Central composite design application in the optimization of the effect of pumice stone on lightweight concrete properties using RSM
Among other lightweight aggregates, one choice is pumice stone (PS), which can replace coarse aggregates in concrete. In this study, the natural coarse aggregates are replaced by various per-centages of PS in concrete to examine its physical and mechanical properties. Experiments are designed to pre...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Published: |
Elsevier
2023
|
| Subjects: | |
| Online Access: | http://eprints.um.edu.my/38453/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Among other lightweight aggregates, one choice is pumice stone (PS), which can replace coarse aggregates in concrete. In this study, the natural coarse aggregates are replaced by various per-centages of PS in concrete to examine its physical and mechanical properties. Experiments are designed to prepare mixtures (by varying ratios of PS) using response surface methodology (RSM). Slump and compaction factor tests are used to investigate the fresh properties, such as the workability of concrete. In contrast, the hardened concrete properties are determined using compressive strength (CS), flexural strength (FS), and split tensile strength (STS). The experi-mental, optimization and validation of the model results show that up to 30% of the PS can be replaced in the lightweight aggregate concrete for 7 - 56 curing days in which the CS is greater than 15 MPa, STS is 7-12% of the CS, and FS is 9 - 11% of the CS. Therefore, the optimum CS for the current study ranges from 17 to 20 MPa, in which the STS is 9.39 - 10.54% of the CS, and FS ranges from 17.49% to 21.12% of the CS. The quadratic model was proposed, and the ANOVA and model fitness was produced through RSM, with the coefficient of determination (R2) above 0.99 (99%) for all three responses representing the high importance of the suggested model. Besides, the difference among the R2 and adjusted R2 was negligible, and the P-Value for CS, STS, and FS was below 0.05 (<0.0001), indicating the model's high relevance. RSM was used to improve the model, which was then verified in further lab trials. |
|---|