Can engineers really design for a 9/11 style attack on steel framed buildings- Dave Parker reports.
Do engineers really know how tall steel framed buildings will behave in an intense fire?
To judge from the views expressed at last week's conference, held in Washington to discuss the official inquest into the collapse of the World Trade Center towers on 11 September 2001, the answer is no.
In October 2002 the United States National Institute for Science & Technology (NIST) was asked to determine exactly what happened to the World Trade Center towers on 9 September 2001.
It was also asked to make recommendations about how to minimise the risks of similar calamities in the future.
NIST engineers must have known that whatever they discovered and concluded would not satisfy everyone.
Last week, on its campus just outside Washington DC, NIST set its conclusions and recommendations before an informed, and potentially sceptical, international audience of academics and fire engineering consultants.
The draft final report took nearly three years of work by 200 researchers. It cost $16M (£8.8M) and contains 10,000 pages. The verdict on the shocking collapse of WTC1 and WTC2 (the twin towers) is simple.
They absorbed the enormous kinetic energy of 100t Boeing aircraft striking them at 800km/h and could not cope with the fires that followed. There was no evidence of deliberate demolition or missile strikes, or internal explosions.
NIST investigators identified and located the 14 grades of steel used in the structures, and fed their unique properties into the computers. They took ultrasound measurements of the thickness of the steel tubes that make up the Boeing's landing gear.
They viewed 300 hours of videotapes of the tragedy, and correlated their massive computer models against them.
They performed simulations that took weeks to run.
They can explain almost every moment on every video - except why aircraft landing gear components went right through the towers and out the far side.
It was a mammoth undertaking, without precedent, and NIST deserves congratulations for the depth and intensity of the investigation. But at the end of it all, engineers around the world are entitled to ask what it all means.
Yes, we have a reasonably convincing model of what happened on 9/11 - but alternative models exist (see News). And yes, we have some obviously sensible recommendations on means of escape which will ensure that future tall buildings will have much more robust stairwells, and probably hardened elevator access for fire-fighters as well.
But what remained unchallenged last week was the acceptance that the traditional prescriptive approach to specifying fire resistance and more advanced performance based computer fire engineering techniques both have their weaknesses.
Many engineers now believe that current standard fire testing and the data it produces is no reliable guide to how structures and structural members behave in real fires. It would seem that this will still make it difficult for engineers to produce convincing designs for intense fires.
NIST's re test programme for the WTC investigation was just one piece of evidence to highlight the inadequacy of current test methods which generally rely on scale models.
Connections between structural elements are almost never tested at full scale, yet these are often the most vulnerable parts of a structure. The uncertainties that still remain after the NIST investigation ought at least to persuade the US government and perhaps its British counterpart that more money should be spent on full scale testing in future. Only then will the work done by NIST have created a truly worthwhile legacy.