Two Kilos per Cubic Metre: Getting Internal Curing SAP Into the Mix

Two kilos of polymer per cubic metre went into the first truck of the day. The pour was a bridge deck, 42 m long, and the crew added it dry at the end of the mixer cycle because that was the only window the batch plant could give them. Nine cubic metres went down. By the time the screed passed, there were pale nodules on the surface the size of peas, and a few of them were still powdery inside when a core was cut a week later.

Internal curing is one of the few SAP applications where the polymer is not there to absorb anything at the time of use. It is there to hold water and give it back slowly, from inside the concrete, while the paste dries and shrinks. That changes what you measure, what you dose, and above all how you get the powder into the truck.

Dose is small, and the water matters more than the powder

Typical internal curing doses run from 0.2 to 0.5 percent of cement weight, which on a 380 kg/m3 mix is roughly 0.8 to 1.9 kg per cubic metre. The polymer itself is not the active ingredient in the engineering sense. The water it carries is. Each kilogram of SAP at 25 g/g in the mix water brings about 25 kg of extra water into the paste, and that water has to be counted in the effective w/c ratio or the mix design drifts.

That is the first trap. A mix designed at 0.38 w/c with 1.5 kg of SAP per cubic metre and no water adjustment behaves like a 0.42 mix, because the polymer took water that the paste needed for workability and locked it inside a network. Slump drops, the crew adds water at the chute, and the strength result at 28 days is blamed on the polymer. The polymer did its job; the batching did not.

The correction is simple arithmetic: 1 kg of SAP at 25 g/g removes 25 kg of water from the free water column. Add that water to the batching sheet, mark it as absorbed, and keep the free w/c where the mix design intended.

Desorption, not absorption, is the specification

A curing agent that holds water perfectly is useless, because the water has to leave the gel and enter the hydrating paste. The property that predicts performance is the desorption curve: how much of the absorbed water the gel releases as relative humidity falls from 100 percent down to about 90 percent, which is the range a drying paste passes through.

A grade that releases less than half its water in that window is a reservoir that never opens. A grade that dumps everything at 98 percent RH gives the water back too early, while the paste is still plastic and the bleed water is still present. You want a release profile concentrated between 98 and 93 percent RH, and you want it measured, not assumed from the free swell figure.

This is why hygiene-grade polymer, tuned for retention under load, is a poor default for concrete. High retention is the opposite of what internal curing needs. Grades supplied for construction are usually tuned the other way, with lower surface crosslinking and a release curve that has been run in a humidity chamber.

Getting it into the mix without fish eyes

Dry addition to a wet mix is where most site failures start. A granule dropped into wet concrete begins to hydrate on contact. If it is not dispersed within a few seconds, the outer layer gels and the granule becomes a sealed bead, which is the pale nodule the crew found on that deck. The inside stays dry for days.

Three approaches work, in descending order of reliability. Pre-soak the polymer in a measured volume of mix water for at least 20 minutes, then add the swollen gel as part of the water. Dry-blend the powder with a portion of the cement or the fine aggregate before it goes into the truck, so every granule is separated by solid particles. Or dose into the mixer through a high-energy stream while the drum turns at full speed, which works but needs a trained operator and a controlled feed rate.

What does not work is dumping a 20 kg sack into the hopper and hoping the drum will do the rest. Even at full speed, a 20 kg slug of powder produces clumps that survive the whole transit. If you must dry-add, feed at no more than about 1 kg per 10 seconds into the moving stream, and never into a standing load.

What to check before the first pour

Run a trial batch at the actual dose and measure four things: slump immediately after mixing, slump at 45 minutes, the mass of gel visible on a washed sieve sample, and the 7 day shrinkage on a prism. The fourth is the point of the exercise; the first three tell you whether the placement will go ahead.

Also check the mesh. Coarse granules are easier to disperse but slow to release; fine powder disperses badly in a dry blend but releases faster. Many concrete grades sit around 50/100 mesh as a compromise, and the choice interacts with the mixing energy available at your plant, not with the polymer alone.

Before the next pour, take one sack, pre-soak 200 g in 5 litres for 20 minutes, and look at it. If the gel is uniform and there are no hard cores, the grade and the soak time are compatible. If you find beads, no admixture schedule will save the deck. There is a short note on dispersing SAP into a wet mix that covers the same order of operations.