measuring

concrete mid-print

Problem statement

Additive construction by extrusion (ACE) lays a cementitious mortar bead by bead into walls, and every structural property of the finished element — strength, stiffness, where it cracks — depends on what the material was doing in the seconds after it left the nozzle: how stiff it was, how quickly it built up yield stress, and how well each new layer wet and bonded to the one below. NIST's assessment is that "the concrete design and engineering community lacks sufficient knowledge about the performance of 3DCP structures subjected to designed loading scenarios," in part because "the role of the printing process in determining the failure mode" is not understood, and that "new metrologies are needed to measure the complex rheological properties of AC materials during the construction process." Today those properties are mostly inferred from off-line lab rheometry, flow-table tests and post-hoc cores; research prototypes of in-line monitoring exist (nozzle-pressure sensing, camera-based filament geometry, near-nozzle accelerator feedback), but no validated, standardised in-process measurement links what happens at the nozzle to a printed wall's interlayer bond and anisotropy, which therefore remain unknown until the element is tested destructively.

Why this matters

The layer interface is exactly where printed elements are weakest — NIST notes that "the interface between layers and the connection between the webbing and perimeter layers are likely to be pathways for crack propagation," and that layered deposition "introduces anisotropy, and the layer interface bonding could provide a preferential pathway for crack propagation." Because printing happens largely during cement's dormant hydration period, the chemistry that determines whether two layers fuse or merely touch is changing minute by minute with temperature, humidity, pump pressure and time-between-layers, and NIST flags the need for "new analytical techniques to assess cement chemistry during" that window. Without in-process measurement there is no basis for a performance-based acceptance standard — the goal NIST, ERDC-CERL, ASTM and ACI committees have organised two "Path to Standardization" workshops (September 2023, August 2024) and an ACE consortium around — and code officials are left with either blanket conservatism or one-off destructive testing per project.

What’s been tried and why it hasn’t worked

Research groups characterise printable mixes with rotational rheometers, penetrometers, slump/flow tables and unconfined compression of fresh "green" cylinders — tests that describe a sample in a lab minute, not the extruded filament under nozzle shear, ambient drying and the weight of layers stacked above it. Buildability is checked empirically by printing until the wall slumps; interlayer bond is checked by sawing prisms from printed specimens days later. These approaches fail as quality control because printability is time- and environment-dependent (a mix that prints at 9 a.m. can fail at 2 p.m.), because a lab rheogram does not predict interlayer bond under a given layer-cycle time, and because destructive post-tests cannot be run on every wall of a house. NIST's program statement is precisely that "new metrologies are needed to measure the complex rheological properties of AC materials during the construction process" and that "developing these in-situ rheology measurements could provide important feedback into the construction robot to eliminate manufacturing defects" — i.e., research-scale monitoring and feedback demonstrations exist (a 2021 review of real-time extrusion quality monitoring; 2024 near-nozzle secondary-mixing feedback control), but a validated closed loop does not, because the correlation between an in-process signal and structural outcome has not been established or standardised. This is a measurement-science gap first and a standards gap second: ASTM/ACI cannot write an acceptance test around a quantity nobody can measure at the nozzle.

What would unlock progress

An in-line, non-contact proxy for fresh-state yield stress and structuration (bead geometry and sag under laser or stereo vision, extrusion pressure/torque signatures, ultrasonic or dielectric readings of early hydration, infrared surface temperature/moisture of the receiving layer) — validated against interlayer bond strength measured on the same specimens — would let a printer adjust speed, layer time or accelerator dosing in real time and give a standards body a measurable quantity to specify. Adjacent fields have the pieces: metal and polymer additive manufacturing have in-situ melt-pool/bead monitoring feeding qualification work, and commercial in-truck slump monitoring exists for ready-mix concrete. None has been transferred to cementitious extrusion as a validated, standards-grade practice; the existing 3DCP monitoring work is research-scale and not yet tied to an acceptance metric.

Entry points for student teams

A team with a small gantry or robot-arm mortar printer (or a syringe-scale analogue) could instrument the print head with a low-cost extrusion-pressure sensor and a camera measuring bead width/height and sag, print specimens across a matrix of layer-cycle times and ambient conditions, then saw and test interlayer tensile/shear strength — producing an open correlation between an in-process signal and bond. A sensing team could evaluate whether IR thermography or surface-moisture imaging of the receiving layer predicts weak interfaces. A standards-oriented team could draft, from the data, a proposed in-process acceptance metric and compare it to the language in draft ACI/ASTM ACE documents. Relevant skills: cement/concrete materials, mechatronics, computer vision, statistics.