An analytical model suggests that porous solids can appear to become stiffer faster than expected as they become denser partly because their internal networks change along the way. The framework treats density and topology — the arrangement and connectivity of the network — as coupled structural features, rather than assuming that only the material’s local struts determine the modulus–density relationship.
That distinction matters when a modulus–density scaling exponent — the number describing how quickly stiffness rises with relative density — is larger than the value expected from the mechanics of an individual strut. The authors interpret such an apparent excess as something topology evolution may partly explain, without requiring a change in the local way those struts deform.
A network can add its own contribution
The study separates the scaling into two pieces. One is the intrinsic exponent associated with local strut mechanics. The other comes from a density-dependent prefactor that represents the changing network. To obtain the apparent exponent, the model takes logarithms of the modulus relation and differentiates with respect to logarithmic relative density, allowing the topology-related contribution to be written separately from the local-mechanics term.
The framework examines three descriptors of that network. Mean coordination number captures how connections are arranged; the load-bearing backbone fraction represents the share of solid that participates in carrying load; and tortuosity describes the efficiency of the paths through which load travels.
In the coordination-number version, the apparent exponent combines the intrinsic term with a connectivity-sensitivity term based on how threshold-adjusted coordination changes with density. In plain terms, a network whose effective connectivity evolves as the material becomes denser can add to the slope of the modulus–density curve.
Three routes to a steeper curve
The backbone model gives a similar result through a different structural measure. When the load-bearing backbone fraction increases with density, the apparent exponent rises above the intrinsic exponent. If that fraction remains constant, the model returns the intrinsic exponent instead, leaving no added backbone contribution.
The tortuosity model reaches the same broad conclusion from load-path efficiency. The apparent exponent is reduced by the logarithmic density derivative of tortuosity; under the model’s stated trend, tortuosity increases as density falls, so the resulting apparent exponent is higher than the intrinsic one.
Taken together, the three modeled topology effects can produce an apparent exponent above the intrinsic value, especially at low relative density. The result is a proposed explanation for why a porous network may show a steeper modulus–density relationship than local strut mechanics alone would suggest.
What the curves do — and do not — establish
The paper uses illustrative parameter sweeps to show how network sensitivity can alter the shape of predicted modulus–density curves. In one backbone illustration, the intrinsic exponent is held constant while a topology-sensitivity parameter is varied, producing curves with different shapes. The exercise shows how the same local-mechanics value can coexist with different apparent scaling behavior when topology changes.
Those curves are model outputs, not measurements from a named material. The study reports no physical specimen sample or empirical dataset, and it does not establish a modulus-scaling exponent for any particular porous material or material class. Its results are analytical statements about the proposed equations and the assumed trends in the structural descriptors.
The density relationships used for coordination, backbone fraction and tortuosity are modeled or illustrated rather than estimated from a defined dataset. The framework treats those structural features as coupled, but the supplied analysis does not provide a material-specific test of how accurately their combined behavior predicts modulus.
The next test is empirical
Applying the framework to real disordered networks would require measurements or fitted relations linking each descriptor to relative density, followed by comparisons between its predictions and measured moduli across different porous-material classes. The supplied analysis identifies those comparisons as unresolved, including how simultaneous coupling among the descriptors should be represented.
No statistical uncertainty, predictive error or confidence interval is reported, and no independent empirical validation is supplied. The preprint therefore does not establish that topology explains any individual experiment, or that topology has replaced a change in local deformation mode as the explanation for a particular result.
The document is an arXiv preprint, version 1, dated 20 Aug 2026, listed as arXiv:2608.20003v1.
Paper data and sources
Original title: Role of topology in scaling laws for studying mechanics in open-porous solids: Moving beyond classical Gibson-Ashby scaling
Authors: Ameya Rege
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-20
DOI: Not available
Original paper · Full text