Shrinkage and Cracking in Printed Concrete

Shrinkage is the reduction in volume a cementitious material undergoes as it hydrates and exchanges moisture with its surroundings. On its own it is harmless: a specimen free to move simply gets smaller. Cracking occurs when that movement is restrained — by an adjacent element, by reinforcement, by a foundation, or, in a printed object, by the layers already hardened below. For 3D concrete printing this matters more than for most cast concrete, because printable mixes are paste-rich, low in water-to-binder ratio, and placed without formwork onto restraint they created themselves minutes earlier. Shrinkage is not one phenomenon but several, with different drivers and timescales, and mitigating the wrong one wastes effort.

What it is / Why it matters

Chemical shrinkage is the underlying cause: the hydration products of cement occupy less absolute volume than the water and clinker that formed them. Before setting this shows up as a bulk volume reduction; after setting the solid skeleton resists it and the deficit becomes empty pore space.

Plastic shrinkage occurs in the first hours, while the material is still soft. Water evaporates from the exposed surface faster than bleed water can replace it, menisci form between particles, and the surface contracts and tears. Plastic settlement is the related early movement: solids sink under gravity while water rises, producing subsidence cracks over obstructions.

Autogenous shrinkage is self-desiccation. Hydration consumes water from the pore network, and with no external supply the finest pores empty first. Curved menisci in partially filled pores put the pore water in tension, and the resulting capillary depression pulls the solid skeleton inwards. It depends strongly on water-to-binder ratio: small in ordinary concrete at moderate w/b, dominant at low w/b with high binder content — precisely where printable mortars sit. It begins at setting and much of it develops within days.

Drying shrinkage is moisture loss to the environment. It develops from the surface inwards, so it is slow, non-uniform, and creates a moisture gradient that itself causes curling. Carbonation shrinkage is a slower, surface-confined process in which atmospheric carbon dioxide reacts with hydration products and the resulting phase changes cause additional contraction. Thermal movement is not shrinkage but contributes to the same crack pattern: the heat of hydration raises the material's temperature, and the subsequent cooling under restraint produces tensile stress.

3D printing amplifies most of these. A printable mortar carries a very high paste and binder fraction to get the rheology it needs, which raises both autogenous and drying shrinkage. Its water-to-binder ratio is low, pushing autogenous shrinkage up further. Its aggregate is fine sand, so there is less stiff, non-shrinking skeleton to restrain the paste internally. There is no formwork, so every surface is exposed to evaporation from the first minute rather than after demoulding. Printed walls are thin, giving a high surface-to-volume ratio. And each new layer is deposited onto material that is already stiffening, so the fresh layer is restrained along its entire underside. The characteristic outcomes are cracks along the interlayer plane and cracks running through several layers where global restraint dominates.

How it is measured

Free length change of hardened specimens is measured under ASTM C157, which follows the length of a prism with reference studs over time, and by EN 12617-4 for repair and similar products. Both give free shrinkage — the material property, not the cracking risk.

Autogenous strain is measured with the corrugated tube method of ASTM C1698, in which sealed paste or mortar is followed from final setting onwards. Sealing removes moisture exchange, so the measurement isolates self-desiccation from drying.

Restrained shrinkage is what actually predicts cracking. The restrained ring test, ASTM C1581, casts a ring of material around a steel ring and monitors the steel strain until cracking, giving a time to cracking and an indication of stress rate. It is the most widely used comparative test for crack sensitivity.

Embedded strain gauges and surface DEMEC mechanical strain gauges are used for real elements and for printed specimens, where moulded prisms do not represent the material. For plastic shrinkage, a common laboratory approach is a restrained slab panel under controlled wind and temperature with crack area recorded; this is a research and comparison method in the 3DCP literature rather than a harmonised standard.

The distinction between free and restrained measurement must be kept explicit. A low free shrinkage does not guarantee low cracking if the material is stiff, brittle and heavily restrained; a higher free shrinkage may crack less if the material creeps and relaxes stress. In printed specimens shrinkage is also likely to be direction-dependent, so the measuring direction relative to the layers should be recorded.

Typical ranges

All values are indicative and drawn from the published concrete and 3DCP literature in general terms. They are not Concreef measurements.

What changes it

Paste content is the first lever. Shrinkage happens in the paste; the aggregate restrains it. Improving particle packing so the same rheology is achieved with less paste reduces shrinkage directly, which is why grading and filler selection matter as much as admixtures.

Water-to-binder ratio works in two directions at once. Lowering it reduces drying shrinkage, because there is less evaporable water, but increases autogenous shrinkage. There is no single optimum; the balance depends on which mechanism dominates in the element's exposure.

Internal curing attacks autogenous shrinkage at its source. Pre-saturated lightweight aggregate or superabsorbent polymers hold water released as the pore humidity falls, keeping fine pores filled and suppressing capillary depression. The water must be proportioned against the binder's chemical shrinkage and distributed finely enough to reach the paste.

Shrinkage-reducing admixtures lower the surface tension of the pore solution, reducing capillary depression for a given pore emptying. They affect both autogenous and drying shrinkage and are dosed per supplier data under ASTM C494 classification.

Fibres change the consequence, not the cause. Polypropylene fibres at low dosage are effective against plastic shrinkage cracking; structural fibres limit crack widths later. Neither reduces free shrinkage.

Supplementary cementitious materials cut clinker content and heat of hydration, which usually helps thermal and early-age cracking. Silica fume, by refining the pore structure, tends to increase autogenous shrinkage rather than reduce it.

Curing and protection are the highest-value intervention in printing because they are immediate and cheap. Covering the element, controlling wind across it, misting, and keeping the workshop humidity up all reduce the evaporation rate during the hours when plastic shrinkage cracking is decided. Joint and geometry design handles what remains: movement joints, controlled crack inducers, and avoiding abrupt section changes that concentrate restraint.

Failure modes

Plastic shrinkage cracks appear within hours as short, shallow, roughly parallel surface cracks, most often on the top of the last-printed layer and on faces exposed to draught. They are an evaporation and curing failure.

Plastic settlement cracks follow obstructions — embedded reinforcement, inserts, sharp geometry changes — where solids settle while water rises.

Interlayer cracking develops where a fresh layer shrinks against an already hardened one and the interface, the weakest plane anyway, cannot accommodate the differential. Through-layer cracking appears where the element is restrained externally and the crack propagates across several layers, usually at the point of greatest restraint or section change.

Curling and warping arise from the drying moisture gradient: the surface shrinks while the interior does not, and a thin printed shell distorts. Crazing is fine surface map cracking from rapid surface drying or carbonation of a weak surface layer.

Two diagnostics help in practice. Crack timing separates the mechanisms — hours means plastic, days autogenous, weeks to months drying. Crack orientation relative to the layers separates interface problems from global restraint.

Concreef context

Concreef is a Bulgarian 3D concrete printing company based in Sofia. It runs material trials in its workshop on a Crane WASP printer, currently printing with a commercial premix while developing its own dry mix. It has no published test data, and nothing here should be read as a Concreef measurement or performance claim.

In workshop practice, shrinkage is the property that shows up after the print looks successful. A piece that extruded cleanly and stood up can still craze, curl or crack along an interface over the following days, and the cause is usually decided in the first hours after deposition rather than in the mix design alone. Because printed pieces have no formwork and a large exposed surface, Concreef's working practice treats immediate covering and protection from draught as part of printing rather than an afterthought, and treats observed cracking as diagnostic information — recording when a crack appeared, where it ran relative to the layers, and the ambient conditions. Mix-side measures such as internal curing, shrinkage-reducing admixtures and reduced paste content are part of the dry-mix development, and any results will be reported with the test method, age and restraint condition stated.