6o Kg rail on introduction of 30 tonne Volume - I 65 axle load works out to be 27.70 Kg/mm2 for 90 UTS rails. Permissible Stresses are 72 UTS 90UTS 19.25 (LWR) 25.25 (LWR) It can be seen from the above that even in 60 kg Rail stresses are higher than the permissible limit, in other words 60 kg 90 UTS Rail also not fit for 30 t axle loads. depth of ballast cushion for 30t axle load is 25cm with sub ballast of 15cm for speed up to 100 Kmph. However a clean ballast cushion of 300mm may prove best solution for running 30 t axle loads. 4.4 Obviously, with introduction of 30 tonne axle load rolling stock, in most cases, formation stresses would exceed the bearing capacity of the formation. This would necessitate provision of a blanket layer of adequate thickness to improve the bearing capacity just beneath the ballast bed. Provision of blanketing, in accordance with the recent guidelines issued by RDSO in June 2003 vide Guideline No. GE: G-1, appears to be the only solution for stabilising weak formations. It is obvious that a yielding formation will result in rapid deterioration of track geometry, which will make it unsafe of higher axle load trains in addition necessitating increased and frequent maintenance efforts. The contact stresses for BOY, BOB and BOXNHA wagons would be as under. The diameter of wheel of Casnub bogie is taken as average of new wheel and worn out i.e. (1000+925)/2=962.5 mm. For 72 UTS rail the maximum allowable shear stress will work out to 21.60 Kg/mm2 and for 90 UTS rail, it will be 27Kg/mm2. It therefore implies that 90 UTS rail will be required for running 30 tonne axle load. The world’s longest rails are now manufactured in India for a length of 120m. With the availability of 120 m long rails, there will be drastic reduction of weld population in Indian rail tracks (from 160 welds per track km presently to 17) resulting enhance safety and cost reduction. 4.2 SLEEPERS As computed above with 60 kg rail on 60 kg sleepers at 43 cm spacing only fit for running of 30 Ton axle load rolling stock from contact pressure criterion on PSC sleeper. The 60 Kg Rail fails in bending stress criteria and 71 kg rail suits as next option. 71 kg rail is having foot width of 160 mm against availability of 162mm width of MS insert. This poses a very severe restriction, which would have far reaching implications and hence needs to be examined thoroughly before introduction of 30 t axle loads considering difficulty involved in maintaining track with such high density of sleepers i.e. 2326 sleepers/Km. 4.3 BALLAST BED The contact pressure on ballast for 60 Kg sleepers at spacing of 43 cm works out to be 0.1635 Kg/mm2. The permissible contact pressure on the ballast bed is taken as 0.50 N/mm2. For the present track structure, stresses on the ballast bed would be within the permissible value when 30 tonne axle load rolling stock is introduced. As per Railway Boards letter No.95/w1/Genl/0/39 dated: 9.10.96, the Volume - I 66 FORMATION 4.5 BRIDGES The bridges are required to of HMLS designs for carrying 30t axle load. 5.0 ALSO THINKING HAS TO GO IN FOR 5.1 MODIFICATION IN WAGONS Introduction of 3-axle bogie for increasing the pay load carrying capacity, while keeping the stresses on the track structure within permissible limits Increase in number of axle (3-axle bogie) is another strategy, which can be thought of for increasing the pay Higher tare loads of the wagons running on the Indian Railways as compared to those running on developed countries is one area where a possible solution to the problem of running increased axle loads lie. To cite an example of BOXN has the payload to tare ratio of 2.61. In most of heavy haul routes, this ratio varies from 3.5 to 5. TRANSWERK, South Africa has developed 104 t gondola tippler coal wagon with a 4.2 tare ratio and 120t tippler iron ore wagon with payload tare ratio. These options will be apparently more beneficial compared to the resource intensive up gradation of the track, bridges and formation, on account of introduction of heavy axle loads wagons. We have opted for Broad Gauge (5 feet 6 inches between rails against the standard gauge world wide of 4 feet 8.5 inches) yet our moving Volume - I 67 dimensions are highly restrictive. Similarly, productivity of our wagons in terms of tare to pay load ratio is probably one of the poorest in the world. We carry 450 kg of dead weight for moving every tonne of traffic as against 170 kg in developed countries. The wagons should be redesigned to increase the cubic content and the load carrying capacity to fall in line with the international norms. thickness as per RDSO’s specifications for which extra cost of Rs. 10 Lac/Km will be involved. For running 30t axle load in the existing track this will be most challenging job due to field difficulty in carrying out the work. 6.8 Each bridge in the dedicated corridor should be evaluated regarding safety vis-à-vis its physical condition and check for HMLS loading. Impose speed restriction, if necessary particularly in arch bridges. 6.9 The cost of the maintenance of the track with increase in axle loads from 20.32 t to 30 t is expected to increase by 3 times depending on the formation and track quality as per AASHO test. This is still ambiguous since Swiss model studies shows 3% increase in maintenance cost after doing rail grinding and lubrication. No tippling for unloading should be resorted to. Wagons should be equipped with end of train telemetry. Coupling height should be 851 mm instead of 1105 mm at present. Lowering of wheel diameter and coupling height substantially increases the volume available for the payload Bogie Mounted Electronic Brake System should be adopted. 6.10 Maintenance inputs are required to be increased. Rail grinding is to be carried out at predetermined intervals. Also lubrication must be carried out regularly. 6.0 CONCLUSIONS 6.1 To carry more freight, cost effectively, presently the Indian Railway has only option to increase loading per axle. 6.2 From the consideration of bending stresses and contact shear stresses, even 60 kg, 90 UTS rail is not able to sustain the increased stresses due to 30 tonne axle loads. 68.5 Kg AAR or 71 Kg UIC rails seems to be a realistic solution. 6.3 Contact pressure between rail and sleeper would be higher than the permissible value even on a track with 60 Kg rail on 60 Kg sleepers (PSC-6) at 60 cm spacing (1660 sleepers/Km). 6.4 It is found that track with 60 Kg rail on 60 Kg sleepers at 43 cm spacing (2326 sleepers/Km would only be fit for running of 30 tonne axle loading from the consideration of contact stresses. Else, PSC sleepers will require redesigning considering the difficulties involved in maintaining a track with such high sleeper density. 6.11 For running the higher axle loads in existing track, increased frequency of inspections, patrolling and USFD testing are required. Normal USFD will be required to be done at periodicity of less than 8 GMT and for gauge face testing to detect corner cracking. 6.12 As an alternative strategy, use of wagons with high payload to tare ratio and increased number of axles may also be considered. Wagon dimensions must be changed with reduced wheel diameter. Signalling system requires to be upgraded. 7.0 SUGGESTIONS & RECOMMENDATIONS (i) Head hardened rails should be used. And use of 120m long panels rolled by steel plants. (ii) Extra ballast profile should be given at curves. 300 mm clean ballast cushion would be required for running 30 tonne axle load and the deep screening and temping needs to be done at closer intervals. (iii) Lubrication of gauge face should be done at closed intervals. (iv) Provision of a blanket layer of adequate thickness to improve the bearing capacity just beneath the ballast bed. 6.6 Therefore, the track structure for new track should be 68.5 kg. AAR or 71 kg. UIC rails laid on prestressed concrete sleepers, 2326 sleepers per km. with minimum 300mm of clean ballast cushion. (v) Codal provisions of tolerances of flat wheel require to be changed. Monitoring of flat wheels should be done closely and en route detachment of wagons with flat wheels should be done. 6.7 With the introduction of 30 Tonne axle load, in most of the cases, formation stresses would exceed the bearing capacity of the formation. This would necessitate provision of a blanket layer of adequate 6.5 Volume - I 68 The US studies revealed that the defect size more than 15% have direct implication due to wheel flat and failure rate is more. The USFD should be carried out within the periodicity of 8GMT Volume - I 69 so that defects of sizes more than 15% shall be detected in time. (vi) Strict tolerances for the track parameters should be kept. Maintenance inputs are required to be increased. Use of superior materials to increase the life cycle should be used. Mechanised maintenance should be adopted. Deep screening and temping should be done at closer interval than existing provisions. (vii) Rail Grinding A better solution to increase the life of the rails on Heavy Haul Routes is rail grinding. Such grinding will remove the plastic deformation on railhead thereby removing the surface cracks before they propagate further into rail section. It also helps in progressively lowering the point of maximum shear stresses thereby increasing the life of the rail and prevention of sub surface cracks due to fatigue. (viii) Thick web switches should be used in the turnout taking out of curve. CMS crossing should be used instead of built up crossing. Maintenance of fittings should be of highest order. (ix) Frequency of Ultrasonic Testing of be increased especially gauge face side testing. Ultrasonic testing of welds. (x) Handling & care of welds. (xi) Special attention to Points & Crossings & SEJs. (xii) Keeping bridges under continuous watch. (xiii) Modification in wagons – Decreasing tare load and improved suspension system. Volume - I 70 8.0 REFERENCES 1. Esveld Coenraad, Modern Railway Tracks. 2. RDSO Guideline No. G-1, Guidelines For Earthwork in Railway Projects. 3. RDSO Civil Engg. Report No. C.55, Investigations on Determination of Intensity of Pressure on Railway Formations- BG Tracks. 4. Eisenmann Dr. Ing. J., Vehicle- Track Panel- Ballast Stresses. 5. Harvey A.F., Sleeper Spacing and its effects on the Maximum Permissible Axle Load. 6. Mohan M.S., Track Structure on heavy Mineral Lines. Volume - I 71 7. International Heavy Haul Association, may 2001- Guidelines to Best Practices for Heavy Haul Operations: wheel and Rail Interface Issues. 8. Tech. Paper No. 245: Report of Bridge Sub-Committee on track stresses, 1925. 9. Tech. Paper no.323- Stress in Rly track, by Venkatramaya, 1950 10. Tech. Monograph No. 12: Track loading fundamental by C. W. Clarke,1959. 11. Civil Engg. Report No. C-100: Dynamic augments of track loads, 1971. 12. 53rd TSC Report, 1977, Item No. 716. 13. IHHA-1997, Strategy beyond 2000. Sixth International Heavy Haul Railway Conference Cape Town, south Africa,Pg.1155-1158. 14 . STRATEGIES FOR MEETING TRANSPORT DEMAND Role of Railways by Ashutosh Banerji for RAILWATCH 15. Technical Report No.4, on load carrying capacity of masonry arch bridges, By Sh. R.R. Jaruhar, Member Engg. Rly Board. CHECKING THE SUITABILITY OF EXISTING ROLLER-ROCKER BEARINGS FOR HIGHER AXLE LOAD RAVINDRA KR. GOEL*, HARI OM NARAYANA** SUJEET NATH GUPTA*** 1.0 INTRODUCTION Conventional roller and rocker bearings are provided in open web girder bridges since long. These bearings have been designed as per the loading standard (mostly BGML or RBG) prevailing at the time of construction of the bridges. As the traffic volume is increasing and extra revenue is proposed to be generated, a policy decision has already been taken to make some of the routes fit for running 25t axle load. The existing bridges are required to be checked for carrying this increased axle load, on restricted speed, if required. The implications of higher axle load are also to be examined on existing bearings with respect to original design parameters of BGML/RBG loading. It has been seen that zonal railways are not having complete awareness about the various design parameters involved in design of rockerroller bearings and replacement of bearings are proposed without exercising adequate design check. The paper describes in detail the relevant provisions for analysis of such bearings and comments on the adequacy of existing bearings of standard spans. 2. LOADS AND LOAD COMBINATIONS 2.1 LOADS The loads to be considered for design/analysis of bearings are to be taken from Clause 2.1 of Bridge Rules. The following loads are to be invariably considered. It is seen that the lateral horizontal forces due to wind (not due to Earthquake) are governing in steel girders. Therefore, the effect of wind load is to be considered unless some special conditions are found prevailing at the site of bridges. a) Volume - I 72 Volume - I Dead Load * Director, B&S, RDSO, Lucknow-226001 ** Sr. Section Engineer/Design, RDSO, Lucknow-226001 *** Section Engineer/Design, RDSO, Lucknow-226001 b) Live load including raking force c) Impact load d) Longitudinal load e) Wind load f) Forces due to curvature and eccentricity a) DEAD LOAD “In case of bridges having open deck provided with through welded rails, rail-free fastenings and adequate anchorage of welded rails on approaches (by providing adequate density of sleepers, ballast cushion and its consolidation etc., but without any switch expansion joints) the dispersion of longitudinal force through track, away from the loaded length, may be allowed to the extent of 25% of the magnitude of longitudinal force and subject to a minimum of 16t for BG and 12t for MMG or MGML and 10t for MGBL. This shall also apply to bridges having open deck with jointed track with rail-free fastenings or ballasted deck, however without any switch expansion or mitred joints in either case. Where suitably designed elastomeric bearings are provided the aforesaid dispersion may be increased to 35% of the magnitude of longitudinal force.” Dead load can be taken from the design sheets or drawings of the existing bridge to be checked. In the absence of reliable documents the dead load has to be estimated on the basis of detailed survey of the different bridge components and connections. b) LIVE LOAD At present it is customary to use the concept of EUDL for working out the live load effect on bearing. EDUL for shear for the particular span (effective) is to be considered for calculating live load on each bearing. Annexure 2 of RDSO letter No.CBS/ Golden/Q/Strength dated 8-8-06 can be referred for the purpose. This EUDL table has been prepared for 2WAG-9/2WDG4 with BOXN wagons of 25t axle load. c) As per this clause the dispersion in longitudinal force (tractive effort or braking force) may be permitted upto 25% by ensuring adequate anchorages of welded rails on approaches. Longitudinal force can be taken from Annexure 5 of RDSO letter No.CBS/Golden/Q/Strength dated 08-08-06 as the maximum value of tractive effort or braking force for effective span under consideration. These values are again for 2WAG-9/2 WDG4 loco with BOXN wagons of 25t axle load. IMPACT LOAD For working out the impact load impact factor CDA can be derived from Clause 2.4.1 of Bridge Rules. Impact factor can be reduced for reduced speed. Proportionately if the bearings are to be checked for operation of traffic at restricted speed as per the following expression. d) Note: Length of approach for the above purpose shall be taken as minimum 30m e) Wind load is to be taken from the existing design sheets which is generally for a wind load intensity of 150 kg/m2. As per A&C Slip No.34 of Bridge Rule, the wind load shall not be considered acting alongwith live load if the wind load intensity is more than 100 kg/m2. Therefore, the wind load for analysis of existing bridges for permitting higher axle load shall have to be modified accordingly. Wind load has following two effects: LONGITUDINAL FORCE It is an important design parameter, considerably affecting the design of bearings. The longitudinal force actually applied at the bearing depends upon the maximum tractive effort of loco running over the route, maximum braking force applied by the train, dispersion of longitudinal force through the girder and track. Clause 2.8.3.2 of Bridge Rules (as per A&C Slip No.22) has to be referred in this regard which is reproduced below for ready reference. Volume - I 74 WIND LOAD Volume - I i) To apply overturning effect on bearings causing increased vertical load on leeward side bearings. ii) To apply a lateral force on bridge simultaneously with longitudinal force. Resultant of the two forces is to be resisted in shear by saddle bolts and anchor bolts. 75 f) Cement Concrete: FORCES DUE TO CURVATURE OR ECCENTRICITY OF TRACK Usually, the girder bridges are on straight. However, whenever they are on curve the extra load on one girder should be calculated as per Clause 2.5.3 (a) of Bridge Rules and the extra horizontal load due to centrifugal force should be taken as per Clause 2.5.3 (b) of Bridge Rules. Eccentricity effects are usually not considered in girder bridges unless they are large enough. 2.2 As laid down for permissible bearing pressure in Plain concrete in Table III and III(a) of the IRS Concrete Bridge Code-1962. Reinforced Concrete: As laid down for permissible stress in direct compression for the specified crushing strength at 28 days for ordinary Portland cement (or the equivalent period of time for other cement) given in Table III and III(a) of IRS Concrete Bridge Code-1962 LOAD COMBINATIONS Following load combinations are to be considered: a) The above-mentioned limits may be exceeded by 331/3 per cent for combinations mentioned in clauses 3.2.2 and 3.2.3 For checking bed plates and bearing pressure on bed blocks iii) DL + LL + Impact The centre of pressure under flat bearing plates attached to the girders shall be assumed to be at one-third of the length from the front edge.” iv) DL + LL + Impact + Overturning effect due to wind (Permissible stresses are increased by 16.67%) v) b) DL + LL + Impact + Overturning effect due to wind & longitudinal force. (Permissible stresses are increased by 33.3%) i) ASSESSMENT OF EXISTING BEARING 3.1 CHECK FOR ADEQUACY OF MAXIMUM BED PLATE PRESSURE This check is conducted to ensure that the concrete/bed stone is safe against crushing under the vertical loads. The maximum pressure calculated should be with in the allowable pressure which can be taken from Clause 3.16 of Steel Bridge Code. This clause is reproduced below for ready reference “Allowable Working Pressure under Bearings or Bed Plates The area of bearings or bed plates shall be so proportioned that when the eccentricity of loads due to combination mentioned in Clause 3.2.1 the maximum pressure on material forming the bed shall not exceed the following limits: Granite … Sand Stone… 36 kg/cm2 MAXIMUM PRESSURE ON CONCRETE Maximum pressure on concrete can be calculated from the formula: Resultant horizontal force comprising longitudinal force and wind load. (Permissiblestresses are not increased in this combination) 3.0 Volume - I 3.2 For checking of anchor bolts and saddle bolts in horizontal shear Where V 3.3 = Vertical load with overturning effect of wind La = Lever arm, in the height from the center of knuckle to the bottom of base plate or top of pier/abutment L = Length of base plate B = Width of base plate LF = Maximum longitudinal force WL = Wind load CHECK FOR ANCHOR BOLTS Anchor bolts are provided at the base plate to restrain the bed plate against any horizontal movement in X & Y direction. These bolts are checked basically to resist resultant force comprising of longitudinal force and lateral force due to wind effect. The shear stress on each bolt can be calculated by the following formula: (33 tons/ft2) 29.5 kg/cm2 (27 tons/ft2) 76 Volume - I 77 Where 3.4 4.0 n = No. of anchor bolts provided d = Diameter of anchor bolts CHECK FOR SADDLE BOLTS These bolts are provided to connect the bottom chord of open web girders with the saddle plate. These bolts provide an interface for the transfer of horizontal force in X, Y direction. These bolts are checked for shear stress which is equal to ADEQUACY OF BEARINGS OF STANDARD SPANS The adequacy of existing roller-rocker bearings of standard open web girder span of BGML & RBG loadings have been checked by following above provisions and the results are shown in Table-1 (a) & (b). The analysis has been done by taking into consideration wind load of 100 kg/m2 alongwith live load as per A&C Slip No.34 of Bridge Rules with and without 25% dispersion of longitudinal forces as per clause 2.8.3.2 amended vide A&C Slip No.22 of Bridge Rules. The adequate anchorage of welded rails on approaches has to be ensured accordingly. The checking of these bearings has been done keeping in view, the speed for which the standard span has been cleared by RDSO vide letter No.CBS/Golden/Q/Strength dated 08-08-06. Where 3.5 n = No. of bolts d = Dia of saddle bolt RELAXATION OF PERMISSIBLE STRESSES Permissible stresses for retaining existing bridges can be relaxed as per clause 3.20.2 of Steel Bridge Code which is reproduced below: “Mild Steel, Wrought Iron and Early Steel Girders Bridge spans other than open web girder spans may, if they are kept under regular observation by the Bridge Engineer and his staff, be retained in use, provided that if the impact effect-specified in clause 3 of the Bridge Rules (Revised 1964) for the maximum permissible speed over the bridges is allowed for the calculated stresses for various combinations of loads as laid down in relevant clauses do not exceed the working stresses specified for those combinations by more than 11 percent. Under the same conditions, permissible shear and bearing stresses on rivets may be increased by 25 per cent. This increase in rivet stresses shall not be allowed if the stresses are calculated by the method given in APPENDIX- E. Under the conditions specified above, open web girder spans may be retained in use, provided that the calculated tensile and compressive stresses do not exceed the specified working stresses by more than 5 per cent. The permissible shear and bearing stresses on rivets may be increased by 10 per cent.” Volume - I 78 Volume - I 79 5.0 CONCLUSION 5.1 It is seen that the design of existing roller-rocker bearings is governed significantly by longitudinal forces. Forces on bearing are considerably reduced by allowing dispersion of longitudinal forces. Railways may thus ensure proper anchorage of track on approaches to take advantage of clause 2.8.3.2 of Bridge rules in this respect. 5.2 The existing bearings can be safely retained by changing saddle bolts with higher class (Property Class 6.6 or 8.8) and providing extra anchor bolts as found necessary. 5.3 As the longitudinal forces are to be resisted at the fixed end only, no action is required to be taken at roller end, provided the physical condition is other satisfactory. Volume - I 80 6.0 REFERENCES 6.1 IRS Steel Bridge Code (1962), Research Designs & Standards Organisation, Ministry of Railways, Lucknow (U.P.). 6.2 IRS Bridge Rules (1964), Research Designs & Standards Organisation, Ministry of Railways, Lucknow (U.P.). 6.3 IRS Drawings and Design Documents of relevant Standard Spans, Research Designs & Standards Organisation, Ministry of Railways, Lucknow (U.P.). 6.4 RDSO letter CBS/Golden/Q/Strength dated 08-08-2006 with Annexures. Volume - I 81 In this paper, it has been tried to build the confidence by sharing the statistics of arch bridges of other railways, sharing the test results of arch bridges done in the past and arriving at the viable conclusion. STRENGTH POTENTIAL OF ARCH BRIDGES FOR HIGHER AXLE LOAD RAMA KANT GUPTA* Most of the Arch Bridges on Indian Railways are of initial construction i.e. having life more than 100 years. These Arch Bridges were designed for loading standard of at that time, which was much lighter than the present day loading. While allowing the higher axle load, particularly keeping in view its age as well, field engineers are in doubt about its strength particularly when elastic theory of Arch Bridge analysis reveals that enormous overstressing is coming in case of allowing the higher axle load. In this paper, firstly, it has been tried to build the confidence by giving example of other World Railways followed by Load Testing experience of Indian Railways. Subsequently, it has been tried to brief the modalities for allowing the higher axle load on arch bridges. 2. POPULATION OF ARCH BRIDGES ON INDIAN RAILWAYS Arch Bridges still constitute a good percentage of the total bridge stock of Indian Railways. Bridge statistics of Indian Railways based on old zones is given in Table No-1 * This paper has been written based on the past experience of the author as Executive Director, Bridge & Structures at RDSO, Lucknow 1. INTRODUCTION Arch Bridges are existing all over the world since ages. It is one of the oldest bridge configurations known to the civilization. Durability-wise, no other bridge configuration is at par with the arch bridge configuration. This is so since arch is the configuration that remains under compression and as far as compression members are concerned, they are having much more life than the tension members/members subjected to compression and tension. Quite an appreciable number of bridges on Indian Railways are still arch bridges. They are mostly 100 years and above of age. Those arch bridges were also designed for loading standard of at that time, which was much lighter than the present day loading standard. Whenever, question of running of higher axle load comes, first doubt goes towards safety of such arch bridges. Actually, arch bridges are not so week as we mostly perceive. Main issue is about having fair idea of the strength potential of arch bridges. Volume - I General Manager/Bridges, IRCON, New Delhi – 110066. From the above table, we see that Central & Eastern Railways (Based on its old jurisdiction) are having more %age of arch bridge than the average figure of 17.5% of Indian Railways. Other Railways are having arch bridge population almost in the same ratio as that of Indian Railways. 3. ARCH BRIDGE STATISTICS OF SOME OF THE EUROPEAN RAILWAYS Arch Bridges are not only predominantly available on India Railways, but the same are available even on higher percentage on other world railways. Statistics of some world railways mostly spreading over Europe is available. The same is given herewith: 3.1 BRIDGE POPULATION From Table No.-2, it is shown that populations of arch bridges on European Railways are more than the average arch bridge population of 17.5% of Indian Railways. Volume - I 83 replacement. Codal life is just for guidance and cannot be taken as replacement criteria. 3.3 CONDITION OF ARCH BRIDGES OF EUROPEAN RAILWAY (Based on details of SNCF, RFI, NR, REFER, DB, RENFE & CD Railways) The above figures reveal that majority of the arch bridges are either in good or medium condition. Only small percentages of bridges are having poor to very poor condition. Almost similar situation is with the arch bridges of Indian Railways as well. 3.2 From the above figures, we further conclude that as far as condition and age profile of arch bridges of European Railways and Indian Railways is concerned, the same are almost matching. However, populations of arch bridges on European Railways are higher than that of Indian Railways. Only one major difference is there that European Railways are mostly catering for the passenger trains while Indian Railways is catering for passenger as well as goods trains both. AGE PROFILE OF THE ARCH BRIDGES OF EUROPEAN RAILWAYS (Based on details of SNCF, RFI, NR, REFER, DB, RENFE & CD Railways) Table No. 3 It is further worthwhile to point out that European Railways are not thinking of replacing those old arch bridges. Rather, they had initiated one project under the umbrella of UIC to device rational way of strength assessment of arch bridges including effective way of its maintenance. Fortunately, Indian Railways is also member of that Arch Bridge Working Group Committee of UIC The above figures reveal that major chunk of arch bridge population on European Railways are in between the age group of 50 to150 years. Even about 12% of bridges are having age more than 150 years. Position of Indian Railways is not so bad as far as percentage of bridge population having life in between 100-150 years is concerned. World over, only age is not the criteria for replacement of a bridge. Age cum condition is the rational criteria for replacement of a bridge. If a bridge even being old and completed the codal life, but is in sound condition and structurally safe, there is no need of its Volume - I 84 4. INITIATIONS TAKEN FOR RETENTION OF ARCH BRIDGES EVEN DESIGNED FOR LIGHTER AXLE LOAD As explained above, in Europe as well, arch bridges represent even good percentage of total bridge population. Regarding retention of such old arch bridges even designed for lighter axle load, first comprehensive step was taken in UK. On behalf of the Ministry of Transport and the Ministry of Supply, Building Research Station (BRS) of UK had conducted series of Volume - I 85 tests on various types of spans ranging from 14 ft to 54 ft and the conclusion is summarised as below: b. (i) Dispersion of load through the fills over the arch ring. “For a single span of arch bridge up to 45 ft when tested with a single axle load of 20 tons, the deflection at crown should not be more than 0.05 inch (1.27 mm) and change in span (spread) should not be more than 0.015 inch (0.38 mm). It was further assumed that if these values are not exceeded, a load of at least 40 tons could be carried, the load being either the weight of a tracked vehicle or the load from the rear bogie of a wheeled vehicle. If the deflection is more than 0.05 inch and the change in span more than 0.015 inch, the arch is considered to be suspected and requires further investigation.” Perhaps, it was the first research work of its kind. Series of research works were followed after that. In this process, Chettoe and Henderson conducted further study on arch Bridges with bogie load arrangement for varying loads from 20 tons to 90 tons. Based on the results, they came to a conclusion that with a better load distribution for 2 axles, the deflection due to 40 tons bogie is only 50% more than the single axle load of 20 tons. Although, BRS criteria state that the change in span (spread) should not be more than 0.015 inch, Chettoe and Henderson, after the series of tests came to conclusion that the limit can be relaxed. The relevant Para of their report is reproduced below: (ii) Strength contribution of the fill, the parapet and the spandrel walls. (iii) Ratio of the load causing initial crack to the intrados of the arch and the load causing ultimate failure. (iv) The mechanism and behaviour of arch under test at ultimate failure. (v) Effect of abutment movement etc. Large numbers of arch bridges were tested and reports published. References of some reports have been given in the Bibliography. 6. TESTING OF ARCH BRIDGES PERFORMED IN INDIA In India as well, large number of arch bridges were tested as a part of research work. Among such testing work, RDSO, Lucknow and Highway Research Station, Chennai did the pioneering work. Before start of the testing work, objectives were defined by RDSO and the same are given on next page: a. Volume - I To study how far the results of British studies could be applied to the Railway Bridges in India. 86 SUMMARY OF TEST RESULT OF ARCH BRIDGES OF THE RELEASED GANDAK VIADUCT OF EC RAILWAY Before coming on the test result part, let us share the salient features of the arch nos. 71 & 73 (both identical) of Gandak Viaduct, as given below: “It is not considered that changes in span are likely to be cumulative in sound arches having adequate backing and sound filling. The amount of movement to be expected might be estimated by comparison with these and BRS test results. In order to allow for the unpredictable nature of this movement, a value of, say 0.02 inch to 0.04 inch might be assumed for single span arches of moderate span, having good backing, although there are cases where it might be more. For bridges with weak backing, a considerable increase in abutment movement and reduction in strength is to be expected”. 5. To get further information on: 6.1 (i) Span : 15’-0” (Semi circular) (ii) Thickness of arch ring : (iii) Depth of cushion : (iv) Barrel length : 14’-0” (v) Fill : Earth impregnated with stone ballast and brick bats. (vi) Haunch fill : Lime concrete. (vii) Masonry : Brick in lime mortar. (viii) Pier widt : 3’-9” (ix) Pier height : 8’-0” above ground level. (x) Pier length : 14’-0” (xi) Foundation soil : Alluvial sand. 1’-10½” Uniform. 3’-0” from the extrados at crown up to the bottom of sleeper. FIELD OBSERVATIONS Detailed observations were taken and reports, reference of some of them is given in the bibliography were published. From convenience point of view, only summarised conclusions and readings in Table No.-5 is given herewith. Volume - I 87 CONCLUSIONS a. The load deflection curves are regular up to 100 tons for crown loading with the fill, spandrel walls and parapets in position and b. The load deflection curves are regular up to 70 tons for crown loading when fill, spandrel walls and parapets removed. Loads for limiting values of deflection and spread as laid down in BRS criteria and the maximum load achieved while testing 6.2 DISCUSSIONS ABOUT FIELD-TEST RESULTS Thorough review of the field test results were made and observations are given as below: 6.2.1 COMMENT ABOUT BRS CRITERIA These tests show that the limiting values of deflection and spread are reached at loads much higher than the test load specified in the BRS criterion. It was further seen that the loads causing limiting deflection (0.05 inch) and spread (0.015 inch) are well within the elastic limit of the arch. 6.2.2 CONTRIBUTION OF AXLE LENGTH, PARAPET WALL, FILL AND SPANDREL WALL The following conclusions were drawn: (i) BG axle is giving more safe load on the same arch than the MG axle (ii) Fill, spandrel wall and parapet wall add strength to the arch Conclusion No.-(i) gives more confidence about those MG standard arch bridges whose fitness for BG is required 6.2.3 OBSERVATION OF DEFLECTION ON THE ADJOINING ARCH WHEN THE MIDDLE ARCH WAS LOADED Observations reveal that the load on the arch under test gets transferred to the adjoining arches due to the possible continuity effect offered by the presence of the lime concrete haunch filling. This continuity effect also makes it difficult to work out theoretical deflections. Due to the existence of the unpredictable phenomenon of the increase in span of the bridge and the part played by the spandrel walls, haunch filling and fill in strengthening the arch, there are considerable errors in theoretical calculation of the deflections of the vaults under different conditions of loading. Conclusions drawn Volume - I 88 Volume - I 89 are as under: (iv) (i) The deflection is fairly proportional to load. (ii) The deflection in both the adjoining vaults are fairly equal. Thus, it is observed that continuity effect is present in the arches under the load test. Parapets, earth fill, spandrel walls and haunch fill all removed i.e., loading on the bare arch. Under the second series i.e., loading beyond the elastic limit and up to the failure of the arch, loading was done on bare arch. The details and sequence of tests are given in Table No. 6 TESTING OF ARCH NO. 76 OF THE SAME VIADUCT FOR FURTHER CONFIRMATION AND GETTING SOME ADDITIONAL INFORMATION Deflections of the arch at crown and spreads of piers at the springing level were measured with the load at crown. Arch No. 76 is having the same details as that of arch nos. 71 & 73. This was tested with the aim to get confirmation of the earlier test results as well as to get additional information. Test conducted on this bridge span is given as below: Cracks were first observed on fresh dabs of plaster of Paris at the intrados of the crown at 40 tons of load when the maximum deflection was 0.056 inch. As the load increased, cracks became wider and deeper. At a load of 60 tons, cracks at the crown widened and became clearly visible extending over the entire length of the barrel. At this load, crown deflection was 0.0920 inch. At a load of 90 tons, cracks were observed to a depth of 4½ inch on the extrados, near both the quarter points of the arch. Cracks on the intrados at crown widened further. At this load, crown deflection was 0.1620 inch. At a load of 100 tons, another cracks at extrados were noticed. Crack at crown intrados widened further. At this load, crown deflection was 0.1980 inch. At a load of 120 tons, cracks on the extrados near quarter points developed fully extending to the full depth of the arch ring. At this load, deflection at crown was 0.2320 inch. When load was increased to 130 tons, it was observed that - 7. (i) Loading on crown up to elastic limit of arch. (ii) Loading beyond elastic limit and up to the failure of the arch. Under the first series i.e., loading of the arch up to the elastic limit, tests were conducted under four different conditions of the arch constituents as given below: (i) With parapets, earth fill over the arch ring, spandrel walls and haunch fill in position. (ii) Parapets removed, while earth all over the arch, spandrel walls and haunch fill in position. (iii) Parapets, earth fill over the arch and spandrel walls, all removed i.e., only haunch fill (of lime concrete) in position and 7.1 OBSERVATIONS RECORDED DURING THE TEST (i) There was a splitting sound. (ii) The cracks at the extrados at near quarter points widened considerably. (iii) The cracks at the intrados at crown tended to close. (iv) The crown arch voussoir showed crushing. (v) The load could not be maintained at this stage as it started dropping. All dial gauges were removed which took about five minutes and hence, no proper deflection readings could be obtained. Efforts were made to maintain the load at 130 tons by pumping oil into the jacks, but the arch failed. Masonry between the cracks near quarter points fell down. A clear mortar face was seen near the quarter point. From the graphs of load deflection curves plotted for bare arch, it was observed that the curve is linear up to 40 tons. At this load, first crack Volume - I 90 Volume - I 91 7.2 was observed. From 40 to 80 tons, even though the deflection curve is linear, the rate of deflection is higher. As already stated, bare arch completely failed at 130 tons. Thus, the ratio between the load causing final failure and the load causing the first crack is 3.25 to 1. (iv) According to BRS criteria, if the deflection exceeds 0.05 inch under a test load of 20 tons single axle, the bridge is suspect. This limit was reached at a central load of 84 tons. Contribution of haunch in strength addition = 10 tons Contribution of fill and spandrel wall in strength addition = 30 tons SPREAD OF ARCH Safe load as per deflection = 80 tons Safe load as per spread = 49 tons Overall safe load = 49 tons If we follow the MEXE Method or Survey & Tabulation Method or go by Elastic Theory of Analysis, this much safe load will not come. Although, the following observations are not part of the report but can be inferred from Table No.6: Spread of the arch had been recorded by three sets of three dial gauges A1 – E1, A2 - E2 and A3 – E3. Generally, spread increases with the load. Movements of both the piers were not equal. Furthermore, it was also observed that increase in span was not uniform along the length of the barrel. According to BRS criteria, if the spread exceeds 0.015 inch under a test load of 20 tons single axle, the bridge is considered to be suspected. The loads at which this limit was reached for the three sets of dial gauges are given below: Sets of dial gauges Load in tons A1 – E1 at one ft from the end of barrel 52.50 A2 - E2 at the centre of the barrel 50.00 A3 – E3 at one ft from the other end of the barrel 45.00 In a multi-span arch bridge, loading on one span affects adjacent spans. 8. SERIES TEST OF SINGLE SPAN OF 20 FT ARCH BRIDGES For better confirmation as well as to ascertain about so many additional parameters, many single span 20 ft arch bridges were tested by RDSO, details of which is given in Table No-7 Thus, average axle load producing 0.015 inch of spread = 49.2 tons 7.3 EFFECT ON ADJOINING ARCHES Loading on test span affects adjacent arches. Deflections of the adjoining arches were measured. Conclusions drawn are given as below: 7.4 CONCLUSIONS The conclusions drawn are as under: (i) The arch behaves elastically up to certain limits as discussed above. (ii) The limits of deflection (0.05 inch) and spread (0.015 inch) as per BRS criteria are reached at loads in excess of BRS test load viz, 20 tons single axle load. (iii) Volume - I The various arch constituents viz. parapets, earth fill, spandrel walls and haunch fill contribute to the strength of the arch. 92 8.1 CONCLUSIONS DRAWN AFTER THE TESTS The conclusions drawn after series of the aforesaid tests are given as below: (i) Volume - I Load versus deflection graph is generally linear for loads up to 80 tons indicating that the arches behave elastically. 93 (ii) Spread and deflection readings in the arch bridges are very small. Even under a load of 80 tons (four times the load specified according to the BRS Criteria), deflections and spreads are smaller than those specified in the BRS criteria. Conclusions drawn after conducting the dynamic tests are given as below: (i) Tests indicate that natural frequency of vibration of the arch ring is very large (30 to 43 cps) as compared to frequency of excitation forces provided by the rolling stock (1 to 6 cps). Due to this reason, resonance cannot take place on arch bridges. (ii) There is no residual strain or deflection after passage of normal working live load indicating that the arch behave as an elastic structure. (iii) Lateral dispersion through the fill may be assumed to be at an angle of 45° (iv) Deflection observations were found to be vitiated due to temperature effects. (v) Spread observations were found to be vitiated due to temperature and wind effects. (vi) Strain gauge readings as recorded could not be used for assessing the strength of arch bridges. 10. (vii) Subsequent tests have shown that allowances should be made for temperature effect and wind effect. The wind effect may be eliminated by making the trestles rigid and closing the barrel of arch by tarpaulin. Allowances for temperature effect will have to be carefully assessed by experiment. From the above test result, it is seen that BRS criteria of UK is giving much conservative strength potential of arch bridges than the actual one. Based on the European studies, UIC has published UIC code 778-3 R .It dictates about the MEXE method developed in UK. This method omits many factors that add to the strength of the arch bridges but mostly, not taken into design consideration. After the series of tests conducted by RDSO glimpse of which has been given above, RDSO had concluded the following factors which are enhancing the strength of arch bridges: 9. DYNAMIC LOAD TEST OF ARCH BRIDGES Till now, testing of arch bridges for static load up to the elastic limit as well as up to the collapse load under various conditions have been discussed and relevant conclusions have been drawn after field observations. A necessity was felt about dynamic test as well. In this category, RDSO had tested bridge nos. 41C and 42B of Kota- Bina Section and published its reports as Civil Engineering Report Nos. C-75 and C-76 respectively. Brief description of the bridges is given in Table No-8 COMPARISON OF THE WORK DONE BY RDSO AS COMPARED TO OTHER WORLD RAILWAYS (i) Spandrel wall (ii) Parapet wall (iii) Fills above the arch (iv) Haunches provided on the arch (v) Availability of track in Railway Bridges and Metal Surface on road bridges, etc Accordingly, contributions of aforesaid factors were finalized. Based on that, “SURVEY AND TABULATION METHOD” was evolved for arch bridge strength assessment by Dr. S. R. Agrawal under leadership of whom, lot of testing works were performed. (Dr. S. R. Agrawal is an IRSE officer retired as Additional Director General, RDSO, Lucknow) This method was circulated to all the Zonal Railways for strength assessment of arch bridges. Subsequently, RDSO desired to legalise this procedure after discussion through BRIDGE & STRUCTURES STANDARD COMMITTEE meeting. Unfortunately, on account of non-submission of results by some of the nominated railways, item lingered on and then after dropped from the BSC agenda. Author of this article is of the strong opinion that this method should Volume - I 94 Volume - I 95 invariably be used for strength assessment of arch bridges, which gives better result nearer to the load testing result. As a member of UIC Arch Bridge Working Group, author of this paper presented the testing results during the 6Th Working Group Meeting on Arch Bridges of UIC held at Vienna in October 2004. Working Group Members appreciated a lot regarding the work done by RDSO under the umbrella of Indian Railway. 11. IRS CODAL PROVISION FOR LOAD TESTING OF ARCH BRIDGES Provision of load test exists in Arch Bridge Code of the Indian Railway, relevant extracts of the same is reproduced below: 12. SHARING OF SOME EXPERIENCE AFTER LOAD TESTING OF ARCH BRIDGES After experience gained thorough field investigations, large numbers of bridges were made safe which were found to be having overstressing even in between 400 to 500 per cent. Even with such a high extent of over stressing arrived after doing the theoretical calculation based on elastic method, some of such bridges were load tested and all of them were found safe. Details of some of the bridges of Mughalsarai-Allahabad section along with its theoretical stress and load test results is given in Table No-9: (Tests were performed during 1979-80) The criteria for arriving at the safe load shall be: (i) Under the proposed load, the crown deflection and spread do not exceed 1.25mm and 0.4mm respectively. (ii) There is no residual deflection or spread after release of load; and (iii) There is no crack appearing on the intrados of bridge. Note: (i) The above criteria will be applicable to segmental and nonsegmental arches of span 4.5m to 15.m provided and span/ rise ratio lies between 2 and 5. (ii) The load test shall be conducted on distressed bridge only after completepressure grouting of the masonry. Thus, we see that theoretical analysis is highly inadequate in strength assessment of arch bridges. Actually, load testing is the practical way of strength assessment and it should be followed invariably in case of any doubt as well as a confidence building measures. 13**. FINALIZATION OF LOAD TEST CRITERIA For determination of the test load, influence line diagram for causing the maximum bending moment at crown is to be prepared. The maximum bending moment caused by the type of loading permitted or proposed to be run on the arch bridges on the section is to be determined based on Clause 5.1.2 of Arch Bridge code. The bogie load required to cause this bending moment should be worked out from the influence line diagram. The arch FILE PHOTO OF ARCH BRIDGE UNDER TESTING WITH THE HELP OF SIX NOS. OF STEAM LOCOMOTIVES Volume - I 96 Volume - I 97 should be tested with this bogie load and if the deflection of the arch is less than 0.05 inch and the change in span is less than 0.015 inch, then arch is to be considered safe to carry the type of loading, which is to be permitted on the arch. However, as railway arches are to be tested with bogie loads greater than 40 tons the change in span that occurs in some cases would be more than 0.015 inch and, therefore, this limit may be relaxed to 0.03 inch. Main factors to be considered when the limit is required to be relaxed are(i) Condition of masonry and its behaviour under test load. CONCLUSION (ii) Type of foundations and nature of soil under which it is founded. (iii) Recovery in side sway after test load, which should be complete. Arch bridges are having much more strength potential than its theoretical value. If theoretical analysis reveals that the arch bridge is unsafe for proposed axle load, then Survey And Tabulation method should be followed. Final confirmation can be obtained by load testing, if required ** It is as per remarks given in RDSO report. 14. arch bridges strengthened by CINTEC method and tested in TRL laboratory of UK. As ED/B&S, RDSO and Member or UIC Arch Bridge Working Group, author of this paper analysed the TRL report and shared with other Working Group Members of UIC. Based on comments of RDSO as well as other observations of Working Group Members, such method was not approved. For kind information of the readers, comment of RDSO is given in the annexure. (Selected paras is only given on account of space constraints) COMMENTS ABOUT CINTEC, HELIFIX AND OTHER SIMILAR METHOD OF ARCH BRIDGE STRENGTHENING A few foreign firms in the names of CINTEC, Helifix, etc are in the market claiming to be having authority in innovative method of arch bridge strengthening. Author of this paper is neither against any new technology nor against any particular firm. But, before adopting any technology, some clarifications should be taken from such firms. Some of the recommended clarification may be as below: BIBLIOGRAPHY (i) Civil Engineering Report No. C-72 –Investigations on Strength of Masonry Arches (July 1969) – Report on the results of investigations on Arches Nos. 71 & 73 (semi-circular 15 ft span brick masonry arches) on abandoned viaduct approach over the River Gandak near Sonepore on North Eastern Railway. (ii) Civil Engineering Report No. C-73 – Investigations on Strength of Masonry Arches (April 1969) – Report on the results of tests conducted on Arch No. 76 (semi-circular 15ft span brick masonry) on viaduct approach over the river Gandak Sonpur on North Eastern Railway. (i) What is the strength of existing arch bridge? (ii) What will be the strength after strengthening? (iii) Design calculation in support of the above (iii) (iv) Has their technology been approved by any internationally recognized institution? Civil Engineering Report No. C-74 – Static Tests on 20ft span masonry arch bridges on Kota Bina Section Western Railway (March 1973). (iv) (v) Has their technology been adopted by any World Railways and if so, its performance certificate? Civil Engineering Report No.C-75 – Investigation on Strength of Masonry Arches (February 1969) – Interim report on dynamic tests on arch bridge no. 41/C on Kota-Bina Section of Western Railway. (vi) What is the likely life after strengthening? (v) (vii) For how much axle load, the strengthened arch bridge is fit? (viii) Cost comparison in rebuilding of a bridge as compared to strengthening cost? Civil Engineering Report No. C-76 – Investigation on Strength of Masonry Arches (February 1969) – Dynamic tests on arch bridge No. 42/B on Kota – Bina Section of Western Railway. (vi) As Executive Director, Bridge & Structures, RDSO, Lucknow, author of this paper tried to get some answers from representatives of such firms. Rational answers have not been supplied. RDSO has got test result of some Civil Engineering Report No. C-77 – Report on Tests on Arch Bridge No. 270 on Poona – Miraj Section of South Central Railway (May 1969). (vii) Civil Engineering Report No. C-79 – Report on Tests on Arch Bridge No. 347 on Poona Miraj Section of South Central Railway (April 1969). Volume - I 98 Volume - I 99 (viii) Civil Engineering Report No. C-80 – Report on Arch Bridge No. 513 over River Nira on Poona – Miraj Section of South Central Railway (May 1969). (ix) Civil Engineering Report No. C-83 – Investigation on Strength of Masonry Arches (August 1969) – Static-cum-Destruction Tests on Arch Bridge No. 41-C on Kota-Bina Section of Western Railway. (x) Civil Engineering Report No. C-84 – Report on Bridge No. 376 (Victoria Bridge0 12 X 50’0” Arches on Rajkot – Jamnagar Section of Western Railway (MG) (August 1969). ANNEXURE Extract of Relevant Paras Prepared by the Author in his Capacity as ED/B&S, RDSO and Member of Working Group of UIC Arch Bridge Committee Para 3: (i) (xi) (xii) Civil Engineering Report No. C-109 – Second Report on tests on Arch Bridge No. 270 on Pune-Miraj Section of South Central Railway after strengthening of the Bridge for BG Traffic (April – May 1970) (October 1971). TEST ARCH CONSTRUCTED IN THE LAB NOT REPRESENTING THE ACTUAL FIELD ARCH (a) Test arch was constructed with 10mm mortar thickness. Normally, arch bridges are having very thin mortar thickness. Even in building construction, mortar thickness is normally provided as 6mm. World over, mortar thickness of 10mm and that too, in arch bridge construction may not be there. Providing more mortar thickness will result unnecessarily more displacement, since it is comparatively more compressible material than the bricks and will give erroneous test results, as the same will be clear from later part of the reply. (b) Arch was constructed by providing sand layers in between the rings for resembling the same as ring-separated arch. Ring separation is one of the defects, but such a ring-separated arch, wherein all the rings are totally separated, is difficult to get in the field. (c) Modeling of the rings of the arch as separated on account of mortar strength loss and not allowing such thing to the mortar available at other locations (like in between the bricks of the same ring) does not seems to be justified. Environmental decay does not follow the selective approach like the same was adopted in test arch. (d) World over, load testing on the arch bridges reveals that there are so many factors contributing strength to the arches but mostly, not included in the design. This is the reason arch bridges are having much more load carrying capacity than its design value. I had elaborated the fact while my presentation during the 6Th meting held at Vienna. Some of the factors Civil Engineering Report No. C-110 – Second Report on tests on Arch Bridge No. 347 on Pune-Miraj Section of South Central Railway after strengthening of the Bridge for BG Traffic (April – May 1970) December 1971). (xiii) Survey and Tabulation Method of Assessment of Strength of Arch Bridges finalized by Dr. S.R.Agrawal and issued by RDSO to all the Zonal Railways. Article can also be seen in the Journal of Institution of Engineers(India) ,January 1973 issue (xiv) UIC Code 778 – 3 R Recommendations for the assessment of the load carrying capacity of existing masonry and mass concrete arch bridges. (xv) UIC Report: Assessment, Reliability and Maintenance of Masonry Arch Bridges: Full Report of January’2004 of UIC (xvi) Arch Bridge Code of Indian Railways (xvii) IRC Special Publication No.37: Guidelines for Evaluation of Load Carrying Capacity of Bridges. Volume - I 100 This office got copy of the TRL report from Indian representatives of M/s Gifford. Comments about the testing work assigned by M/s Gifford to TRL are given as below: Volume - I 101 of that order might had occurred due to providing unnecessarily much more mortar thickness and providing sand layer in between the rings, which is not representing the real arch bridges available in the fields. contributing strength to the arch bridges are reiterated as below: (ii) Provision of haunches Fill on the arches Spandrel wall and parapet wall Track or road surface for which the bridge is there Type of loading whether the same is axle load or single point load Maximum 1.25mm Crown deflection and 0.4mm spread (change in span at the springing point) If many factors, contributing strength to the arch brides was not studied, then the test result is not representing the actual arch bridge strength potential. Residual deflection after the load test should be zero. (b) Hence, as per Arch Bridge Code of Indian Railways, the above tabulated test results are not showing the safe load in terms of vertical deflection as well as residual deflection, both. Codal provisions of other World Railways is not known to me. Opinion of other WG members may kindly be taken in this regard. DESIGN CRITERIA NOT DISCUSSED Arch bridge was strengthened by providing 12 numbers of 55mm dia CINTEC anchors. How the necessity of providing such number of anchors required was not discussed. System should be rational as far as input requirement regarding extra strength demand is there. To work out the demand, firstly, strength assessment of existing bridge needs to be done. All such aspects are missing in the report. (iii) (c) Spread (change in span at the springing point) is one of the important criteria to be measured while load testing. The above test result is silent about this. (d) Instead of providing masonry arch pier, RCC pier was provided which does not represent the field condition. Providing stronger and more elastic support than the masonry will certainly affect the result. TESTING OF THE ARCH WAS DONE UP TO INELASTIC LIMIT, VIOLATING SERVICEABILITY CRITERIA Load test result mentioned in the TRL report is summarized as below: (iv) From the test result, it was concluded that after strengthening, strength of the test arch bridge was enhanced from 2.05 to 2.25 times. Regarding the test results, comments are as under: (a) Volume - I I had gone through the various test results performed in India, in TRL itself and in USA, most of the details can even be seen from the Proceedings of First International Conference on Arch Bridges held at Bolton, UK on 3-6.9.1995. In most of the cases, maximum displacement remains very less and not of the order as shown in the table. It is likely that maximum displacement 102 As per Arch Bridge Code of Indian Railways, while doing the load test, safe load is considered to that load which results: Strength of arch based on elastic method/modified MEXE method was not done and hence, it is not possible to compare the result. This office has got one report from M/s Gifford bearing No.B 1660 A/V 10/R 02 Rev C August, 2003. Vide this document, TRL test result was compared with output result of ELFEN software and claimed that TRL test result and the result obtained from ELFEN software are within the range of 2% and thus, almost matching. Even after so many lapses to the test arch, if ELFEN software is claimed to be giving very close result, then adequacy/ reliability of this software also needs to be examined. Para 10: For better illustration M/s Gifford should submit some design calculation so that design aspect can be well appreciated. Note: M/s Gifford is the Technical Hand / Partner of M/s. CINTEC Volume - I 103 IDENTIFICATION OF CIVIL ENGINEERING STRUCTURES BY VIBRATION ANALYSIS FOR HEALTH MONITORING AND DAMAGE DETECTION S.C. GUPTA* ABSTRACT The paper describes main methods used in recording vibration signatures of the structures and its use in damage detection of bridges. This also clarify that simply recording of vibrations of the structure does not give any useful information until and unless the information received from such testing is used in correcting the basic Finite Element model. 1. INTRODUCTION Technical systems are damaged by overloading, fatigue, ageing and environmental influences. If structures of civil engineering are planned for a finite life time, monitoring with respect to damage is one chance to uarantee safe functionality. The life time of a structure can be split into three main phases, which are the design phase, the construction phase and the utilization phase. With regard to their functionality these three phases differ as follows: • The design of a structure deals with the specification of the type of structural system, of loads and other influences. Such specification depends, of course, on the demands made on the structure, especially under safety and economical aspects. • The construction phase covers the quality examination of the building materials, the safety of the planned construction and the safety and examination of the various stages of a structural building, in order to realize the goals defined during the design phase. • Volume - I The utilization phase starts with the release of the structure and then the structure is exposed to manifold influences, e.g. ageing and fatigue processes, as well as further planned and non-planned external events.[1] *Director/B&S/Testing RDSO/Lucknow An appropriate instrument to guarantee structural safety and economic efficiency is the monitoring of a structure with comparatively little cost for maintenance and monitoring in contrast to the high cost for structural repair or maintenance work, which would then be avoided. Some decades ago, the major concern of Structural Engineers was the development and automatic application of new and powerful numerical methods for the analysis (static and dynamic) and design of large Civil Engineering structures. In this context, the fast development of the finite element techniques accompanied by the tremendous technological progress in the field of personal computers allowed the structural designer to use currently excellent structural analysis software packages, which enable to accurately simulate the structural behaviour. However, the design and construction of more and more complex and ambitious civil structures, like dams, large cable-stayed or suspension bridges, or other special structures, made structural engineers feel the necessity to develop the appropriate experimental tools that might enable the accurate identification of the most relevant structural properties (static and dynamic), providing reliable data to support the calibration, updating and validation of the structural analysis numerical models used at the design stage. Beyond that, the continuous ageing and subsequent structural deterioration of a large number of existing structures made structural engineers gradually more interested in the development and application of effective vibration based damage detection techniques supported by structural health monitoring systems, in which the regular identification of modal properties also plays an important role. Therefore, the first and natural tendency of Civil Engineering researchers was to take some profit from important previous developments made in System Identification and Experimental Modal Analysis in Electrical and Mechanical Engineering, trying to accurately identify the main dynamic properties of civil structures by applying well established input-output modal identification techniques.[2] The difficulty to excite large civil structures in a controlled form, as well as remarkable technological progress registered in the area of transducers and analogue to digital converters, made however feasible to open a new and very promising road for the modal identification of large structures, exclusively based on the measurement of the structural response to ambient excitations and application of suitable stochastic modal identification methods. Volume - I 105 2. METHODS USED IN VIBRATION RECORDINGS To experimentally identify the dynamic characteristics of a structure, also referred to as a system identification two methods are available: 2.1 i) Forced Vibration Testing (FVT) also known as input-output modal ii) Ambient Vibration Testing (AVT) also known as output only modals FORCED VIBRATION TESTING For linear systems, the Frequency Response Matrix is diagonal. This means that it suffices to either determine one row or one column of this matrix (Fig.4). The choice is to either keep the excitation point constant and rove the response points over the structure or vice versa. Because it is not so easy to move the exciters used in civil engineering investigations, the first method is preferred here. In mechanical engineering, where the structures to be tested are comparatively smaller and easy to excite, e.g. with a hammer, the latter of the procedures mentioned is more common. 2.1.1 BASICS With Forced Vibration Testing (FVT) the structure to be identified is artificially excited with a forcing function in point i and its response y k(t) to this excitation is measured together with the forcing signal xi(t) (Fig.1). Transformation of these signals into the frequency domain and calculation of all Frequency Response Functions (FRF’s) Hik between the response and the forcing function time signals yields the Frequency Response Matrix, also referred to as Transfer matrix, H(iω) (Figs.2 and 4). The frequency Response Matrix contains all the information necessary to determine the dynamic natural properties of the structure under investigation (natural frequencies and the associated mode shape and damping coefficients). Dedicated software packages are available on the market to extract these modal parameters from the results of a Forced Vibration Test. 2.1.2 EQUIPMENT & TEST PROCEDURE Simplification of the equation is given in fig 3 The Conventional Model Testing is based on the estimation of a set of Frequency Response Functions (FRFs) relating to the applied force and corresponding response of several pairs of points along the structure, with enough high spatial & frequency resolution. The construction of FRF requires the use of an instrumentation chain for structural excitation, vibration measurement, data acquisition & signal processing. 2.1.3 EXCITATION The excitation can be introduced by impulse hammer or shakers see figure 5. Generally speaking, the means of excitation has to be chosen such as to • Figure 3 Frequency Response Function Volume - I 106 Volume - I excite all natural frequencies of interest, 107 • be significantly larger in effect than any other “unwanted” excitation (because: the processing procedures are based on the assumption that the measured, artificial excitation is the only source of excitation during the tests). 2.1.4 RESPONSE The type of sensor chosen for the response measurement has to fit the requirement concerning sensitivity & frequency range. This is usually measured with accelerometers as they are much easier to apply and rove over a structure. Measuring displacement in many points is a very cumbersome task for Civil Engineering Structures. Velocity transducers are well suited for structures exhibiting fundamental natural frequency f> 4.5 Hz, but the most Civil Engineering structure exhibit lower frequencies. For frequency >1 Hz piezoelectric type otherwise for f<1 Hz Force balance type accelerometers are the most suited. Broad-band vibration generators excite all natural vibrations of the structure in the frequency band at the same time. Examples are impulse hammers and servohydraulic or electro-dynamic shakers generating random or swept-sine type forces. Narrow-band vibration generators excite one specific frequency at a time. Mechanical devices using counter-rotating masses can be mentioned here. Of course, hydraulic or electric shakers can also be used as narrow-band exciters. Fig 5(i)-Impact Hammer The data acquisition and storage of measurement data involves the use of an analogue-to-digital (A/D) converter inserted in a digital computer. The digital raw data must be preliminary analysed and processed, considering operations of scale conversion, trend-removal and decimation. Afterwards, the acceleration time series can be multiplied by appropriate time windows (Hanning, Cosine-Taper, etc.), in order to reduce leakage effects, and subdivided in different blocks for evaluation of average spectral auto and cross spectra estimates, using the FFT algorithm[4]. The automatic evaluation of FRFs requires appropriate software for analysis and signal processing, which is already available in commercial Fourier analyzers. Broad-band exciters are very time effective, but they have to have (relatively) more energy disposable than narrow-band exciters. These devices distribute their energy on many frequencies at a time. Using a narrow-band exciter is very time consuming, but such a device concentrates all the energy available into a specific frequency. To excite Civil engineering structures, hydraulic and electric shakers are better suited than hammers: As a next point, the measurement directions and the measurement point grid density have to be chosen. The basic rule here is: information on mode shapes is available in measured points and direction only. These information can be made available from a preliminary finite element modeling of the structure. So the FE analysis is the first and most important step of the procedure, because major goal of the experimental system identification was to update the preliminary FE model based on experimental results. Compared with mechanical structures, the fundamental natural frequency of a civil engineering structure is low. The average value, e.g. for some 200 highway bridges in Switzerland is f=3 Hz[4]. This updated FE model could subsequently be used as a basis to identify problem solution performing parameter studies. AMBIENT VIBRATION TESTING 3.1 BASICS No artificial exciter is used with Ambient Vibration Testing (AVT), also referred to as output-only modal analysis or natural-input modal analysis. The response of the structure to ambient excitation is fig-5(ii):- servo-hydraulic shakers Volume - I 3. 108 Volume - I 109 measured instead. With civil engineering structures, ambient excitation can be wind, traffic or seismic micro-tremors. The more broad-band the ambient excitation, the better the results. Otherwise, there is some risk that not all natural frequencies of the structure are excited. more than one reference point unless the structure to be tested is very simple. The risk of the reference point sitting in a node of one or more modes can thus be reduced significantly. If response measurements are three-dimensional, at least one 3D-point has to be chosen as a referenc point. Generally speaking : The information resulting from the force input signal xi(t) with FVT investigations is replaced with the information resulting from the response signal yR(t) measured in a reference point R (Fig.7). Spectrum of such modal is shown in fig.6. As a rule of thumb, the length of the time windows acquired should be 1’000 to 2’000 times the period of the structure’s fundamental natural vibration. This is a simple but very important rule of thumb. Experience shows that many investigators do not care about this. But: You can not harvest feathers from a frog! Therefore: we should make sure that our time windows are long enough! The first software package to extract modal parameters from AVT investigations has been developed by a civil engineer in the early nineties of the last century. Today there are several packages on the market making use of the frequency domain procedures shown schematically in Figure 10. One of them offers more sophisticated method like FDD (Frequency Domain Decomposition) and EFDD (Enhanced FDD), the latter also including estimation of damping values [1] . Figure 7 Ambient Vibration Testing relationship Scheme; R is a reference point, K is a roving point 3.2 EQUIPMENT & TEST PROCEDURES Modern force balance accelerometers (figure 9) specially conceived for measurement in the range 0-50 Hz and virtually insensitive to high frequency vibration, have contributed very significantly to the success of ambient vibration test. Number of points needed is conditioned by the spatial resolution needed to characterize appropriately the shape of the most relevant modes of vibration (accordingly to the preliminary FE modeling), while the reference point should be far from the nodal point. Fig.6 – Output only model However, the most recent signal processing tools are not based on an analysis in the frequency domain. Stochastic Subspace Identification (SSI) is a method working completely in the time domain. This method has especially been developed for AVT investigations. Concerning response measurement requirements, the same basic rules apply as for FVT investigations. In addition, it is wise to use Volume - I Figure 8 Calculation of the cross relationship between the reference point R and roving response point k signals 110 3.3 MODEL IDENTIFICATION METHODS There are several model identification methods available now a days and it is easy to use any of these techniques with the help of computer programme readily available. Schematic representation of model identification methods is shown in Fig.10. Details of these methods has not been discussed here as it is beyond the scope of this paper. Volume - I 111 Figure-9 force balance accelerometers Numerical techniques used: FFT fast Fourier transform SVD singular value decomposition LS last squares fitting EVD eigenvector decomposition QR orthogonal dcomposition Fig.10 – Schematic diagram of model identification techniques Figure-9 force balance accelerometers Volume - I 112 Volume - I 113 4.0 FORCED VERSUS AMBIENT TESTING The main advantage of AVT is the fact that no artificial excitation is necessary. This makes such tests comparatively cheap. In addition, AVT investigations can be performed without embarrassing the normal user. This fact is very important e.g. for highway bridges. Ambient excitation is of the so-called multiple-input type. Wind, traffic and micro-tremors are acting on many points of a structure at the same time. In the contrary, a forced vibration is usually of the single-input type. For small, structures, this difference is not important. For large and complex structures, AVT has an advantage on the excitation side. AVT offers multipleinput excitation “free of charge”. Ambient excitation being non-controllable usually results in a lack of stationarity. This may lead to problems due to the non-linearity of the structure (no civil engineering structure behaves in a really linear way). In case of the excitation amplitude being significantly different for each of the setups, a certain scatter in the results may occur. This is not the case for FVT where the structural vibrations induced can be kept stationary. EXAMPLES 4.1 FVT INVESTIGATION For a short RCC/PSC Arch Bridge on the Aare River at Aarburg. The results of a single input FVT analysis is shown in Fig10 & 11 [20] shows the comparison between an FE analysis & FVT investigation results. A similar study done on long bridge in Berlin (eight span continuous beam). The details as shown in fig12&13and results also discussed [21]. Figure 13 Westend Bridge Berlin. Figure 11- Bridge on the Aare River at Aarburg Volume - I 114 Volume - I 115 Figure 13 is showing : Frequencies and shapes of the first five (out of nine[3]) modes of the Westend Bridge as derived from the FVT in vistigation and an up dated FE-model. (MAC for the modes shown in the MAC=0.74 ... 0.83 range.) It can be seen that : a) the freqency of F1 is too low to properly be excited, b) the second torsional mode f2 is not properly excited at the right-hand side of the bridge, c) the bending modes are well excited in the area of the shaker but not at the right-hand side of the bridge. d) the natural vibration dictated by the 31 mtrs. middle span is properly excited(f 5) 4.2 EXAMPLES OF AMBIENT VIBRATION TESTING A long bridge : Ganter Bridge [23] A two lane high way bridge with a total length of 678m has eight spans with a length between 35m & 174m, height of the tallest piller is 174m. Accelerometers were placed inside the box girder. Three 3D & three 1D sensors were roved in pairs along the structure test took ten days. A total of 25 modes could be identified in the frequency band f=0.40, …..3.55 Hz. AVT proved to be a very good method to identify dynamic parameters of such a large structure exhibiting very low natural frequencies. 5. FINITE ELEMENT CORRELATION & UPDATING The modal identification of bridges and special structures plays a relevant role in terms of experimental calibration and validation of finite element models used to predict the static or dynamic structural behaviour, either at the design stage or at rehabilitation. After appropriate experimental validation, Figure 13(A) Volume - I Figure 14 Ganter Bridge 116 Volume - I 117 finite element models can provide essential baseline information that can be subsequently compared with information captured by long-term monitoring systems, in order to detect structural damage. The correlation of modal parameters can be analyzed both in terms of identified and calculated natural frequencies and in terms of the corresponding mode shapes, using correlation coefficients or MAC (Model Assurance Criterion) values. Beyond that, modal damping estimates can also be compared with the values assumed for numerical modeling. The accurate identification of the most significant modal parameters based on output-only identification tests can support the updating of finite element models, which may be a very interesting task in order to overcome several uncertainties associated to the numerical modeling. Such updating can be developed on the basis of a sensitivity analysis, using several types of models and changing the values of some structural properties in order to achieve a good matching between identified and calculated modal parameters. This type of procedure has been recently followed to study the dynamic behaviour of a stress-ribbon footbridge at FEUP Campus ( Fig.16) [2]. For that purpose, initial finite element models were developed idealizing the bridge deck as a set of beam elements with the geometry considered at the design stage or measured through a topographic survey. Later on this model was totally changed after AVT considering partial rotations between beam elements to simulate the lack of sealing of the joints and reducing the area and inertia of the beam elements to simulate the effects of cracking and lack of adherence between precast and in situ concrete. After all these iterations, very good level of correlation between identified and calculated natural frequencies and mode shapes was achieved, as extensively described in ref. (16) Volume - I 118 Beyond this type of sensitivity analyses, more automatic finite element updating techniques can also be used [17]. In this context, a drawback of output-only modal identification seemed to be the impossibility to obtain mass normalized mode shapes. However, this inconvenience can be overcome [18] by introducing appropriate mass changes. 6. CONCLUSION Vibration signature techniques may be used under normal operation conditions & can provide a solid basis for: (i) the development of finite element correlation analyses, (ii) the finite element updating and validation by recording dynamic parameters through vibration i.e. mode shapes, natural frequency and damping; (iii) the definition of a baseline set of dynamic properties of the initially non-damaged structure, that may be subsequently used for the application of vibration based damage detection techniques; (iv) the integration of output-only modal identification techniques in health monitoring systems; (v) the implementation of vibration control devices. Ambient Vibration Testing can be effectively used in the condition assessment of Civil Engineering structures like bridges. However, this technique shall be used in a method as described in the paper, otherwise simply recording vibrations of a bridge does not give any useful data. REFERENCES 1. Identification of Mechanical Systems by Vibration Analysis for Health Monitoring and Damage Detection – by Armin Lenzen, University of Applied Sciences, Germany. 2. From Input-output to output-only modal identification of Civil Engineering Structures – by 1) Alvaro Cunha, Faculty of Engineering University of Porto(FEUP, Portugal and 2) Elsa Caetano, Faculty of Engineering, University of Porto (FEUP) Portugal 3. Experimental methods used in system identification of civil engineering structures – by Reto Cantieni, rci dynamics, Structural Dynamics Consultants, Switzerland. 4. Maia, N.et at. “Theoretical and Experimental Modal Analysis”, Research Studies Press, UK, 1997. 5. Cunha, A., Caetano, E. & Delgado, R. “Dynamic Tests on a Large Cable-Stayed Bridge. An Efficient Approach”, Journal Bridge Engineering, ASCE, Vol.6, No.1, p.54-62, 2001. Volume - I 119 6. Mc Lamore, V.R., Hart, G. & Stubbs, I.R. “Ambient Vibration of Two Suspension Bridges, Journal of the Structural Division, ASCE, Vol.197, N.ST10, p. 2567-2582, 1971. 18. Brincker, R. & Andersen, P. “A way of getting Scaled Mode Shapes in Output Only Modal Testing”, Proc.21st Int. Modal analysis Conference, 2003. 7. Abdel-Ghaffar, A.M. “Vibration Studies and Tests of a Suspension Bridge”, Earthquake Engineering and Structural Dynamics, Vol.6,p.473496, 1978. 19. Cantieni, R. “Dynamic Load Tests on Highway Bridges in Switzerland – 60 Years Experience of EMPA”. EMPA Report No.211, (1983). 20. 8. Felber, A. “Development of a Hybrid Bridge Evaluation System”, Ph.D.Thesis, University of British Columbia (UBC), Vancouver, Canada, 1993. Cantieni, R., Deger, y., Pietrzko, S., “Modal Analysis of an Arch Bridge: Experiment, Finite Element Analysis and Link”. Proc. 12th International Modal Analysis Conference (IMAC), (1994) 425-432. 21. 9. Prevosto, M.”Algorithmes d’Identification des Caracteristiques Vibratoires de Structures Mecaniques Complexes”, Ph.D.Thesis, Univ. de Rennes I, France, 1982. Deger, Y., Cantieni, R., Pietrzko, S.J., Rucker, W., Rohrmann, R., “Modal analysis of a Highway Bridge : Experiment, Finite Element Analysis and Link”. Proc. 13th International Modal Analysis Conference (IMAC), (1995) 1141-1149. 10. Correa, M.R. & Campos Costa, A. “Ensaios Dinamicos da Ponte sobre o Rio Arade”, in “Pontes Atirantadas do guadiana e do Arade” (in Portuguese), ed.by LNEC, 1992. 22. Felber, A.J.Cantieni, R., Introduction of a new Ambient Vibration System – Description of the System and Seven Bridge Tests, EMPA Report No. 156’521, (1996). 11. Brincker, R.,Zhang, L. & Andersen, P. “Modal Identificatiion from Ambient Responses using Frequency Domain Decomposition”, Proc. 18thInt. Modal Analysis Conference, Kissimmee, USA, 2001. 23. Felber, A.J., Cantieni, R., Advances in Ambient Vibration Testing : Ganter Bridge, Switzerland, Structural Engineering International (6), Number 3, (1996) 187-190. 12. Brincker, R.,Ventura, C. & Andersen, P. “Damping Estimation by Frequency Domain Decomposition”, Proc. 19th Int. Modal Analysis Conference, San Antonio, USA 2000. 13. Rodrigues, J., Brincker, R. & Andersen, P. “Improvement of Frequency Domain Output-Only Modal Identification from the Application of the Random Decrement Technique, Proc.23 rd Int. Modal Analysis Conference, Deaborn, USA, 2004. 14. Cunha, A. & Calcada, R. “Ambient Vibration Test of a Steel Trussed Arch Bridge”, Proc. Of the 18th Int. Modal Analysis Conference, San Antonio, Texas, 2000. 15. Caetano, E, & Cunha, A. “Ambient Vibration Test and finite Element Correlation of the New Hintze Ribeiro Bridge”, Proc. Int. Modal Analysis Conf., Kissimmee, USA, 2003. 16. Caetano, E. & Cunha, A. “Experimental and Numerical Assessment of the Dynamic Behaviour of a Stress-Ribbon Bridge”, Structural Concrete, Journal of FIB, 5, No 1, pp.29-38, 2004. 17. Teughels, A. “Inverse Modelling of Civil Enginering Structures based on Operational Modal Data”, Ph.D. Thsis, K.U.Leuven, Belgium, 2003. Volume - I 120 Volume - I 121 USE OF POLYMER MODIFIED MORTARS AND CONCRETES (PMM/PMC) IN REPAIRS & REHABILITATION OF STRUCTURES M.M. GOYAL* INTRODUCTION Polymers are long chain of organic molecules (having high molecular weight) formed by combination of single units called monomers. A polymer consists of numerous monomers, which are linked, together in a chain like structure, and the chemical process, which causes these linkages, is called polymerization. Polymers are classified as either thermoplastics or thermosets. They form the basis of all plastics and elastomers. In general polymers are inert materials having higher tensile and compressive strength than conventional concrete. However, polymers have a lower modulus of elasticity and a higher creep, and may be degraded by heat oxidizing agents, ultraviolet light, chemicals, and microorganisms; also certain organic solvents may cause stress cracking. Many of these disadvantages can be overcome by choosing a suitable polymer and by adding substances to the polymer, e.g., antioxidants to suppress oxidation and light stabilizers to reduce ultraviolet degradation. Polymers are added to conventional Portland cement mortar/concrete to enhance their properties, such as reducing permeability, increasing bond with the substrate, improving resistance to chemicals and damage from freeze-thaw cycles. The material used for repairs is described as ‘concrete’ where it is practical to compact it properly. In situations where it is not possible to adequately compact the material, it is described as mortar. This is one of the reasons why many repairs are carried out with a mortar or shotcrete. TYPES OF POLYMER CONCRETE There are three principal classes of concrete containing polymers: The distinction between these three classes is important to the design engineer in the selection of the appropriate material for a given application. The salient characteristics of the three types polymer concrete are described below. Polymer Impregnated Concrete: It is produced by impregnation under pressure of hardened cement concrete with a low viscosity monomer, methyl methacrylate (acrylic plastic) and styrene (elastomer - synthetic rubber) that polymerizes in situ to form a network within the pores. Polymerization is achieved by thermal-catalytic means. Impregnation results in markedly improved strength and durability in comparison with conventional concrete. Principal applications include storage tanks for seawater, desalination plants and distilled water plants, sewer pipes, tunnel lining and swimming pools. The disadvantage is the relatively high cost of the polymer and complicated production process. However, partial impregnation of concrete members may be economically viable. For example, the shear capacity of RCC beams, without shear reinforcement, may be increased by about sixty percent and resistance to anchorage stresses is increased. Partial impregnation of bridge decks will increase their flexural strength, reduce deflection and improve water tightness and surface durability. Polymer Concrete (PC): It consists of a polymer binder and mineral fillers such as aggregate, sand and crushed stone dust. In PC, Portland cement as a binder is replaced entirely by a synthetic organic polymer. Early PCs were made with epoxy and polyester resin systems. Epoxy is a high strength adhesive compound formed as a result of polymerization of resin at ambient temperature in presence of a specified proportion of hardener. Ambient-cured epoxy systems (thermosetting resins) are a mixture of two components: the epoxy resin (component A), and the curing agent (component B). Neither component is stable or commercially useful separately. The components are manufactured separately and not combined until ready for use by the person applying the material. Epoxy resins react with the curing agents to yield the desirable flexible, semi-rigid, or rigid thermosetting plastics. The curing agent, also known as the hardener, chemically brings about the change from liquid, paste, or mortar consistency to a solid plastic. It is in this state that the system is usually used, there being limited usage in the uncured, non-cross-linked state. (a) Polymer impregnated concrete, (b) Polymer concrete, and (c) Polymer modified cement mortar/concrete (PMM/PMC). PC is substantially more costly than conventional concrete and should be used only in applications in which the higher cost can be justified by superior properties in a particular job. It is also known as resin concrete. PC * Add. Member (Project), Retd. Railway Board Volume - I Volume - I 123 has higher strength, greater resistance to chemicals and corrosive agents, lower water absorption and higher freeze-thaw stability than conventional Portland cement concrete. The monomers and prepolymers most widely used to produce PC are methyl methacrylate (acrylic), unsaturated polyester, epoxy and furan-based polymer. Polymer Modified Cement Mortar/Concrete (PMM/PMC): In these materials, a part of the cement binder is replaced by polymer latex. The process of making PMM/PMC is similar to that of the conventional cement concrete. Most polymers, such as latexes are a colloidal dispersion of a rubber resin in water, which coagulates on exposure to air. These are initially mixed in water in required proportion and then added to the cement mortar or concrete. The latex-modified mortar or concrete is placed in a manner similar to normal concreting. Polymer modification results in formation of continuous polymer film when dried. This improves substantially the characteristics of PMM/PMMC in both (a) the fresh state (e.g. workability, water retention property), and (b) the hardened state (e.g. strength, creep, shrinkage, impermeability and chemical stability) at a reasonable cost. For repairs to spalled and disintegrated concrete and corroded reinforcement, PMM/PMC are most often used these days and of the three types of composite concretes, PMM/PMC is generally the preferred choice of the material in a repair and rehabilitation job. This will be the main focus of our attention in this paper. are white, milky white liquids consisting of very small diameter particles (0.05 – 5 µm) emulsified (colloidal dispersion) in water. See Figure 1 for molecular structure of latex. They form continuous polymer films when dried. Latex reduces permeability, and increases bond strength with the substrate. Polymer latexes are copolymer systems of two or more different monomers and their total solid content including polymers, emulsifiers, stabilizers, etc. is limited to 40 ± 3 %. Higher percentage would adversely affect concrete compressive strength. All latex systems should ensure controlled foaming (air entrainment). Commercially available polymers must contain proper amounts of antifoaming agents. One should be able to use them directly without addition of anti-foaming agent during cement modification at site. CHEMISTRY OF POLYMER MODIFICATION Formation of Polymer Film: Most of the polymer latexes for cement modification form continuous polymer film when dried. Latex modification of cement mortar and concrete is governed by both cement hydration and polymer film formation process in their binder phase. The cement hydration process generally precedes the polymer film formation process. In due course, a co-matrix phase is formed by both cement hydration and polymer film formation processes. This yields a monolithic interwoven matrix of solidified polymer and its continuous film with hydrated cement; this binds the aggregates strongly. The aggregates used for normal concreting operations are recommended for latex mixes. The aggregates should be clean, sound, and of proper grading. The basic principles in regard to design of concrete mixes apply to the design of PMM/PMC as well. It is important to understand the reactions that take place during cement hydration and polymer film formation. Immediately after mixing, the concrete when placed is a mixture of unhydrated cement particles, polymer particles and aggregates with interstitial spaces filled with water (water phase) as shown in Figure 2 (A). Polymer particles partially deposit on the mixtures of unhydrated cement particles and cement gel in the first stage process of hydration as shown in Figure 2 (B), completely enveloping them and forming a membrane in the binder phase (Figure 2C). As hydration proceeds and water is further removed from the pore solution, continuous polymer film forms encapsulating the cement hydrates (Figure 2D) providing a strong binder matrix having substantially reduced permeability. Polymers: Polymer modifiers most widely used to enhance the properties of the repair material are latexes based on poly methyl methacrylate also called acrylic latex, and SBR (Styrene Butadiene rubber) latex. Latexes The continuity of the polymer phase through the binder matrix is more pronounced with polymer cement (p/c) ratio in the range of 0.1-0.2 by weight of cement. Volume - I Volume - I MATERIALS EMPLOYED IN PMM/PMC Cement and Aggregates: The materials used in PMM/PMC are the same as those employed in normal mortar and concrete but for the polymer, which is used as a modifier. Ordinary Portland Cement (OPC) is widely used as the basic binder. Air-entraining admixture is not used in PMM/PMC because air entrainment occurs due to latex addition. 124 125 Reduction in Pore Size Distribution: Concretes modified with polymer latexes generally show a different distribution of shape of pores. In conventional Portland cement concrete the structure consists of large pores surrounded by a number of small pores, whereas concrete with latex tends to show only spherically shaped pores, usually at a smaller average pore diameter. Capillary water uptake depends upon capillary pore diameter and is effectively stopped at about 0.1 mili-micron. Polymeric modifiers are designed to reduce the content of these over-sized pores. The net effect on the impermeability characteristics of latex modified concrete is very favorable, particularly regarding its resistance to chloride ion penetration. IMPROVEMENT IN THE PROPERTIES OF PMM/PMC Workability: Polymer latex modification of cementitious mixtures improves workability and water retention property compared to conventional mortar/ concrete. There is better resistance to bleeding and segregation even though they have better flowability. Resistance To Crack Propagation: Micro cracks occur easily in the ordinary stressed/hardened cement paste. This results in poor tensile strength and fracture toughness. In the latex modified mortar/ concrete, it appears that the micro-cracks are bridged by the polymer film or membrane, which prevents crack propagation and simultaneously, a strong cement hydrate-aggregate bond is developed. Increase in Tensile Strength: PMM/PMC with SBR latexes have a noticeable increase in tensile and flexural strength but there is hardly any improvement in its compressive strength compared to ordinary mortar/ concrete. An increase in the polymer content (defined as the weight ratio of the amount of total solids in polymer latex to the amount of cement) leads to increase in flexural tensile strength and fracture toughness. However, excessive air entrainment and polymer inclusion cause discontinuities of the formed monolithic network structure, whose strength is reduced. Chemical And Abrasion Resistance: This depends on the type of polymer, polymer cement ratio and type of chemicals used in the PMM/ PMC. Most PMM/PMC with SBR latex are attacked by strong organic and inorganic acids and sulphate but they resist well alkalis and salts. Their resistance to chlorides, fats and oils is also rated good, while they have poor resistance to organic solvents. PMM/PMC have better abrasion resistance than conventional mortar/ concrete. Volume - I 126 Temperature And Shrinkage: PMM/PMC show rapid reduction in strength with increase in temperature. Most thermoplastic polymers have glass transition temperature (tg) of 80 – 100°C, the temperature at which a reversible change occurs in an amorphous polymer when it is heated to a certain temperature and undergoes a rather sudden transition from a hard, glassy, or brittle condition to a flexible or elastomeric condition. Below tg, molecules have little mobility. Drying shrinkage of PMM/PMC may be larger or smaller depending on the type of polymer and polymer cement ratio used. More is the polymer ratio less is the drying shrinkage. Permeability And Durability: PMM/PMC have a structure in which the larger pores are filled by polymer. The sealing effect due to the polymer film formation in the structure also improves water tightness (impermeability) as well as resistance to chloride ion penetration, carbonation and oxygen diffusion, chemical resistance, and durability against freezing and thawing. Such an effect is promoted with increasing polymer – cement ratio up to a point. Generally these materials are bio non-degradable after total polymerization takes place. However, certain polymers such as styrene tend to disintegrate under any form of energy like ultra violet rays, heat, etc. Acrylate based materials (acrylic polymers) are reported to be robust and bio non-degradable. Adhesion or Bond Strength: A very useful property of PMM/PMC is their improved adhesion or bond strength to various sub-strata compared to conventional mortar/concrete. MIX PROPORTIONS Polymer Modified Mortars: Mix proportion for most PMMs is in the range of 1:2 to 1:3 (cement – fine aggregate ratio). The polymer latex (solid contents): cement ratio ranges from 5 to 20% by weight. The water cement ratio is of the order of 0.3 to 0.5 depending upon the requirement of workability. The standard mix proportion for different usage is indicated in Table 1. Polymer Modified Concrete (PMC): The mix proportion for latex modified concretes cannot be easily determined in the same manner as that of latex modified mortars. Because of many factors in design, normally, the polymer latex: cement ratio ranges from 5% to 15% and water cement ratio from 0.3 to 0.5. Volume - I 127 Table 1 Typical Applications and Standard Mix Designs of Latex Modified Mortars c) For resurfacing, flooring and patch repairing all loose and nondurable materials including laitance must be removed by sand blasting, wire brushing and blowing with compressed air. The cleaned surface should be thoroughly wetted well before placement of PMM/PMC. Before application, surface should be in saturated dry (wet but no standing water) condition; d) The choice of PMM/PMC depends on the thickness of coating to be applied; e) It is advisable to finish the surface by trowelling 2-3 times. Over trowelling should be avoided; f) PMM/PMC should never be placed below 5° and above 30° C. The surface of the newly placed material should be protected from rainfall or other source of water. The surface should be immediately covered with burlap or plastic sheet; g) In large area of application, it is advisable to provide joints 15 mm width at intervals of 3 – 4 m; h) Polymers and latexes are non-toxic and safe for handling. They should be stored in a cool dry room and should not be kept in exposed areas. Curing: Curing of PMM/PMC is different from that of conventional concrete because the polymer forms a film on the surface of the product retaining some of the internal moisture needed for continuous cement hydration. Because of the film-forming feature, curing under water immersion or under wet condition is detrimental to PMM/PMC. Moist curing of the latex modified mortars/concretes is generally shorter than for conventional products. Optimal conditions for strength development are moist curing for 3 days followed by dry curing at ambient temperature. FIELDS OF APPLICATION OF PMM/PMC GENERAL GUIDELINES FOR USE OF PMM/PMC Polymer should be mixed with cement slurry or mortar in the proportion recommended by the manufacturer for various uses. The following precautions may be borne in mind while using PMM/PMC: a) The speed and time of mixing should be properly selected to avoid unnecessary entrapment of air; b) The PMM/PMC have excellent adhesion even to metal and hence all equipment should be washed immediately after use; Volume - I 128 Structural Repairs to RCC: PMM/PMC are used to make up the damaged concrete or lost cover due to their better bond with substrate, and the reinforcement. Ultra Rapid Hardening Polymer Modified Shotcrete: It can be classified into two categories depending on usage as under: (a) Volume - I Repairs to Leaky Liquid Retaining Structures: In this system, polymerisable monomers are used that react with Ordinary Portland Cement at ambient temperature to form protective cover 129 for concrete structures with leaking and flowing material. It uses magnesium acrylate monomer and its setting time can be controlled within few seconds or less. (b) Urgent Construction And Repair Works: Ultra Rapid Hardening Cement Concrete is used with SBR latex for urgent construction and repair works subject to heavy traffic. Polymer Ferrocement: For the purpose of improving the flexural behavior and durability of conventional Ferrocement, polymer-Ferrocement have been developed using latex modified mortars instead of ordinary cement – sand mortars. Use of SBR and EVA modified mortars is found to be quite effective in improving their flexural behavior, impact resistance, drying shrinkage and durability. Incorporation of short fibers such as steel and carbon fibers in the latex modified mortars is found to be further effective in improving such characteristics. d) Excellent elongation, flexibility and crack resistance, e) Good waterproofness, and f) Resistance to carbonation and chloride ion penetration. An innovative water proofing material, which solidifies in water has recently been developed in Japan for tunnels and dams. It has potential as a shock absorbing, waterproofing backfill material. Bond Coats (Structural Adhesive) And Grouts: Polymer modified cement mortars (PMMs) as well as slurries are used as bond coats and grouts due to their very good adhesive qualities on cementitious as well as metallic surfaces. Anti Washout Underwater Concrete: Major requirement for such concrete is anti washout or segregation resistance, flowability, self-leveling ability, and bleeding control. Anti-washout admixtures are water-soluble polymers and are added in the polymer-cement ratio of 0.2% to 2.0% during mixing of the ordinary cement concrete. They bond to a part of the mixing water by hydrogen bonds in the concrete and disperse in a molecule form in the mixing water. As a result, the mixing water is confined to the network structure of the dispersed polymer and becomes very viscous. The very viscous water envelops the cement and aggregate particles to impart antiwashout character to concrete. Protective Waterproofing Membranes and Repair Materials: Polymer modified pastes or slurries with very high polymer: cement ratio of 50% or more have been widely used as liquid applied waterproofing membrane and repair material. The constituents normally comprise Portland cement, silica sand, water and polymer latexes such as SBR, EVA, PAE, SAE, epoxy and asphalt latexes besides some other additives. Thickness of such waterproofing membrane is 1.5 to 2 mm; they are generally available as pre packaged products. The performance advantages of such membranes are as under: a) Safe application due to no organic solvent system, b) Convenience of application as it does not require the surface to be dry, c) Good adhesion with the cementitious, metallic and most other substrates, Volume - I 130 Volume - I 131 LONGITUDINAL LOADS ON RAILWAY BRIDGES SHYAM SUNDER GUPTA* ABSTRACT Increasing demand of traffic on existing infrastructure leads to higher intensity of loads on bridges. To allow higher intensity of loading on existing bridges, the alternatives are: efforts and braking force or due to temperature changes. In this paper longitudinal force due to tractive effort and braking force only are discussed. HISTORY OF PROVISION OF LONGITUDINAL LOADS IN BRIDGE RULES Prior to 1892, bridges were built according to the British Board of Trade Rules. The Bridge rules 1892 did not specify longitudinal loads. Reference to the tractive and braking forces first appeared in 1923 Rules. A historical note on longitudinal load provision as per Bridge rules is given in annexure-I. The history of loading standards for bridges on Indian Railways is also given in annexure-II. a) To exploit extra potential available on account of conservative designs of early days of steam engines. b) To allow heavy axle loads on existing bridges on physical condition basis on theoretically overstressed structure. As per 1923 rules, traction and braking forces were taken about 1/7th of train loads. Detailed provisions were made in 1926 Bridge Rules, as given in annexure-I. Since these provisions on braking force and tractive efforts were copies of BESA rules, it was suggested in 1932 that there formulae be amended to make them applicable to the standards of loading that were in use at that time in India. Accordingly provisions were modified in 1933 rules. c) To revise codal provisions based on instrumentation, field testing, latest knowledge and actual requirements. It may be noted that provision for dispersion of longitudinal loads were made for the first time in Bridge Rules of 1933. d) Strengthening/ rebuilding with minimum disruption to traffic at least cost. In Bridge Rules of 1941 printed in 1960, a ‘note’ reproduced below is appearing. While alternative (a) and (b) are already being followed, the alternate (c) and (d) are being discussed in this paper. For alternative (c), history of codal provisions regarding longitudinal loads and retention of some deleted provisions and amendments to existing provisions are discussed. For alternative (d), use of STU’s for permitting higher longitudinal loads on existing bridges is discussed. “Note – In the case of piers and abutments of RCC bridges with ballasted floors, no longitudinal force need be taken into account for spans upto and including 20 ft.” INTRODUCTION Bridges on Indian Railways were constructed to lighter loading standard in the earlier stage of developments of Railways, as the speeds as well as the loads were low. Longitudinal forces were not even considered for design of bridges prior to 1923. As the speed and loads were increased, the longitudinal forces assumed importance. The first reference to provision of longitudinal forces on Railway bridges in India is made in the Bridge Rules of 1923. By this time more than 60% of the bridges were built on Indian Railways. The provision for longitudinal forces in earlier days (prior to 1975) was made on empirical basis for design of bridges. On bridges, longitudinal forces are caused by train running, i.e. tractive Volume - I *Chief Bridge Engineer, North Western Railway, Jaipur This note is appearing as clause 2.8.2.3 of Bridge Rules 1964 but was later deleted and does not exist in current Bridge Rules. This was an important clause, which needs review. In 1964 Bridge Rules, HM loading was deleted. It is also seen that in 1964 Bridge Rules Tables of longitudinal forces were modified by adding dispersion allowance so that table gives longitudinal loads (without deduction for dispersion). This is why the values of TE and BF for BGML bridges are more than the MBG bridges for smaller spans. The provision of TE and BF given in Bridges rules prior to 1975 are based on empirical formulae. BGML loading accounts for a maximum tractive effort of 47.6 t, which was raised to 75 t in RBG loading introduced in 1975. Modified Broad Gauge loading (MBG-1987) introduced in 1988 caters for 100 T of tractive effort for coupled locos with TLD of 8.25 t/m. Volume - I 133 After the introduction of longitudinal force provision in Bridge Rules, mainly three loading standards have been used on Indian Railways for bridges (BG). There are BGML, RBG and MBG standard of loading. The comparative statement of longitudinal loads for these loadings is given in annexure-III. b) Continuous track 660+1.25 (L-30) in kN (May be distributed over a length of 30m) For L>300m : 1000 kN constant Where L is the loaded length PROVISION OF LONGITUDINAL LOADS ON OTHER RAILWAY SYSTEMS The provision regarding longitudinal forces i.e. tractive and braking forces on foreign Railway systems are given in appendix-IV. Maximum dispersion permitted as per BS 5400: 1978, is up to onethird of the longitudinal loads for bridges supporting ballasted track, provided that no expansion switches or similar rail discontinuities are located on or within 18m of either end of the bridge. For normal type RU loading which consists of four 250 kN concentrated loads preceded and followed by a uniformly distributed load of 80 kN/m, maximum tractive force is 750 kN for loaded length exceeding 25m. As per AREA manual (1991) for Railway Engineering: a) The longitudinal force (LF) from Train shall be taken as 15% of the live load without impact and b) Where the rail are continuous (either welded or bolted joints) across the entire bridge from embankment to embankment, the effective longitudinal force shall be taken as L/1200 (where L is the length of the bridge in feet) times LF given in (a) above but the value of L/1200 shall not exceed 0.80. So the dispersion of longitudinal force will be (1-L/1200) x100%, Subject to a minimum value of 20%. UIC 776-1R As per this code the term “accelerating force” has been used in place of tractive force. The value of accelerating force is as under: For L< 30m L > 30 m a) discontinuous track 33 kN/m b) Continuous track 22 kN/m a) discontinuous track 1000 kN (Uniformly distributed over a length of 30m) Volume - I Braking force (as per UIC 776-1R) acting on a structure is 0.25 times the vertical force in the UIC loading calculated over the loaded length L. The braking force shall be evenly distributed over the whole of the loaded length. It is seen that no dispersion of braking force has been considered as per UIC 776-1R Code. Also accelerating force for L> 30 m has been considered to be distributed over a length of 30m whereas for braking force, it is considered to be distributed over whole loaded length L. DISPERSION OF LONGITUDINAL FORCES Longitudinal forces generated due to traction/ braking of train are partly dispersed to approaches and balance transferred to sub-structures. Amount of dispersion of longitudinal forces to approaches depends on the stiffness of approaches, type and condition of track structure, provision of guard rails and their condition, type of braking, number and length of span of the bridge, vertical load over the bridge etc. Tests conducted in India and abroad under different conditions are briefly summarized below: TESTS CONDUCTED IN INDIA It is seen that no distinction is made between tractive force and braking force as per this manual. For It is seen that for continuous track, part of the accelerating force is considered to be dispersed away from loaded length. 134 Some tests were carried out by North Western Railway in the early thirties to determine the dispersion of tractive force through the rails to the approaches. These tests were conducted with steam locomotives. The actual force applied to the piers in the case of bridges with one, two and three spans of 40feet were measured. It was found that about 70% of the TE is dispersed for 1x40 feet span and for 3x40 feet span, 50% of the TE goes through the rails and balance 50% to piers. For a length of about 400-500 feet, it was found that nearly whole of the TE is transferred to piers. Based on these observations it was concluded that in bridges not exceeding 130 feet in total length, 50% only of the specified tractive and braking forces need be considered as passing through the girders to the piers and abutments. Volume - I 135 sequence. The last one to come to a stop is the Loco or the front vehicle. Due to the time gap between stopping of different vehicles, the peak braking force is not the sum of individual axle/ vehicle peak braking force. Due to the short duration of the peak braking force at the moment of stopping, some vehicles would have stopped while other would still be in motion, so the peak force is not generated along the entire train at the same instant but travels like a wave from rear to the front. The maximum braking force will therefore, get distributed over adjacent spans. These tests indicate that length and number of spans affects the amount of dispersion to approaches. Braking force tests on South Bassein Bridge (60x 60 feet) confirm that in a long bridge, the whole of the LF is ultimately distributed amongst the piers. RDSO was entrusted with the work of investigation of dispersion of longitudinal forces. The tests were conducted on Sone bridge of 14x76.3m span with two shore spans of 30.5m near Chopan on Chunar- Chopan section, Varuna Bridge No 32 of 6x 18.3m span near Varanasi on Varanasi-Zafrabad section, Bridge No 3 of 2x 12.2m span near Shikohabad on Shikohabad – Farrukhabad section and Kolab bridge No. 543 of 10x45.7m through spans on SE Railway. During the longitudinal force trials on Varuna Bridge of 6x18.30m span (RDSO’s Civil Engg Report No C- 153), it was observed that dispersal through track away from span under braking condition is much less as compared to the dispersion under starting condition. Simultaneously, during trials on Kolab Bridge (10x45.7m through span) on SE Railway (RDSO’s Civil Engg Report No C- 235), it was observed that in case of multi-span long bridge with spans supported on rocker and roller bearings when entire bridge is covered under load and all the axles are simultaneously braked, the dispersion of braking force away from some of the intermediate span is negligible. TESTS CONDUCTED BY UIC A number of tests have been carried out by office of Research and Experiments (ORE) of the UIC. The results are reported in various reports (RP1-15) on question D 101 “Braking and acceleration forces on bridges” and some of the important conclusions are: i) ii) iii) Volume - I Guard rails transfer considerable longitudinal forces to approaches. The sharing of longitudinal load between rails and other elements of bridge is affected by horizontal stiffness of track on approaches and on bridge. If track on approaches is stiffer than on the bridge, a large share of horizontal load is taken by approaches. On the other hand, if track on bridge is stiffer as compared to approaches, a large share of load goes to bearings. All the vehicles/ axles don’t stop at the same instant. The rear most vehicle/ axle stops first, followed by the other axles in a 136 iv) Track in the approaches up to 30m absorbs the longitudinal forces and thereafter, there is hardly any affect on the approach track. v) Ballast in ballasted deck bridges does not transfer more than 3% of the horizontal load to approaches. From the experimental results of tests concluded by UIC, it was found that percentage of longitudinal force in case of tractive/starting forces, going to substructure was less as compare to braking. TEST CONDUCTED BY AAR The Association of American Railroad (AAR) tested [10] a single span (50feet) open deck steel girder to measure the longitudinal forces. The test bridge is located on 1% grade and was built in 1908 for copper E-55 design loading. The rail is continuously welded on the bridge and its approaches also. The rail anchoring conditions were varied on the bridge and approaches to study the effect of dispersion of longitudinal forces. The AC locomotives used in the test have a maximum tractive effort per loco unit in excess of 1,50,000 pounds, while maximum dynamic brake effort is limited to about 80,000 pounds. 3- Unit set of six axle AC locomotive were used. During tests, it was noted that: (i) The force into the bridge increases as the TE applied to the rail on the bridge increases (ii) The maximum measured longitudinal force into the bridge is roughly 25% of the locomotive weight applied to the bridge. The highest forces measured during any of the tests were for the condition when the rails on the approaches are minimally anchored. A force into the bridge of nearly 100kips was measured for an applied TE of 135 kips. This force was noticeable reduced when the rail on the approaches are Volume - I 137 tightly anchored. This allows more of applied TE to be dispersed to approaches. “It was found that the measured longitudinal forces (i.e. 100 kips) into the bridge are considerably higher than the design values (<4 kips for 50 feet steel span) recommended by AREA for steel bridges. IMPORTANT CLAUSES OF PRESENT AND PAST CODES NEEDING ATTENTION Arch Bridge Code: Clause 2.2.3 states “Horizontal loads on the arch: The effect of the tractive effort and braking force may be neglected in designing or analyzing arch covered by this code”. Out of about 1,20,000 bridges on Indian Railways, about 18% are arch bridges. As per Clause 2.2.3 of Arch Bridge Code, longitudinal force due to TE and BF may be neglected. So as far as arch bridges are concerned, there is no problem of longitudinal forces. Bridge Rules 1933: In Clause No 22, it is mentioned that, “In the case of single span bridges up to 80 feet long, the dispersion allowance may be increased by 50%”. The word “single span” was later deleted vide Correction Clip No 9/36 and the extra dispersion up to 50% for bridges up to 80 feet long was deleted in Bridge Rules 1941. It brings out an important issue to be considered that is whether for single span bridges, dispersion of forces is more as compared to multi span bridges. This appears to be logical and needs further testing and instrumentation to establish that dispersion in case of single span bridges is more than multi-span bridges. Bridge Rules 1941, Reprinted in 1960 (Incorporating CS- 1 to 13): Note to clause 23 states, “Note: In the case of piers and abutments of RCC bridges with ballasted floors, no longitudinal forces need be taken into account for span up to and including 20 feet.” The above note was reproduced in clause 2.8.2.3 of Bridge Rules 1964, but clause is missing in Bridge Rules of later years. This is an important clause which is not finding place in current Bridge Rules. It appears to be logical not to consider longitudinal forces for RCC bridges of spans up to 6 m as most of the longitudinal forces will be transferred to approaches. Steel girders up to 6.1 m span are already being replaced by RCC/ PSC slabs. There is a need to establish by instrumentation the span up to which longitudinal forces can be neglected. Volume - I 138 NEW V/S EXISTING BRIDGES It is pertinent to note that whereas new bridges are to be designed for next 100 – 150 years for higher standard of loading expected in future, the existing bridges need only to be checked for permitting present loads knowing fully well, the existing condition at site. For new bridges, loading standards catering for future expected loads for design life of bridges with higher factor of safety needs to be considered, whereas for existing bridges, lower factor of safety can be permitted taking into account the actual condition of bridge structures by removing uncertainties with instrumentation and testing. Higher percentage of dispersion of longitudinal forces to approaches can be considered while checking existing bridges than what is provided in current Bridge Rules whereas for design of new bridge, no dispersion of LF be considered. It is necessary that detailed guidelines are made available for checking strength of existing bridges. SINGLE V/S MULTI-SPAN BRIDGES It needs to be kept in mind that percentage of dispersion of longitudinal forces in case of single span bridges will be different than in multi-span bridges of same length because part of the longitudinal forces is already transferred to piers before dispersion to approaches in case of multi-span bridges. Experiments in nineteen thirties in North Western Railway and investigations by RDSO between 1967-80 gives evidences that dispersion of longitudinal forces in case of single span bridges is more as compared to multi-span bridges. This area needs further investigation to permit higher dispersion of longitudinal forces in case of single span bridges than what is permitted as per present codal provisions which does not distinguish between single span and multi-span bridges. TRACTIVE EFFORT V/S BRAKING FORCE Longitudinal forces on bridges are generated on account of tractive effort of locomotive during starting of train and braking force during braking of trains. It is the maximum of tractive effort or braking force which is considered for design of bridge structure for longitudinal loads. The maximum tractive effort which can be generated is fixed by the design of locomotive. Volume - I 139 Single WDG-4 locomotive can generate a maximum tractive effort of 52.9t which is even more than MBG standard of loading which caters for a maximum tractive effort of 50t for a single loco. The maximum braking force which can be generated depends on train load, speed, braking distance required, capacity of brake blocks and wheel to absorb energy, loose or tight couplings etc. During slow movement caused by temperature change, creep and shrinkage, the silicone is able to squeeze through the annular space. A suddenly applied load causes the transmission rod to accelerate through the silicone putty within cylinder. The acceleration quickly creates a velocity where the silicone putty can not pass fast enough around the piston. At this point, the device locks up usually within half a second. Under normal operating conditions, the large longitudinal force occurs when pulling a heavy train upgrade or when braking a heavy train downgrade. Therefore, bridges in the vicinity of significant grades are more likely to experience a large longitudinal force. LUD/STU’s can be used for strengthening of existing bridges or economical design of new bridges. It is important to note that for bridges longer than a set of locomotives, the higher longitudinal forces may be caused by train braking. For shorter bridges, tractive effort will govern because of higher adhesion of the locos, so braking force governs in case of longer spans, whereas tractive effort governs in case of shorter spans. Due to increase in longitudinal loads (tractive effort/braking force), many existing bridges which are found unsafe for increase tractive effort/ braking force require either strengthening or re-building. STU/LUD’s provides a convenient means of strengthening existing bridges by load sharing between pier/abutment during suddenly applied loads i.e. tractive effort and braking force. This load sharing will reduce net longitudinal force going to substructure, with the result LF due to increased tractive effort/ braking force can be carried by redistribution of load and without any need for strengthening of pier/ abutment/foundation. Also if any pier is found damaged (under water), the longitudinal force on this pier can be reduced by providing suitably designed STU’s on adjacent piers. USE OF SHOCK TRANSMISSION UNIT (STU) A shock transmission unit (STU) or “Lock up device” (LUD) is a fabricated component which is designed to link or connect separate elements of bridge structure so as to form a rigid link under rapidly applied (short duration) loads such as longitudinal forces due to traction or braking, seismic forces etc. but move freely under slowly applied load, such as temperature. This mechanism helps in load sharing of suddenly applied short duration loads. After removal of these sudden dynamic loads, the device returns to its original position and structure behave in normal manner. Longitudinal force due to tractive effort and braking force are short duration rapidly applied horizontal loads so STU/LUD’s can be advantageously used to control the longitudinal forces coming on any pier/ abutment within safe limits and train with higher TE/BF can be permitted to run. PRINCIPLE STU/LUD operates on the principle that rapid passage of viscous fluid through a narrow gap/orifice generates considerable resistance, while slow passage generates only minor resistance. STU/LUD consists of a cylinder with a transmission rod that is connected at one end to the structure and at the other end to loose fitting piston inside the cylinder. The annular space is filled with silicone putty. Volume - I 140 STRENGTHENING EXISTING BRIDGES On Indian Railway, the use of suitably designed STU/LUD’s can go a long way in not only solving the problem of strengthening of existing multispan bridges, for increased longitudinal forces due to increased TE/BF, but also can be used for permitting LWR on existing bridges. DESIGN OF NEW BRIDGES The concept of distribution of longitudinal loads by means of STU/ LUD’s can be advantageously used in designing economical structures resulting in saving in pier/ foundations. The new Bassein Creek Bridge located on NH 8 crossing the sea at Mumbai, India is designed to take advantage of STU’s for distribution of seismic loads. The use of STU’s reduced the size of well foundation resulting in substantial cost reductions. Use of STU/LUD’s and its effectiveness on Railway bridges in India needs to be established considering safety and economical aspect and if found appropriate, the use may be permitted after making suitable provisions/ amendments in clauses in relevant Codes like Bridge Rules. Volume - I 141 RECOMMENDATIONS & CONCLUSION Present clause pertaining to longitudinal forces in Bridge Rules must be reviewed to: a) effect on longitudinal loads is mainly due to braking force of wagons, considering the same locos. This will affect only large span bridges, where use of STU/ LUD’s may be thought of. Permit more dispersion for single span/ short length bridges as compared to multi-span bridges. REFERENCES b) Disallow dispersion for new bridges 1. Tenth Report of the Bridge Standard Committee, 1931. c) Include a clause for neglecting longitudinal forces while checking existing bridges in case of arches, pipes, slab bridges box culverts and possibly all minor bridges with span up to 6.1m. 2. AREA Manual for Railway Engineering, 1991. 3. UIC Code:776-1 R “Loads to be considered in Railway Bridge Designs”, 1994 Clearly defining clauses for new bridges and existing bridges separately for considering dispersion and longitudinal forces. 4. BS 5400: 1978 (Part-II) 5. Civil Engg. Reports C-64, C-153, C-235, RDSO. 6. Fourth Report of Bridge Standard Committee, 1927. 7. RDSO Report on BS-26, “Index of the Report of the Bridge and Structure Committee”, 1999. 8. Agrawal, S.R.,“Dispersion of Tractive and Braking forces in Railway Bridges”, PhD Thesis, University of Roorkee, 1973. 9. Patel, D.J., “Shock Transmission Units in Bridge Engineering” Engineering World, October – November, 2000” Since TE required is more on gradients, requirement of TE for level tracks be calculated and wherever, the bridge are on level track, the same can be considered. 10. Otter, D.E. and LoPresti, Joseph, “Longitudinal Forces in an Open – Deck steel- deck plate girder bridge”, Railway Track & Structures. May 1997. On smaller length bridges/single span bridges, more dispersion can be permitted than what is provided in current Bridge Rules. So investigations must be carried out to find out percentage of dispersion of LF to approaches. 11. IRS Bridge Rules of year from 1923 to 2005. 12. IRS Bridge substructure and Foundation Code 2003. 13. IRS Arch Bridge Code 2000. d) Since provision of guard rails and its condition, strength of approaches helps in dispersion of LF for approach; these aspects must be kept in view to increase dispersion of LF to approaches. Since tractive effort as well as braking force is maximum at zero speed, any speed restriction imposed on a bridge is detrimental from longitudinal force consideration. So for any work in progress in substructure/foundation or temporary arrangements, speed restriction of dead stop must be prohibited. The main reason for lesser dispersion of braking force in case of longer span/ longer bridges is due to trailing loads. It appears reasonable to separately consider/ investigate dispersion of braking force due to loco and due to trailing loads. Since TE govern short length bridges and BF governs longer bridges, if a bridge is found to be unsafe for longitudinal loads on account of braking, then STU/LUD’s may be provided on the bridge to distribute the longitudinal load. If a bridge is found unsafe due to tractive effort, then for actual condition at bridge site like grades, detailed analysis be done for required TE. Due to increase of axle load of wagons to CC+6+2 or CC+8+2, the Volume - I 142 Volume - I 143 (b) In case of lines worked solely on electrical multiple unit system, the amount of the tractive effort on one track shall be ascertained by multiplying the sum of the actual wheel loads on the span by a factor equal to 3 plus 0.10 L- 10 Annexure-I HISTORICAL NOTE ON LONGITUDINAL LOAD PROVISION AS PER BRIDGE RULES Prior to the formulation of the Bridge Rules in 1892, bridges were built to comply with the British Board of Trade Rules. These rules specified the permissible stresses for wrought iron and steel but not the longitudinal loads. After 1892, bridges were built according to the standards laid down in the British Rules (which were revised first in 1903 and again in 1923, 1926, 1933, 1941 and 1964). Reference to the tractive and braking forces first appeared in 1923 Rules. According to these rules “Traction and braking loads are to be taken as horizontal forces, acting at the rails of each track in the direction of moving train and are to be computed as directed in clause 9 of Part-3 of the British Standards Specification for girder bridges” .These forces were approximately 1/7th vertical loads. In 1926 Rules, the provision for longitudinal forces was made in clause 30 and 31 which are reproduced below: CLAUSE NO 30 & 31 OF 1926 BRIDGE RULES FOR THE LONGITUDINAL FORCES: “30: Longitudinal forces: Where a structure carries a railway provision shall be made for the stresses due to the tractive effort of the live load and the braking effect resulting from the application of the brakes to such load while passing there over, these forces being considered as acting at rail level. 31 (a) For railway worked by steam or electric locomotives, the amount of the tractive effort on one track shall be ascertained by multiplying one and three quarter times the maximum end shear due to the live load on that track by a factor equal to 20 , L + 75 Where L= the span in feet. The factor shall not exceed 0.15 . The braking effort shall be similarly determined using a factor equal to 12 plus 0.75 L + 90 Also limited to a maximum of 0.15 Volume - I 144 Where L= the span in feet. The factor shall not exceed 0.20. The braking effort shall be similarly determined by using a factor of 0.20 for all spans” In the 1933 Rules the provisions for longitudinal forces were further modified and a special clause was introduced to allow for the dispersion of the horizontal forces through the track to the approaches of the bridges on the basis of the experiments carried out with steam locomotives on bridges. The relevant clause is reproduced below: “22. Longitudinal forces: Where a structure carries a railway provision as under shall be made for the stresses due to the tractive effort of the live load and to the braking effect resulting from the application of the brakes to such load while passing there over. These forces shall be considered as acting horizontal through the girder seat where girders have sliding bearings or through the knuckle pin. For span supported on sliding bearings the horizontal forces shall be considered as being divided equally between the two ends; for spans which have roller bearings at one end the whole of the horizontal forces shall be considered to act through the fixed end. (a) Tractive Effort: For railways worked by steam or electric locomotives the amount of tractive force on one track shall be ascertained by multiplying the EUDL on one track taken from the Table of loads for calculating bending moments by 32 . L + 90 Where L= the length of the bridge in feet subject to a maximum of 120 feet for HM loading and 100 feet for ML and BL loadings. For bridges or spans exceeding 120 feet in length for HM loading and 100 feet for ML and BL loadings the tractive effort shall be assumed to be constant. (b) Braking effect: For railways worked by steam or electric locomotives the braking effect on one track shall be ascertained by multiplying the EUDL on the track, taken from the Table of loads for Volume - I 145 the bridge, i.e. the dispersion allowance in tons per track can be taken as one tenth of the weight of rails in pounds per yards. The 50% increased dispersion on single span bridge up to 80 feet applies.” calculating bending moments, by ___32____ plus 0.03 145 + L Where L= the length in feet of the bridge or the length of the train whichever is less. (c) In case of lines worked solely on the electric multiple unit system, the following allowance shall be made: In 1941, 50% extra dispersion of longitudinal force to approaches incase of bridges up to 80 feet span provided in 1933 Bridge Rules deleted. Tractive effort: 25% of the sum of the driving axles on the bridge. In Bridge Rules of 1941, printed in 1947, dispersion clause was changed as under: Braking effort: 20% of the sum of all axles on the bridges. HM loading – In all case the amounts calculated under heads (a), (b) & (c) shall be reduced to allow for dispersion of the horizontal forces through the track at the ends of the bridge by amount as given in the following table: Standard L< 100’ varying uniformly from 17.25 ton for L=0 to 11.5 ton for L= 100’ ML Loading – through the track 7.5 tons ML 9.0 tons HM 11.5 tons L > 100’ 9.0 ton L < 100’ Varying uniformly from 13.5 ton for L=0 to 9.0 ton for L=100’ Allowance for dispersion of horizontal forces BL L> 100’ 11.5 ton BL loading L > 100’ 7.5 ton L< 100’ varying uniformity for 11.25 ton for L= 0 to 7.5 ton for L=100’ For TE, L shall not be taken to exceed 120 feet for HM and 100 feet for ML and BL loading. In the case of single span bridges up to 80 feet long, the above allowances may be increased by 50 percent. The word single span was deleted in 1936. In all cases the net horizontal forces after deducting dispersion shall be assumed to be distributed equally amongst the spans in the length L as defined.” In 1934, an additional clause 137 was inserted as follows: “For the purpose of calculating tractive and braking force on existing Bridges Rules 22 shall be held generally to apply excepting that the maximum tractive forces can be calculated as 25% of the axle loads of the drivers of the actual engine under consideration and braking force as 20% of actual braked engine axle loads and 10% of other braked axles running or proposed. In addition where the table of dispersion allowance specifies BL, ML and HM, these shall be interpreted in terms of weight of rail in use on Volume - I 146 It is also noted that in Bridge Rules of 1941 printed in 1947, value of net longitudinal forces (after deducting dispersion allowance) were specified in Table given in annexure –F of Bridge Rules 1941 (Printed in 1947). In Bridge Rules 1941, Printed in 1960, A note to clause 23 as given below is added. “Note: In the case of piers and abutments of RCC bridges with ballasted floors, no longitudinal forces need be taken into account for span up to and including 20 feet.” Clause 33 was also added for existing bridges as follows: Tractive effort- 25% of axle load of coupled wheels of actual engines under consideration. Braking force- 20% of actual braked engine axle loads and 20% of other braked axle loads. Volume - I 147 Dispersion to approaches in tons if no rail expansion joint in track would be (a) 10 % of weight of rail in lbs/ yd for loaded length L equal to or exceeding 100’ and (b) increasing uniformly from 10% for L = 100’ to 15% for L= 0 for loaded length less than 100’ HM loading was deleted in 1964 Bridge Rules. Dispersion of tractive and braking force problem was investigated by field trials conducted on various bridges during 1967-1980. Based on these investigations, dispersion provisions of longitudinal forces was modified as under: “In case of bridges, provided with through welded rails, rail-free fastenings and adequate anchorage of welded rails on approaches (by providing adequate density of sleepers, ballast cushion and its consolidation etc. but without any switch expansion joints) dispersion of longitudinal load P through the rails away from the loaded length may be allowed to the extent of 25% of the value of longitudinal load P as obtained through Appendices VII , VIII and VIII(a) of Bridge Rules subject to minimum pf 16t for BG, 12 t for MGML and 10t for MGBL. This will also apply to bridges with jointed track with rail free fastenings but without any switch expansion of mitred joints. Where suitably designed elastometic bearings are provided the aforesaid relief may be further increased by 40% thereof”. Modified Broad Gauge loading of 1987 introduced in 1988 envisaged a maximum tractive effort of 100t with coupled operation with 25% braking load for locomotives and 20% braking load for trailing load of 8.25t/m which could be on both sides of locos. Tractive and braking force values for various spans for the standard loading have been given in the Bridge Rules. The provision of dispersion of longitudinal forces remains unchanged. The 20% braking force of Train Load was revised to 13.4 % in 1993. Heavy Mineral loading -1995, introduced in 1998 for heavy minerals routes, caters for maximum tractive effort of 135t and braking force of 13.4% of train load. The total dispersion under above clause shall not exceed the capacity of the rails for transferring the longitudinal load to the approaches nor should it exceed the capacity of the anchored length of the track on the approaches in resisting the longitudinal load. After dispersion of longitudinal forces, distribution of longitudinal forces in different supports is considered as given below: For span supported on sliding bearings, the horizontal loads shall be considered as being distributed between different supports as below: (a) 2 supports directly under the loaded span; each 40% of the horizontal load due to tractive/ braking effort after deducting force dispersed as stipulated in clause above . (b) Other two adjacent supports each 20%. For spans which have roller bearings at one end, the whole of the horizontal load shall be considered to act through the fixed end. Volume - I 148 Volume - I 149 Volume - I 150 Volume - I 151 CONDITION MONITORING OF BRIDGES FOR RUNNING HIGHER AXLE LOADS IN SOUTH EAST CENTRAL RAILWAY P. BARAPATRE*, AMIT GOEL** SYNOPSIS Railway Board has approved many sections for running of higher axle load trains. Five such Pilot Projects have been approved in SEC Railway also. Adequacy of existing infrastructure for carrying heavier loads has to be verified for these sections. The bridges, being generally very old, require special verification by checking of their designs, by verification of their physical condition and by instrumentation & condition monitoring of sample bridges. This paper deals with the various works done so far by SEC Railway for checking the bridges in identified sections. The paper also discusses the instrumentation, data recording and results of the first quarter test train run with 25 ton axle load BOBS wagons, on one Plate Girder Steel Bridge in Durg-Dallirajahara section. 1.0 GENERAL The growing demand of transport is one of the most important features of the developing economy. In developing countries like India, Railways attract major share of growing traffic, being the largest transporter. The volume of traffic on Indian Railway has increased manifold and a quantum increase is expected in the passenger and freight traffic in Indian Railways in the next few years. The growing traffic demands not only necessiate expansion of the existing infrastructure but also, the optimum utilisation of the existing infrastructure. One of the major steps taken for optimum utilisation of the existing infrastructure was to permit the heavier axle loads on the existing system. This also necessitated checking the capacity of the existing system and to strengthen it, if required. Indian Railways has taken up several pilot projects in recent past to upgrade many routes for running of higher axle loads. 2.0 ROUTES IDENTIFIED FOR HIGHER AXLE LOADS IN SEC RAILWAYS Railway Board has approved many routes for taking up the pilot projects for running of higher axle loads on identified routes on South East Volume - I 152 Volume - I * CBE/SEC Railway/BSP ** Dy CE/Bridge/ SEC Railway/BSP Central Railway as follows: 3.0 2.1 Indian Railways has gone for modernization and upgradation of various assets from time to time. In the field of Rolling stock the development of ABB loco, development of Electrical multiple unit with Air break suspension, high speed modern stainless steel coaches with very modern suspension system etc. have brought the Indian Railways almost at par with International standards. Similarly in the field of Signaling & Telecom, the Indian Railways have been pursuing the most modern systems with high import content and foreign collaborations almost on the sustained basis. Even in operations, Indian Railways is going in for large-scale implementation of freight operation information system. Upgradation of track has also been receiving considerable attention for the last few years. However, nothing concrete has been done in the field of Bridges mainly due to long life of the bridges and also due to the fact that upgradation of the bridges was always considered to be a very difficult task under the heavy traffic conditions. DURG- DALLIRAJHARA SECTION This route was approved in May 2005 for running of BOBS wagons with axle load of 25 tons. This line is about 88 kms. long and in a length of about 78 kms. between Marauda and Dallirajhara BOBS wagons are running presently with axle load of 22.9 tons. This line carries iron ore from iron ore mines, which are situated on one end of the section to Bhillai Steel Plant situated on the other end of this line. Running of increased axle load of 25 tons on this section is yet to be commenced. 2.2 JHARSUGUDA- KIRODIMAL NAGAR SECTION This section was approved in February 2006 for running of CC+8+2 loaded BOXN wagons. This section of about 80 kms. length is situated on Group ‘A’ route of Howrah- Mumbai. Running of increased axle load was commenced in April 2006 with iron ore traffic originating from Chakradharpur Division of South Eastern Railways. 2.3 KORBA-CHAMPA-BILASPUR-ANUPPUR-NKJ SECTION This section was approved in February 2006 for running of CC+6+2 loaded BOXN/BOBRN wagons. This section of about 404 kms. length is situated partly on Group ‘A’ route of Howrah- Mumbai and partly on other important routes. Running of increase axle load was commenced in April 2006 mainly with coal traffic originating from sidings near Korba in Bilaspur Division of S.E.C. Railway for further transportation to W. C. and other Railways. 2.4 DUMRI KHURD- KANHAN- KAPERKHEDA/KORADI SECTION This section was approved in February 2006 for running of CC+6+2 loaded BOXN/BOBRN wagons. This section of about 50 kms. length is situated partly on Group ‘A’ route of Howrah- Mumbai and partly consist of Assisted sidings. Running of increase axle load was commenced in April 2006 mainly with coal traffic originating from sidings near Dumrighurd in Nagpur Division of S.E.C. Railway for further transportation to two Power plants near Nagpur. 2.5 WHY TO CHECK BRIDGES Most of the bridges in the Indian Railways were constructed at the time of construction of the line and have never undergone any replacement, except for superstructure of some bridges. A large number of bridges in Indian Railways are having life close to hundred years. The older age, the need to permit higher density of traffic and also to permit heavier axle loads on these bridges necessitated rigorous analysis and checking of these bridges. In the approved routes of South East Central Railway also a very large number of these bridges are very old, with a large population of bridges constructed almost 100 years back. These bridges were constructed mainly with BGML loading standard. Checking the adequacy of existing bridges necessitated broadly following items: (1) Frequent physical inspection of the bridges with elaborate check list (2) Checking the design of substructure of the bridges for increased axle loads (3) Checking the design of superstructure of the bridges for increased axle loads (4) To identify the different works on short-term/ long-term basis required to be carried out for running of higher axle loads (5) Sanctioning and execution of various strengthening works (6) Identification of the sample bridges for carrying out instrumentation and condition monitoring JHARUGUDA- BILASPUR-DURG-NAGPUR SECTION This section was approved in June 2006 for running of 25 ton axle load. This section of about 615 kms. length is situated on Group ‘A’ route of Howrah- Mumbai. Running of increase axle load is yet to be commenced. Part of this route was sanctioned in the earlier sanctions for running of CC+6+2 and CC+8+2 ton wagons. Volume - I 154 Volume - I 155 3.1 (7) Carrying out instrumentation and condition monitoring for seeing the realistic response of some sample bridges for the increased axle loads (8) To assess the residual life of the bridges based on fatigue consideration CHECKING THE SUBSTRUCTURE OF THE BRIDGES FOR INCREASED AXLE LOADS The trains with increased axle loads have been approved at a sped of 60 kmph. The forces were calculated for Locos Double headed WDM2, Double headed WAG5 and Single headed WAG7, which are sanctioned for running in SEC Railway. Basic wind pressure was considered as 150 Kg/m2. It was found that EUDL was lesser than the EUDL considered as per BGML standards. Hence, checking of substructure was not considered necessary. However, sample checking was done for a few bridges. Subsequently, while processing the application to CRS for the proposed running of Double Headed WAG7 loco, it was found that only for the bridges of 80 feet and above clear spans, the longitudinal forces are exceeding the design longitudinal forces for BGML loading standard. Hence, for the bridges of clear spans more than 80’ replacement of bearings and jacketing of piers and abutments was proposed for sanction to cater for future requirements. 3.3 (b) For 20’ spans the bridges mainly are either steel girders or RCC flat top bridges of BGML standard or PSC slabs of MBG standards. Most of the steel girder bridges have been replaced either with PSC slabs or are proposed/sanctioned for replacement. (c) For the spans below 20 feet bridges are mixed type i.e. RCC flat top or small arches or PSC slabs or steel girders or some very small span bridges (span upto 3feet) with pipeline culverts. PHYSICAL INSPECTION OF THE BRIDGES For taking up the Pilot Project each and every bridge was considered as an individual entity. Divisions were advised to carry out the initial physical inspection of all the bridges and also to carry out physical inspections at an increased frequency and also to report the bridges warranting attention of the Headquarter and also to propose the works required to be carried out on the basis of these physical inspections. Divisions were also advised to keep a special watch on the weighment records and the over speeding of the trains. Headquarter had identified some sample bridges for instrumentation and condition monitoring. Divisions were also advised to suggest bridges for the condition monitoring basing upon their physical inspections. 3.2 Only some bridges are arch type or PSC girders with MBG standard. For the bridges of spans below 20 feet covered in item (c), the design checking was not done and continuance of these bridges has been permitted on the basis of their physical condition and with close monitoring. For the bridges of spans of 20 feet with PSC slabs covered in item (b), the design checking was not done since, 20 feet PSC slabs are of MBG standards permitting the increased axle loads. For the bridges of 20 feet span with steel girders, the design checking was done and the bridges were found suitable for increased axle loads. For the arch bridges of spans 20 feet and above covered in item (b) and (a), the design checking was not done and continuance of these bridges has been permitted on the basis of their physical condition and with close monitoring. For the steel girder bridges of different spans checking was done either by manual calculations or by development of in-house software or with the help of the guidelines provided by RDSO for different type of bridges. Development of Software for checking of steel girder bridges: For checking the bridges for increased axle load SEC Railway has developed software. This software is having following features: (a) It includes the calculation of EUDL for various formation of trains, locos and wagons (b) It can check the bridges for rolling stocks as per RDSO’s general criteria CHECKING THE SUPERSTRUCTURE OF THE BRIDGES FOR INCREASED AXLE LOADS (c) It can check Plate Girders for permissible stress criteria SEC Railway is mainly having following type of bridges in the identified routes for higher axle loads: (d) It can check the adequacy of Piers and Abutments (e) For Open web and under slung type bridges, it can only check stringers but it cannot check other members. Other members have to checked through conventional manual calculations (a) Volume - I For the spans above 20 feet the bridges are mainly steel bridges. 156 Volume - I 157 The development of the software was mainly done adopting following methodology/principles: 3.3.1 CHECKING OF BRIDGE FOR ROLLING STOCK FROM RDSO’S GENERAL CRITERIA (a) CHECK FOR MAXIMUM ALLOWABLE SPEED 3.3.2 CHECKING OF GIRDER FROM PERMISSIBLE STRESS CRITERIA 3.3.2.1 CHECKING OF PLATE GIRDERS (a) Calculation of dead loads including the self-weight of girder, weight of bracings, track load, weight of the fastenings. (b) Calculations of live load for given span, train formation and loco combination including dynamic augment. For the calculation of bending stresses, EUDL for BM and for the calculation of shear stresses, EUDL for SF should be adopted. (c) Calculation of maximum bending moment developed at mid span and maximum shear force developed near support. If W1 is the total of loads in (a) & (b) above for BM & W2 is the total of loads in 1&2 above for SF, then: This includes the calculation for maximum allowable speed on the basis of EUDL for bending moment and EUDL for shear force according to the formula Vm = [LLD+(LLDxID)-LLT] Vt/(LLTxID) Here, LLD is EUDL for BM/SF for which the bridge has been designed. ID is CDA for 125 Kmph speed. LLT is EUDL for BM/SF for which these bridges are required to be checked. With above calculations, the maximum allowable speed is to be taken as 90 % of the minimum of Vm obtained for Bending Moment and Shear Force. The maximum allowable speed should be greater than the proposed speed. (b) BM=W1xL/(2x8)=M SF=W2/4=V (d) Calculation of section properties of girder such as its moment of inertia (Ixx), gross area (Ag) and net area (An) (e) Calculation of bending stresses on the basis of gross area and net area REQUIRED PERCENTAGE STRENGTH OF BRIDGE This includes the calculation of %age strength required for the span to move the train at a speed of 10% more than the proposed speed on the basis of EUDL (BM) and EUDL (SF) according to the following formula: 6g=(Mxy/Ixx). 6n= (6gxAg)/An (f) Comparison of net area stress with permissible tensile stress for fatigue criteria. Permissible stress from fatigue criterion can be obtained from the appendix –G of Steel Bridge Code on the basis of no. of cycles and the ratio of fmin./fmax..Here fmin. Is the stress developed in girder without live load and fmax. Is the stress developed in girder with live load. AVAILABLE IMPACT FACTOR (g) CDA available shall not be less than 0.1.This can be calculated by the formula [LLD+(LLDxID)-LLT]/LLT. Calculation of shear stress by: ô= V/ (dxt). Here d & t are the depth and thickness of web of the girder. (h) Maximum CDA available is taken as minimum of the values obtained on the basis of EUDL (BM) & EUDL (SF). Comparison of shear stress calculated above with the permissible shear stress. (i) Checking for wind forces. % Strength required = [LLT+(LLTxIT)] x 100 / [LLD+(LLDxID)] Here, IT is CDA for speed 10 % more than the proposed speed. Maximum % strength of the bridge is to be taken as maximum of above two values. (c) (d) LONGITUDINAL FORCE CRITERIA 3.3.2.2 CHECKING OF OPEN WEB GIRDERS AND UNDER SLUNG TYPE GIRDERS It includes the comparison of induced tractive effort/braking force (max. of the two) with the longitudinal force for which the bridges have been designed. As per clause 2.8.3.2 of IRS Bridge Rules, dispersion may be allowed to induced longitudinal force. Volume - I 158 (a) Checking of stringer and cross girder can be done as per the criteria given above in Para 3.3.2.1 (b) Checking of the other members of Open web and Underslung type girders can not be done by the software. These were Volume - I 159 checked manually as per following methodology: (b) Calculation of direct load, bending moment about X-X axis and bending moment about Y-Y axis on the basis of dead load, live load, longitudinal forces due to water currents, forces due to wind and seismic forces etc. (c) Calculation of combined maximum and minimum stress at the base of pier due to direct load and bending moments calculated at (b) above. (d) Comparison of stress calculated in step (c) above with the permissible stresses. (j) Drawing of influence lines for all the members of open web girder. (ii) Calculation of dead loads for unit length, live loads for unit length, longitudinal forces, wind forces including portal and sway effect. Span for calculating EUDL should be decided on the basis of influence lines. (iii) Total dead load, live load, may be obtained by multiplying the corresponding loads for unit length by the corresponding area of influence lines. 3.3.3.2 CHECKING OF ABUTMENTS (iv) Calculation of actual axial stresses developed in compression members on the basis of gross area. (a) Calculation of vertical loads, earth pressure, surcharge loads and moments of all these forces about toe of abutment base. (v) Calculation of actual axial stresses developed in tension members on the basis of net area. (b) Checking against overturning. Factor of safety against overturning >2. (vi) Permissible compressive stresses for compression members may be calculated on the basis of its slenderness ratio or fatigue criterion, whichever is critical. (c) Checking against sliding. Factor of safety against sliding>1.5. (d) Checking for maximum and minimum base pressure. (vii) Permissible tensile stresses for tension members may be calculated on the basis of fatigue criteria or normal permissible tensile stresses, whichever is critical. (viii) Comparison of actual compressive stresses with permissible compressive stresses and tensile stresses with permissible tensile stresses. (ix) Calculation of bending stresses in the top chord members in case of that type of under slung girders in which sleepers are resting directly on the top chord members. (x) Checking for combined direct and bending stresses according to: (6c’/6c)+(6b’/6b)< or = 1 3.3.3 CHECKING OF SUBSTRUCTURE BY SOFTWARE The software can also be used for checking of Piers and Abutments. For checking of Piers and Abutments following principle/methodology was followed in the software: 3.4 RDSO GUIDELINES FOR RUNNING 25 TON AXLE LOAD In the month of August 2006 RDSO has issued EUDLs for different spans for 25-ton axle loads and has also given clearance to certain standard RDSO type spans for running of 25-ton axle loads at a speed of 60 kmph. Hence, for running of 25-ton axle loads between Jharsuguda- Bilaspur-Durg and Nagpur, which was approved at a later stage, the standard spans cleared by RDSO were not further checked. Only, those spans, which were not included in the RDSO list, were checked at Zonal level. 3.5 RESULTS OF CHECKING OF DIFFERENT SPANS Basing upon the checking of different type of girders either through software or manually following results were found for different type of spans: (i) 40’ plate girders confirming to RDSO Drg. No. BA 1056,BA11003 have been found safe. (ii) 60’ plate girders confirming to RDSO Drg. No. BA 1057,BA11004 CE’s Drg. No. 10291, CE’s Drg. No. 11237 have been found safe. (iii) 80’ plate girders confirming to RDSO Drg. No. BA 1058 has found to be safe. 3.3.3.1 CHECKING OF PIERS (a) Volume - I Calculation of sectional properties of pier. 160 Volume - I 161 (iv) (v) 4.0 100’ girders confirming to RDSO Drg. No. BA 1059(Plate Girder), BA 6081 (Under Slung Type) has been found safe. 100’ plate girder confirming to RDSO Drg. No. BA 1518 has been found safe for 50 Kmph speed in 2 million cycles for BOBS wagons running in Durg- Dallirajhara section. This type was found unsafe for 10 million cycles. These types of Plate Girders have been found safe by RDSO for BOX-N wagons at a speed of 70 kmph. 150’ open web girders confirming to RDSO Drg. No. BA 5025,BA11101 have been found safe. Cross Girders and Stringers for these girders have also been found safe. BRIDGE WORKS PROPOSED Basis upon the design checking and other items like type of bearings provided on the existing bridges, following works have been mainly planned in SEC Railway: 4.1 First Pilot project in this Railway was approved in Durg- Dallirajhara section in May 2005 for running of BOBS wagons with axle load of 25 tons. Following bridge works were sanctioned in Pink Book 20052006 for this section: (a) 4.2 Regirdering of Br. No. 105 with Spans 5 x 100’ and 2x40’. This bridge is provided with Plate girders and as per the theoretical calculations, 100’ girder failed from fatigue consideration. Hence, all the spans were sanctioned for regirdering. However, as per the instrumentation and data recording for the first quarter, actual stresses are measured to be much lesser than the theoretical and fatigue stresses. It appears at this stage that with further analysis after getting second quarter report and after getting the report for residual life analysis, this bridge, may not require replacement and should be able to give service for many decades with increased axle load. (b) Strengthening of Plate Girder bridges of 40’ span by providing thicker flange plates. (c) Minor works of strengthening on small span bridges. On Jharsuguda- Bilaspur- Durg- Nagpur section running of 25 ton axle load wagons has been approved in June 2006. On this route, bearings are provided as per BGML standards and no replacement or strengthening was done for running of Double headed WAG7 wagons. For the bridges of spans 80’ and above replacement of bearings and Volume - I 162 jacketing of Piers and abutments has been proposed for sanction in PWP 2007-2008. 5.0 NEED FOR INSTRUMENTATION OF BRIDGES The theoretical checks done are based on many assumptions and therefore the actual stresses are best checked by means of instrumentation. There are many items of concern, which are not even amenable also to theoretical analysis. Similarly, the inspection carried out in the conventional way are more of subjective in nature and to have an objective assessment of the deterioration in the condition of bridge, or otherwise, due to running of enhanced axle loads, it is required that condition monitoring of sample bridges is done by suitable instrumentation. Instrumentation is a monitoring mechanism to assess the effects of increased longitudinal loads & axle loads with respect to measurement of settlement of foundations, tilting of piers/abutments, loads on bearing, deflections & stresses at critical points. The monitoring mechanism consists of using various instruments & equipments to record desired data at different intervals and analysis of the same. 5.1 IDENTIFICATION OF THE BRIDGES FOR CARRYING OUT INSTRUMENTATION AND CONDITION MONITORING Due to the increased axle loads, the existing bridges will be subjected to additional vertical & longitudinal loads. In order to assess the actual effect of these increased loads on the bridges, Instrumentation and measurement of actual stresses developed in the different members is essential. This is also required for assessing the residual life of the bridges & for planning the Rehabilitation/Strengthening of the same. As per the different instructions issued by the Railway board sample bridges containing different types/spans and also the vulnerable bridges were to be selected for instrumentation and condition monitoring. Bridges for instrumentation were identified mainly on following considerations: a. Type of girders b. Type of spans c. Age of the bridge d. Condition of the bridge e. Distribution of selected bridges in the different routes identified for higher axle loads Volume - I 163 5.2 BRIDGES SELECTED FOR INSTRUMENTATION AND CONDITION MONITORING IN SEC RAILWAY (a) Type and detailing of structure (b) Type of loading (1) Bridge No. 105 in Durg- Dallirajhara section (5x 100’ + 2x40’ Steel Plate Girders) Numerical model is necessary for having an idea about the following: (2) Bridge No. 154 Up in Jharsuguda- Kirodimal Nagar section (4x 20’ Arch + 1x100’ Steel Plate Girder) (3) Bridge No. 91K in Bilaspur- Anuppur section (7x60’ Steel Plate Girders) (4) Bridge No. 175 K Up in Anuppur- Katni section (2x100’ Under slung type Steel Girders) (5) Bridge No. 34 Up in Kamptee- Kalumna section (2x 60’ Steel Plate Girders + 7x150’ Open Web type Steel Girders) 7.2 6.1 6.2 7.3 7.0 THE STAGES/ACTIVITIES INVOLVED IN INSTRUMENTATION & MONITORING 7.1 PREPARATION OF REALISTIC COMPUTATIONAL NUMERICAL MODEL Realistic computational numerical model is prepared by using finite element analysis system (structural analysis system, SAP) depending upon the following aspects: Volume - I 164 (b) To have an idea about the theoretical values of various dynamic characteristics of the structure. (c) The model is used to identify the critical components/ locations for doing the instrumentation. CARRYING OUT STRUCTURAL ANALYSIS ON NUMERICAL MODEL (a) Theoretical maximum stresses in bending (b) Theoretical maximum stresses in shear. (c) Theoretical maximum deflection CARRYING OUT INSTRUMENTATION & RUNNING OF TEST TRAIN ON BRIDGE This is done to know about: One tender was initially invited for Bridge no. 105 in Durg – Dallirajhara section in March 2006. This was awarded to M/s Sharma Associates in April 2006. So far, Data Recording for two quarters has been done in this contract and report has been received for the Ist quarter testing. Second tender was subsequently invited for four different bridges in May 2006. This was also awarded to M/s Sharma Associates in August 2006. So far, Data Recording for first quarters has been done on one bridge in this contract and on the second bridge it is planned to be done in second week of December 2006. To have a knowledge about the theoretical values for stress/ strain, deflection etc. at various points of the structures. This is done for both static and dynamic load cases under IRS Bridge Loading Standards and proposed heavier axle loads and increased longitudinal load cases. Basing upon this, values for the various parameter like: 6.0 FIXING AGENCY FOR INSTRUMENTATION Railway board had issued some guidelines in the month of June 2005.With these guidelines Board had pinpointed some specialized agencies for different zones. Since no particular agency was suggested for SEC Railway, a Special List containing all the nine specialized agencies suggested by Railway Board in their different letters was prepared and Special Limited Tenders were invited from these nine tenderers. Summary for the tendering is as follows: (a) 7.4 (a) Actual maximum bending strains/ axial strains at different critical locations.. (b) Actual maximum deflection at centre of girder. (c) Actual maximum shear stress at support in case of plate girders. (d) Actual longitudinal strains in rails and on bearings (e) Actual settlement and tilt of pier. (f) Strains in pier (g) Forces in members showing signs of distress/ failure etc. DATA COLLECTION, RECORDING AND STORAGE This process will be done quarterly. In this process the data is recorded using data loggers of required channel capacity. In SEC Railway, recording has been done using data logger for 32 channels. The raw data is stored in a computer or transmitted to the server. The data is then processed using suitable softwares. After analysis and interpretation, reports are prepared which give the value of various parameters and their exceedance, if any recorded. Volume - I 165 7.5 ASSESSMENT OF RESIDUAL LIFE OF THE BRIDGE Residual Life Analysis of the Bridge is done based on selection of appropriate SN Curve, which depends on the physical and metallurgical properties of the material. In the SN curve N is the number of cycles, of the varying stress. Value of ‘N’ for any given material depends on ‘S’ (Stress range), which is difference of maximum and minimum stress in fatigue cycle. Due to introduction of higher axle loads, the stress range for any member will increase, which will reduce the fatigue life of the member. By strain gauging of the various members of the steel girder, the number of cycles due to any trainload passage and the stress pattern get recorded. Using this information, the fatigue damage done by one train passage is arrived at. The traffic history over the bridge is required to be obtained to get the fatigue damage already done for the steel structure and this is used to get the residual fatigue life for various members. 7.6 (iii) Weighment of the test wagons was done in advance with the help of On line weigh bridge at Marauda. 25 ton axle load on each wagon was esured. (iv) Instrumentation was commenced 4 days prior to the test runs. (v) Special sanction of CCM and General Manager was obtained for overloading of the wagons without any punitive charges. Sanction of COM was taken in advance for arrangement of special test trains and for arrangement of traffic blocks. (vi) For helping the test trains to run at proper speeds and to carry out proper braking, Sectional PWI and Loco Inspectors were associated with and deputed with the test runs. (vii) RDSO has been associated with the recording for the second quarter on this bridge and onwards for further testing and analysis. (viii) For measurement of the vertical forced the test train was run at following speeds: SUBMISSION OF REPORTS Perform numerical structural modification studies to determine costeffective rehabilitation/ strengthening scheme where instrumentation analysis indicates abnormality in the structure. This item has been kept as an optional item for execution. 7.7 length test train (running in this section) of 48 numbers of BOBSNM1 wagons loaded with 100 ton (25 ton axle load) was used. In addition data recording was done for the normal revenue and passenger running trains for 3 days. Continued monitoring for two years. This item has also been kept as an optional item for execution. (a) 5 Kmph (considered as static loading) (b) 15 Kmph INSTRUMENTATION AND DATA RECORDING OF BRIDGE NO. 105 IN DURG – DALLIRAJHARA SECTION (c) 30 Kmph For Bridge no. 105 in Durg-Dallirajhara section testing/ data recording with test trains for the first quarter was done on 25.05.2006 and 26.05.2006 and on 17.11.2006 and 18.11.2006 for the second quarter. Following are the salient features of these recordings: (e) 60 Kmph 8.0 (i) The section involved is a non-electrified section. For the first quarter recording Double headed WDM2 locomotives were used. For the second quarter recording Double headed WDG3A locomotives were used. (ii) For the recording of vertical forces test train consisting of 10 numbers of BOBSNM1 wagons loaded with 100 ton (25 ton axle load) was used. For the recording of longitudinal forces full Volume - I 166 (d) 45 Kmph (f) 75 Kmph (ix) For measuring the longitudinal forces i.e. Braking forces and Tractive Efforts, the test train was run with Double Headed Locomotives with 48 numbers of BOBSNM1 wagons loaded with 25 ton axle load. The test was done as per the guidelines issued by RDSO. (x) Summary of Instrumentation: Instrumentation was done for both type of Plate Girders i.e. for 30.5 m. span and 12.2 m. span. A summary of the instrumentation done for the first quarter recording is as follows: Volume - I 167 9.0 DISCUSSION OF RESULTS Detailed results in tabular and graphical form for the data recorded for the first quarter recording have been received in the report. However, for easy understanding a comparison of the recorded stresses with respect to the result of numerical modeling and with theoretical and permissible stresses has been shown in the following tables: (a) Bending Stresses at Mid Span : (b) Shear Stresses : In the testing for the second quarter bearings were also instrumented for measurement of longitudinal forces. Volume - I 168 Volume - I 169 l Numerical analysis of the bridge under MBG loading indicates that even considering full CDA, the bridge has reserve capacities for both the 30.5 m and the 12.2 m span. 10.1 SUPERSTRUCTURE BEHAVIOR l Initial review indicates that the bridge spans are performing as expected under the test train. Dynamic behavior is consistent with that of similar bridges. l Review of primary bending behavior of the 30.5 m spans indicates that the measured values of bottom flange strain represent about 63 - 68 % of the theoretical value of strain, and that the shape of the time history is consistent with the expected theoretical time history. l For the 12.2 m spans, the measured values of bottom flange bending strain represent about 82 . 85 % of the theoretical value of strain, and the shape of the time histories was consistent with the expected theoretical history. l Similarly, for shear strain at .d. from the support, the measured shapes were consistent with the expected theoretical shapes, and the measurements were about 62% and 79% of the theoretical value for the 12.2 m and 30.5 m girders, respectively. l At higher speeds, increases in strain and deflection levels were noted. These dynamic augment percentages are significantly lower than the design values. l The peak stresses measured under the test train (25 ton axle loads) were significantly lower than the allowable stresses for the superstructure. 10.2 PIER/SUBSTRUCTURE BEHAVIOR 10.0 SUMMARY OF FINDINGS AS PER THE FIRST QUARTER RECORDING FOR BRIDGE NO. 105: As per the first quarter recordings of the Bridge No. 105, following conclusions have been submitted by the specialized agency: l Volume - I l It was indicated that the piers and corresponding substructure are in good condition and experienced no significant movement under train passage. l Vertical pier deflections, pier strains and pier tilt were all very small and no significant movement was noted. l The piers and substructure performed well under the test train with 25-ton axle loads. 10.3 LONGITUDINAL FORCE EFFECTS l Bridge 105 was instrumented, tested and analyzed, to evaluate its suitability for carrying heavier axle loads. 170 Volume - I No notable changes in pier behavior were observed during the braking and acceleration runs also, implying that no significant longitudinal forces were being transferred to the substructure. 171 l Review of forces measured on the rail indicate that peak forces of about 15 tons were generated under normal rail operations and peak forces of about 29 tons were generated during the traction and braking runs. ASSESSMENT OF LOAD CARRYING CAPACITY OF MASONRY ARCH BRIDGES l It was also observed that the forces generated were being transferred by rail off the bridge and that the proportion transferred through the bearings was very small, as expected. CHAHATEY RAM*, SANJAY BANERJEE** SYNOPSIS 11.0 CONCLUSION (1) Results of first quarter report of the instrumentation and Bridge monitoring for Bridge No. 105 are encouraging. The recorded actual stresses are in general lower than the theoretical and permissible stresses. It appears at this stage that with further analysis after getting second quarter report and after getting the report for residual life analysis, this bridge, which was proposed and sanctioned for replacement, may not require replacement and should be able to give service for many decades even with increased axle load. (2) Due to the changing scenario in the Indian Railways and in order to optimum utilisation of assets, adoption of higher axle loads is a beneficial preposition. To analyse the existing assets, carrying out such scientific studies is very useful and essential. Such rational studies will save huge assets from avoidable and premature condemnations and replacements of assets. This will not only save a lot of unnecessary expense of efforts and money but, will also help Indian Railways in mopping up additional revenue by effective and optimum utilisation of existing resources. Indian Railway has recently gone for an increase in the axle load of freight trains on nominated routes to enhance throughput which has become critical in view of fast growing economy of the country. It is therefore essential that all existing bridges are assessed for their load carrying capacity. Masonry arch bridges form a substantial percentage (about 20%) of all the bridges in Indian Railways. There are no analytical methods available in Indian Railways on date for assessment of these bridges. This paper brings out briefly the problem which differentiates masonry arch bridges from other bridges, lists various methods of analysis and describes two methods in details viz. RING 1.5 & Modified MEXE Method. INTRODUCTION Indian Railways is one of the oldest Railway systems in the world, its origin dating back to 1850s. Over the years, Indian Railways has undergone tremendous changes in geographical and spatial spread, quantum of passenger and freight traffic, trailing load density and speed. One of the oldest civil engineering assets of Indian Railways are its bridges which vary widely in span, type, material of construction and age. Although a large number of the bridges have been built after independence, a vast majority are more than 100 years old. Among the bridges which are more than a 100 years old, stone and masonry arch bridges are the most common. This paper brings out the results of the analysis carried out over a variety of stone / masonry arch bridges and compares the results with their physical condition based on actual field inspections. PURPOSE O F THE ASSESSMENT The present loading standard of Indian Railways is Modified Broad Gauge(MBG),1987, which stipulates a design axle load of 25 tonnes. Volume - I 172 Volume - I * Chief Bridge Engineer, Eastern Railway ** Deputy Chief Engineer/Bridge, Eastern Railway However, the actual axle load of BOXN being run at present is 20.32 tonnes. The analysis was prompted by a recent Railway Board decision to permit certain degree of overloading over specified routes, which means an axle load of 22.83 tonnes. Also in the very near future, an actual axle load of 25 tonnes is to be introduced. WHAT IS ASSESSMENT Assessment is the procedure by which a Civil Engineering structure is subjected to systematic and scientific study by appropriate method so as to quantitatively determine the safe load carrying capacity. This also includes adequate understanding about the behavior of the structure and its material of construction. It also involves judging the effects of the assumptions made prior to the analysis so that the end result is as free from any bias as possible. Assessment also involves correlating the physical condition of the structure with the results of the analysis in order to achieve a realistic prediction of the load carrying capacity. Assessment is also a tool which is used to diagnose the structural inadequacies to arrive at the extent, nature, technique and quantum of repair / rehabilitation required. being non- homogeneous, the properties of the masonry arches depend on the properties of the masonry unit (i.e. brick / stone), the mortar (i.e. lime/ surkhi/cement) and most importantly the properties of the bond at the interface of the mortar and the masonry unit. It is extremely difficult to apply the principles of elastic theory on such non- homogeneous material. Moreover, arch bridges are indeterminate structures to third degree. Besides, the shape of the arch also plays a crucial role in determining its final load carrying capacity. In addition to the above, presence and depth of backing and the depth of fill considerably modifies the load dispersion over the arch barrel. Since this load dispersion finally determines the actual load intensity coming on the arch, the more accurate is the assumed load dispersion mechanism in the analysis, the more realistic will be the final output. However, there is no consensus world wide regarding the actual load dispersion mechanism of the arch. Finally, the effect of parapet walls and track, which significantly increases the strength of the arch do not figure in any of the analyses. Because of the above mentioned facts the results of analysis of masonry arch bridges are very much difficult and debatable. DIFFERENT LEVELS OF ASSESSMENTS OF MASONRY ARCH BRIDGES WHY ASSESSMENT OF ARCH BRIDGES ARE SO DIFFICUL AN D DEBATABLE ? The first step of assessment of any civil engineering structure is analysis of the structure. Depending on whether the structure is determinate or indeterminate, several rational methods are available for analysis. Determinate structure is easy to analyze as they offer easy solution by simple application of the equation of statics alone. However, indeterminate structures pose a greater level of difficulty as they cannot be analyzed by application of the equation of statics alone. They are required to be analyzed by other methods like double integration methods, strain energy methods, elastic centre method, column analogy methods etc. All these above rational methods are based on certain assumptions e.g. a) The material of the structure is homogeneous and isotropic. b) Elastic theory is applicable. c) Plain sections remain plain before and after bending. The different levels of assessment of masonry arch bridges along with the methods used and the use the results are put to are given in the table below :- Masonry arch bridges do not conform to the above assumptions. Firstly, masonry arch bridges are neither homogeneous nor isotropic. Secondly, Volume - I 174 Volume - I 175 STAGES The project commenced with the detailed field inspections of all the bridges falling over the routes specified for this Railway by the Board. The requisite data for the analysis were either obtained from the available completion plans or were accurately obtained from field inspections. In all 40 arch bridges were analyzed. Volume - I 176 Volume - I 177 abutments are fed in. The number of blocks mentioned herein is the number of layers of brick/stone masonry which constitutes the pier/abutment along its height. The span tab is used to enter the values of clear span, rise of the arch above springing and the number and thickness of arch rings. SOFTWARE The analysis were carried out by two methods A) RING Software version 1.5. B) Modified MEXE Method. DISCUSSION ABOUT THE SOFTWARES A) The next tab pertains to material properties through which data relating to material properties are fed in. This involves data related to soil in the cushion, the properties of the stone/ brick masonry i.e. the structural component of the arch bridge and the properties related to interaction between soil and brick/stone masonry. RING Software Version 1.5 - This Software has been developed by Sheffield University, U.K. The same is available in downloadable form through internet at www.shef.ac.uk/ring. The software is available for use free-of-cost and may be downloaded and used directly by any user having access to an internet connection. The higher version of this software i.e. RING version 2.0 is presently under development as part of an on-going UIC project on masonry arch bridges and is likely to be available for use to users by 2007. The next tab relates to loading. MBG loading as defined in IRS Bridge Rules has been used for the purpose. More details have been provided in the subsequent paragraphs. The next part involves the analysis based on the data fed in so far. By clicking on the analysis tab, the software performs iterations for the various load cases and gives load position at failure and the minimum load factor at that condition. There is an option of viewing the report on the analysis, which when clicked generates the report after analysis about the Bridge in MS Word. This report also contains the diagram of the condition of the bridge at failure. The load factor indicates the factor of safety against such failure for that particular load case. Downloading of the software automatically creates a link on the desktop of the PC in which the software is downloaded. By double clicking on the said link, one window opens which prompts the user to choose from among the following three options: 1. Create a new Bridge Project. 2. Open the existing Bridge project. B) 3. Open the recently access Bridge project. Choosing the first option creates a new window, which is basically used for entry of data regarding the bridge, loading etc. The other options are meant for opening existing bridge projects already analyzed by this software as the name suggests. The first tab pertains to geometry of the bridge through which the dimensional details of the bridge is entered. By clicking on the geometry tab, the first window opens which is used to enter the global parameters like number of spans, whether abutments and piers shall be considered in the analysis, whether effects of backing would be considered, width of the bridge to be analyzed and the depth of fill. The parameter “width of the bridge to be analyzed” is described more elaborately in the subsequent paragraphs. Once these values are entered, the next tabs are clicked thereafter one-by-one and data related to height, bottom and top width and number of blocks of piers and Volume - I 178 Modified MEXE Method – The first engineering method for the assessment of masonry arch bridges was developed by Pippard and Ashby (1939)and Pippard (1948) and was used extensively during the second world war for military purposes for quick assessment of load carrying capacity of these bridges. Full scale tests were conducted in 1950s and the resulting knowledge was consolidated. As a result of this research, MEXE method was established. A computerized version of this method finds mention in UIC Code 778-3R, Appendix – 4. Details of this method are described in subsequent paragraphs. ASSUMPTIONS The various assumptions made in the analysis are given below:RING Software Version 1.5. – The input file consists of four parameters viz. i) Geometry, ii) Material, iii) Loading and iv) Advanced. Volume - I 179 • • • • Volume - I In the Geometry, one very important parameter is the “width of the bridge to be analyzed”. Since the load imposed over the fill of the arch ring shall be distributed under the sleepers, it has been assumed that there is a 1:1 dispersion in the transverse direction, i.e. perpendicular to the track and the width at top of arch ring extrados has been assumed to be the width for which the bridge is to be analyzed. The other input data are quite evident. However, all the analyses have been done without considering the effect of backing which the software caters to in sufficient details. A) d = thickness of arch h = fill from top of arch L = clear span rc = rise of arch above springing rq = rise of arch above springing at quarter point. Fsr = Span/Rise factor. Graphs are available for calculating Fsr values for Span/Rise ratios more than 4. For Span/Rise ratios less than and equal to 4, Fsr = 1.0. In case of Materials, the input parameter seeks to input the data required for the masonry and the backfill material. For masonry, the default values of unit weight, coefficient of friction (both tangential and radial) have been adopted. For arch bridges which appeared to be physically sound upon inspection the default value of crushing strength of masonry of 10 kn/m2 have been assumed but for those which inspection revealed some sort of physical distress, a value of 5 KN/m2 have been assumed. The default values of solution convergence tolerance and maximum number of iterations have also been retained without change. For backfill, the default values of unit weight, limiting fill friction along with Boussinesq type load distribution with a limiting angle of 0.524 radians have been retained. A classical horizontal pressure distribution has been assumed with a coefficient of earth pressure of 0.33 with the option of automatic selection of passive pressure zones. Fp = Profile factor. If rq/ rc value is less than equal to 0.75, then Fp = 1.0. If rq/ rc value is more than 0.75, then Fp = 2.3{(rc - rq)/ rc}0.6. Fb = Barrel factor. This can be obtained from a given table which gives a value depending on the material of construction of the arch and also the physical condition. Ff = Fill Factor. This can be obtained from a given table which gives a value depending on the material of the fill. Fw = Width factor. This can be obtained from a given table which gives a value depending on the width of joints in masonry. Fmo = Mortar factor. This can be obtained from a given table which gives a value depending on the condition of the mortar in masonry. Fd = Depth factor. This can be obtained from a given table which gives a value depending on the condition of joints in masonry. In case of loading, a load vehicle has been created as per MBG1987, as per the details of axle load and its distribution as in Annexure – I. This load vehicle has been assumed to cross the bridge and several load cases depending on the span of the bridge have been created by incrementing the load vehicle position by 300 mm. There is very critical parameter to be considered while creating the load case, i.e. the width of the load. After careful consideration, since the load will act on the fill through the sleepers, it was decided to adopt the value of the width of the sleepers across the track i.e. in the transverse direction as the width of the load. Fj = Joint factor. This factor is a product of the above three factors specifying the effect of joints. Fj = Fw x Fmo x Fd. Fcm = Condition factor. This can be taken between 0.0 and 1.0 (lower limiting value excluded) at the discretion of the engineer depending on the condition of the bridge as a whole, the highest value to be adopted for a sound arch. A provisional value of permissible axle load over the bridge PAL is calculated from the following formula, PAL = 740 {(d+h)2/L1.3} or 70 tonnes whichever is less. The modified value of the axle load PALm is calculated as follows, PALm = PAL x Fsr x Fp x Ff x Fj x Fcm. The default values of advanced properties regarding constraints and block weights/pressures have not been touched. 180 There is no scope for any assumptions in the Modified MEXE Method as all the input parameters are well defined. The input parameters are as follows:- The tables mentioned above are given below for ready reference: Volume - I 181 # Interpolation between these values is permitted, depending upon the extent and position of the joint deficiency. Volume - I 182 Volume - I 183 ANALYSIS RESULTS substructure making the slenderness effects more pronounced. Decrease in depth of fill and width of bridge shall increase the intensity of applied load over the arch leading to decrease in load factor. The results of the analysis is available at Annexure – II. The summary of the results are given in tabular form below:- DISCUSSION & COMMENTS A) RING Software Version 1.5 – After analysis of 35 arch bridges by this software, it is seen that total 4 nos. bridges are having load factor less than 1.0. The failure mechanism after analysis of three of these four bridges are enclosed as Annexure – III (a to c). It is seen that the failure mechanism in all these cases pertains to tilting failure of abutments/piers while the actual arch ring maintains continuity even at failure. The values of the critical parameters of these four bridges are given in tabular form below:- From the above, it is seen that these bridges are all having span equal to or greater than 3.66 m. Bridge Nos. 3 is having quite high abutments which is the principal cause of low load factor. Although other three bridges are not having very high abutments, their principal cause of failure are low depth of fill and low value of the width of the bridge where the load is acting or a combination of these two factors. Hence it may be concluded that this software is extremely sensitive to the three values of height of abutments, depth of fill and width of the bridge over which the load acts. This is of course most consistent with the general expectations of actual behavior of arch bridges. Increase in height of abutments will adversely affect the stability of the Volume - I 184 The calculations have been performed without taking into account the effect of backing. Also, RING software does not take into consideration the effects of parapet walls and track, all of which significantly increase the strength of the arches to the extent of nearly 30 to 40%. Moreover, the tilting failure of abutments is extremely unlikely in actual scenario and at the present moment, none of these bridges show any visual signs of distress. It is felt that these bridges shall offer sufficient advance warning before failure which can be safely detected in the periodicity of inspections of these bridges. Hence these bridges are considered as safe. From the above discussions it may be concluded that the program in its present version suffers from the following limitations: • Modeling of piers and abutments – While data feeding, provision of feeding batter angles must be available to ensure that the exact profile of the abutment / pier both at the front and back is ensured. This is not possible at the present moment as the resultant diagram always gives uniform batter angles to both ends which is not always so for the actual bridges. • All material properties should be available in tabular from to ensure that the user makes a judicious choice on the basis of field observations and avoids using the default values arbitrarily. For example, soil properties may be given in tabular form against the various types of soil to enable the user to make an informed choice about the engineering properties from the basic input from field i.e. soil type. • The effects of multiple tracks over a single arch and parapet walls needs to be added with their corresponding input parameters to make the result more realistic and reliable. This is so as considering the effects parapet walls will serve to increase the strength of the arch, multiple tracks over the same arch shall increase the intensity of loading due to load coming from the adjacent track and shall considerably reduce the load factor. In fact this is one of the most serious limitations of the software. Volume - I 185 • If the height of the fill is large and if the number of blocks in a ring exceeds the limit set in the program, it is seen that no convergent solution can be found. This places a severe restriction on the range of applicability of the program and hence these input parameters need considerable upward revision to ensure that all arch bridges irrespective of input parameters may be analyzed by this software. This method do not take into consideration the effects of 3D, track, backing and parapet wall which significantly increases the strength of arches. Moreover, 8 out of 10 of these bridges do not show any sign of distress and are hence considered safe. It is also mentioned that all these 10 bridges are having load factor much greater than 1.0 by the RING software. Only two bridges out of these 10 have some visual signs of distress. • A database is required to be developed containing the data of the bridges analyzed supported by full scale load tests over the same bridges to check and verify the reliability of assessment by this software. Apart from the above, the method has the following serious limitations: B) Modified MEXE Method - After analysis of 35 arch bridges by this software, it is seen that 10 Nos. bridges are having permissible axle load less than 25 tons. “Modified MEXE Method” being a totally empirical method, no failure mechanism is obtainable and the method merely gives a rough assessment of the permissible axle load over the arch bridge. The values of the critical parameters of these ten bridges are given in tabular form below:- • The method completely ignores the effects of piers and abutments in its calculations. This makes the results obtained by the method doubtful as the effects of the piers and abutments on the overall assessment of the load carrying capacity of arch bridges are significantly large to ignore. • The method is applicable only to single span arch bridges. The effect of multi span which considerably modifies the behavior of the arch bridges being completely ignored again renders the assessment of results liable to grave doubts. So the results obtained by this method for Bridge Nos. 515, 501, 491, 482 & 480 are most doubtful. • This method is also silent about the arch bridges carrying multiple parallel tracks. In fact, width of the bridge which is important in view if the fact that it gives the intensity of load is completely ignored. Here in case, results obtained for Bridge Nos. 515 to 480 are also most doubtful. SUMMING UP 1. It is quite evident from the above table, that the combination of these four factors i.e. span, arch ring thickness, rise of arch and permissible axle load are the most critical in this method. This is in line of what is expected. Although these input parameters are exact, other input parameters are the condition factors which involve heavy subjectivity and may vary considerably from user to user leading to doubts about reliability. Volume - I 186 Volume - I Although these methods provide an excellent tool for first level assessment of arch bridges to the extent that the load carrying capacity is quickly assessed, too much importance should not be given on the results obtained. The results need to be interpreted taking into account the assumptions made in the analysis. They should also be correlated with the physical condition of the bridge found on inspection. The assessment should in NO WAY be an alternative to VISUAL INSPECTION which will remain to be of primary importance to the bridge engineer. These methods in their present form may be used as an aid to decision making regarding adoption of repair techniques and their extent and decision for the frequency of inspections. 187 2. Doubts resulting from discrepancy between results obtained by these methods and inspection should be cleared after close study with or without adequate instrumentation supplemented by non-destructive methods which are available as tools for the bridge engineers of today. 3. After analyzing a large number of masonry arch bridges the following important facts emerge – • A fill of about 1 metre above crown of extrados upto bottom of sleeper is a must for satisfactory performance of arch bridges. If the fill is about 60 to 70 cm, no amount of repairs can hold good. • Clean ballast cushion is very important over arch bridges. There should be no fish plated joint over such bridges. If it is unavoidable than it should be located just over the piers. Deep screening over masonry arch bridges is required to be done more frequently say once in 3/5 years. • As all the masonry arch bridges are more than 100 years old and hardly any maintenance, repairs have been carried out, it is desirable to take up the work of thorough pointing and cement grouting of all masonry arch bridges from one end of the section to the other. • RCC jacketing of arch bridges should not be thought of unless other measures have been exhausted. The quality of RCC jacketing which is normally done manually leaves much to be desired and in real terms does not improve performance of arch bridges. At best it gives only a psychological satisfaction. If RCC jacketing is required, the use of batching plant and concrete pump must be made compulsory for making its use effective. These measures are expected to enhance life of masonry arch bridges substantially. Volume - I 188 Volume - I 189 ANNEXURE – II ANNEXURE – III(a) Bridge No. 3, DME – KPK Span 1 X 3.66 m Analysis result Critical load factor = -0.83 (load case 6) Volume - I 190 Volume - I 191 ANNEXURE – III(c) ANNEXURE – III(b) Bridge No. 4, DME – KPK Span 1 X 4.57 m Bridge No. 14, BQT – DSEY Span 1 X 3.66 m Analysis result Critical load factor = 0.12 (load case 5) Analysis result Critical load factor = -0.86 (load case 5) Volume - I 192 Volume - I 193 FATIGUE ASSESSMENT CRITERIA FOR DESIGN AND ANALYSIS OF STEEL GIRDER BRIDGES FOR HEAVY AXLE LOAD OPERATIONS 2.0 PLAMGREN-MINER LINEAR DAMAGE RULE 2.1 Palmgren has proposed a damage model on the basis of constant energy absorption per cycle. The energy absorption per cycle leads to linear summation of damage. Miner has subsequently, represented this concept in mathematical form. Palmgren-Miner rule states that the fatigue damage contribution by each individual load spectrum at a given stress level is proportional to the number of cycles applied at a stress interval, ni, divided by the total number of cycles to failure at the same stress level, Ni. It is obvious that each ratio can be equal to unity if the fatigue cycles at the same stress level would continue until failure occurs. The total damage, in terms of partial cycle ratios or damage, can be written as – 2.2 The Palmgren-Miner rule, described above is considered a simplified and versatile tool for determining the total life of the structure under study. It is apparent that the detail under consideration is said to have failed if the Total Damage becomes unity (1.0). The rule does not account for the effect of load sequence and load interaction on damage accrued, and have an over simplified assumption of linear summation. However, Palmgren-Miner rule is still widely used to estimate life of a structure, on account of its ease of application. 3.0 TRAIN LOADS AND TRAFFIC MODEL 3.1 Railway Board vide its letter No.2006/CE-II/TS/2 dated 26-10-06 has advised the axle load and Track Loading Density (TLD) for design of foundation and bridges. As per this letter axle load of 32.5t and TLD of 12 t/m has to be adopted for design of bridges. Based on this information, the freight train compositions including their length, weight and GMT etc. have been developed for light, medium and heavy traffic classifications. 3.2 The wagon details for DFC are yet to be finalised, therefore the wagon details (axle spacings) of IRS HM loading have been taken in the above analysis. No passenger trains have been considered and the PIYUSH AGARWAL* , R.K. GOEL** 1.0 INTRODUCTION 1.1 Railway bridges are subjected to heavy fluctuating dynamic loads. These fluctuations cause fatigue failure of members or connections at lower stresses than those at which would otherwise fail under static load. IRS provisions, which are based on stress ratio concept does not taken into account the phenomenon of fatigue adequately. World over, such effects are taken into account following the Palmgren Miner cumulative damage rule based on stress range concept. British Code BS-5400 part 10 and more recently Euro Code EN-1993-1-9: 2002 address the fatigue effects in a rational manner by taking design parameters such as route GMT, type of traffic, design life and detailed category of connection. However, these provisions are specific to loading conditions prevailing in their countries and the design parameters as given in these codes can not be straightaway applied for design of bridges in traffic conditions prevailing in India. 1.2 Recently, Indian Railways has taken a leap forward to cope up with the increased demand of freight transportation by deciding to have a dedicated freight corridor (DFC) for which a new set of standard designs would have to be prepared. This would be a big challenge for Bridge Designers to develop the new designs, based on rational criteria for fatigue in accordance with established international practices. Development of a fatigue load model for such a dedicated freight route is the first design input required by the designers. In this paper an attempt has been made to develop such a load model based on the anticipated axle load, type of locos and the train lengths. This model with modifications can be subsequently used for analyzing the existing bridges on feeder routes. * Executive Director/Bridges & Structures /RDSO/Lucknow, Volume - I ** Director/Bridges & Structures /RDSO/Lucknow. Volume - I 195 existing combinations of HM loading have been modified to develop the above load model for fatigue assessment. The traffic classification and load model developed have been shown in Table–1 & Table-2 respectively. The salient features of the load model are described as under: tractive effort of 180 t has been proposed with a combination of 3 locomotives. The comparison of locomotives and their tractive efforts for double, triple or quadruple traction has been shown in table given below: 3.2.1 LOADING For 32.5 t, none of the existing loadings given in IRS Bridge Rules is suitable. Therefore, the train formations have been tentatively developed on the basis of the details of locomotives and wagons for HM loading as available in Bridge Rules with following modificationsi) The number of train formations, have been reduced to 12 from 17 by removing the similar type of formations. ii) Trailing Load density: The Gondola wagon as taken in HM loading has been adopted with same axle spacing and increased axle load of 32.5 t. It gives a trailing load density of 12.79 t/m. Railway board vide its letter no.2006/CE-II/TS/2 dated 26-10-06 has instructed to design the bridges for TLD of 12 t/m. During discussion with Wagon Directorate of RDSO, it has been learnt that wagons of BOXN type are being modified for heavy axle load and these would be giving a TLD of 12.33 t/m. The proposed TLD is therefore, slightly on higher side and the designs of standard spans shall be safer. The locomotives of HM loading have been adopted with their axle spacings unchanged. However, the axle loads of all the locomotives have been proposed as 32.5 t for stress analysis. This will take care of any possible increase in axle loads of the locomotives. iv) Braking Forces: The braking force of the locomotive has been indicated against the train formations in Table-2. The braking force of the train load has been taken as 13.4% of train load as has been done in the case of HM loading. Therefore it is in accordance with prevailing practice. 3.2.2 SPEED It has been advised by Railway Board that the maximum permissible speed of freight trains of heavy axle load would be 100 kmph. It was also learnt that speed trials are required to be done with 10% extra speed. Further keeping in view the possibilities of 10% increase in future, design speed of 125 kmph has been proposed. It is also in conformity with the Coefficient of Dynamic Augment (CDA), given in IRS Bridge Rules which has been developed for a speed of 125 kmph. iii) Tractive Effort: In the HM loading the tractive effort of locomotives has been observed as 60t, 45t and 30.5t depending upon the number of locomotives coupled together. The maximum tractive effort of HM loading is 135 t with three locomotives of WAG6C & WAG6B. It has been learnt that increasing the axle load increases the tractive effort of locomotive. As discussed with concerned directorates there is no possibilities of increasing the tractive effort of an electric loco beyond 75t. In such a case only 2 locomotives would be normally sufficient for heavy haul. For heavier traction, three or four locomotives of lesser tractive effort can be coupled. In the proposed loading the maximum Volume - I 196 Volume - I 197 Volume - I 198 Volume - I 199 Table - 2 Train Formations Considered in Load Model Sheet 1a/3 Volume - I 200 Volume - I 201 Table - 2 Train Formations Considered in Load Model Sheet 2a/3 Sheet 1b/3 Volume - I 202 Volume - I 203 Table - 3 Train Formations Considered in Load Model Sheet 3a/3 Sheet 2b/3 Sheet 3b/3 4.0 RELEVANCE OF LOAD MODEL WITH DESIGN LIFE OF BRIDGE 4.1 Following factors primarily affect the fatigue strength of a typical connection- 4.2 i) Type of joint details ii) Stress range at the location under consideration. iii) No. of cycles of stress range The type of joint detailing is decided keeping in view the methodology of fabrication to be adopted. Once the type of joint detail is finalised, the allowable stress-range, to which the connection can be subjected to, can be obtained from relevant S-N curve for ‘N’ number of cycles. ‘N’ is usually taken as 2 million. S-N curves developed for the purpose are shown in Fig. 1. Figure-1 : Fatigue strength curves for director stress ranges Volume - I 204 Volume - I 205 4.3 It implies that the connection of a particular category is able to safely withstand 2 million cycles of the allowable stress-range. As practically observed, the different components of the structure undergoes different no. of cycles of different stress-ranges. Therefore, every connection detail, over a period of time, is subjected to a stress-range histogram consisting of number of stress-ranges and corresponding number of cycles. A typical stress-range histogram is shown in Fig. 2. can be adequately assessed over a period of time. The load model will have to specify the distribution of train types and their frequencies with respect to their cumulative GMT and traffic volume. These parameters in turn will have to be taken as design input for assessing the fatigue strength of connections. 5.0 FATIGUE ASSESSMENT CRITERIA 5.1 It is well known that the fatigue provisions of IRS Steel Bridge Code are based on stress ratio concept, which is quite obsolete. Revision of fatigue provisions is already being done with active support of IIT/ Roorkee. The new provisions are to be based on stress range concept using the Palmgren- Miner cumulative damage rule. It has been observed that the provisions of British Code BS-5400 are based on stress range concept and are well understood by practicing design consultants. Therefore, the fatigue assessment criterion has been framed based on BS-5400. Accordingly, the fatigue assessment to be done in accordance with Palmgren-Miner summation rule by assessing accumulated damage as described below. 5.2 From the train configurations given in fatigue load model (Table-2) stress histories shall be determined at the structural detail (including such secondary effects as would be relevant) and stress-ranges and their numbers shall be evaluated by rain flow method or reservoir method. Fatigue damage analysis shall be done by applying appropriate partial safety factors to stress ranges and characteristic S-N curves in accordance with relevant provisions of BS:5400. 5.3 The damage summation shall be performed as per Clause 8.4, 9.2 and 11.1 of BS: 5400 Part-10 as under: Figure-2 : Typical stress range histogram 4.4 The actual damage to the connection detail is the cumulative effect of all such stress-ranges that are included in the stress-range histogram. The concept of design life comes into picture at this stage, as the cumulative damage should be equal to unity, at the end of design life. The stress-range histogram, to which the detail is subject to, is a function of type of trains, frequency of trains, speed and the GMT etc. In practical scenario, it is a complex phenomenon of cumulative fatigue damage, which will be very difficult to model unless some kind of standard of load-frequency distribution is assumed. Therefore, In order to standardize the stress-range histogram, it is necessary to standardize the load models so that the cumulative fatigue damage Volume - I 206 Where 5.4 n is the number of cycles associated with stress range modified with appropriate partial safety factor for load N is the number of cycles corresponding to the design S-N curve modified with appropriate partial safety factor for material. Failure shall assumed when Dd≥1.0 Volume - I 207 5.5 Partial safety factor for load, γfL & partial safety factor for material strength, γm shall be taken as per BS:5400 Part-1 & Part-3. 5.6 Following parameters shall be considered in assessment of fatigue damage: 5.6.1 Fatigue load Model and traffic classification as per Table-2. Bridges shall be designed for Medium traffic for annual GMT of 100. 5.6.2 S-N curves for Direct stress range as per BS: 5400 Part -10. 5.6.3 Joint detail classification as per Table-17 of BS: 5400 Part -10. The design shall be in conformity with the description and requirements of the connection detail chosen. 5.6.4 Appropriate Correction factors for stress concentration as given in Appendix-H of BS: 5400 Part-10 for the detailed classification shall be applied. 5.6.5 Design life of 120 years shall be considered. 5.6.6 Maximum design speed 125 kmph shall be considered. 6.0 CONCLUSIONS 6.1 A rational approach, in accordance with latest international practice has been suggested for design of new steel girder bridges for heavy axle load operations. The approach follows stress-range concept and Palmgren-Miner cumulative damage rule, which forms the basis of fatigue provisions of British Standards and Euro Codes. The approach can also be applied for fatigue assessment of existing bridges for heavy axle load operations. 6.2 6.3 A fatigue load model has been developed keeping in view the requirements of heavy haul on Dedicated Freight Corridor, which is to come in near future. The load model takes into account the parameters of locomotives and wagons that are necessary inputs required for design of steel girder bridges. Fatigue assessment criteria has been proposed in accordance with British Standard BS: 5400, keeping in view the fact that the provisions are well understood and practiced by leading design consultants in India. It is expected that a rational fatigue assessment procedure would be followed in developing the future designs of steel bridges on Indian Railways. Volume - I 208 7.0 REFERENCES 7.1 Revision of Fatigue Provisions in IRS Steel Bridge Code (2004), First Interim Project Report submitted by Department of Earthquake Engineering, Indian Institute of Technology, Roorkee to Research Designs Standards Organisation, Ministry of Railways, Lucknow (UP)226011. 7.2 Stress Spectra for Fatigue Design of Railway Bridges (1991), Project report submitted by Department of Civil Engineering, Indian Institute of Technology, Kanpur to Research Designs Standards Organisation, Ministry of Railways, Lucknow (UP)-226011. 7.3 Technical Documents on Traffic Details for Revision of Fatigue Provisions of IRS Steel Bridge Code (1989), file No.CBS/PSB, Research Designs Standards Organisation, Ministry of Railways, Lucknow (UP)226011. 7.4 IRS Bridge Rules (1986), Research Designs & Standards Organisation, Ministry of Railways, Lucknow (U.P.). 7.5 British Standards BS 5400 : Steel, Concrete and Composite Bridges, Part-1 General Statement, British Standards Institution, 1980. 7.6 British Standards BS 5400: Steel, Concrete and Composite Bridges, Part-3 Code of Practice for Steel Bridges, British Standards Institution, 1980. 7.7 British Standards BS 5400 : Steel, Concrete and Composite Bridges, Part-10 Code of Practice for Fatigue, British Standards Institution, 1980. 7.8 EN 1990:2002 (Eurocode – Basis of Structural Design) – (For safety, comfort, deformation including twist and deflection) 7.9 EN 1991-2:2003 (Eurocode I – Action on Structures, Part 2 – Traffic Loads on Bridges) – (Natural frequency range and Loading for fatigue estimation) 7.10 EN 1992-1:2004 (Eurocode 2 – Design of Concrete Structures, Part – I – General Rules and Rules for Buildings) 7.11 EN 1992-1-1:2004 (Eurocode 3 – Design of Steel Structures, Part I – 1 – General Rules) – (Classification of cross sections) Volume - I 209 7.12 EN 1993-1-8:2002 (Eurocode 3 – Design of Steel Structures, Part 1-8 – Design of Joints) – (Classification of HSFG Bolts) SUITABILITY OF BGML AND RBG STANDARD BRIDGES FOR HIGHER AXLE LOADS 7.13 EN 1993-1-9:2002 (Eurocode 3 – Design of Steel Structures, Part 1-9 – Fatigue Strength of Steel Structures) RAMA KANT GUPTA* 7.14 EN 1993-2:2004 (Eurocode 3 – Design of Steel Structures, Part 2 – Steel Bridges) – (Requirements for fatigue assessment, Road and Rail Bridges) 7.15 EN 1994-2:2003 (Eurocode 4 – Design of Composite Steel and Concrete Structures, Part 2 – Rules for Bridges) – (Width of effective flange, shear connectors) Most of the bridges of Indian Railways are still of RBG, BGML or even prior to that standard of loading. While permitting Higher Axle Load, it is necessary to ascertain the safety of the bridges. In this paper, Author first tries to share the input required for upgrading the bridges to MBG standard of loading. Simultaneously, it has been tried to suggest the via media, if found safe, to permit the powerful locomotive/rolling stocks even before up gradation of the existing old bridges. * This paper has been written based on the past experience of the author as Executive Director, Bridge & Structures at RDSO, Lucknow 1. INTRODUCTION For survival of the Indian Railways in the competitive market, it is necessary to reduce the transportation cost. Since major share of the earnings of Indian Railways is from goods traffic, hence it is but natural that first consideration goes to the freight stocks, particularly regarding how to operate them with higher axle load, so that with the same configuration and same frequency, more load can be transported. (As far as coaching stocks are concerned, these are having much lighter TLD as compared to freight stocks and as such, don’t pose any problem.). In this paper, it has been tried to explain the modalities how to upgrade the existing bridges of lighter loading standard with bare minimum input and as a interim measure even before upgrading the bridges, how to check the adequacy of operation of any particular locomotive/rolling stocks on a particular section. 2. PERCENTAGE POPULATION OF THE BRIDGES WITH DIFFERENT STANDARDS OF LOADING Before discussing anything more, first let us have an idea bout the population of bridges conforming to different standards of loading along Volume - I 210 Volume - I * General Manager/Bridges, IRCON, New Delhi – 110066. with load intensity of various standard loadings. The same is given in Table No. -1: Since, currency of RBG standard of loading was for a short time, hence number of bridges conforming to RBG standard are less. 3. SEVERITY OF THE PROBLEMS OF OLD BRIDGES IN CONTEXT TO PRESENT DAY LOADING We are now familiar about BGML, RBG and MBG loading. Let us have an idea about the loading standard even prior to introduction of BGML i.e. prior to 1926. The same is given in Table No. – 2. eras still exist. But, the capacity of Arch Bridges to take the load is much more than its designed load. As such, Arch Bridges are also not posing any problem provided the same are in sound condition. From the above Table, it is also clear that there was no consideration of longitudinal forces in olden days. Actually, longitudinal forces were made part of BRIDGE RULES in 1923. It does not mean that prior to 1923, Engineers were not aware about the longitudinal forces, particularly Tractive Efforts and Braking Forces. But, its magnitude was so small and majority of the bridge foundations being of gravity type of masonry structures, there was no any problem to the Tractive Efforts/Braking Force of the locomotives of that period. Problems only started after increase in axle load over a period of time, warranting requirement of powerful locomotives to haul the heavier trains not only having comparatively higher axle loads but also having more length of the trains as well It is worthwhile to mention that for bridges, longitudinal forces are creating more severity than the vertical forces, and most of the bridges are becoming unsafe on account of increased longitudinal forces of present day locomotives. For further appreciation, comparative statement of forces for Bending Moment and Shear Force for different standard of loading has been given in Annexure-I. Similarly, for Tractive Effort and Braking Forces for different loading standard, the same is given in Annexure-II. 4. ACTION TAKEN BY RDSO After revision of loading standards, RDSO came into action to check the adequacy of the existing bridges whether the same are safe for revised loading standard or not? In most of the cases, it was found that the input required to upgrade the bridges of BGML and RBG standard to MBG standard of loading is not much. The same is discussed as below: 4.1 INPUT REQUIRED FOR UPGRADATION OF SUPERSTRUCTURES OF THE BRIDGES TO MBG LOADING STANDARD From the above Table, it is clear that over a period of time, axle load has increased tremendously. Fortunately, Steel Bridges of those eras are not existing at present. Those were of EARLY STEEL category and had been replaced on account of designed for lighter loading and hence unsafe or on account of its EARLY STEEL category. Only Arch Bridges of those Volume - I 212 After thorough checking, it was found that majority of the BGML standard bridges are safe, except in most of the cases, bearings require strengthening. Wherever such requirements were felt, supplementary drawings were prepared and issued to all concerned. In few cases only, overstressing was found in BGML standard of bridges but that too, were within the codal provisions of keeping such Volume - I 213 bridges under observation and as such, here as well, requirement of regurdering avoided. Similarly, in case of RBG standard of bridges, most of them were found safe, except strengthening of bearings and keeping the few bridge members under observation. Only one standard bridge span of 76.2m of RBG standard was found having overstressing of 6.5%, if MBG standard of loading is to be introduced. It is further worthwhile to mention that a few members that are having such an extent of overstressing can either be strengthened or bridge span as a whole can be replaced as per the situation to make such bridges safe. Complete position about the input requirement for making old standard bridges fit for MBG loading is given in Table No.-3 All other standard steel spans of BGML and RBG not included in the above table are safe for MBG loading. Volume - I 214 Volume - I 215 4.2 SUBSTRUCTURES For substructures as well, RDSO did the remarkable work. Here, required profiles of all standard spans were suggested. Many factors are there for consideration, starting from different types of materials used, piers and abutments including its shapes and overall, varieties of standard spans existing on Indian Railways. The alternatives tried is given as below: Profiles for abutments for spans 6.1 m (PSC slabs and Plate Girders), 9.15 m (PG), 12.2 m (PG), 18.3 m (PG), 24.4 m (PG), 30.5 m (PG), 30.5 m (US), 30.5 m (OW), 45.7 m (OW), 61.0 m (OW) and 76.2 m (OW) Profiles for piers for spans 6.1 m (PSC slabs and Plate Girders), 9.15 m (PG), 12.2 m (PG), 18.3 m (PG), 24.4 m (PG), 30.5 m (PG), 30.5 m (US), 30.5 m (OW), 45.7 m (OW), 61.0 m (OW) and 76.2 m (OW) With the help of above-mentioned exercise of RDSO, one has to just compare the profile of existing bridge substructure with the required profile. If the required profile is less than the existing profile, then but natural, the substructure is safe. Strengthening of the substructure will only be required when the required profile is more than the existing profile. This is subject to sound condition of the substructures CONSTRUCTION MATERIALS Brick Masonry in Lime Mortar 1:2 Brick Masonry in Cement Mortar 1:4 Coarsed Rubble Masonry in Lime Mortar 1:2 Coarsed Rubble Masonry in Cement Mortar 1:4 Mass Concrete Substructures of M-10 Grade concrete 5. TYPES OF CUT WATER Semi circular ends Triangular cut and ease water, angle included between the faces being 90 degree. Arcs of cut and ease water intersecting at 90 degrees named as ‘Standard Cut Water’ GEO-TECHNICAL PARAMETERS The following Geo-technical parameters have been considered: Till now, we have discussed the modalities required in upgradation of old standard bridges to MBG standard. Actually, loading standard defines the maximum loading which is likely to come based on which civil engineering structures are deigned. Furthermore, future rolling stocks are designed in such a way to keep its loading profile well within the envelop of that particular loading standard. In upgradation of the bridges, it is possible that the same may take time. Now, we will discuss the action required in checking adequacy of the bridges of modern locomotives having higher Tractive Efforts (TE) Angle of internal friction = 35 0 Density of Backfill material = 1.76 t/m2 Front slope of abutment = 1 in 15 Length of abutment = 6.1 m In addition to above, 100% overstressing in masonry structure as per Sub-Structures Code of Indian Railways have also been considered. It is worthwhile to point out that earlier, Factor of Safety for Masonry Structures was taken as 6. Hence, even after allowing 100% overstressing, there remains Factor of Safety of 3 which is adequate as per Material Science knowledge of present day, provided the masonry structure is in sound condition. Based on the considerations/factors mentioned above, profiles for the following Standard Spans have been developed: Volume - I CHECKING ADEQUACY OF THE BRIDGES FOR MODERN LOCOMOTIVES HAVING HIGHER TRACTIVE EFFORTS (TE) EVEN WITHOUT CONVERTING TO MBG STANDARD 216 Process start with the Speed Certificate issued by RDSO. RDSO issue the Speed Certificates for standard span only with the condition that condition of the bridge is sound. In addition to standard spans, there might be so many non-standard spans existing on a particular Railways, whose adequacy also need to be ascertained for safe operation by the concerned Zonal Railways. Furthermore, process of checking for any new Locomotive/ Freight Stock is based on the EUDL generated by them and comparing the same with that particular standard of loading for which the bridge was designed. It is the most conservative way and inadequate in many aspects. But, sitting in RDSO and without having appropriate field data, it is not possible for RDSO to do better. Actual practice should be to work out the Volume - I 217 BRIDGE RULES permit operation of Coupled Locomotives. From the above Table, we see that in case of Coupled Locomotives; total Tractive Efforts generated by the locomotives in many cases will become more than the designed value of Tractive Efforts taken in BGML and RBG loading standards. Normally, worry to the Bridge Engineer from this consideration. We will see in the subsequent paras about the reality of such worry. actual Tractive Efforts required for operation of that particular load on particular section. It is possible that bridges of BGML and RBG standard might be safe even before strengthening them to MBG standard. Here, we are frequently talking about MBG loading. This is so, since, it is the loading standard which is enveloping almost all rolling stocks except a few aberrations. As already briefed, firstly, one should go through the Speed Certificate issued by RDSO to ascertain the restricted spans, if any and then after, check the availability of that particular span on the section under consideration. In case any span has been restricted and that is available on the section under consideration, one has to calculate the actual Tractive Effort required to haul the desired load on that particular section. Such exercise is very much section dependent since gradients and curves demand extra Tractive Efforts while negotiating them. Furthermore, locomotives are designed with extreme upper limit requirement of Tractive Effort, so that the same can work in most of the situations. It is further worthwhile to point out that locomotive exerts that much Tractive Effort only that is required in haulage of particular load in particular situation. Exertion of extra Tractive Effort may cause slippage to the wheels and thus, may not be operative. To understand the modalities, let us proceed step by step, starting from locomotive and freight stock characteristics. 5.1 5.2 LOAD PARTICULARS OF FREIGHT AND COACHING STOCKS Load particulars of freight and coaching stocks, which are plying on Indian Railways, is given in Table No. – 5 LOAD PARTICULARS OF LOCOMOTIVES Load particulars of various locomotives are given in Table No. – 4 * Often, Mechanical and Operating Departments ask the Engineering Department that when BOBs and BOYs with 22.9t of axle loads are plying since long on Indian Railways, why not the same axle load of other stocks, particularly BOXN is being permitted? In this regard, it is worthwhile to point out that in addition to axle load, it is equally important to know the TLD that rolling stock is producing. From the above table, it is clear that although BOBs and BOYs are having 22.9t of axle loads but, its TLDs are 7.9 and 7.68 t/m respectively i.e. closer to the values prescribed for BGML and RBG loading. Even permitting CC+10 for BOXN wagons having axle load of 22.82 t/m, i.e. slightly less than that of BOBs and BOY, but it generates the TLD of 8.52 t/m which is not even more than that of BOBs and BOYs but, even more than the prescribed TLD of 8.25 t/m of MBG loading standard. Volume - I 218 Volume - I 219 5.3 FORMULA TO BE USED FOR CALCULATING TRACTIVE EFFORT REQUIRED FOR HAULING A GIVEN LOAD Tractive Effort (TE) required for hauling a load “T” tonnes on one in “G” grade and “S” degree curve is give by: TE (kg) = T1 + T2 + T3 + T4 Where T1 = RV x Train load gives Train Resistance in kg RV is running resistance of stock in kg/t At start RV is taken as 4 kg/t for BOXN Wagon For loaded BOXN in running condition, RV = 0.6438797 + 0.01047218 V + 0.00007323 V2 For Empty BOX N in running condition, RV = 1.333973 + 0.21983 V + 0.000242 V2 T2 = R1 x Loco Weight gives locomotive resistance in kg R1 is specific resistance of locomotive in kg/t At start R1 taken as 6 kg/t In running condition, R1 = 0.647 + (13.17 / W) + 0.00933 V + (0.057 / WN) V2 Where N is number of Axles. W is axle load of the locomotive in tonnes V is speed in Kmph T3 is Grade Resistance for train and loco in kg. T3 = (1 / G) x 1000 x (Train load in tonnes + loco wt. in tonnes). T4 is curvature resistance for train and loco in kg. T4= 0.4 x S Degree of curvature X (Train load in tonnes + loco wt. in tonnes) Based on the above formula, it was tried to calculate the total Tractive Efforts required for different load combinations and for different degrees of curvatures of the curves and different gradients. The same are summarised in Table No. - 6 Volume - I 220 Volume - I 221 Since v and s are known, we can calculate acceleration required i.e. ‘a’ Similar exercise can be done with other locomotives as well. Difference is not likely to be much since weight of the locomotives is the only variable. After knowing the required acceleration, force (TE) can be worked with the following formula: From Table No. – 6, it is clear that in most of the cases, it is likely that starting Tractive Efforts are not very much in most of the cases. Here, the Author tries to say that it is not an appropriate decision to go by the capability of a particular locomotive regarding its Tractive Efforts. Actual thing that matters is the requirement of Tractive Efforts for a particular load on a particular Section having particular steepest gradient and particular sharpest degree of curvature. After working out the Tractive Effort requirement, the same should be compared with the designed longitudinal force of that particular bridge to ascertain whether the bridge is safe for that particular Tractive Effort or not? 5.4 FURTHER COMMENTS ABOUT THE ACTUAL TRACTIVE EFFORTS REQUIRED FOR HAULING OF THE LOAD On the advice of Railway Board, one exercise was done on VaranasiAurihar Section of Varanasi Division of NE Railway in association with Electrical, Mechanical and Operating Departments of NE Railway and RDSO. That particular Section was chosen since the same was having steepest gradient where operation of WDG-2 Locomotive was planned for operation but the bridges were not coming safe with respect to its designed longitudinal forces, when compared with the Tractive Effort capability of WDG-2 locomotive. In the exercise, many shortcomings crept in on account of some fellow Departments and as such, actual conclusion could not be drawn. However, one finding, which B&S Directorate of RDSO noticed that during the operation, it has been tried not to accelerate the train to the extent so that it may pick up that speed after covering the prescribed distance, as given in Working Time Table. Not speeding up means not providing proper acceleration to achieve a particular speed over a particular stretch. From running point of view, it is necessary. This is as per requirement of the Working Time Table of any particular Division of any railway. However, this force can be worked out as suggested below: Let v be the speed to be achieved after traveling a distance s, then, as per Newton’s Law, F(which actually in this case is TE) = ma This force (TE) can be added to the TE required for hauling the train at a particular speed to know the total TE required for haulage of the proposed train on the proposed section having its own gradients and curves . If the total TE is less than the designed longitudinal force of the bridge, then, it is safe; otherwise, its strengthening/rebuilding will be required as the case may be. CONCLUSIONS It is not necessary to re-build each and every bridge to make it fit for MBG standard. Even old bridges conforming to BGML and RBG standards can be upgraded to MBG standard with the small input as suggested in the paper. In case any span is restricted for operation of any particular rolling stock, actual Tractive Efforts required on that particular Section should be worked out. If the same is less than the designed longitudinal force of that particular bridge, then even without any strengthening input, that particular rolling stock can safely be permitted for its operation. REFERENCES 1. RDSO letter No. CBS/PBR/RLS dated 05.08.1988 regarding ‘MBG Loading – 1987 – Position of Designs and Drawings of Bridges’ circulated to all the Zonal Railways. 2. RDSO letter No. CBS/PBR/RLS dated 24.04.1989 regarding ‘Guidelines for checking and strengthening of bridges for MBG loading of 1987’ circulated to all the Zonal Railways. 3. RDSO letter No. CBS/DOW dated 28.10.1993 regarding ‘Suitability of BGML Fixed End Bearings of Open Web Girders for MBG Loading – 1987 for Spans 30.5m (Through Type) and 45.7m up to Seismic Zone – III, circulated to all the Zonal Railways. 4. RDSO letter No.CBS/DOW dated 31.08.1994 regarding ‘Suitability of v 2 = u2 + 2 a s When the locomotive starts from rest, u = 0 Then, the equation modifies to v2 = 2 a s Volume - I 222 Volume - I 223 BGML Fixed End Bearings of Open Web Girders for MBG Loading of 1987’ circulated to all the Zonal Railways. 5. RDSO letter No. CT/DG/LW/HAW dated 21.03.1998 addressed to CRB regarding ‘Higher Axle Load Wagons’ with a copy to all the Zonal Railways. 6. RDSO letter No. CBS/DPA dated 09.11.1998 regarding ‘Guidelines for Checking Suitability of Substructures of Existing Bridges for MBG Loading – 1987’ circulated to all the Zonal Railways. 7. RDSO letter No. CBS/Golden/Q/Strength dated 30.12.29\004 regarding ‘Strengthening of Golden Quadrilaterals, its Diagonals and other identified Routes for 100/75 kmph Goods Trains Operation’ circulated to all the Zonal Railways. This letter contains RDSO letters mentioned from Sl.1 to Sl.6 as a part of Annexure. Volume - I 224 Volume - I 225 END Volume - I 226 Volume - I 114