Performance Assessment of One-Part Self-Compacted Geopolymer Concrete Containing Recycled Concrete Aggregate: A Critical Comparison Using Artificial Neural Network (ANN) and Linear Regression Models

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Fig. 1: It is recommended to add hidden layers as well as an activation function to Fig. 1. ANN models are generally developed by considering hidden layers to enhance their accuracy.

Section 2.1: The authors should mention if the RCA are treated before being added to the concrete mixture and discuss how treating can affect the concrete properties. The following reference might be useful:

• Ismail, S., & Ramli, M. (2013). Engineering properties of treated recycled concrete aggregate (RCA) for structural applications. Construction and Building Materials, 44, 464-476.

The authors are recommended to add histograms of the input parameters. Histograms clearly illustrate the distribution of the data.

The database used for developing the ML models is significantly small. This prevents the model from being accurate and generalized. The authors should discuss this matter in the manuscript.

Do the values reported in Tables 5 and 6 correspond to the training database or the testing database? The performance metrics for both training and testing should be provided and discussed for a better evaluation of the models.

To facilitate easier comparison of the models, the authors can add a Taylor diagram which aids in the accuracy assessment of different models. The following reference could be helpful regarding Taylor diagrams:

• Dabiri, H., Clementi, J., Marini, R., Mugnozza, G. S., Bozzano, F., & Mazzanti, P. (2024). Machine learning-based analysis of historical towers. Engineering Structures, 304, 117621.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript is interesting and contains a good study. However, the innovation part has not been properly highlighted or deficient. The paper has been made too complex for audience and lacks coherence. How can this study be used further? Here are some comments 

Improve the wording of manuscript for understanding of broad audience.

Innovation part is not properly highlighted in abstract.

Improve the quality of Fig.18

Which software was used. Describe the methodology as well

Has the author validated the results with other studies?

On what basis, the input parameters were selected?

 

 

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The paper evaluates the “Performance Assessment of One-Part Self-Compacted Geopolymer Concrete Containing Recycled Concrete Aggregate: A Critical Comparison of Artificial Neural Networks (ANN) and Linear Regression Models”.  The paper could be of interest to the readers of the Recycling Journal. Some amendments are suggested to improve the manuscript:

1.       Typically, the components of mix design are regarded as the input parameters for model development, as indicated in Table 1. However, this study has considered only a few selected parameters (Section 2.3). Please provide a rationale for the selection of these specific parameters. Could you provide a correlation plot to justify the exclusion of certain parameters?

2.       The validation data performance curves for the ANN model, based on the Bayesian regularization (BR) algorithm, are not provided. It is recommended to include these curves, as they will offer a clearer understanding of the model's accuracy and reliability.

3.       The fly ash classification can significantly influence the properties of the resultant concrete mixes. However, this information is missing in the manuscript. Please include the fly ash classification and provide a related discussion on the test results.

4.       The properties/composition of the constituent materials used (fly ash, micro fly ash, GGBS, recycled aggregates, natural aggregates, basalt fibers) need to be incorporated.

5.       What was the source of the Recycled Aggregates? Furthermore, the characteristics (shape, size, gradation, water absorption) of the natural and recycled aggregates need to be provided.

6.       What was the moisture condition of the recycled aggregates at the time of mixing? This is an important parameter to consider, as recycled aggregates generally exhibit higher water absorption than natural aggregates. If added dry, recycled aggregates can lower the water-binder ratio, ultimately affecting the resulting concrete properties.

7.       The results of the Rheological properties of the SCGC mixes, including flowability, viscosity, and passing ability, are not provided.

8.       There are inconsistencies in the layout and formatting of the figures that require correction.

 

9.       It is recommended to carefully check English grammar, sentence formation, and spelling mistakes to enhance the quality of the manuscript. Currently, there are some mistakes. 

Comments on the Quality of English Language

It is recommended to carefully check English grammar, sentence formation, and spelling mistakes to enhance the quality of the manuscript. Currently, there are some mistakes. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf