Casting flaws found in failed cast iron beam

But until the two halves of the beam are tested to destruction and the infill concrete above analysed, engineers will be no nearer to pinning down the cause of the sudden failure.

Until it fractured near midspan, the 6.6m span, 370mm deep beam had been part of a system of jack arches supporting a first floor level terrace behind a row of five storey town houses in Hyde Park Gardens (NCE 8/15 August).

'We spotted what looks to be a blowhole in the bottom flange which had been filled either with lead or 'Beaumont's Egg', a mixture of lampblack, beeswax, iron filings and 'Fiddler's Rosin' - resin from the Balsam Pine. This was frequently used to cover up such defects, ' explained consultant Cameron Taylor Bedford's technical director Clive Richardson. The firm is investigating the failure on behalf of the Church Commissioners, which owns the building.

'This could be where the failure began. We were also surprised by the unusual degree of surface corrosion, with up to 2mm of flaky rust all over.'

By Tuesday morning a demolition contractor had cleaned up the rubble from the collapse and gingerly opened up a hole from terrace level right through into the space beneath. This revealed that the depth of construction above the beam was much greater than originally assumed (see diagram).

'Above what could be either lime or gypsum concrete infill there's what was known as 'tile creasing' - basically layers of ordinary roof tile stuck together with Roman cement. And above that there's more concrete. This would explain why the terrace didn't fall into the hole below when the beam failed - but it still doesn't explain why the beam failed when it did, ' said Richardson.