The first paper, by Hewlett et al. (2014), should be translated into every European language and widely circulated. It is likely to become a classic. It explains how the forensic lessons learned from the UK temporary works failures of the mid-1970s and published in the Bragg Report in 1976 (HSE, 1976) were successfully embedded in BS 5975 (BSI, 1982), resulting in a profound reduction in the number of very serious falsework and excavation collapses. It systematically compares Stephen Bragg's recommendations with BS 5975 and its replacement EN 12812 (BSI, 2008) and raises concerns about the philosophy of this new standard and what may happen if it is used alone without modification. A primary concern is that EN 12812 does not require the appointment of a temporary works co-ordinator (TWC) as the controlling mind of the process and that this omission will contribute to a reduction in the system's safety with poor communications between all parties leading to a less robust design. The authors consider the imminent absence, owing to retirement, of enough experienced designers and site supervisors with memories of the pre-Bragg collapses may create a need for larger factors of safety in future works. I recall this generation effect was remarked upon in the mid-1970s with respect to the high alumina cement conversion that occurred in some swimming pool roofs at the time. The authors argue there is a need for more specialised training for their replacements because the EN 12812 approach uses more second-order computerised calculations and so will not compensate for reduced practical insight, inexperience and the inability to recognise counterfeit proprietary equipment on site.
Most of us recognise that some failures occur because of the unfortunate combination of several minor defects, which singularly are not critical, but combined can induce a system failure. They quote the analogy of the holes in slices of Swiss cheese aligning. I prefer to think of each of these influences as a piece of coloured string lying on a sheet of paper. The thickness of the string is a measure of its criticality in the system, and if some of these forensic strings overlap at one point such their combined height above the paper is beyond some threshold probability value then a forensic event occurs. Not having the TWC there to monitor these interactions and tease the strings apart is a concern. Another interesting point made by Hewlett et al. is the observation that BS 5975 has ‘reasonableness and a sense of judgement at its heart’; whereas EN 12812 favours ‘highly analytical methods and lower factors of safety’.
We have been there before. Some nineteenth-century lock gate designers in major ports adopted a robust approach to specifications and their gates were able to survive impacts in the fog, the floating detritus wedged in closing gates and the associated stress concentrations; and they were repairable. They were rarely out of service. Other designers were publicly critical of this approach and developed, with the available theory, a more refined design with less material and plenty of arithmetic. Their gate's reliability was not as good and the commercial loss for their employers outweighed the cost of thicker oak frames and pitch pine planks (Scholfield and Smith, 1999). Solzhenitzin (1974) in The Gulag Archipelago quotes the Russian proverb, ‘Dwell on the past and you will lose an eye; forget the past and you will lose two eyes.’ We must keep our eye on the lessons from Bragg.
When the original editorial panel first met there was a remarkably quick accord on what we wanted to do. Most of us had at some time designed, built, researched and taught engineering in some way. Somewhere along the line to varying degrees we had been involved in forensic matters. All of these experiences coloured our ambitions for Forensic Engineering. We wanted a truly international panel and a broad content beyond mainstream civil or structural engineering, to establish the unity of the forensic discipline, if it exists as such, and many of us were keen to have the occasional scholarly paper on the discipline's origins. We are doing all these things, for example, having had papers on system dynamics applied to German power distribution networks in the past, and looking forward soon to case studies in expert witness work in building services failures. My own working definition for forensic engineering is whatever this journal publishes over a decade is the discipline.
The invited paper by Jarvis (2014), therefore, comes from the panel's learned society ambition to seek the origins of the subject. It is for a classical scholar looking at the original sources to tell us why Hammurabi's legal code was more lenient to shipbuilders than house builders, but Jarvis takes us a considerable way from the European medieval conclaves of builders holding enquiries into the quality of cathedrals that did not fall down, through to the regulatory bodies the British government set up to make safer shipping and railways in the nineteenth century. He shows artfully that the forum for technical debate existed in the fourteenth century, and reminds us that in dealing with well-minuted history we can look at the political, economic and personal drivers that influence forensic issues and matters of state then in a way we cannot always do today. Certainly, there was a forensic process involved in the full-scale loading of a warship in a dry dock, Lloyds insurers developed a permissible design life for their insured vessels and Mallet wrote on the corrosion of sea cocks, which is still a cause of yacht sinking today. Corrosion and remaining design life is a theme in the third paper in this issue by Roffey et al. (2014). What we would see as a database established by the UK Board of Trade at the end of the nineteenth century for wrecks is similar in principle to the UIC's (International Union of Railways) database for railways, referred to in the fifth paper by Lane and Thompson (2014). So the systematic records we have today have their origins in the nineteenth-century prototypes. There is a de facto continuing forensic tradition and some remarkable published work on 1000 exploding steam engines that established how petty savings on maintenance was ill advised. There are also signs that the forensic cycle was applied in terms of design upgrades following some steam engine failures.
Jarvis also indicates how, in a rapidly expanding rail industry with the slow communications of the time, the application of known forensic issues, including those related to what we would now call asset management, did not happen. This need for communications was brought out by Hewlett et al. in the first paper with respect to retaining the TWC. So the historical imperative is there. For many perception is all, and Jarvis describes how the Royal Commission into the Dee Bridge disaster established a precedent in terms of having a research budget as part of a major public and scientific forensic investigation. This leads into the discussion of the third paper in this issue which is on the Forth Road Bridge.
Roffey et al. (2014) describe how the early application of forensic methods, as part of the asset management of a major bridge, can give continuing confidence in its structural capacity. As a result of advanced scientific techniques, including electron microscopes, materials analysis and fracture mechanics, applied to samples of the suspension cables, a retrofitted waterproof elastomeric wrap and dehumidification system has successfully retarded corrosion, and continual monitoring will provide the designers with feedback on the material's behaviour as part of the forensic cycle. The authors identify the need for further research into fatigue testing of cables in an environment closer to that experienced in service conditions. Thus the forensic process is driving forward the need for knowledge as it did in the previous two centuries as described in Jarvis's paper. There is a continuity in the discipline and a link to explicit asset management. The authors also identify that there is a combination of mechanisms at work and so one might argue the dehumidification scheme is teasing apart the forensic string mentioned earlier.
The fourth paper, by Abé et al. (2014), is a celebration of successful forensically informed asset management. The paper describes how a national rail network's bridges are managed against natural forces and vehicle impacts. Graphs show how the focused investment on accident countermeasures following forensic and statistical analysis has reduced disaster-related losses. They also demonstrate that as economies prosper such investment protects increasingly valuable lives. Is the corollary true? A case study is presented in which an out of service train was blown off a high bridge prone to significant wind loads. The bridge was, following wind tunnel testing, rebuilt to permit a wind shield to be carried so train service levels were maintained. This was affordable commercial robustness.
Retrofits for scour effects on bridge foundations and monitoring using inclinometers has been introduced on 150 susceptible bridges with a view to reducing fatalities after forensic evidence showed that inclination of supports increased the risk of fatal accidents. Similarly 40% of the bridges with clearances lower than 4ṡ5 m have had protection frames retrofitted to protect them against vehicle impact. So forensic methods can recognise good practice as well as bad and this paper, like the preceding one in this issue, links the forensic methods to asset management.
Japan is the home of judo, and my first experience of a forensic approach was over 50 years ago as I studied a book that analysed the traits of winning judoka from photographs at international competitions. It was most enlightening and as I recall those who had a good breakfall technique were very successful. Preparing to fail was a key to success. And forensic methods were certainly used in the fore-running jujutsu styles of the Tokugawa period, perhaps some 200 years earlier when poor technique was fatal (Leggett, 1978). The point is that the forensic process has existed, in principle at least, a long time, and outside engineering and science. It is not ours alone and it is central to the reflective practice of any calling. What this paper shows is that philosophically forensic events can be turned to our advantage, introducing better practice. Many of these approaches can be considered for rail networks in other countries too.
The fifth and final paper in this issue, by Lane and Thompson (2014), is similar to Abé et al. but approaches the topic from the viewpoint of the regulator whose origins were identified by Jarvis. Forensic events have been collated into a European database that indicates some parts of Europe having eight times the derailment rate as the safest; they have influenced standards and the safety risk model. There is common cause with many of the scenarios identified in the paper by Abé et al. and the authors address the issue of structural vulnerability to train impact. A holistic approach is advocated looking at the interaction of elements, the possibilities of risk mitigation and the use of event-tree analysis to look at combinations of events. Teasing out the forensic string, as it were, to create a reasonable level of robustness without generating costs disproportionate to the benefits achieved by any strengthening works. The forensic information is back where it belongs serving the design process.
Concluding this editorial, there are perhaps three strong threads: the warning in the first paper, the historical theme's lessons, and the fact that three of the papers impinge on asset management. The Institution of Civil Engineers now has journals in history, forensics and asset management and these make for relevant reading for those managing our built environment.
