Heat was the biggest threat to the success of the casting operation. Cold, even the bitter cold of a Baltic winter, was less of a problem once the decision was made to cast all the segments under cover. 'But at peak there would be two segments, each containing 2,700m3 of high cement content concrete, gaining strength inside,' points out 0TC concrete production manager Lars Lundberg. 'That's a lot of concrete, giving off a lot of heat.'
Minimising the temperature rise caused by the hydrating cement was just one of the parameters controlling the mix design. For the sake of long term durability water/cement ratio had to be low - 0.38 was specified at first, later relaxed to 0.39. And, as there would be no other waterproofing measures, the concrete would have to be watertight.
Strength was less of a problem. A characteristic cylinder strength of 50Mpa would meet all specification requirements. Previous experience among the joint venture partners suggested that a triple cementious blend of Portland cement, ground granulated blastfurnace slag and condensed silica fume would reduce heat gain, improve plastic properties and produce a much less permeable concrete.
'But we needed high early strengths, at least 20Mpa at two days,' Lundberg says. 'So in the end we decided on a blend of low alkali sulphate resisting Portland cement, plus pulverised fuel ash and microsilica in a 80:15:5 ratio.'
Although reinforcement density was generally remarkably low, there were a few problem areas, particularly under the central service gallery. High workability would also facilitate the use of fixed concrete pumps for placing. Low water/cement ratio and high 'pumpability' is a paradox that can only be resolved by high admixture use, and careful choice of aggregates.
'We also had to meet a tight specification on hardened density, 2400kg/m3, to make sure the buoyancy calculations worked,' Lundgren reports. 'The answer was to use Norwegian-sourced materials, a nice rounded natural granite sand and a mix of natural and crushed granite coarse aggregate.'
Combining superplasticisers and stabilisers to achieve a pumpable, placeable mix was one thing, fine tuning it for the very large complex pours involved was another. 0TC turned for advice to Dutch specialist Intron, whose specialist software had already been used in the design of the segments (see main story).
Lundberg says the challenge was to devise a way of pouring the segments such that any given layer would not go off before fresh concrete was poured on top of it. A pouring sequence which minimised the time interval between layers was a start, but not enough on its own to reduce the risk of leak- prone 'cold joints' to acceptable levels.
'Intron finally recommended that we vary retarder dosage throughout the pour with heavier dosage in the first concrete poured,' says Lundberg.
Thermocouples cast into the segments monitored the temperature rise during the first few days. 'The specified maximum adiabatic increase was 43degreesC, maximum concrete temperature was 65degreesC.
'In summer this meant the highest fresh concrete temperature we could allow was 22degreesC, which wasn't easy.'
With so little water in the mix, simply batching with chilled water or even crushed ice was not enough. Eventually crushed ice was added to the aggregate bays and allowed to melt and drain away. This did the trick, even though 'everything was right on the limit sometimes', according to Lundberg.
This particular problem obviously disappeared along with the warm weather - to be replaced by the risk of thermal shock when the segments eventually left the sanctuary of the casting shed and faced the Copenhagen winter. A simple insulating blanket around the segments took care of that.