Recent Developments for the Bridge Industry
Report on the SIMoNET seminar, 6th Dec 2000 at UCL NDE Centre.
by MWB Lock, Cranfield University
|
|
Recent Developments for the Bridge IndustryReport on the SIMoNET seminar, 6th Dec 2000 at UCL NDE Centre. |
Introduction
The chairman, Professor Bill Dover of UCL, welcomed the delegates to the second SIMoNET seminar. This meeting was to contain both summaries of the problems inherent in monitoring bridges old and new and some possible solutions, current and future.
Keynote Address
Brian Bell from Railtrak established the problem areas for the meeting. Railtrak, he said, had 40,000 bridges with 60,000 spans Their ages averaged out at 136years for masonry, 28yrs for concrete and 63yrs for steel construction and cost £700M pa to maintain.
On the whole they all got an annual visual inspection with about a 6yr interval for 'touching' inspections: these were generally done by tradesmen specially trained to both inspect and report.
There exists a computer aided planning tool, the Structures Marking Index, which grades problems according to fracture mechanics principles and can help decide where new inspection techniques might be used. The principle was to use monitoring only where existing systems were doubtful but a major dilemma was where to draw the line.
Brian highlighted some areas of concern: 3 bridges were hit every day by lorries etc, 1 bridge was lost every year through scour and there was the old perennial problem of whether to replace or repair rivetted connections.
Fortunately, historically bridges were built with large safety margins; Brunel's bridge at Maidenhead was carrying 10x the design load.
A Bridge Bash Too Far
David Hughes of Ferranti Technologies introduced their Bridge Guard system.
A bridge, having been hit (3 a day), had a 5mph speed restriction imposed until inspection and any repair was carried out. About 90% of collision caused 'no damage' and only 3% needed repair. Clearly, the need was to minimise delay. Typically, the 210,000minutes annual delay cost over £5M.
Bridge Guard measures shock using 3-axis accelerometers under the bridge feeding to a cpu which was connected with a local control room or signal box. Comparison of the collision signal with that of typical train signatures was made and a level of 10% taken as a 'safe' level.
Eight systems were currently being tested in situ and it was hoped that a production standard would be agreed by April 2001 and that the system would be in service later in the year.
Remote bridge strike reporting and information
Velautham Sarveswaran of Maunsell Ltd. introduced his company's remote impact detection and evaluation system, RIDE.
This system was also based on accelerometers together with cctv. A PC controlled the logging and analysis of up to 32 channels of data and any collision information or alarm situation was monitored in their Birmingham office. Sampling frequency up to 1KHz was possible with rejection of false alarms caused by train, electrical or radio interference.
The system, being autonomous was quick to install, typically one day, with the sensors/cctv being simply attached to the structure. No trackside wiring was necessary as ISDN links were used as required.
Ground radar for bridge inspection and monitoring
Simon Brightwell, Aperio Ltd., presented the possibilities of a ground radar approach to detecting and evaluating thickness, embedded rebars, voids, disbonds and washout etc using changes in the dielectric nature, radiation velocity and conductivity changes of the materials investigated.
Their new digital system was found to be more reliable and very repeatable (better than 1% relocation). Signal processing was easily applied to remove the beam-width induced parabolic effect and to compare current and past results. A feature 'straightener' was also incorporated.
The presence of water enhances signal velocity and thus damp stone can be differentiated from dry. The technique was inherently less effective on highly conducting materials such as engineering brick, wet clay or sea water.
Bridge monitoring using ACSM
Chris Dearden of TSC Inspection Systems presented their novel stress measuring technique, AC stress measurement. This utilises the changes in magnetic permeability which occurs when magnetic domains change shape when subject to stress.
Results were presented for a model 3pt bend test rig and differences due to the nature of the steel under test considered. Recent development showed the separation of actual stress from residual stress was a possibility.
Field experience of rating the acoustic energy in materials to characterise structural integrity
Alan Bloor of Rock Mechanics Technology introduced a new instrument (the Acoustic Energy Meter or AEM) which they have developed and used to carry out surveys in mines and in lined and unlined tunnels although the principle could be applied to bridges also.
The rate of decay of reverberant energy in a material or structure, resulting from a sharp impact, is related to the size and nature of the contacts between the material and its surroundings. This principle has been used as a basis for testing the structural integrity of lined and unlined rock surfaces with the objective of using the method to identify areas of potential instability.
The instrument has proved effective in locating back-fill problems, delamination in tunnel roofs and checking reinforcement, a loose surface covering giving a distinctly longer acoustic decay time compared with the solid.
In bridges, sand infiltration, voids and bulges due to weak material would be amenable to detection.
Reliability-based assessment of bridges: opportunities and challenges for using SIM data
Professor Marios Chryssanthopoulos of Surrey University, representing the ASRANet Bridge Group, first described the change in bridge monitoring from the early research on masonry behaviour in the 1920s, through finite element analysis (c60s) and design codes which considered stress and limit state formulation (c70s) to the assessment methods and standards of the 80s up to the current bridge management concern with safety levels and reliability assessments.
Probabilistic and reliability methods based on available statistics including loadings and material properties lead to codes which are then tested for reliability to produce new safer codes.
The author suggested that intervention was more costly than getting it right at the design stage. Key factors were
One difficulty with the current situation is that an existing structure is assumed OK as is compared with a new one which has to comply with new codes.
With the peak of bridge building occurring in the early 70s, most bridges are at around 25% of their design life. One therefore has to allow for changes in load and deterioration. In practice this means performing a spatial analysis of, say, a deck joint with run-off, salt spray etc assessed and the effects predicted. This implies site specific data is to be gathered. Acceptance criteria would then be established.
Remaining problems include finding good models for durability and fatigue life, progressive collapse and the time variance of performance due to deterioration mechanisms and the uncertainty of future requirements.
Acoustic Emission for Periodic Proof Testing
Gordon Drummond, IMES Group, described the PUMA (Passive Ultrasound Monitoring and Analysis) system.
This is a knowledge based state of health system which monitors a structure whilst under load. It was stressed that only active, ie not benign, sources might be detected.
The author emphasised that an AE system such as this was unobtrusive and could be used to monitor large structures.