Open3DCP: A Public Data Schema for Performance-Based 3D Concrete Printing Records

Nicholas Sonnentag
Sunnyday Technologies
May 2026

Abstract

Three-dimensional concrete printing (3DCP) is a coupled material-process problem. A printable cementitious system is defined by composition, rheology, equipment, toolpath, environmental exposure, curing, specimen extraction, test orientation, and measured performance. Current 3DCP publications often report many of these facts, but they are distributed across prose, tables, figures, and supplementary files using inconsistent names and unit bases. That fragmentation makes cross-study comparison, meta-analysis, and machine-learning workflows harder than the underlying measurements alone require.

Open3DCP is an open, flat schema for recording 3DCP mix-design and test records. The public v1.5 schema defines a broad column vocabulary spanning binder and SCM composition, aggregates, eight core fiber families plus cellulose compatibility, admixtures, pigments, fresh-state rheology, hardened mechanical properties, durability indicators, process parameters, interlayer bond, specimen/test metadata, environmental conditions, and provenance. All material quantities are recorded as mass percent of total wet mix. Missing, unreported, not-applicable, or unmeasured values are represented as NULL; 0 is reserved for explicit source-reported zero dosage or absence.

The contribution of Open3DCP is not a new test method and not a substitute for ASTM, EN, ACI, ICC, RILEM, or other standards. It is a reporting layer: a public column vocabulary that allows 3DCP records to be assembled into interoperable datasets suitable for scientific review, digital-twin reconstruction, ICME-style process-structure-property analysis, and downstream modeling where sufficient validated data are available. This paper describes the rationale, scope, and limits of the schema and identifies where additional public data and instrumentation would improve future work.

Keywords: 3D concrete printing; data schema; ICME; digital twin; rheology; interlayer bond; FAIR data; machine learning

1. Introduction

3D concrete printing has moved from laboratory demonstrations toward construction-scale experimentation. The technical literature now includes work on printable mortars, alkali-activated binders, fiber-reinforced systems, recycled aggregates, interlayer bond, anisotropic strength, rheology, buildability, and durability. The public record is large enough to support comparative analysis, but not yet consistently shaped enough to make that analysis straightforward.

The central problem is not that researchers fail to report data. Many papers report substantial detail. The problem is that the detail is not represented in a common structure. A material may appear as “GGBFS,” “slag,” “ground granulated blast-furnace slag,” or a supplier-specific name. A dosage may be reported as kg/m3, percent of binder, percent of dry mass, or percent of total batch mass. A compressive strength value may refer to a cast cube, a printed prism, a cored specimen, or a specimen loaded across the layer interface. These distinctions matter scientifically, but they are often not preserved in a machine-readable way.

Open3DCP addresses this reporting layer. It defines canonical column names, units, types, and engineering context for 3DCP-shaped records. It is intentionally flat: every analysis-relevant feature is a named column rather than a nested JSON object. This choice is not because flat tables are theoretically richer than graph schemas; it is because most researchers, dataset curators, and ML practitioners can inspect, export, merge, and model a table immediately.

The schema is best understood as an ICME substrate. Integrated Computational Materials Engineering depends on linking composition, processing, structure, properties, performance, and provenance. In 3DCP, the “processing” portion is unusually visible and unusually important: pump, hose, nozzle, motion, layer time gap, environmental exposure, and curing all participate in the final material state. A useful digital-twin record for 3DCP must therefore include both the formulation and the process by which the formulation became a printed specimen.

2. Scope and Non-Scope

Open3DCP is a schema specification. It defines how to record data; it does not provide the data itself.

Open3DCP is:

• A public column vocabulary for 3DCP mix-design and test records.
• A mass-percent, SI-unit reporting convention for material quantities and measured properties.
• A standards-aligned cross-reference layer for common material classifications and test methods.
• A flat structure intended for CSV, SQL, Parquet, dataframe, and repository-deposit workflows.
• A citable artifact with DOI and Apache-2.0 licensing.

Open3DCP is not:

• A dataset or benchmark.
• A database service or API.
• A structural design method.
• A code-compliance path.
• A replacement for ASTM, EN, ACI, ICC, RILEM, ISO, or jurisdiction-specific requirements.
• Evidence that any particular mix is safe, durable, printable, or construction-ready.

The schema can record data used in qualification or research workflows, but any construction use still requires appropriate laboratory validation, professional engineering review, and approval under the governing jurisdiction.

3. Why 3DCP Needs a 3DCP-Native Record

Conventional concrete datasets usually focus on composition, age, and one or more hardened properties. That is useful for cast concrete, where placement and compaction are often treated as standardized conditions. 3DCP makes that assumption unsafe. The manufacturing process is part of the material definition.

At minimum, a 3DCP record needs to distinguish:

• What was weighed into the mix.
• How the mix was prepared and modified over time.
• How it was pumped and extruded.
• What geometry was deposited.
• How much time elapsed between adjacent layers.
• What the environmental conditions were during deposition.
• How the specimen was cured and extracted.
• Which direction the specimen was loaded relative to the layer interface.
• Which test method or local protocol produced each value.
• Whether each value was measured, calculated, estimated, or merely reported.

Without these attributes, two identical-looking compressive strength values may describe physically different experiments. A cast cube and a printed wall coupon loaded across interlayers should not collapse into the same data point simply because both are reported in MPa.

4. Schema Design Principles

4.1 Flat Schema

Every stored feature is a named column. This preserves compatibility with SQL, CSV, spreadsheets, and dataframe tools. It also reduces ingestion ambiguity: if a source reports nozzle diameter, the target column is nozzle_diameter_mm; if it reports layer time gap, the target column is layer_time_gap_s.

Nested JSON remains useful for archival metadata, rich provenance graphs, raw sensor files, or linked image datasets. It is not the right primary container for material proportions, process parameters, or measured properties that will be used in analysis.

4.2 Mass Percent of Total Wet Mix

All material quantities are recorded as mass percent of total wet mix, water included. This convention keeps each formulation on a common 0-100 basis and avoids silently assuming density when a source reports kg/m3 without a measured unit weight.

Conversions should be documented. If a record begins as kg/m3, the total wet mass should be computed, each component converted to mass percent, and the original basis recorded in notes. If density assumptions are required, they should be explicit.

4.3 NULL Is Not Zero

Open3DCP distinguishes missingness from absence.

Use NULL when a value is:

• Unknown.
• Not reported.
• Not applicable.
• Not measured.
• Not recoverable without an assumption.

Use 0 only when the source explicitly reports a true zero or absence. For example, steel_fiber = 0 is appropriate if a paper states that no steel fiber was used. steel_fiber = NULL is appropriate if the paper simply does not discuss steel fiber. This distinction is critical for statistics and model training because false zeros can bias means, correlations, feature importance, and learned absence effects.

4.4 Standards Alignment Without Standards Substitution

Column names and descriptions reference established standards where they define material classes or test methods. Examples include ASTM C150 for Portland cement types, ASTM C618 for fly ash classes, ASTM C989 for slag, ASTM C1240 for silica fume, ASTM C33 for aggregate grading, ASTM C39 and EN 12390-3 for compressive testing, and RILEM TC 304-ADC terminology for printed-specimen orientation.

These references are interoperability hooks. They do not imply endorsement, certification, or compliance. A column can record that a result was produced under a given test method, but the schema itself cannot verify that the test was performed correctly.

4.5 Version Boundary

This paper describes public Open3DCP v1.5. Future releases may refine material subclasses or additional process telemetry, but those additions should be documented only when they are part of a public release that readers can verify from the repository and citation record.

5. What v1.5 Records

Open3DCP v1.5 documents a broad public column vocabulary across the main schema and companion tables. The exact canonical list belongs in Open3DCP_SCHEMA.md; this paper summarizes the record categories.

5.1 Composition

Composition includes Portland and blended cements, calcium aluminate and calcium sulfoaluminate cements, fly ash, slag, silica fume, metakaolin, limestone, pumice, bottom ash, rice husk ash, alkali activators, nanoscale modifiers, mineral powders, recycled sand, pigments, aggregates, fibers, admixtures, clay rheology modifiers, water, and derived ratios.

The schema preserves chemically meaningful distinctions instead of forcing all cementitious material into a single “binder” field. Fly ash class, slag, silica fume, limestone, and metakaolin are not interchangeable in hydration, packing, rheology, or long-term performance. They should remain separately recorded even when downstream analysis later derives aggregate features.

5.2 Fibers

Open3DCP v1.5 records eight core fiber families:

• steel_fiber
• pp_fiber
• pva_fiber
• glass_fiber
• basalt_fiber
• carbon_fiber
• nylon_fiber
• aramid_fiber

The schema also includes cellulose_fiber for natural-fiber and ICC-related compatibility where it is reported. All fiber dosage columns use the same mass-percent basis as other materials. Fiber geometry is captured separately through fiber_length_mm, fiber_diameter_mm, fiber_aspect_ratio, and fiber_tensile_strength_mpa.

This design leaves room for future fiber materials. New materials should be added as schema extensions only when they are distinct enough to affect analysis and have enough reporting frequency or technical rationale to justify a canonical column.

5.3 Fresh-State and Rheology

Fresh-state fields include slump, spread, J-ring, V-funnel, L-box, initial and final setting time, air content, fresh unit weight, bleeding, yield stress, static yield stress, dynamic yield stress, plastic viscosity, thixotropy, structuration rate, open time, and green strength.

These fields are essential for 3DCP because printability is not a single property. Pumpability, extrudability, shape stability, layer support, and open time can move in different directions as water, superplasticizer, VMA, accelerator, aggregate grading, and ambient conditions change.

5.4 Process Parameters

Process fields include print speed, layer height, layer time gap, nozzle diameter, nozzle shape, nozzle area, filament width, layer width, extrusion rate, number of layers, path length, infill pattern, contour count, print direction, pumping parameters, mixing parameters, and environmental conditions during printing.

These are part of the processing leg of the ICME chain. A formulation that is printable at one layer height, one speed, and one time gap may behave differently when any of those conditions change.

5.5 Specimen and Test Context

The schema records specimen preparation, geometry, dimensions, extraction method, curing conditions, test age, test method, number of specimens averaged, and test orientation. Test orientation is especially important: printed concrete can show anisotropic properties depending on whether loading is parallel to the print path, transverse within the layer plane, or perpendicular to the layer interface.

5.6 Hardened Properties, Durability, and Interlayer Bond

Hardened-property columns include compressive, tensile, splitting tensile, flexural, modulus, bond, fracture energy, toughness, impact, fatigue, density, and Poisson’s ratio. Durability fields include chloride transport, carbonation, shrinkage, creep, freeze-thaw, sulfate expansion, ASR expansion, permeability, absorption, sorptivity, scaling, corrosion indicators, thermal properties, and fire resistance. Interlayer fields record bond, shear, void area fraction, deposited air content, surface roughness, surface moisture state, and surface treatment.

The schema does not assert that every dataset should measure all of these properties. It provides stable homes when those measurements exist.

5.7 Provenance and Quality

Each record should carry enough provenance to be traced and interpreted: DOI, source citation, lab name, measurement confidence, provenance notes, quality flags, and exposure classifications. Provenance is not clerical overhead. It is the bridge between a row in a dataset and the physical or bibliographic source from which it came.

6. Digital Twin and ICME Framing

A full digital twin of a 3DCP process would include at least four layers of information:

1. Material definition: composition, supplier/product identity where available, particle characteristics, water and admixture basis, and fiber geometry.
2. Process definition: mixing, pumping, nozzle geometry, motion, extrusion rate, layer schedule, environmental exposure, and curing.
3. State and structure: fresh rheology, thixotropic recovery, interlayer surface condition, porosity, moisture state, temperature, hydration, fiber orientation, and microstructure.
4. Performance and provenance: mechanical properties, durability indicators, test method, specimen geometry, orientation, lab, confidence, and source.

Open3DCP v1.5 covers much of layers 1, 2, and 4, and a useful subset of layer 3. It should therefore be described as a practical digital-twin record for reported 3DCP experiments, not as a claim that the schema alone captures a complete real-time twin. Full process twins will require time-series machine logs, synchronized sensor data, imaging, and richer links to raw files.

This distinction matters. Open3DCP records attributes needed to reconstruct and compare many important aspects of reported 3DCP experiments, but a scalar schema is not a complete live twin of the physical process.

7. Data Gaps Worth Stating Publicly

Several data gaps are scientifically important to state clearly because they limit cross-study comparison across the field.

First, rheology is method-dependent. Vane, parallel-plate, slump flow, and field printability tests do not produce interchangeable measurements. The schema can record the result and method, but cross-method calibration remains a research need.

Second, process telemetry is underreported. Pump pressure, screw speed, motor current, toolhead acceleration, pauses, turns, and actual volumetric output can affect printed geometry and material state, but they are rarely published as synchronized datasets.

Third, interlayer state is hard to measure. Surface moisture, roughness, degree of hydration, contamination, and time gap influence bond, but standardized in-situ measurements at the moment of deposition remain limited.

Fourth, microstructure and image data do not fit comfortably in a scalar table. SEM, XRD, MIP, micro-CT, optical petrography, and thermal images should usually live as linked files or separate datasets, with Open3DCP rows carrying identifiers and summary fields.

Fifth, legacy literature is incomplete. The schema should not hide this. Missing fields should remain null, and reconstructed values should be marked with appropriate confidence.

8. Adoption Workflow

A laboratory, repository, or review team can adopt Open3DCP without changing its experimental methods.

1. Map local column names to canonical Open3DCP names.
2. Convert material quantities to mass percent of total wet mix and document the original basis.
3. Preserve missing values as NULL; use 0 only for explicit zeros.
4. Fill provenance fields before analysis fields whenever possible.
5. Record test method, specimen geometry, test age, and orientation for every mechanical result.
6. Deposit the resulting dataset in a public repository when publication rights allow.
7. Cite the schema and the original test methods.

The recommended citation is:

Sonnentag, N. (2026). Open3DCP: Open Data Standard for 3D Concrete Printing. Sunnyday Technologies. DOI: https://doi.org/10.5281/zenodo.19647471

9. Discussion

Open3DCP is deliberately limited in what it claims. It does not solve printability, certify structures, validate a model, or create data where none exists. It gives the field a common way to record what was done and what was measured.

That common structure matters because 3DCP is inherently coupled. Composition affects rheology; rheology affects pumping and deposition; deposition affects interlayer structure; interlayer structure affects mechanical response and durability; test orientation affects the measured result. A schema that records only composition and 28-day strength cannot represent that chain.

The most useful next step for the field is not necessarily a larger model. It is better public data: rows where composition, process, rheology, orientation, interlayer condition, and measured properties are reported together with provenance and missingness preserved honestly.

10. Conclusion

Open3DCP v1.5 provides a public, Apache-2.0, flat schema for 3DCP mix-design and test records. Its value is not that every column will be populated in every dataset. Its value is that when a variable matters, the field has a shared place to put it.

The schema supports ICME-style analysis by preserving composition, processing, property, condition, and provenance data in a single interoperable structure. It supports digital-twin reconstruction to the extent that public experimental records contain the necessary attributes. It also makes current limitations visible: method heterogeneity, sparse telemetry, incomplete interlayer-state measurements, and missing legacy metadata.

Open3DCP should therefore be positioned as an enabling data standard. It is a way to make 3DCP experiments more findable, comparable, and reusable. It is not a substitute for physical validation, standards compliance, or engineering judgment.

Manuscript License

Copyright (c) 2026 Sunnyday Technologies LLC and Nicholas Sonnentag.

This manuscript is licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0): https://creativecommons.org/licenses/by/4.0/

The Open3DCP schema, SQL definitions, machine-readable metadata, and repository artifacts remain licensed under the Apache License, Version 2.0, unless a specific file states otherwise. Third-party standards, publications, figures, trademarks, and cited works remain the property of their respective rights holders and are not relicensed by this manuscript.

References

• Abrams, D. A. (1918). Design of Concrete Mixtures. Bulletin 1, Structural Materials Research Laboratory, Lewis Institute.
• Bos, F. P., Wolfs, R. J. M., Ahmed, Z. Y., and Salet, T. A. M. (2016). Additive manufacturing of concrete in construction: potentials and challenges of 3D concrete printing. Virtual and Physical Prototyping, 11(3), 209-225. https://doi.org/10.1080/17452759.2016.1209867
• Buswell, R. A., Leal de Silva, W. R., Jones, S. Z., and Dirrenberger, J. (2018). 3D printing using concrete extrusion: A roadmap for research. Cement and Concrete Research, 112, 37-49. https://doi.org/10.1016/j.cemconres.2018.05.006
• Chen, Y., Jansen, K., Zhang, H., Romero Rodriguez, C., Gan, Y., Copuroglu, O., and Schlangen, E. (2020). Effect of printing parameters on interlayer bond strength of 3D printed limestone-calcined clay-based cementitious materials: An experimental and numerical study. Construction and Building Materials, 262, Article 120094. https://doi.org/10.1016/j.conbuildmat.2020.120094
• De Schutter, G., Lesage, K., Mechtcherine, V., Nerella, V. N., Habert, G., and Agusti-Juan, I. (2018). Vision of 3D printing with concrete: Technical, economic and environmental potentials. Cement and Concrete Research, 112, 25-36. https://doi.org/10.1016/j.cemconres.2018.06.001
• International Code Council. (2026). ICC 1150-2026 Standard for Automated Construction Technology for 3D Printing Walls. ICC. ISBN 978-1-971077-70-3.
• Roussel, N. (2018). Rheological requirements for printable concretes. Cement and Concrete Research, 112, 76-85. https://doi.org/10.1016/j.cemconres.2018.04.005
• Sanjayan, J. G., Nematollahi, B., Xia, M., and Marchment, T. (2018). Effect of surface moisture on inter-layer strength of 3D printed concrete. Construction and Building Materials, 172, 468-475. https://doi.org/10.1016/j.conbuildmat.2018.03.232
• Wangler, T., Roussel, N., Bos, F. P., Salet, T. A. M., and Flatt, R. J. (2019). Digital concrete: A review. Cement and Concrete Research, 123, Article 105780. https://doi.org/10.1016/j.cemconres.2019.105780
• Wilkinson, M. D., Dumontier, M., Aalbersberg, I. J., Appleton, G., Axton, M., Baak, A., Blomberg, N., et al. (2016). The FAIR Guiding Principles for scientific data management and stewardship. Scientific Data, 3, Article 160018. https://doi.org/10.1038/sdata.2016.18
• Wolfs, R. J. M., Bos, F. P., and Salet, T. A. M. (2019). Hardened properties of 3D printed concrete: The influence of process parameters on interlayer adhesion. Cement and Concrete Research, 119, 132-140. https://doi.org/10.1016/j.cemconres.2019.02.017