There were many good reasons for the structural designers to aim for as low a reinforcement content as possible in the immersed tube elements. Not least was cost. A saving of only a small percentage adds up to a lot of money when around 40,000t of rebar is involved, especially at Scandinavian steelfixers pay rates.
Less reinforcement also means faster production. Most importantly though, a low density of reinforcement would encourage full compaction of the plastic concrete - a key requirement if the completed elements were to be waterproof.
Doing away with the need for sand ducts, cooling tubes and temporary legs was a good start. Casting the whole segment in one pour cut down the need for reinforcement to resist early age cracking. But it was adopting the principle that each element should be individually designed that really made the difference, says Symonds immersed tube design team leader, John Busby. 'The biggest variable was the increase in hydrostatic pressure towards the centre of the crossing.
'The key was to develop a detailing system which avoided surplus reinforcement in the shallower elements.'
He adds that the design conditions for the segments and elements could be divided into four distinct phases: in the casting yard, float out, lowering, in service. Of these the most complex were while the elements were still on dry land and the concrete was a long way from its mature strength.
However, one of the main tasks of the temporary works designers was to ensure that none of the handling stresses in the casting yard required extra reinforcement above that needed to meet the in-service conditions (see main story). Apart from simple hydrostatic pressure, load cases for these included the risk of a sunken ship coming to rest on the crossing, and of it being struck by a falling anchor. 'The irregularities of the gravel bed foundation made calculating the resulting stresses quite difficult,' Busby comments.
Developing a reinforcement design that facilitated pre-fabrication became the main priority. Some unusual options were selected. Bars up to 21m long were shipped in by sea, greatly cutting down the number of laps. Such laps as were still necessary were generally confined to compression zones. And where access openings through the internal dividing walls were required, the contractor originally opted to cut these through the standard cage after prefabrication and add trimming bars as required.
'Later this was switched to detailing the vertical bars around the opening but still cutting the horizontals,' Busby reports.
Making up the cages from a series of sub-assemblies contributed to high production rates, he adds. 'Everything was based on a module of 25mm diameter bars at 300mm centres. If extra bars were needed, they were first added between the basic bars, reducing the module to 150mm.
'Then, if still more bars were needed, we started bundling, so that nominal centres didn't go below 150mm and the cage was still very open to facilitate pouring.'
By dint of measures like these the design team got the average reinforcement density down close to 80kg/m3 - a total saving of nearly 3,000t over the original estimate.