article · Structural Concrete
Self-compacting concrete flows freely through reinforcements to fill formwork completely without mechanical vibration. Understanding its 28-day compressive strength based on how it behaves during flow is critical for its effective use. Artificial neural networks were trained using both empirical flow test parameters and fundamental rheological values. Specifically, the models took inputs including slump flow diameter, the L-Box H2/H1 ratio, V-Funnel flow time, plastic viscosity, and yield stress. By evaluating various network designs through numerical and experimental investigation, an optimal artificial neural network architecture was identified. The most effective configuration featured two hidden layers structured as 5-50-50-1, which successfully predicted the 28-day compressive strength of self-compacting concrete with a strong Pearson correlation coefficient of 97.58 percent.
Accurately estimating the strength of specialised building materials ensures structures are safe and durable. By relying on early fluid measurements rather than waiting weeks for standard curing tests, engineers can evaluate concrete performance more rapidly. This helps construction teams better understand how the flow characteristics of self-compacting mixtures relate directly to their ultimate hardened strength.
This research provides a computational method that construction quality control teams, materials testing laboratories, and ready-mix concrete suppliers could use to forecast compressive strength from standard fluid and rheological tests. Given that the findings are based on an experimental and numerical study presenting an optimal trained model architecture, the approach represents early-stage applied research that would require integration into practical testing software or workflows before market adoption.
AI-generated from the published abstract. Always read the original work before citing.
Abstract Self‐compacting concrete (SCC) is a fluid concrete designed to flow freely through reinforcements in order to completely fill the formwork. The appearance of this type of concrete increases the need to precisely characterize its compressive strength as a function of their behavior during flow. This article summarizes the use of artificial neural networks for the modelization of compressive strength, at 28 days, of SCC based on rheological parameters found during empirical tests (slump flow diameter, H2/H1 ratio of L‐Box, and V‐Funnel flow time) and the values of plastic viscosity and the yield stress. The objective of this numerical and experimental study is to find an optimal model to modelize the compressive strength. Thus, the results obtained after training of several models are showed that the architecture of the optimum with two hidden layers model is 5‐50‐50‐1 with a Pearson's correlation R = 97.58%.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.1002/suco.202100796
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
