Evaluating the impact of demolished concrete aggregates on workability, density, and strength with predictive modeling
Abstract Concrete production heavily relies on natural aggregates, leading to environmental concerns due to resource depletion and extraction impacts. This study explores the feasibility of using demolished concrete aggregate (DA) as a sustainable alternative. Despite being a major global waste prod...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2025-04-01
|
| Series: | Discover Civil Engineering |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s44290-025-00230-y |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Concrete production heavily relies on natural aggregates, leading to environmental concerns due to resource depletion and extraction impacts. This study explores the feasibility of using demolished concrete aggregate (DA) as a sustainable alternative. Despite being a major global waste product, DA’s effects on concrete properties remain underexplored, particularly in regions with limited natural aggregates. This research examines DA’s impact on workability, density, and compressive strength, alongside machine learning-based performance predictions. Concrete mixes with 0–30% DA replacement was evaluated. The results showed decrease in workability as the DA increases, with a maximum slump reduction of 72.7%. Bulk density showed a slight decline, while compressive strength decreased significantly, reaching a 56.7% reduction at 30% DA replacement. ANOVA results (F = 12.97, p = 1.84E− 07) confirmed significant strength differences, with post-hoc tests indicating no significant effect at 2–7% replacement but notable reductions at ≥ 10% (p < 0.05). Three machine learning models: linear regression, polynomial regression, and artificial neural networks, were developed to predict compressive strength. Linear regression (R2 = 0.801, MAE = 0.663, RMSE = 0.757, SI = 0.060) outperformed polynomial regression and ANN models. The findings underscore DA’s potential as a sustainable aggregate alternative, emphasizing the importance of optimizing replacement levels to maintain concrete performance. |
|---|---|
| ISSN: | 2948-1546 |