seminars index

Structural Integrity Monitoring - Current and Future Report on the SIMoNET seminar, 23rd Nov 2001 at UCL NDE Centre. by MWB Lock and JV Sharp, Cranfield University

Introduction

Bill Dover welcomed everyone to the meeting, and briefly explained the background to SIMoNET.

1) Paint based thick film piezo-electric strain sensors (Jack Hale, University of Newcastle-upon-Tyne)

Dr.Hale outlined the background to the development of these sensors, which were based on powdered piezo-electric materials in a matrix. Glass had good properties as a matrix but a high temperature cure was required. Polymers had less good properties but with many other advantages and were being used. Films of thickness of between 50-100 µm were sprayed onto surfaces using a paint sprayer, which was very convenient.

The characteristics of the sensor had been obtained using a test rig and the strain measurement was shown to be approximately linear between 40-4000 µstrain. The frequency response was also good, between 1-1000Hz. Field trials are being undertaken on the new Millennium bridge over the Tyne in Newcastle, which showed good promise and in terms of weathering there were no problems over the summer period (installed May 2001); winter trials would be undertaken over the following months.

2) Life Extension of Pressure Vessels and Pipelines in Corrosive Service through continuous corrosion monitoring (John Lilley, AEA Technology)

John Lilley described the monitoring method using UT thickness measuring transducers, which can be interrogated remotely. In general pressure vessels and pipelines are retired when the minimum allowable thickness (MAT) is reached; the MAT is normally derived from codes and tends to be conservative.

The overall process involved identifying the presence of corrosion, investigating all potential causes of failure from corrosion, performing corrosion maps, and then permanently attaching corrosion probes at the identified locations to monitor corrosion rates. Fitness for service assessments would then be undertaken, based on the observed corrosion rates, and compared with the specified minimum criteria.

The probes can work up to 150ºC continuously. Cable runs can be difficult but radio communication of data is now being investigated. Data is usually recorded on a weekly basis, by connecting a laptop and downloading data. On the basis of the data recorded it has been found possible to defer major inspection programmes, with a considerable benefit in costs.

The technique can also be used on buried pipelines. The remaining life assessment is usually initially based on codes, followed by fracture mechanics assessments using known defect data. Use of local wall thickness data has allowed a considerable extension in life with some pipelines.

3) Recent developments in the use of on-line structural monitoring for the assurance of the integrity of fixed offshore installations. (Dave Sanderson, ABS Consulting & Alan Dougan, Fugro Structural Monitoring Ltd)

Dave Sanderson said that the idea for the joint industry project had come from the first SIMoNET seminar. It now had six sponsors (5 oil companies and HSE). He outlined the background to current inspection offshore, based on flooded member detection (FMD) + GVI/CVI every 2-3 years. This was low cost and with low risk to divers, but based on the premise that some level of damage is acceptable. This approach fitted with using on-line monitoring (OLM) as a tool. In this respect FMD and OLM are complementary techniques. OLM is based on measuring changes to the stiffness matrix, and correlating these with member failure. Accelerometers are placed on each corner of the topside structure, and wave loading on the structure provides the driving force. Mass variations on the topside can be removed by comparing east/west variations with north/south changes in frequency. OLM requires a degree of structural robustness, which can be determined for each platform. In the study the changes in frequency have been correlated with robustness. This showed the need to improve on the measurement of frequency change compared to background and work by Fugro using several different transformation methods has enabled the sensitivity of the measured frequency change to be improved significantly ( by a factor of between 2 and 3).

In the study 13 different platform types are being investigated, and the change expected for member loss compared with measurable stiffness changes, with account taken of robustness and platform geometry. It would appear that for many 4 legged platforms OLM is a viable technique, with most member failures detectable. For these FMD is often considered as a marginal technique because of limited redundancy.

4)Classification Society Perspective on Structural Integrity Monitoring. (David Davenport Jones American Bureau of Shipping)

David Davenport Jones introduced this topic as 'part of the regime that maintains ships' integrity'. It is necessary to understand the strengths and weakness of design using advanced methods, eg.

• analysis of static and dynamic loads and the resistance to failure modes
• monitor construction especially critical areas, ie NDT and survey
• monitor condition and capability of the vessel and the role of coatings
• link the previous three phases by information management (IT).

He highlighted the necessity for an excess of capability over demand and the fact that Classification Societies have prescriptive rules to define requirements.

Three main phases were described.

1. Design phase with a safe hull regime to define loads, understand failure modes, set criteria for lifetime safety and, after identification of critical areas for robustness and steel weight distribution, to supply a graphical display of acceptable thickness.
2. Construction phase where the safe hull concept identifies critical areas, requires a monitoring plan from the shipyard and ensures rigorous implementation of this during construction.
3. Operational phase where periodic survey is a guide by analysis of service experience using accumulated information from many sources.

Inherent in this is a need for robustness and here Mr Jones introduced the concept of residual resistance factor and showed how this interacted with OLM and changes in the structural vibrational frequencies. Several frequency transforms were tried and the simple Fourier was found to be most suitable giving f to 0.7%.

5) A Better Way to Monitor Hull Structures. (John Waite, The Salvage Association).

The Salvage Association ( a trading style of BMT Salvage Ltd) specialises in damage surveillance for insurance purposes. Many excellent slides from a wide variety of sources were a feature of this presentation which included:

• Cracking in the cross bar supports of the StÝreboelt bridge sections
• The difficulty of photographing grooving in inaccessible places even though it was readily visible to the eye.
• Small cracks in fruit juice containment vessels which, being on a chilled ship were perhaps not so significant and were on a neutral axis so not as subject to bend loads
• Bow fatigue where a large chunk was missing; the problem being that selective gauging was not at relevant places
• A centreline girder crack where fracture through the depth of the web and over 70% wastage precluded load bearing.
• Use of ABS, Safenet and Timeline graphic representations were described.

The importance of monitoring coatings in maintaining structural integrity, especially where pitting was a problem. Increasing use of IT was called for to 'put the information where it's needed'. This must include vessel drawings, maintenance modules, current structural condition and most importantly, where the monitors are.

The author's conclusion were:

• Whereas classification originally assessed risk, a risk-based approach can now assess and limit it
• SIM was the sole responsibility of the Classifier and not the owners
• IT should give everyone the information they need.

6) ALERT/TURBOLIFE™ - Integrated Condition Monitoring System for Gas Distribution Networks. (David Green, Serco Assurance)

The size of the problem was made clear first of all: 12000 miles of pipe at 73bar, 20 gas turbine stations with 60 turbines at about £20M per turbine making these a £1Bn job with another ten times more than that in pipeline costs.

An on-line management display system had been developed for the turbines which monitored:

Advantica Technologies Ltd and Serco Assurance had cooperated in the project and developed many software modules. Serco Assurance produced those for creep/fatigue assessment incorporating stress synthesis, strain synthesis, cycle recognition and damage mechanics and Advantica Technologies Ltd the remainder.

The graphical interface shows the current state of the compressors etc and compares actual with optimal performance. Calculation of the state is deterministic i.e. it utilises modern methods of creep fatigue to estimate damage and does not rely on an artificial neural net assessor requiring a large data base of operating experience.

Continuous blade assessments are possible using an incremental approach based on consecutive time paths: an elastic FE program is used to calibrate the real time blade stress algorithm from speeds, temperatures, pressures and gas velocity. Real time strain is derived from plasticity corrections to elastic analysis and creep assessment.

Other features taken into account include, damage mechanics from intergranular effects due to creep or fatigue and hysteresis.

The benefits of the system are:

• Blade replacement costs substantially reduced
• Greater productivity
• More power
• Integrated management facilitated
• Flexibility in fleet operation.

7) FOINAVON Umbilical Monitoring System. SIM-VIV Considerations. (Geoff Lyons, Mechanical Engineering Dept. UCL in collaboration with MIT Norwegian Inst of S&T and BP)

Vortex-induced vibrations (VIV)occur on the Foinavon flexible risers at up to 600m off the Shetlands. The problem was to measure the bending and tensile stress at the vessel in respect of the following topics:

• did the expectations match the realities
• measurements of responses to mooring/wave/VIV frequencies were expected to be minutes/4-20 secs/ up to 10 Hz.
• measurement of accumulated damage was required
• by implication one would expect similarities with the risers themselves
• examine VIV and establish feedback for future design
• measure asset integrity.

A multi-cored curve sensor was incorporated with the riser which passed through a bend constrictor in the hull bottom. The curvature was sensed throughout the riser and tensions and curves associated with heave, roll and pitch were determined.

Planar plots were made and the curvature converted to displacements by FE analysis.

VIV measurements were made for varying currents at known depths.

The conclusion was that VIV can be reasonable modelled - both current and vessel motions reduce VIV. A fibre optic approach was suggested for the future.

8) Using GPS to Measure the Deflections of Structures. (Gethin Roberts, Inst of Engineering Surveying and Space Geodesy, Nottingham University)

Using the 24 satellite, NAVSTAR system as a basis, Gethin Roberts used a relative system (RTK) having a fixed and a roving station to determine positions within a few mms.

The equipment was used on the Humber bridge, centre span 1.4km to measure displacements for known loads and compare them with FE calculations. The results agreed within 20mm for both displacement and the associated oscillating frequencies.

The system put 6 receivers on the bridge with one reference control. Five lorries of 160 tons were driven across an empty bridge with a 148m/s wind. The results were post-processed with NTOV and PNAV.

The lorries were driven across in three ways: first all one way, then all the other and finally with two in each direction simultaneously.

Problems were experienced with multi-pathing from reflections from satellite antennae and variable delays in the troposphere: slow variations were eliminated with filtering.

It was found that mm precision is possible, a major problem being multi-pathing due, essentially to aerial size.

9)Fibre Bragg Gratings and their Application to Structural Integrity Monitoring. (David Jackson, School of Physical Sciences, University of Kent)

This now well developed system consists of using Bragg reflection from gratings placed anywhere along a continuous fibre. By varying the grating spacing and thus their reflected frequency response, one can interrogate chosen parts of structures to which these are attached. Changes in grating spacing due to local strain provides a proportional change in reflected frequency. Wavelengths of 8-1550 nm have been used.

A great advantage is the 'digital' nature of the system: it being insensitive to source intensity.

Many combinations of optical devices are usable. The simplest using a broad band source interfaced with a Fabry-Perot detector to isolate grating reflections of differing frequencies. Systems with up to 16 sensors at distances up to 10km have been used. For short distances it is possible to use a spectrometer to analyse all reflections simultaneously.

Several examples of current usage were described including:

• Concrete beam test at TRL where fibre pairs (strained and unstrained) were used to eliminate temperature effects (typically 10pm/ºC). The frequency shifts were differenced to provide strain measurements.
• Trucks passing over an I10 bridge in Mexico were monitored by the Naval Research Lab with 70+ FBGs.
• Twin hull surface effects were measured on a ship: sensitivities down to 2nm being attained.

The final conclusion was that, whilst not cheap, FBGs were exceedingly stable, flexible in application and capable of great accuracy.

10) SIMoNET: The Future. (Bill Dover, UCL NDE Centre).

Bill Dover outlined the future of SIMoNET. He reminded the members present that this network has proved to be successful, having organised four seminars with over 200 attendees. It has nearly 400 members, an active website and has also produced three newsletters. The existing network is funded by EPSRC and has an active committee of 15 members from a range of industrial and government organisations. Its aims are to bring together people involved in structural integrity monitoring (SIM), whether users, providers or consultants, for the dissemination and exchange of information on the most up to date methods and developments.

A second phase is now proposed, which aims to build on the existing SIM network, in order to move to a higher level for communication, reach a wider number of sectors of industry, strengthen links with other overseas networks, tackle problems of standardisation, identify future research requirements and to reach full independence.

Phase 2 funding is sought from existing and new sponsor members, the EPSRC, EU and individual sponsors. The proposed rate for companies is £600 p.a. for a three year period, and each of these can supply a member for the Steering Committee. A total of £18,000 is required for the period 2002-5. He was pleased to report that over half of this is already promised. An approach had also been made to the EPSRC, requesting additional funding to extend the project a further three years. With their contribution the total budget would be ~£50,000 over 3 years. EPSRC had been informed that the network was planning to be financially independent after 2005.

The European Oil and Gas Innovation Forum (EUROGIF) has launched a thematic network on information and communication technologies, known as GOTECH. This would include structural and downhole monitoring. A proposal has now been submitted to the EU, with UCL (on behalf of SIMoNET) as one of the potential members.

Tasks for Phase 2 include:

• Task 1 - to broaden the contact base, particularly internationally
• Task 2 - to develop the seminars, widen the topic base
• Task 3 - to develop the web site to be more interactive
• Task 4 - to develop approach to standardisation and identify research priorities.

It is planned to hold 2/3 seminar each year, provide regular web based newsletters and to set up a news group on the internet. The contact base for phase 2 will include existing contacts plus the network GOTECH, US based structural health monitoring workshops and new sectors of industry (aircraft, road vehicles etc).

He concluded by reminding all present that we are still looking for sponsors, with the intention of launching phase 2 in the New Year.

seminars index