Assembling the evidence…
The use of excavated earth is a circular and carbon neutral solution gaining growing interest. However, its high variability in properties makes it difficult to master as raw materials for construction. The objective of this paper is to highlight the key parameters allowing overcoming the influence of the variability of soil composition on the earth-based concrete properties. We show that the solid volume fraction of clay paste is the key parameter allowing predicting the fresh state properties of the clay binder, whatever the constitutive clay type, whereas the clay activity and the dry density of the material control the compressive strength of the material. These promising results, validated on sludge, pave the way to optimized mix design strategies for robust earth concrete products.
We show that the solid volume fraction of clay paste is the key parameter allowing predicting the fresh state properties of the clay binder, whatever the constitutive clay type
same-data model fitting and mineralogy-dependent divergence below 0.75 ϕ/ϕm limit the claimed across-clay prediction
the clay activity and the dry density of the material control the compressive strength of the material
cross-clay correlations confound activity with mineralogy and particle characteristics rather than independently identifying strength control
for a solid volume fraction higher than 0.75 ϕ/ϕm, YODEL model is able to predict the yield stress of all tested clays with the same pre-factor m1 value
the concentration-qualified relationship fits the tested clays, but parameters are calibrated on the same observations without quantified prediction error
Derived from the full evaluation — not a separate score.
Strengths
Nine clays are characterized by mineralogy, particle-size distribution, density and methylene blue value, allowing the proposed descriptors to be examined across materially different binders.
↳ Sections 2.1 and 4.2–4.3; Tables 1–2
Fall-cone measurements are supplemented by vane rheometry where the paste is too soft for the penetration method. Standardized procedures also support material characterization and compression testing.
↳ Sections 2–3; Section 4.2; Figures 2–3
Nine sludges provide a separate material set for examining whether the clay-derived trends extend beyond the characterized powders. This supports a potential screening pathway while remaining a laboratory-scale check.
↳ Sections 2.3 and 4.4; Figure 7
Limitations
The central relationships are assessed through graphical comparisons without reported fit statistics or prediction uncertainty. Repeat counts do not establish how many independently prepared batches underlie the plotted observations.
↳ Sections 3.2–3.3 and 4.2–4.4; Figures 1–7
The abstract's across-clay prediction wording omits the divergence below 0.75 ϕ/ϕm. Its strength-control wording also assigns more authority to the cross-material correlations than the confounded design establishes.
↳ Abstract; Sections 4.2–4.3; Section 5
The data-availability statement says no data were used despite the reported measurements, leaving their availability unresolved. Separately, the text and Figure 4 caption give conflicting definitions of the normalization parameter.
↳ Data availability, page 8; Section 4.2; Figure 4 caption
The nine-clay comparison and additional sludge set support a useful organizing framework for fine-fraction earth binders. Standardized characterization and complementary yield-stress methods support the experimental program, but repeated measurements do not establish independent batch replication. Sections 4.2–4.4 present fitted graphical relationships without quantified predictive accuracy, limiting the authority of the abstract's universal prediction and strength-control language. The contradictory data-availability declaration and conflicting normalization definitions require clarification without themselves demonstrating that the measurements are invalid.
Nabu’s assessment, alongside the field’s view.
Are you an author of this paper?
Sound3.3
Confidence mediumThe comparison across nine clays and extension to nine sludges meaningfully connects geotechnical descriptors with binder performance. The advance remains bounded by fitted relationships, unquantified predictive precision and an incompletely delineated increment over earlier work.
Standardized characterization, complementary yield-stress methods and repeated mechanical measurements support the experimental program, but independent experimental units and prediction uncertainty remain undefined. The contradictory data-availability declaration is surfaced separately through the reliability flag and does not cap this merit-based score.
The materials-to-modeling sequence is followable, but conflicting normalization definitions and incorrect cross-references obstruct reproducibility. The abstract omits the concentration boundary that materially restricts its headline finding.
The paper connects earth construction, geotechnical characterization and concrete rheology while explicitly limiting the study to fine fractions. Transfer from oven-dried micromortars to service-condition earth concrete is less fully developed.
Concerns4 of 4 checks
The normalization parameter ϕm has conflicting definitions in the results text and a figure caption. Central abstract wording also extends beyond the concentration-dependent and correlational findings described in the results.
The data-availability declaration contradicts the reported experimental study and leaves the underlying measurements unauditable from the supplied material. Competing interests are declared, and the materials-only design does not require human or animal ethics approval.
Flags: 1 declared / 5 total
84 references in manuscript 65 of 68 checkable references found in an index 16 are books, websites or datasets — counted, but not index-checkable
No retraction notice found in Retraction Watch.
Sources: Retraction Watch ✓
Where this paper’s evidence sits on the path from initial observation to real-world use.
The evidence supports laboratory proof of concept rather than deployment-ready guidance. Complete earth-concrete components and service conditions are not tested.
AI-generated, human-governed. Something look off? Contact us to request a review.