The 365-Day Cube

The two most consequential roller-compacted concrete dams in Indian practice are not the largest in the country. They are Ghatghar in Maharashtra, built between 2002 and 2007 as the first RCC dams ever constructed in India, and Middle Vaitarna in Maharashtra, built between 2010 and 2012 for the Municipal Corporation of Greater Mumbai. Middle Vaitarna stands 102.4 metres high. It used 1.215 million cubic metres of roller-compacted concrete, placed over 15.2 months, ranking it as the tenth-fastest RCC dam ever constructed worldwide. Ghatghar is smaller in scale and central to the story, because Ghatghar was where Indian engineers first met the puzzle this issue is about.

The trial sections at Ghatghar used a mix called G-85, identified by a cementitious content of 85 kg per cubic metre of concrete. Two test sections were placed. Their engineering log, recorded in the design-mix paper that subsequently informed Middle Vaitarna, reads in the language a field engineer recognises immediately.

Test section-1: G-85. Segregation, insufficient strength and weak in joints.
Test section-2: G-85. Insufficient strength and weak in lift joints.

Ashtankar V.B. and Chore H.S. (2014). Development of design mix roller compacted concrete dam at Middle Vaitarana. Advances in Concrete Construction, Vol. 2, No. 2, pp. 125-144. Techno-Press. Table 1.

A first reading of those two entries says the trial sections failed. They did, in the operational sense that the as-placed concrete had inadequate strength and weak bond at the lift joints. But the deeper question, which Ghatghar's engineers had to settle and Middle Vaitarna's engineers inherited, was a different one. Was the mix wrong, or was the acceptance criterion being applied to it wrong?

The answer turned out to be a mix problem, but the path to it ran through the acceptance-criterion question. When Middle Vaitarna ran its design-mix trials a few years later under MERI Nashik, the engineering team produced a strength-versus-age table for several candidate mixes. The numbers tell the story without commentary.

For the mix designated G-75 (75 kg of cement plus 145 kg of fly ash per cubic metre, fly ash replacing roughly two-thirds of the cementitious content), the 7-day cube strength was between 1.25 and 3.46 megapascals. At 28 days the strength was 4.95 megapascals. At 56 days, 7.89. At 91 days, 10.82. At 180 days, 17.31. At 365 days, 24.50 megapascals. The design target for the dam was 17.5 megapascals at 365 days. The mix approached design at six months and exceeded it by a full year.

Apply 28-day conventional concrete acceptance criteria to a 7-day RCC cube reading 1.25 megapascals, and the placement is condemned. Apply RCC mass-concrete acceptance criteria, anchored at 365-day strength, and the same number is exactly where the engineering predicted it would be. The cube does not lie. The acceptance criterion does, when it is the wrong criterion for the mix.

That is the Middle Vaitarna and Ghatghar lesson on test-result interpretation. The criterion has to match the engineering. The G-85 trial-section failures at Ghatghar were not a low-strength puzzle; the segregation and the weak lift joints were real and traced to aggregate gradation and admixture dose, and were resolved by moving to G-75 with higher fly-ash content and better gradation. But the surrounding question, the one that hangs over every QC engineer running cube tests on a mass-concrete pour during monsoon this year, is the test-result interpretation discipline.

The acceptance protocol for a properly run RCC pour is not the 28-day cube. It is a four-point check at the time of placement, anchored by strength verification at one year. Vee-Bee consistency is measured to confirm the mix is workable enough to compact without segregating. Density is measured with a nuclear gauge against 97.5 per cent of the theoretical air-free density as the acceptance threshold. Lift-joint preparation is logged for every joint, because joint quality is the dominant failure mode in RCC, and "weak in lift joints" is precisely the language a mix earns when its segregation manifests at the joint. Strength is then verified at 91 days on cylinders and at 365 days on the final acceptance specimen.

The Bhakra investigation in 2020 makes the same lesson visible from the other end of the life cycle. When the Central Water Commission and the National Council for Cement and Building Materials extracted core samples from Bhakra Dam under the Dam Rehabilitation and Improvement Project, the Type B mass concrete designed for 15.85 megapascals returned equivalent cube strengths ranging from 18.32 to 35.98 megapascals, 16 to 127 per cent above design after more than fifty years in service. The cores demonstrated that the slow pozzolanic reaction of the calcined-clay pozzolana used in the original 1957 mix continued for decades. The puzzle that triggered the investigation was unusual deflection observed in the structure; the cores ruled out concrete deterioration as the driver.

The lesson: The strength curve does not lie. The acceptance criterion has to match the engineering. Roller-compacted concrete mixes with high fly-ash replacement are designed for 365-day strength, and 28-day cube criteria applied to a 7-day RCC sample produce the wrong verdict on the right concrete. Bhakra's half-century-old cores tested up to 127 per cent above design strength because slow pozzolanic gain continued for half a century. None of these are exotic findings. They are the standing discipline of mass-concrete acceptance, written down in the design-mix records of two Indian RCC dams that every QC engineer running monsoon-season pours this year should read before the first low-strength cube turns up at the lab.

Read more: What a 3-Day Cube Tells You That a 28-Day Cube Doesn't →